alces26_142.pdf alces(25)_146.pdf alces27_50.pdf alces29_263.pdf alces28_31.pdf alces26_37.pdf alces21_447.pdf alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces(23)_157.pdf alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces26_9.pdf alces24_1.distinguishedmoosebio.pdf alces28_189.pdf alces21_299.pdf alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces22_323.pdf alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces20_95.pdf alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces(25)_75.pdf alces28_101.pdf alces27_183.pdf alces27_118.pdf alces21_161.pdf alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces24_1.pdf alces24_159.pdf alces24_48.pdf alces20_1preface.pdf alces vol. 20, 1984 alces26_73.pdf alces(23)_301workshopsessions.pdf alces vol. 23, 1987 alces22_437workshopsessions.pdf alces vol. 22, 1986 alces21_55.pdf alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces(25)_15.pdf alces22_181.pdf alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces24_102.pdf alces26_14.pdf alces(25)_167.pdf alces28_203.pdf alces27_65.pdf alces21_475.pdf alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces29_267.pdf alces27_74.pdf alces28_35.pdf alces21_321.pdf alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces(23)_1.pdf alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces20_107.pdf alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces28_111.pdf alces27_193.pdf alces22_345.pdf alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces24_56.pdf alces26_80.pdf alces(23)_303distinguishedmoosebio.pdf alces vol. 23, 1987 alces22_439_nbmoosemodel.pdf alces vol. 22, 1986 alces vol. 22, 1986 alces21_191.pdf alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces(25)_25.pdf alces21_77.pdf alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 48, 2012 henningsen et al. – elaeophora in wyoming moose 35 distribution and prevalence of elaeophora schneideri in moose in wyoming john c. henningsen1, amy l. williams1,2, cynthia m. tate3, steve a. kilpatrick1,4, and w. david walter5 1wyoming game and fish department, po box 67, jackson, wyoming 83001, usa; 3wyoming game and fish departmentwildlife diseases laboratory, 1174 snowy range road, laramie, wyoming 82070, usa; 5usda/aphis/ws national wildlife research center, 4101 laporte avenue, fort collins, colorado 80521, usa. abstract: elaeophora schneideri causes disease in aberrant hosts such as moose. documented e. schneideri infections in moose are relatively rare, yet noteworthy enough that individual cases describing morbidity and mortality have been the norm for reporting. surveillance efforts for e. schneideri in wyoming moose in the 1970s found zero cases, but since 2000 several moose in wyoming discovered dead or showing clinical signs of elaeophorosis have been found infected with e. schneideri. in 2009 we searched for worms in the carotid arteries of 168 hunter-harvested moose from across wyoming to determine the prevalence and distribution of e. schneideri in moose; 82 (48.8%; 95% ci: 41.4-56.3%) were positive for e. schneideri. prevalence did not differ between sexes or among age classes but there was difference in prevalence among herd units (range = 5-82.6%). intensity of infection (range = 1-26 worms) did not differ between sexes, among age classes, or among herd units. our findings indicate that moose do not succumb to the parasite to the extent previously thought. prevalence and intensity were constant across age classes, suggesting that infected moose are surviving and an acquired, immunological resistance to further infection develops. in addition, moose might sometimes act as natural hosts to the parasite, as indicated by 1) high prevalence of infection in moose in areas where sympatric mule deer had much lower prevalence of infection, and 2) preliminary necropsy findings that revealed microfilariae in skin samples from 3 moose. however, negative impacts to moose and moose populations cannot be ruled out entirely, as this study was limited to apparently healthy hunter-harvested animals. while moose appear to often survive infection with e. schneideri, prevalence of ~50% is still cause for concern because it is unknown to what extent this parasite causes subclinical effects in moose that might impact recruitment or productivity. subsequent research on moose herds where e. schneideri occurs should consider the effects of elaeophorosis and attempt to clarify its role. alces vol. 48: 35-44 (2012) key words: alces alces, arterial worm, disease, elaeophora schneideri, elaeophorosis, moose, parasite, wyoming. elaeophora schneideri is a filarioid nematode that lives in the cephalic arteries of mule deer (odocoileus hemionus) (hibler et al. 1970) and black-tailed deer (o. hemionus columbianus) (weinmann et al. 1973). adult nematodes in the arteries of deer give birth to live young microfilariae (hibler and adcock 1971). microfilariae then migrate to the capillaries in the dermis of the host’s face and forehead where they are taken up in the blood meal of the intermediate host. horse flies (family tabanidae) of the genera hybomitra, silvius, and tabanus (hibler et al. 1970, clark and hibler 1973, espinosa 2present address: university of wyoming, department of veterinary sciences, 1174 snowy range road, laramie, wyoming 82070, usa. 4present address: conservation research center of teton science schools, 700 coyote canyon, jackson, wyoming 83001, usa. elaeophora in wyoming moose – henningsen et al. alces vol. 48, 2012 36 1983) are intermediate hosts of the parasite. transmission of the third stage infective to the vertebrate host occurs after e. schneideri larvae develop in the horse fly vector for 2-3 weeks (hibler and adcock 1971, hibler and metzger 1974, davies 1979). development of e. schneideri in the definitive host has been described previously (weinmann et al. 1973, hibler and metzger 1974). less is known about development of e. schneideri in aberrant hosts but pathogenesis usually stems from the parasite’s delayed migration through the host body or complications from circulatory impairment (i.e., elaeophorosis; adcock and hibler 1969, hibler and metzger 1974, anderson 2001). gross clinical signs of infection among aberrant hosts range from dry gangrene of nose and ear tips and antler malformations, to blindness, central nervous system damage, and death. documented aberrant hosts of e. schneideri are moose (alces alces), elk (cervus elaphus), white-tailed deer (o. virginianus), bighorn (ovis canadensis), and domestic sheep (boyce et al. 1999, anderson 2001). e. schneideri is widespread across north america occurring in mule deer in nebraska (mckown et al. 2007), south dakota (jacques et al. 2004), utah (pederson et al. 1985), texas (pence and gray 1981), colorado, and new mexico (davies 1979). it has been documented in white-tailed deer in arizona (hibler 1982), texas (waid et al. 1984), and several southeastern states (prestwood and ridgeway 1972). infected elk have been reported from oklahoma, new mexico, arizona, colorado, and wyoming (hibler 1982). the first documented e. schneideri infections in moose were in montana (worley et al. (1972). subsequently, its presence in small numbers of moose was documented in utah (jensen et al. 1982), colorado (madden et al. 1991),washington (pessier et al. 1998), and wyoming in 2000 (w. e. cook, wyoming game and fish department [wgfd], unpublished report), and oregon in 2010 (matthews, 2012). in wyoming moose both the prevalence of infection and the parasite’s geographic extent appear to have undergone a recent, notable increase. in 1973-74 worley (1975) examined 74 apparently healthy, hunter-harvested moose: 69 from teton and fremont counties in northwestern wyoming, and 5 from park and gallatin counties in southwestern montana. no wyoming moose and only 3 of 5 montana moose were infected with worms in the carotid arteries. presumably, low prevalence of e. schneideri in wyoming led hibler (1982) to believe elaeophorosis was of minimal importance to wyoming elk and moose. whereas e. schneideri has been documented consistently in small numbers of mule deer and elk in wyoming since 1967 (h. e. edwards, wgfd, unpublished data), the first infected moose was not identified until much later. within 2 weeks in january 2000, 2 moose were euthanized by wgfd field personnel in fremont county (central wyoming) because they were lethargic or walking in circles and showed signs of impaired vision; illness in each of those cases was attributed to elaeophorosis (w. e. cook, unpublished report). in 2008 a 3-yr-old cow moose in western wyoming was euthanized because of its abnormal behavior (lack of fear, blindness, and loss of motor skills). upon gross examination, a heavy load of worms (30-50) was found in the carotid arteries and its clinical signs were attributed to elaeophorosis (c. m. tate, wgfd, unpublished report). the wgfd and the wyoming state veterinary lab (wsvl) increased opportunistic surveillance of moose in 2008. animals found dead, euthanized due to illness, and road-kills were examined for e. schneideri; several were found infected with e. schneideri. most positive cases were from animals discovered dead or showing clinical signs of illness, and pathology associated with e. schneideri was implicated in several cases. in order to survey moose for e. schneideri alces vol. 48, 2012 henningsen et al. – elaeophora in wyoming moose 37 more uniformly across wyoming, a rigorous plan was developed to establish baseline data on prevalence and distribution of e. schneideri by surveying hunter-harvested moose during the 2009 hunting season. to our knowledge, this was the most comprehensive and widespread effort to date for surveillance of e. schneideri in moose. study area brimeyer and thomas (2004) described the history and status of moose in wyoming through the early 2000s. moose in wyoming occupy 3 distinct ranges: 1) bighorn mountain range in north-central wyoming, 2) the snowy range and sierra madre ranges of southeast and south-central wyoming, and 3) western wyoming among comparatively connected mountain ranges from the utah border north through yellowstone national park (fig. 1a). moose in wyoming are managed as 11 herd units (herds) comprising discrete populations for which migration among adjacent herds is thought to account for <10% of a herd population. these herds are further divided into 43 hunt areas to provide flexibility for hunting seasons; begin and end dates of hunting seasons vary among areas. the wgfd has a statewide population objective of 14,630 moose (post-hunt), yet population estimates are considered relatively unreliable or completely lacking in most herds (thomas 2008). a b fig. 1. a) moose herd units that hunter-harvested moose were collected and sampled for elaeophora schneideri in 2009 in wyoming, usa. b) intensity of elaeophora schneideri found in carotid arteries categorized as none (●), low (○), moderate (○), and high (○) in hunter-harvested moose by herd unit in 2009 in wyoming, usa. elaeophora in wyoming moose – henningsen et al. alces vol. 48, 2012 38 methods we examined hunter-harvested moose for the presence of immature or adult worms in the terminal portion of the common carotid arteries and in some instances the proximal portion of the internal maxillary arteries (hereafter field examinations). many field examinations were conducted at hunter check stations and during opportunistic field checks of successful hunters. field examinations also took place when hunters brought heads of harvested moose to wgfd regional offices, taxidermists, or meat processors. incisor teeth were collected from harvested moose for aging by cementum annuli. the specific ages obtained via cementum annuli are reported in whole years (i.e., yearling is 1). successful development and migration of e. schneideri to the carotid arteries of moose would be expected to take 5-6 months (hibler and metzger 1974), thus nematodes would not be expected in the carotids until typically december. thus field examinations as conducted in this study could not adequately diagnose infections in calves, and surveillance of calves was not included in this study. intensity of infection with worms (bush et al. 1987) was recorded as 1 of 3 categories: 1-6, 7-13, and ≥14. these categories were based on intensities observed previously in moose from wyoming and other states (worley et al. 1972, worley 1975, madden et al. 1991). in addition to examining for the presence of e. schneideri, visual signs of elaeophorosis (e.g., cropped ears, necrotized tissues, lesions, or malformed antlers) were recorded. statistical analysis prevalence was based on the total number of positive individuals; 95% confidence intervals (ci) around these proportions were calculated based on the binomial distribution (rózsa et al. 2000). fisher’s exact test was used to test for homogeneity in prevalence between sexes. chi-square was used to test for homogeneity in prevalence among age classes and herds; fisher’s exact test was used to test for differences between pairs of herds when contingency tables had ≤5 observations in ≥1 cell. because of the small number of examined animals from individual age classes, 4 combined age classes were created (1, 2-4, 5-7, ≥ 8) for statistical comparisons. likewise, some adjacent herds (jackson and targhee, lincoln and uinta) were combined to obtain adequate sample sizes for analyses. some herds were dropped from analyses because they had both small sample sizes and were too geographically separate to justify merging. because intensity was recorded as a categorical variable, chi-square was used to test for differences in intensity among sexes, age classes, and herds. statistical significance was set at p ≤0.05 for all tests. calculations were accomplished using sigmaplot 11 (systat software, san jose, ca.). results prevalence the reported harvest in fall 2009 was 548 moose (394 adult males, 135 adult females, and 19 calves); 126 males and 42 females were examined for e. schneideri from 1 september-14 november (table 1), or 31% of adult females and 32% of adult males in the harvest. e. schneideri was present in the carotid arteries of 48.8% of all moose ≥1 yr of age (95% ci: 41.4-56.3). exactly 50% of males (95% ci: 41.4-58.6%) and 45.2% of females were infected (95% ci: 31.2-60.1%); prevalence did not differ by sex (χ2 = 0.127; p = 0.72). the number of females checked in each herd was small, but prevalence did not appear to differ between sexes within any individual herd. thus, we combined sexes for statistical comparisons of prevalence among age classes and herds. ages were obtained from 151 of 168 moose: 9 were yearlings, 54 were 2-4 years old, 71 were 5-7 years old, and 17 were ≥8 years. the infection rate was 56% in yearlings (95% ci = 26.6-81.2%), 43% in 2-4 year olds alces vol. 48, 2012 henningsen et al. – elaeophora in wyoming moose 39 (95% ci = 30.3-55.8%), 56% in 5-7 year olds (95% ci = 44.8-67.3%), and 41% in those ≥8 years (95% ci = 21.6-64.0%). there were no statistical differences in prevalence of e. schneideri among age classes (χ2 = 2.420; df = 3; p = 0.490). moose from 10 herds were checked for e. schneideri (table 1). to obtain adequate sample size for statistical comparison, the adjacent jackson and targhee herds, and the lincoln and uinta herds were combined; the absaroka, dubois, and lander herds were dropped because of low sample sizes and geographic separation (table 1). prevalence was different geographically (χ2 = 27.082, df = 4, p < 0.001). the lowest prevalence occurred in the bighorn herd (5%; 95% ci = 0-25.4%) and was lower than that in the other 4 herds in the analysis. the snowy range herd had the highest prevalence (82.6%; 95% ci = 62.3-93.6%) which was higher than in the bighorns, lincoln-uinta (43.8%; 95% ci = 23.1-66.8%), and sublette herds (52.2%; 95% ci = 42.0-62.2%), but not different than in the jackson-targhee herds (61.5%; 95% ci = 35.4-82.4%). intensity of the 82 positive cases, we found 44, 26, and 11 moose with low, moderate, and high e. schneideri intensity, respectively (table 1); intensity was not recorded for 1 positive individual. the greatest number of worms counted in any moose was 26. parasite intensity was similar between sexes (χ2 = 0.564; df = 2; p = 0.754) and among age classes (χ2 = 4.177; df = 6; p = 0.653). a low-intensity worm burden was most common in all age classes, ranging from 40-71%. none of the age classes had a large number of high-intensity worm loads. high-intensity infections were not observed in the 7 infected moose in the oldest age class (≥8 years). although prevalence was high in snowy range moose, most (74%) had low-intensity infections (fig. 1b). similarly, most positive individuals in the jackson-targhee herd (88%) had low-intensity infections. the only infected moose found in the bighorns had a moderate-intensity infection (table 1). the pattern of intensity was reversed in the lincoln-uinta and sublette herds; more individuals had moderate and high intensities than low intensities. however, patterns of intensity were not different among herd units no. moose with differing intensities of infection herd no. examined no. infected % infected (prevalence) no worms low (1-6) moderate (7-13) high (≥14) absaroka 1 0 0 1 0 0 0 big horns 20 1 5 19 0 1 0 dubois 1 0 0 1 0 0 0 jackson 10 6 60 4 6 0 0 lander 4 0 0 4 0 0 0 lincoln 13 5 38.5 8 2 2 1 snowy range1 23 19 82.6 4 13 4 1 sublette 90 47 52.2 43 21 18 8 targhee 3 2 66.7 1 1 1 0 uinta 3 2 66.7 1 1 0 1 total 168 82 48.8 86 44 26 11 table 1. prevalence of elaeophora schneideri in hunter-harvested moose, wyoming, usa, 2009. 1includes 1 animal found positive for e. schneideri for which number of worms was not recorded. elaeophora in wyoming moose – henningsen et al. alces vol. 48, 2012 40 (χ2 = 11.950; df = 8; p = 0.153). clinical signs when possible, tissues were examined for gross evidence of damage as a result of infection by e. schneideri. more thorough examinations only occurred after heads had been prepared for taxidermy or when hunters donated their antlerless specimens. of the 31 infected moose that were thoroughly examined, 10 showed visual signs of elaeophorosis: 7 displayed cropped or hardened ears, 1 had antler malformation, and 2 had cropped ears and antler malformation. three of the 10 moose with visual signs had low-intensity infections, 4 had moderate-intensity, and 3 high-intensity infections. discussion prevalence of e. schneideri in wyoming moose was much higher than anticipated. documented infections in moose have been fairly rare and noteworthy enough that individual cases have been the norm for reporting (worley et al. 1972, jensen et al. 1982, madden et al. 1991, pessier et al. 1998). the prevalence reported here is probably biased low because only the main cephalic arteries were examined for worms, yet post-mortem migration of worms occurs (adcock and hibler 1969). furthermore, there was potential for false negatives because the length of the carotid artery was often short and compromised from hunter processing; there was no corresponding risk of false positives. prevalence of e. schneideri in adult mule deer has been 100% in certain local populations in the southwestern united states (hibler and adcock 1971). prevalence has been as high as 93% in elk (hibler et al. 1969, davies 1979); high prevalence in elk occurs only in areas where mule deer also have high prevalence of infection. our study focused solely on moose so we have no analogous surveillance data from deer and moose for comparison. however, opportunistic sampling indicated ~10% prevalence of e. schneideri in mule deer in a portion of the area comprising the jackson, targhee, and sublette moose herds (j. c. henningsen, unpublished data). it may be that prevalence of e. schneideri in mule deer is too low in wyoming to generate >90% prevalence in moose; however, the snowy range had 82.6% prevalence. it has long been believed that deer are the only competent definitive hosts for e. schneideri (anderson 2001). however, some researchers (worley et al. 1972, madden et al. 1991) found gravid adult female worms in moose suggesting that they may be competent hosts. histopathologic and laboratory evidence from 3 different cases in our study support the idea that moose are a competent host for e. schneideri: 1) several microfilariae associated with an adult female worm were in a cross-section of formalin-fixed carotid artery, 2) several microfilariae were in a section of formalin-fixed skin overlying the mandibular artery at the jugular notch of the mandible, and 3) one dead microfilaria was in an overnight saline soak of fresh forehead skin. although not definitive, our evidence suggests that moose are competent hosts for e. schneideri reproduction and transmission. if this is the case, prevalence in mule deer and spatial overlap with infected mule deer could be less influential in determining e. schneideri prevalence in moose. we can only speculate about the increased prevalence of e. schneideri in wyoming moose over recent decades. because elaeophorosis was perceived to have no effect on wyoming ungulate populations, there is inconsistent historical data to make inferences. numerous case reports have expanded our knowledge of the general distribution of e. schneideri, but recent reports have not attempted to describe the ecology of e. schneideri and explain observed prevalence in wildlife (e.g., davies 1979). prevalence of the disease is presumably related to the density of definitive hosts as well as the abundance of tabanid vectors. alces vol. 48, 2012 henningsen et al. – elaeophora in wyoming moose 41 tabanid populations can be highly variable among years depending on weather conditions, because temperature and precipitation influence the timing of fly emergence, seasonal longevity, and total population size (pence 1991). thus gradual climate change has been attributed with observed and predicted increases in the effects of vector-borne parasites (patz et al. 1996, hoberg et al. 2008, laaksonen and oksanen 2009). we might have either conducted our surveillance when stochastic weather conditions were temporarily conducive for high e. schneideri transmission and/or prevalence, or changing conditions over decades has lead to higher prevalence of e. schneideri. determining the vectors of e. schneideri in wyoming and subsequently confirming the impacts of temperature and precipitation on those vectors will require further research. for the same reasons prevalence varies over time, it can exhibit high spatial variability. we found higher than expected prevalence among most herds; the snowy range and bighorns stood out as having especially high and low prevalence, respectively. moose habitat use and behavior could differ across herds in ways that affect sympatry with mule deer or susceptibility to horse flies (davies 1979). domestic livestock grazing adds another layer of complexity. livestock could either increase horse fly populations and exacerbate the transmission potential among wildlife, or dilute the effect because tabanids would prey on domestic animals instead of wildlife (davies 1979). further research is needed to fully understand the spatial dynamics of elaeophorosis in moose and other species in wyoming. on a more basic level, the effects of elaeophorosis on individual moose remain unknown. the high prevalence of apparently healthy infected moose suggests elaeophorosis is often not debilitating to this host. yet e. schneideri has been implicated in morbidity or mortality in several cases (worley et al. 1972, madden et al. 1991, pessier et al. 1998). as was demonstrated in elk (adcock and hibler 1969, hibler and adcock 1971, hibler and metzger 1974), pathogenic effects of e. schneideri on moose are more complex than a simple linear or threshold response by the host to number of worms. complications from infection could arise at a number of critical stages in the life cycle of the parasite. even slightly compromised basic functions resulting from impaired blood flow such as vision, hearing, mastication, smell, and brain function could expose individuals to malnutrition, predation, and ultimately lower survival and reproduction. while the maximum number of worms found in a hunter-harvested moose in our study was 26, preliminary necropsies of symptomatic moose have sometimes revealed double that intensity (j. c. henningsen, unpublished data). additionally, while none of the ≥8-yrold moose had high-intensity infections, this may have been an artifact of low sample size. limiting surveillance to hunter-harvested moose possibly eliminates important cases from consideration. comprehensive surveillance that includes sick and dead moose with subsequent histopathologic examinations will be valuable in elucidating impacts of this parasite on individual moose. prevalence in our study was consistent across age classes. we interpret this to mean that moose of all ages are equally susceptible to infection and that infection does not affect survival differently across ages. constant intensity of e. schneideri across ages of checked moose might additionally indicate a mechanism limiting worm burdens in moose. perhaps individuals that tolerate initial infection acquire some immunity against further infection; hibler and metzger (1974) suggested as much for infected elk. immune protection has been demonstrated in other ungulate-nematode systems involving longlived adult worms. parelaphostrongylus tenuis intensities in white-tailed deer are constrained across age classes (slomke et elaeophora in wyoming moose – henningsen et al. alces vol. 48, 2012 42 al. 1995) and prestwood and nettles (1977) demonstrated white-tailed deer acquire immunity to additional p. andersoni infections. this hypothesis presumes e. schneideri are long-lived; however, it is unknown how long e. schneideri can live in moose. other filarioid nematodes live in their definitive hosts from 2->10 years (review by gems 2000). alternatively, constant e. schneideri prevalence and intensities with increasing age of moose might simply reflect new infections occurring at a rate that essentially replace those mature nematodes that die naturally. under this scenario, immune protection would not be perfect and new infections would continue through life at some rate that is tolerated by the host. on the other hand, pathologies arising from dead nematodes in the vascular system (adcock and hibler 1969) would be inconsistent with a hypothesis where moose can survive unaffected beyond the lifespan of the adult parasite. thus this hypothesis presumes moose can tolerate not only live parasites, but individuals that die within their vascular system. regardless of the mechanism, constant intensity and prevalence across age classes indicate infected moose are surviving, hence mortality caused by e. schneideri is lower than previously suggested. while moose might not overtly succumb to elaeophorosis to the extent previously thought, prevalence of 50% is still cause for concern. at high prevalence, even a moderate proportion of infected individuals suffering from subclinical effects might impact recruitment or productivity at the population level. subsequent research on moose herds where e. schneideri is present should consider the effects of elaeophorosis and attempt to clarify its role in moose population dynamics. acknowledgements the wgfd moose working group was integral in designing the survey, and numerous wgfd field personnel conducted the field examinations. t. cornish was integral in the histopathologic and laboratory work. we are grateful to c. hibler, associate editor b. mclaren, and reviewers e. addison and m. lankester for improving the manuscript. references adcock, j. l., and c. p. hibler. 1969. vascular and neuro-ophthalmic pathology of elaeophorosis in elk. pathologia veterinaria 6: 185-213. anderson, r. c. 2001. filarioid nematodes. pages 342-356 in w. m. samuel, m. j. pybus, and a. a. kocan, editors. parasitic diseases of wild mammals. iowa state university, ames, iowa, usa. boyce, w., a. fisher, h. provencio, e. rominger, j. thilsted, and m. ahlm. 1999. elaeophorosis in bighorn sheep in new mexico. journal of wildlife diseases 35: 786-789. brimeyer, d. g., and t. p. thomas. 2004. history of moose management in wyoming and recent trends in jackson hole. alces 40: 133-143. bush, a. o., k. d. lafferty, j. m. lotz, and a. w. shostak. 1987. parasitology meets ecology on its own terms: margolis et al. revisited. the journal of parasitology 83: 575-583. clark, g. g., and c. p. hibler. 1973. horse flies and elaeophora schneideri in the gila national forest, new mexico. journal of wildlife diseases 9: 21-25. davies, r. b. 1979. the ecology of elaeophora schneideri in vermejo park, new mexico. ph.d. dissertation, colorado state university, fort collins, colorado, usa. espinosa, r. h. 1983. tabanid vectors of the arterial nematode, elaeophora schneideri, in southwestern montana. m.s. thesis, montana state university, bozeman, montana, usa. hibler, c. p. 1982. elaeophorosis. pages 214-218 in e. t. thorne, n. kingston, w. r. jolley, and r. c. bergstrom, edialces vol. 48, 2012 henningsen et al. – elaeophora in wyoming moose 43 tors. diseases of wildlife in wyoming. wyoming game and fish department, cheyenne, wyoming, usa. _____, and j. l. adcock. 1971. elaeophorosis. pages 263-278 in j. w. davis and r. c. anderson, editors. parasitic diseases of wild mammals. iowa state university, ames, iowa, usa. _____, _____, r. w. davis, and y. z. abdelbaki. 1969. elaeophorosis in deer and elk in the gila forest, new mexico. bulletin of the wildlife disease association 5: 27-30. _____, _____, g. h. gates, and r. white. 1970. experimental infection of domestic sheep and mule deer with elaeophora schneideri wehr and dikmans, 1935. journal of wildlife diseases 6: 110-111. _____, and c. j. metzger. 1974. morphology of the larval stages of elaeophora schneideri in the intermediate and definitive hosts with some observations on their pathogenesis in abnormal definitive hosts. journal of wildlife diseases 10: 361-369. hoberg, e. p., l. polley, e. j. jenkins, and s. j. kutz. 2008. pathogens of domestic and free-ranging ungulates: global climate change in temperate to boreal latitudes across north america. review scientifique et technique-international office of epizootics 27: 511-528. gems, d. 2000. longevity and ageing in parasitic and free-living nematodes. biogerontology 1: 289-307. jacques, c. n., j. a. jenks, d. t. nelson, t. j. zimmerman, and m. c. sterner. 2004. elaeophorosis in free-ranging mule deer in south dakota. prairie naturalist 36: 251-54. jensen, l. a., j. c. pederson, and f. l. andersen. 1982. prevalence of elaeophora schneideri and onchocerca cervipedis in mule deer from central utah. great basin naturalist 42: 351-352. laaksonen, s., and a. oksanen. 2009. status and review of the vector-borne nematode setaria tundra in finnish cervids. alces 45: 81-84. madden, d. j., t. r. spraker, and w. j. adrian. 1991. elaeophora schneideri in moose (alces alces) from colorado. journal of wildlife diseases 27: 340-341. matthews, p. e. 2012. history and status of moose in oregon. alces 48: 63-66. mckown, r. d., m. c. sterner, and d. w. oates. 2007. first observation of elaeophora schneideri wehr and dikmans, 1935 (nematoda: filariidae) in mule deer from nebraska. journal of wildlife diseases 43: 142-144. patz, j. a., p. r. epstein, t. a. burke, and j. m. balbus. 1996. global climate change and emerging infectious diseases. journal of the american medical association 275: 217-223. pederson, j. c., l. a. jensen, and f. l. andersen. 1985. prevalence and distribution of elaeophora schneideri wehr and dikmans, 1935 in mule deer in utah. journal of wildlife diseases 21: 66-67. pence, d. b. 1991. elaeophorosis in wild ruminants. bulletin of the society for vector ecology 16: 149-160. _____, and g. g. gray. 1981. elaeophorosis in barbary sheep and mule deer from the texas panhandle. journal of wildlife diseases 17: 49-56. pessier, a. p., v. t. hamilton, w. j. foreyt, s. parish, and t. l. mcelwain. 1998. probable elaeophorosis in a moose (alces alces) from eastern washington state. journal of veterinary diagnostic investigation 10: 82-84. prestwood, a. k., and v. f. nettles. 1977. repeated low-level infection of whitetailed deer with parelaphostrongylus andersoni. journal of parasitology 58: 897-902. _____, and t. r. ridgeway. 1972. elaeophorosis in white-tailed deer of the southeastern usa.: case report and distribution. jourelaeophora in wyoming moose – henningsen et al. alces vol. 48, 2012 44 nal of wildlife diseases 8: 233-236. rózsa, l., j. reiczigel, and g. majoros. 2000. quantifying parasites in samples of hosts. journal of parasitology 86: 228-232. slomke, a. m., m. w. lankester, and w. j. peterson. 1995. infrapopulation dynamics of parelaphostrongylus tenuis in white-tailed deer. journal of wildlife diseases 31: 125-135. thomas, t. p. 2008. moose population management recommendations. wyoming game and fish department, cheyenne, wyoming, usa. waid, d. d., r. j. warren, and d. b. pence. 1984. elaeophora schneideri wehr and dikmans, 1935 in white-tailed deer from the edwards plateau of texas. journal of wildlife diseases 20: 342-345. weinmann, e. j., j. r. anderson, w. m. longhurst, and g. connolly. 1973. filarial worms of columbian black-tailed deer in california 1. observations in the vertebrate host. journal of wildlife diseases 9: 213-220. worley, d. e. 1975. observations on epizootiology and distribution of elaeophora schneideri in montana ruminants. journal of wildlife diseases 11: 486-488. _____, c. k. anderson, and k. r. greer. 1972. elaeophorosis in moose from montana. journal of wildlife diseases 8: 242-244. alces24_112.pdf rodgersar text box alces vol. 45, 2009 glushkov moose population management in russia 43 improving population management and harvest quotas of moose in russia vladimir m. glushkov research institute of game management and fur farming, kirov, russia. abstract: annual harvest quotas for moose and other game species in russia have been based on population estimates derived from traditional winter track counts and hunter surveys. this labor-intensive approach has failed to account for evident changes in population density of moose. specifically, regional differences in survival and mortality data and the impact of increased poaching are not measured or included in population estimates, and overharvest of moose occurs. i propose implementing a standard management approach similar to that used in other countries with moose populations that includes population trend analyses, productivity and mortality data, and a regional management approach. such changes will improve the professional management of moose and other game species in russia. alces vol. 45: 43-47 (2009) key words: alces alces, harvest, management strategy, moose, poaching, population estimate, russia. in russia the population size of most game species including moose (alces alces l., 1758) is estimated from winter track counts along established census routes. these annual counts are conducted in the latter half of winter after the hunting season, and include approximately 45,000 routes, each about 10 km long. the necessity and continuation of this large-scale annual effort and traditional approach stem from the desire to ensure an adequate harvest quota; however, this approach and related data sets have not recognized annual fluctuations in the moose population that are critically important when setting harvest quotas (glushkov 1995). hunter surveys (about 10,000 questionnaires) conducted throughout russia failed to reveal abnormal causes or rates of mortality that could cause population fluctuations in the moose population (glushkov et al. 1989). the relative estimates of moose populations received from hunters at the start of winter under the program of “the harvest service of vniioz” also failed to show any annual fluctuations in the population (fig. 1). an analysis performed with a large sample size of biological data (2045 jaws from harvested moose, 555 female reproductive tracks, observation of 1360 family groups) collected in forests in the south taiga of the european part of russia (kirov region) showed no dynamic changes in birth rate and natural mortality (glushkov 1999). a decline in fecundity of sub-adult females was noted only when population density in a local area approached its maximum value (3.1-3.4 moose/1000 ha forest; 1981-1990), or when a decline in the proportion of females/litter (statistically significant only for females in the 4th age class) reduced the rate of population growth from 0.041 to 0.000. however, this decline was not only the result of self-regulation, but also of poaching that doubled from 1 to 2.1 moose per poaching incident (versus 1 moose/license). however, this population decline that started in 1987 is not reflected in the census data or hunter survey results. a comparison of the data from the population estimates of the census and the hunter surveys at the beginning of winter revealed that both provided similar conclusions about the moose population growth rate; that is, stable moose population management in russia glushkov alces vol. 45, 2009 44 and/or slow one-way growth (growth rate was 0.026). according to odum (1986), this type of growth rate is sigma-shaped (logistic) and is regulated directly by factors that are population density dependent. such growth is described simply by the logistic equation: dn/dt = rn (k – n)/k as the population reaches the upper asymptote k, growth rate (dn/dt) decreases and approaches zero. harvest and lower birth rate further reduce the growth rate and prevent the density of exploited moose populations from reaching their maximum level; this would be illustrated by a low angle or minimal slope of the population growth rate curve. the population estimates estimated from the winter census routes (fig. 1) show some irregularity as the population increased slowly (average growth rate was 0.048). this growth curve was similar to those depicting population changes caused by natural conditions for various species of birds (williamson 1975). migration of moose in the kirov region was reduced during years with late snow cover; presumably there is a relationship between fluctuating population estimates from the winter census data and the occurrence and intensity of migration. a number of aerial surveys were carried out in early and late winter during a 5-year period (1981-1985) that confirmed this assumption. further, it was also evident that the early winter aerial surveys could not identify small population growth rates (0.041 or 4.1 % per year) determined afterward from demographic tables; the aerial censuses indicated stable populations. data from moose populations in finland and canada confirmed 2 fundamentally important features concerning the type of population growth of a given species (i.e., stability and one-way direction; glushkov 2001), but those populations were characterized by higher growth rates and more measurable response to harvest regulations than those in russia. this occurs for two primary reasons: 1) the use of selective harvesting that produces a highly productive population, and 2) the absence of poaching that allows effective use of selec0 5 10 15 20 25 30 35 19 70 19 72 19 74 19 76 19 78 19 80 19 82 19 84 19 86 19 88 19 90 y e ar m o o se p o p u la ti o n (1 00 0s ) 0 0. 5 1 1. 5 2 2. 5 3 3. 5 4 re la ti ve p o p u la ti o n e st im at e l ate w inter es timate early w inter es timate fig. 1. moose population estimates in the kirov region of russia during early winter (relative scale) and from track counts in late winter (1000s of moose), 1970-1990. alces vol. 45, 2009 glushkov moose population management in russia 45 tive harvest quotas to manage populations effectively. analysis of our moose population growth rates and population responses of other hunted species indicated that all species with logistic growth rates (density dependent) are regulated by common internal (population) mechanisms. these include: 1) slow, difficult-to-measure population growth rates, 2) anthropogenic factors that cause measurable population decline (i.e., hunting mortality), 3) extended periods of population growth and decline, and 4) slow recovery when special conservation and bio-technical management techniques are required to restore the population. for conventional purposes, i described such species as “controlled” (glushkov 2008) in contrast to species with fluctuating (trigger) growth rates. it is clear that the continued existence of “controlled” species depends substantially on the intensity of hunting, and implementing conservation and biotechnical measures. effective harvest regulations provide the most reliable tool to control and reduce mortality to conserve hunted populations with logistic growth rates. the harvest quota for moose is set by comparing the difference between birth and natural mortality rates. however, annual juvenile mortality is influenced by variable environmental conditions that affect food resources, weather, and predation. because high calf mortality occurs in the first 3 months, harvest rates must account for calf survival not the actual birth rate (i.e., that is the growth rate at start of winter). this approach will provide the best estimate of the number of animals available for harvest. because natural winter mortality is much lower than calf mortality, it is often ignored when setting harvest quotas. however, in russia the rate of non-selective harvest has resulted in a negligible growth rate because it represents the difference between the population growth rate prior to the hunting season and winter mortality due to poaching. for example, the average population growth rate in the kirov region is 0.190 at the beginning of winter. if this rate were effectively reduced to compensate for winter predation (0.02), natural mortality from diseases, wounds, other unknown causes (0.015-0.020), poaching (0.08), and a population reserve for increased reproduction (0.02), the non-selective harvest should not exceed 5.5% (0.190 – 0.135). however, this broad calculation does not account for partial replacement of certain mortality factors; the overall mortality rate could be lower than the sum of the rates of individual mortality factors (glushkov 2002). therefore, the integrity of the harvest quota is principally dependent upon the accurate estimate of the autumn population and calf survival prior to the hunting season; the relative importance of winter mortality due to poaching and natural factors is magnified by errors in this estimate. ineffective and harmful harvest quotas in russia occur because of inaccurate population estimates, erroneous documentation about migration, lack of regional population growth rates and related mortality (e.g., poaching) data, and the temporal nature and population response to these factors. the introduction of a selective harvest system could increase both the birth and population growth rates of moose and help nullify their current, stagnant growth rate. however, its implementation and resultant change in harvest quotas will be difficult in the current system, and will require adaptive economics, harvest, and scientific management of moose in russia. this task may be simplified somewhat by following the example of foreign game biologists. rather than depend entirely upon absolute, quantitative population estimates from annual data, they typically analyze trends in annual population data to assess and set harvest quotas. for example, harvest quotas are set relative to the previous year’s quota by assessing special indices of population density (responses to the current level of harvest) on moose population management in russia glushkov alces vol. 45, 2009 46 a regional basis. proposed changes in harvest quotas could be delayed 2 years to reassess the status of the current quota and to reduce the potential negative impact of a changed harvest quota on the structure and productivity of the population. the ability of a selective harvest management system to increase population growth rates and harvest has been documented many times. for example, i used examples from scandinavian countries to illustrate moose harvest rates of 35%, or about 5x that in russia. management problems in these countries are also quite different; the scandinavians typically manage their moose population to maximize harvest, avoid overpopulation, and prevent agricultural and forestry damage. in russia, we strive to reduce poaching and hope to restore our moose population to a level sustainable with the natural productivity of the landscape. conclusions 1) the current method of setting moose harvest quotas is principally flawed because of error in estimating regional populations and mortality rates, and the regional and temporal variation of these parameters. harvest quotas based on erroneous and incomplete data reduce the efficiency and economics of moose management, and for both practical and scientific purposes, an improved method is needed to set moose harvest quotas. 2) an improved strategy in setting regional harvest quotas would mimic common approaches in other countries that include an analysis of the population response to the previous year’s harvest quota. if this system was introduced in russia, federal managers should focus on strategic elements including the overall harvest quota and structure, and implementing management changes; tactical elements such as regional/local harvest quotas should be determined by regional management branches. 3) population assessment of moose and other game species at the onset of winter should be done with annual population trend/index data. each administrative district should have one game biologist responsible for such analyses; such an approach will reduce laborious fieldwork and overall costs substantially. 4) biological assessment of population dynamics will need to improve. moose populations need to be managed regionally in order to address variable population growth rates and environmental conditions. standardized methods are needed to index/census populations of game species in order to produce reliable population density estimates. administrative protocols need to be adopted to guide population monitoring efforts. 5) the current system employed to set the moose harvest quota in russia has many weak components including lack of specific population dynamic information and laborious annual fieldwork to estimate population density. i propose a more simplified procedure of calculating harvest quotas for moose by using better estimates of population density, calf survival, mortality factors including the rate of poaching, and establishing population trend analyzes. these changes will make management of moose and other game species more professional and accurate, and provide an improved practical approach in conservation efforts with these valuable species. references glushkov, v. m. 1995. method of winter route census as a factor of irrational use of resources of wild ungulates. pages 5556 in hunting science and nature use. kirov, russia. (in russian). _____. 1999. moose. pages 117-163 in management of game animal populations. collected scientific papers of vniioz. kirov, russia. (in russian). _____. 2001. moose: ecology and management of populations. kirov, russia. (in russian). _____. 2002. ecological bases of population alces vol. 45, 2009 glushkov moose population management in russia 47 management. pages 115-119 in problems in recent hunting science. proceedings of scientific and practical conference, december 5-6, 2002. centrokhotcontrol, moscow, russia. (in russian). _____. 2008. okhota i okhotnichie khozyastvo (is it a rate or a quota)? 12: 1-2. (in russian). _____, v. n. piminov, and b. p. ponomaryev. 1989. winter mortality and reserves of wild ungulate harvesting. pages 81-92 in management of populations of wild ungulates. collected scientific papers of vniioz. kirov, russia. (in russian). odum, e. 1986. ecology, vol. 2. moscow, russia. (translated from english). williamson, m. 1975. the analysis of biological populations. mir, moscow, russia. (translated from 1972 english version). alces(25)_168.pdf alces26_24.pdf wild animal research – new legal requirements in the european union margareta stéen1, katarina cvek2, and petter kjellander3 1swedish center for animal welfare, swedish university of agricultural sciences, uppsala, sweden; 2department of clinical sciences, swedish university of agricultural sciences, uppsala, sweden; 3department of ecology, grimsö wildlife research station, swedish university of agricultural sciences, riddarhyttan, sweden. abstract: the european union agreed on a directive (dir) for the protection of animals used for scientific purposes in 2010 which was implemented by member states at the onset of 2013. the dir applies to animals used for science or education that are subjected to pain, suffering, distress or lasting harm equivalent to, or higher than that caused by a needle. the dir changes the legal framework for wild animal research and requires educational and training standards of staff involved in capturing, planning, or performing research. both wild animals studied in or taken from the wild into captivity are covered by the dir. an animal welfare body must be established that includes a scientific member and at least one person responsible for animal welfare, and they must receive input from a designated veterinarian. the dir will aid and improve wild animal research because standards of animal welfare and research ethics must be met. although similar standards for moose research were employed previously in scandinavia, future moose research and conservation will likewise benefit. alces vol. 49: 127–131 (2013) key words: alces alces, animal welfare, capture, european union, legislation, marking, moose, research. introduction the european union (eu) developed a new directive (dir) for the protection of animals used for scientific purposes in 2010 (directive 2010/63/eu), that was implemented by member states (ms) at the onset of 2013. the principle reasons for the dir were to standardize legislation between ms and to improve the welfare of animals used in scientific research and procedures. this action was actualized by the increasing scientific knowledge about factors that influence animal welfare (aw), as well as the capacity of animals to sense and express pain and suffering. ethical concerns of the general public were also influential, and the desire to replace the use of animals for scientific purposes with non-animal alternatives. it was recognized that animals have an intrinsic value which must be respected, they should always be treated as sentient creatures, and their use should be restricted to benefitting human or animal health, or the environment. thus, ms must ensure that live animals are not used when a scientifically satisfactory method or testing strategy not entailing the use of live animals is available. wildlife research with wild ruminants had been regulated previously by national legislation in scandinavian countries; however, no comprehensive european regulations existed. moose (alces alces) in the wild have been used for scientific purposes in northern europe for decades and been used in theoretical and applied ecological research covering a wide range of topics: predator-prey interactions, herbivore-plant interactions, ecology, behavior, migration, veterinary medicine, disease, and moose-human 127 interactions using a variety of scientific methods and procedures, and technical and analytical approaches. moose provide an excellent case study for educating and explaining the dir to current wildlife researchers. in this paper we aim to inform researchers about the content and intent of the dir, and identify the consequences of its implementation for future moose research. description the dir states that animals taken from the wild (e.g., ruminants) shall not be used for scientific studies unless a competent authority has granted an exemption. the exemption shall only be granted if the purpose of the procedure cannot be achieved by the use of an animal which has been bred for such use. captures must be performed with care by competent persons such that an animal is not caused any avoidable pain, suffering, distress, or lasting harm. if a wild animal is found to be injured or in poor health at or after capture, it must be examined by a veterinarian or other competent person; action shall be taken to minimize the suffering of the animal. competent authorities may grant exemptions from the requirement to minimize suffering of the animal if there is scientific justification. scientists using research animals need to understand the dir legislation to identify which parts apply to and affect their research. this will facilitate experimental design and planning, minimize the number of animals used, and ensure that their research is within the legal framework. the dir applies to all animals that are used for scientific or educational purposes that are subjected to pain, suffering, distress, or lasting harm equivalent to, or higher than that caused by the insertion of a needle, in accordance with good veterinary practice. this includes moose studied in or taken from the wild and kept in captivity if subjected to procedures that cause “pain” or the equivalent to pain; e.g., an immobilized moose fitted with ear tags or radio-collar. all users of moose for scientific purposes must be authorized and registered by a competent authority. research projects must pass an ethical evaluation to receive authorization for use of animals in research or teaching; the use is evaluated and justified relative to the societal, scientific, and/or educational purpose of the project. the project should be designed such that procedures are in compliance with the requirement of the 3r’s (i.e., replacement of the use of animals for scientific studies, reduction of the number of animals used, refinement of procedures; russell and burch 1959). the evaluation of the project shall be performed in an impartial manner, the evaluation process must be transparent, and it shall weigh the predicted societal and scientific benefits, or educational value of the project against the harm and suffering of the research animals. furthermore, the evaluation shall verify that the project is designed such that procedures are performed in the most humane and environmentally sensitive manner possible. the ethical evaluation shall also determine whether the project should be evaluated retrospectively, concerning the aw and outcome of the study. for all projects, a non-technical summary written by the project leader/researcher shall be published by the ms to facilitate communication with the general public. before initiation of research, any staff involved in capturing, handling, planning, or performing research must be formally educated and trained until proven competent to perform their tasks. all users and breeders of animals for research must form an aw body which shall include the person responsible for aw and care, and in the case of a user, a scientific member; the body shall also receive input from the designated veterinarian. the 128 wild animal research – stéen et al. alces vol. 49, 2013 primary task of this body is providing advice on aw issues and the outcome of aw in projects. the body should also foster a climate of care and provide tools for the practical application and timely implementation of recent technical and scientific developments in relation to the principles of the 3 r’s to enhance the life-time experience of research animals. the advice of the aw body should be properly documented and open to scrutiny during inspections. the ms shall ensure that an animal may only be reused in a new procedure provided that the actual severity of any previous procedure was ‘mild’ or ‘moderate’ and that the general state of health and well-being has been fully restored; veterinary advice shall be taken into account regarding the lifetime experience of the animal. the ms may allow animals used in procedures to be returned to a suitable habitat appropriate to the species, provided that the state of health of the animal allows such and there is no danger to public health, animal health, or the environment. appropriate measures shall be taken to safeguard the wellbeing of the animal. application to moose a typical radio-collaring project with moose that is managed by a university or research institute in europe needs to comply with the dir by adopting the following protocol chronologically: 1) the department has to be granted a general permit to use animals in research for a limited period (years); 2) a local group (e.g., aw organization) should be formally appointed at the department level to help oversee aw; 3) the project leader should write an application describing the purpose (what, why, when, how) of the study relative to aw and how the project complies with the 3 r’s; 4) the application is signed by both the project leader and head of the local aw group; and 5) the application passes a review by a national ethics board. a permit to radio-collar a defined number of moose would then be approved for a maximum of 5 years. research studies involving capture and restraint can be stressful and cause measurable harm to moose, and can also influence experimental assumptions and data. the dir permits competent authorities to exempt the requirement to minimize the suffering of wild-caught animals found to be in poor health or injured, given scientific justification. the dir states that the assessment of health and welfare of the animals must be performed by a competent person. assessment of competence is based upon an appropriate level of understanding animal behavior, biology, and ecology of the species, and the ability to recognize poor health, injury, discomfort, pain, and distress. it is important to minimize the disturbance of a study population and understand the potential pathologies related to the capture activity, and how to prevent sickness and take appropriate actions in the case of poor health or welfare of a captured animal. proven competence is required for capture, handling, and restraint techniques including the operation and maintenance of any trapping devices. the idea of wild animals suffering is of concern for the legislating bodies, as well as scientists and the general public. concern regarding aw is related not only to marking methods, but also capture and handling procedures prior to, during, and after marking and release. combined long-term effects associated with these procedures and activity can occur at both the individual and population level. to our knowledge, the first chemical immobilization of a wild moose for research purposes in europe took place in 1975 at grimsö wildlife research area in south-central sweden (sandegren et al. 1987); > 2800 moose have been immobilized throughout scandinavia since (arnemo et al. 2006). moose are usually darted from a helicopter alces vol. 49, 2013 stéen et al. – wild animal research 129 and marking is performed under general anesthesia with typical surveillance including pulseand respiratory rate, body temperature, and sometimes arterial oxyhemoglobin saturation (spo2). chemical capture and anesthesia of free-ranging mammals involves some risk of mortality even in healthy animals. a scandinavian study on immobilization of moose estimated such mortality as 0.7% (n = 2,816) with 0.2% related directly to the immobilization procedure (arnemo et al. 2006). even if mortality is the most apparent negative side-effect of marking wild animals, other subtle, stress-induced biases are critical to identify because of their potential influence. for example, it is recommended to omit data from the initial 5 days post-capture when measuring moose movement and distribution (neumann et al. 2011). to help achieve zero mortality in scandinavian moose research, 3 factors have been identified: 1) use an experienced, trained professional capture team, 2) develop and follow a species-specific capture protocol, and 3) require a mortality assessment after any capture-related death to provide a knowledge-based evaluation of the capture protocol. this approach complies with the intent of the dir and should be followed by all moose researchers. when recapturing and re-marking the same animal, the three r′s should be employed to reduce the risk of impaired aw in moose research, conservation, and management. specific considerations are: 1) replacement strategies by which the required information may be obtained by other means than marking live wild animals, 2) educational strategies to use the fewest animals possible for providing valid information and statistical significance, and 3) refining strategies to use the most humane, least invasive marking techniques with the goal of minimizing pain and distress. these requirements are already in the spirit of most, if not all contemporary moose researchers. in addition to ecological studies with radio-collared moose, both physiological and pathological studies have been performed on captive moose housed indoors. for example, several successive swedish studies of infectious diseases in moose including brainworm (elaphostrongylus alces) and moose wasting syndrome involved calves obtained from the wild, and subsequently penned and raised in stalls (stéen et al. 1997, broman et al. 2002). the dir classifies facilities where wild animals are housed for experimental studies as ‘establishments’ and defines them as any installation, building, group of buildings, or other premises, and may include a place that is not wholly enclosed or covered, or a mobile facility. the ms must ensure that an establishment has installations and equipment suited to the species of animal housed and related research procedures. the dir states that animals removed from the wild shall not be used for scientific studies unless a competent authority has granted an exemption; of concern is the future of research requiring the capture, mark, and indoor housing of wild animals. to grant exemption for future studies of this type, competent authorities will require adequate knowledge of the behavior, biology, and veterinary studies of multiple wild species. if granted, the requirements of the dir are beneficial for all study participants including researchers, veterinarians, and field assistants since they will participate in project planning. this planning would include educational requirements of the various participants and their responsibilities, experimental design, ethical concerns, and the aw of the study animal. the practical implications of the dir should be advantageous for research with wild animals overall 130 wild animal research – stéen et al. alces vol. 49, 2013 and specifically benefit moose research and conservation globally. references arnemo, j. m., p. ahlqvist, r. andersen, f. bernsten, g. ericsson, j. odden, s. brunberg, p. segerström, and f. swenson. 2006. risk of capture related mortality in large free-ranging mammals experiences from scandinavia. wildlife biology 12: 109–113. broman, e., k. wallin, m. stéen, and g. cederlund. 2002. a wasting syndrome in swedish moose (alces alces): the background and current hypotheses. ambio 31: 409–416. directive 2010/63/eu. 2010. directive of the european parliament and the council on the protection of animals used for scientific purposes, 22 september 2010. <http://eur-lex.europa.eu.> (accessed march 2013). neumann, w., g. ericsson, h. dettki, and j. m. arnemo. 2011. effect of immobilizations on the activity and space use of female moose (alces alces). canadian journal of zoology 89: 1013–1018. russell, w. m. s., and r. l. burch, 1959. the principles of humane experimental technique. methuen, london, united kingdom. sandegren, f., l. pettersson, p. ahlqvist, and b. o. röken. 1987. immobilization of moose in sweden. swedish wildlife research supplement 1: 785–791. stéen, m., c. g. m. blackmore, and a. skorping. 1997. cross-infection of moose (alces alces) and reindeer (rangifer tarandus) with elaphostrongylus alces and elaphostrongylus rangiferi (nematoda, protostrongylidae): effects on parasite morphology and prepatent period. veterinary parasitology 71: 27– 38. alces vol. 49, 2013 stéen et al. – wild animal research 131 http://eur-lex.europa.eu wild animal research new legal requirements in the european union introduction description application to moose references alces29_267.pdf alces21_493.pdf alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces21_339.pdf alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces28_123.pdf alces26_44.pdf alces28_41.pdf alces(23)_33.pdf alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces27_208.pdf winter distribution of moose at landscape scale in northeastern vermont: a gis analysis thomas l. millette, eugenio marcano, and danelle laflower geoprocessing laboratory, mount holyoke college, south hadley, massachusetts, usa. abstract: a gis analysis of landscape scale distribution of moose (alces alces) in northern vermont during winter 2010 showed that most moose were located at elevations of 300–600 m, with little discernible elevational gradient. slope and aspect were not correlated with locations as moose were distributed in the study area with the relative amount in each descriptive class. the distribution of >85% moose based on noaa cover types was in deciduous, mixedwood, and coniferous stands relative to their availability; locations in scrub/shrub and wetlands were higher and lower than expected, respectively. higher resolution aims imagery indicated that moose used mixedwoods more and coniferous stands less than available. the most significant landscape characteristic influencing the location of moose was proximity to forest openings/timber cuts that presumably provide important seasonal browse. alces vol. 50: 17–26 (2014) key words: alces alces, distribution, gis, moose, vermont, winter. introduction most landscape analyses of moose (alces alces) have been focused upon coarse-scale location and description of home ranges, or fine-scale seasonal habitat selection and utilization through use of radio-telemetered animals, aerial markrecapture surveys, and/or fieldwork (courtois and beaumont 2002, courtois et al. 2002, potvin and courtois 2004, poole and stuartsmith 2005, scarpitti et al. 2005, dussault et al. 2006, gillingham and parker 2008, van beest et al. 2010). although increased use of gps collars has provided more accurate and plentiful locations of individuals, most studies are limited by animal sample size due to the difficulty and cost associated with monitoring the broader population itself. a recently developed airborne thermal vertical-imaging system integrated with gis has presented the opportunity to identify and map locations of hundreds of moose during mid-winter (millette et al. 2011), with subsequent exploration of landscape attributes associated with their locations. because locations are accurately geocoded with gps by the thermal imaging system, it is possible to use a wide variety of off-the-shelf gis databases to model habitat characteristics. the landscape distribution and winter habitat use examined here is thought to be unique in that no such winter “snapshot” of a regional moose population has had such high sample size of animals. methods study area the study area was 682 km2 within wildlife management unit (wmu) e1located in the northeastern corner of vermont and bordered by new hampshire and quebec (fig. 1). the area is topographically expressive, and heavily forested with expansive maple (acer saccharum, a. pensylvanicum) and american beech (fagus grandifolia), stands of balsam fir (abies balsamea), spruce (picea rubens, p, glauca, corresponding author: thomas millette, geoprocessing laboratory, mount holyoke college, south hadley, ma, usa. 17 p. mariana), hemlock (tsuga canadensis), and eastern white pine (pinus strobus); conspicuous evidence of timber harvesting existed throughout. the estimated moose density based on a rolling 3-year average of moose sightings by november deer hunters was 0.89 moose/km2 (c. alexander, vermont fish and wildlife department). similarly, the estimated density was 0.84 moose/km2 during the aerial thermal census (millette et al. 2011) that produced the data for this study. data the study area was sampled with 35 survey units (su) distributed (relatively) evenly throughout. each was laid out nonrandomly in a gis to account for the variety of topographic settings and land cover types, while avoiding major changes in elevation along flight lines to maintain constant height above ground level (agl); thus, the image swath-width was as constant as possible during flights. the total area surveyed was 131.6 km2 or 20% of wmu e1. to insure that land cover types along flight line transects were representative of wmu e1, a gis overlay analysis was used to compare the proportions along transects with those from the noaa coastal service center land cover data (noaa 2006) (fig. 2). this analysis indicated that the relative proportions of cover types were almost identical between transects and the entire wmu such that no cover type was underor oversampled. details of the sampling design can be found in millette et al. (2011). the data were developed using the aims-thermal airborne imaging system. the sensor array pairs a 16-bit radiometric thermal camera to detect warm targets on a cold background, and simultaneously acquires 8-bit high resolution color photos to identify specific heat sources. unlike most aerial thermal systems used in previous research, the aims-thermal acquires its imagery vertically like a mapping system rather than using a low-oblique viewing angle while panning across the landscape. the vertical orientation of the cameras causes minimal screening effect in coniferous stands that is more typical of systems with oblique look-angles, and preserves uniform scale and spatial resolution throughout each image allowing detailed measurements within an image. a complete description of the aims-thermal system can be found in millette et al. (2011). the aims-thermal system was deployed in january and february 2010 over a 4-week period when 6 flights were flown between 0700 and 1100 hr. in total, these flights produced 94,605 thermal images and 12,530 high-resolution color images under continuous snow cover and sky conditions ranging from heavy overcast to bright sunshine. snow cover never exceeded 45 cm and no restrictive crust layers existed. fig. 1. location of the study area in wmu e1 in northeastern vermont, usa. 18 winter distribution of vermont moose – millette et al. alces vol. 50, 2014 all imagery exposure times and associated flight data were processed into gis attribute tables that support creation of shape‐ files containing photo centers for each exposure from the thermal and natural color cameras, as well as the flight path of the aircraft. this table also provides the framework for the integration of related spatial data such as sampling transects, flightlines, topography, and vegetation and facilitates the landscape-scale analysis described here. all gis database development operations were done using software developed by the researchers; all gis analyses used arcgis software tools. gis analyses an assessment of moose locations relative to landscape attributes was conducted to explore whether relationships or patterns existed that would describe habitat selection during the winter study period. the locations of 112 observed moose were used in a series of gis overlay procedures with the usgs national elevation data (gesch et al. 2002), the noaa coastal service center land cover data (noaa 2006), and the national agriculture imagery program (naip) vermont digital color orthophotography (2009) to examine the distribution of locations relative to elevation, slope, aspect, land cover type, and land management practices. all gis data were generalized to 90 m spatial resolution to limit landscape data heterogeneity. elevation was divided into 7–100 m classes with the majority (90%) of the landscape ranging from 300–700 m. slope was divided into 4 classes of <2.5°, 2.5°–5°, 5°– 10°, and >10°. aspect was divided into fig. 2. comparison of noaa land cover types with imagery transects that indicates the similarity between availability of cover types in the study area and the actual survey area in winter 2010, northeastern vermont, usa. alces vol. 50, 2014 millette et al. – winter distribution of vermont moose 19 the 8 cardinal and inter-cardinal points of the compass. the analysis of locations relative to land cover was done using classifications from the noaa csc land cover data with 30 m spatial resolution, which was generalized to 90 m to reduce artificial heterogeneity in the land cover data derived from landsat tm data, and to allow the spatial resolution of the land cover data to better match the 1.3 ha footprint of each thermal image. a second classification created from the aims-thermal natural color data with 3.2 cm spatial resolution was visually interpreted for each 2.2 ha image containing a moose. the aims-thermal classification differs from the noaa data since it had to be generalized into deciduous, mixedwood, and coniferous cover types due to snow cover which prevented accurate delineation of wetlands, scrub/shrub, and grassland. we performed a series of chi square goodness of fit tests (snedecor and cochran 1989) to determine if the distribution of moose was random among the different classes of elevation, slope, aspect, and land cover: x2 ¼ x o � eð þ2 e ð1þ where o = the number of observed moose in each category and e = the expected number of moose if the distribution were random and determined only by the proportion of the area sampled. for this test we counted the number of pixels in each gis layer (e.g., dem, land cover) that had a moose and those that did not, as well as the total number of pixels in each category. we then estimated the number of pixels that should be expected if the distribution were random. this estimation was adjusted to sum to 112, the number of moose observed. a similar analysis was done to test the randomness of moose distribution relative to the parameter distance to forest openings/cut areas. in this analysis, there is no underlying image to count pixels, so we generated approximately 10 random ground points within each survey unit (totaling 341) using the gis. the distance from these points and the locations of moose to cut areas were then compared in a similar way. results elevation the elevational distribution of moose was not random (p = 0.012); however, locations were not clustered at one elevational range. the majority of moose (78%) were at low to mid elevation (300–599 m), as was most of the study area (71%). at higher elevations (>599 m) use (21%) was less than available (28%), and use declined sharply above 699 m (fig. 3). slope there was no relationship between any slope category and locations (p = 0.444); the proportion of locations was correlated with (similar to) the proportion available in each category. the majority of locations (∼88%) were on slopes of <10° and were evenly distributed (27–32%) among the 3 classes of lower slope. moose were found at the highest slope category (>10°) at a rate of 13%, similar to what was available (11%) (fig. 4). aspect there was no relationship between aspect and location (p = 0.932) with moose located in all aspect categories (fig. 5). proportional use was highest in the east (20%) and lowest in the north (7%). the proportion of locations in north-northeast-east directions (38%) was slightly higher than in southeast-south-southwest directions (35%) with more solar exposure (fig. 5). land cover the land cover analysis indicated that the proportional use of cover types was 20 winter distribution of vermont moose – millette et al. alces vol. 50, 2014 not entirely proportional with availability (p = 0.001), although the proportional use (locations) and availability in the 3 most common cover types (deciduous, mixedwood, coniferous) were similar (88 and 87%, fig. 6), indicating no preference for any forest type. use of scrub/shrub was ∼3 x higher than available (4%), and conversely, use of wetlands was negligible with 4% availability (fig. 6); the scrub/shrub cover type represented young forest openings. there were certain differences between the analyses with the noaa and aims land cover data. availability of the 3 major forest cover types and use of the deciduous cover type was similar in both analyses; however, use was measurably lower (30 vs. 48%) in mixedwood and higher in coniferous (17 vs. 6%) in the noaa analysis than the aims analysis (fig. 7). in part, the difference was due to the reallocation of 13 moose (12% of total moose identified) fig. 3. the distribution of moose observations by elevation class (usgs national elevation data) indicating that most moose (78%) were located at 300–600 m, yet moose were observed at higher elevations similar to availability (χ2 =16.34, p = 0.0121) in winter 2010, northeastern vermont, usa. 36 32 30 14 32 29 27 13 28 28 34 11 0 5 10 15 20 25 30 35 40 <2.5 2.5-5 5.1-10 >10 m oo se o bs er va � on s slope classes in degrees moos e observa�ons by slope clas s percent moos e observa�ons percent study area fig. 4. the even distribution of moose observations by slope class (usgs national elevation data) indicating no relationship (χ2 =2.68, p = 0.444) between slope and location, including steep slopes in winter 2010, northeastern vermont, usa. alces vol. 50, 2014 millette et al. – winter distribution of vermont moose 21 located in the wetland and scrub/shrub categories in the noaa classification into either deciduous, mixedwood, or coniferous classes in the aims classification. further, the more accurate mapping of mixedwood and coniferous stands supported by the 3.2 cm aims imagery reclassified certain moose from coniferous to the mixedwood cover type. distance to timber cuts based upon visual analysis of >100,000 thermal and natural color images asso‐ ciated with the 2010 census, we had strong 8 13 22 13 13 12 15 16 7 12 20 12 12 11 13 14 10 12 17 11 10 13 15 12 0 5 10 15 20 25 north northeast east southeast south southwest west northwes t m oo se o bs er va � on s aspect moos e observa�ons percent moos e observa�ons percent study area fig. 5. the even distribution of moose observations by aspect class (usgs national elevation data) indicating no relationship (χ2 = 3.04, p = 0.9319) between aspect and location in winter 2010, northeastern vermont, usa. fig. 6. the distribution of moose observations by noaa land cover type indicating the uneven use (χ2 =22.39, p = 0.001) of shrub/scrub (higher) and wetlands (lower); major forest cover types were used relative to availability and accounted for the majority of observations (88%) in winter 2010, northeastern vermont, usa. 22 winter distribution of vermont moose – millette et al. alces vol. 50, 2014 anecdotal evidence that moose locations were influenced by the relative distance to forest openings associated with timber harvest. an analysis using vermont orthophotography at 1.0 m spatial resolution was done to measure the radial distance from each location to the nearest forest opening. the distribution of locations relative to the pro‐ ximity of forest openings was non-random (p < 0.0001). the strong, direct relationship was evident as 65% of all locations were within 100 m, 85% within 300 m, and 99% within 700 m of a forest opening (fig. 8). fig. 7. the distribution of moose observations with aims land cover types in winter 2010, northeastern vermont, usa. locations increased in mixedwood and declined in conifer relative to the proportional distributions based on noaa cover types. fig. 8. the distribution of moose observations relative to distance to forest opening/timber cut (vermont 2009 orthophoto data) in winter 2010, northeastern vermont, usa. a strong correlation (χ2 =133.09, p <0.0001) existed between distance and the proportion of locations with the majority of locations at <100 m. alces vol. 50, 2014 millette et al. – winter distribution of vermont moose 23 discussion this gis analysis of the winter distribution of moose at the landscape level indicated that, for the most part, moose were located throughout the study area in proportion with available cover types, and were little influenced by elevation to 600 m, slope, or aspect. we further investigated whether moose distribution between 300–500 m was influenced by availability of cover type and forest openings but found no pattern. the only obvious deviation in use and availability of cover types was in wetlands (lower) and scrub/shrub (higher); forage use is presumably minimal in wetlands during winter whereas scrub/shrub areas were likely regenerating forest providing preferred winter browse in the region (thompson et al. 1995, scarpitti et al. 2005, bergeron et al. 2011, andreozzi et al. 2014). both analyses with the noaa and aims land cover data were reasonably consistent with most locations either in deciduous or mixedwood forest areas (as expected) and fewer in coniferous forest. however, a lower observation rate in the coniferous cover type could possibly reflect reduced sightability due to higher canopy cover. due to the inaccessibility of the survey transects and resource limitations, no independent attempt was made to estimate the sightability or error rate of moose not captured in imagery. therefore, the 93 thermal images containing moose were analyzed to test if the camera lens parallax produced different probabilities of detection inside and outside the image nadir due to screening effects of trees. images with moose were divided into 5 zones, each representing 20% of the image area from the westedge to the east-edge, and the numbers of observations were totaled for each zone. the distribution of moose across these zones indicated that there was no apparent screening effect due to lens parallax since more observations were at the edges of images where parallax distortions are highest (fig. 9; millette et al. 2011). with regard to the noaa land cover assignments (fig. 6), the number of observations in coniferous stands was similar to the available coniferous forest, as it was in the deciduous and mixedwood stands, suggesting that the aims-thermal sensor with its vertical view angle did not suffer from the screening effects of coniferous canopy; overall, coniferous stands (with fig. 9. the distribution of aims imagery parallax observations (93 images with 112 moose) indicating that minimal screening effects probably occurred in the coniferous cover type (see millette et al. 2011). 24 winter distribution of vermont moose – millette et al. alces vol. 50, 2014 and without moose) as seen in the aims color imagery were not considered densely stocked with tight canopy closure. interestingly, in adjacent northern new hampshire, moose were also observed in all cover types with less powerful infrared technology (adams et al. 1997). a substantial number of observations moved from coniferous to mixedwood stands when using the aims versus noaa classification. because the aims land cover classification is the product of fine (3.2 cm) spatial resolution color imagery processed by an experienced photo-interpreter, it is considered to be more accurate than the noaa land cover classification derived from machine-processed landsat tm data that is less sensitive to distinctions between coniferous and mixedwood forest. analysis of most observations that were reassigned from the noaa coniferous cover type to the aims mixedwood cover type indicated that, although conifers were present, they represented <20% of the total forest cover in each image. therefore, we believe that the aims analysis provided more accurate use of cover types. high use of mixedwood forest during winter was also measured in northwestern quebec (courtois et al. 2002) and northern new hampshire (scarpitti et al. 2005). mixedwood forest can be ideal winter habitat if it contains openings that provide preferred winter forage and coniferous canopy that provides thermoregulatory cover if needed. about 20% of moose were observed bedded in the study and sheltered by a conifer in either coniferous or mixedwood stands. the most striking landscape metric identified in this study was the strong relationship between moose locations and proximity to forest openings/timber cuts (fig. 8). although this relationship is widely recognized (see peek 1997), this study is based on a very large sample size of moose that can be analyzed at both the landscape and local scale. because locations were in proportion to the availability of cover types, it is apparent that timber harvesting activity both influenced winter habitat use and was extensive throughout the study area. given that use of regenerating forest is generally temporal (about 10–15 years), regular use of these surveys could identify shifting habitat use and/or sites with high winter fidelity that are often of concern relative to adequate forest regeneration. for example, andreozzi et al. (2014) identified certain 10–20 year old clear-cuts in the study area that had poor regeneration. although this study was conducted principally to provide a population estimate, it also provided gis imagery databases that can be explored for current and temporal analyses of habitat use. specifically, high resolution aerial imagery can be used to produce detailed forest metrics of tree species, stocking density, dbh measurements, and shrub condition in areas of dense winter moose populations. further, given the concentration of wintering moose, it is possible to cost-effectively task additional survey flights at lower altitudes to produce color imagery with sufficient spatial resolution (1.0 cm) for detailed assessments of size, age, and sex of individuals, twinning rates, and potentially health status based on weight and coat condition. future studies that simultaneously measure habitat use and population characteristics with this technology will have the distinct advantages of large sample size and accurate temporal information regarding changes in cover type, land use, and moose population size and distribution. references adams, k. p., p. j. pekins, k. a. gustafson, and k. bontaites. 1997. evaluation of infrared technology for aerial moose surveys in new hampshire. alces 33: 129–139. alces vol. 50, 2014 millette et al. – winter distribution of vermont moose 25 andreozzi, h. a., p. j. pekins, and m. l. langlais. 2014. impact of moose browsing on forest regeneration in northeast vermont. alces 50: 67–79. bergeron, d. h., p. j. pekins, h. f. jones, and w. b. leak. 2011. moose browsing and forest regeneration: a case study in northern new hampshire. alces 47: 39–51. courtois, r., and a. beaumont. 2002. a preliminary assessment on the influence of habitat composition and structure on moose density in clear-cuts of northwestern quebec. alces 38: 167–176. ———,c.dussault,f. potvin, and g.daigle. 2002. habitat selection by moose (alces alces) in clear-cut landscapes. alces 38: 177–192. dussault, c., r. courtois, and j. p. ouellet. 2006. a habitat suitability index model to assess moose habitat selection at multiple spatial scales. canadian journal of forest research 36: 1097–1107. gesch, d., m. oimoen, s. greenlee, c. nelson, m. steuck, and d. tyler. 2002. the national elevation dataset. photogrammetric engineering and remote sensing 68: 5–11. gillingham, m., and k. parker. 2008. the importance of individual variation in defining habitat selection by moose in northern british columbia. alces 44: 7–20. millette, t. l., d. slaymaker, e. marcano, c. alexander, and l. richardson. 2011. aims-thermal a thermal and highresolution color camera system integrated with gis for aerial moose and deer census in northeastern vermont. alces 47: 27–37. naip – 1mclrnaip digital ortho photography (vermont). 2009. vermont center for geographic information, waterbury, vermont, usa. noaa coastal services center (noaa). 2006. c-cap zone 65 2006-era land cover classification of landsat scenes. noaa ocean service, coastal services center, charleston, south carolina, usa. peek, j. m. 1997. habitat relationships. pages 351–375 in a.w. franzmann and c. c. schwartz, editors. ecology and management of the north american moose. smithsonian institute press, washington, d.c., usa. poole, k., and k. stuart-smith. 2005. finescale winter habitat selection by moose in interior montane forests. alces 41: 1–8. potvin, f., and r. courtois. 2004. winter presence of moose in clear-cut black spruce landscapes: related to spatial pattern or to vegetation? alces 40: 61–70. scarpitti, d., c. habeck, a. r. musante, and p. j. pekins. 2005. integrating habitat use and population dynamics of moose in northern new hampshire. alces 41: 25–35. snedecor, g., and w. cochran. 1989. statistical methods, 8th edition. iowa state university press, ames, iowa, usa. thompson, m. e., j. r. gilbert, g. j. matula jr., and k. i. morris. 1995. seasonal habitat use by moose on managed forest lands in northern maine. alces 31: 233–245. van beest, f., m. atle, l. loe, and j. milner. 2010. forage quantity, quality and depletion as scale-dependent mechanisms driving habitat selection of a large browsing herbivore. journal of animal ecology 80: 771–785. 26 winter distribution of vermont moose – millette et al. alces vol. 50, 2014 winter distribution of moose at landscape scale in northeastern vermont: a gis analysis introduction methods study area data gis analyses results elevation slope aspect land cover distance to timber cuts discussion references alces24_62.pdf body temperature of captive moose infested with winter ticks edward m. addison1,2, robert f. mclaughlin1,3, and peter a. addison1,4 1wildlife research and development section, ontario ministry of natural resources, 300 water street, 3rd floor north, peterborough, ontario, canada k9j 8m5. abstract: eighteen captive moose calves (alces alces) were divided into 3 groups that represented 3 levels of winter tick (dermacentor albipictus) infestation (0, 21,000, and 42,000 ticks). a total of 321 body temperatures (tb) were taken on 19 occasions between late november and mid-april. the mean tb of individuals was 38.2 ± 0.4 °c, ranging from 38.0–38.3 °c, and was not different among the control and infested groups (p = 0.816), but varied temporally (p < 0.001) with a significant interaction effect between treatment and time (p = 0.041); these temporal differences are unexplained. the tbs measured in this study are some of the lowest reported for moose and presumably represent the resting tb of free-ranging moose, more so than those measured after pursuit, restraint, and/or immobilization during capture. this was not a definitive test of the effects of tick infestation on wild moose because the captive moose consumed a high quality diet throughout winter and surprisingly low numbers of ticks remained on the animals in mid-april. alces vol. 50: 81–86 (2014) key words: body temperature, alces alces, dermacentor albipictus, moose, tb, winter tick. premature hair loss by moose (alces alces) in winter that is associated with infestations of winter tick (dermacentor albipictus) is well documented (e.g., addison et al. 1979, samuel and barker 1979, samuel 1991) including by mclaughlin and addison (1986) studying the same captive moose reported here. this hair loss might influence body temperature (tb) that is reflective of increased energetic cost and stress in moose. the typical tb of moose reported in the literature is usually measured on individuals that were pursued, restrained, and/or immobilized during capture. many of these values may reflect higher than resting tb since excitability raises tb in moose (franzmann et al. 1984). objectives of this study were to assess the possible effects of winter tick infestation on tb of moose, and to obtain tb from captive animals that more accurately represent resting tb of unstressed free-ranging moose. importantly, animals in this study were young-of-the-year, exceptionally tractable, and readily accepted the measurement procedure. because technological advances in telemetry now allow tb to be measured in free-ranging moose, these data are also valuable for related comparisons. methods the experiments were conducted in algonquin provincial park, ontario (45° 30′ n, 78° 35′ w) where 13 of 18 calves were captured at <2 weeks of age in may 1982; 5 calves were from other areas in central and northeastern ontario (addison and mclaughlin 1993). male and female calves were paired in each of 6 adjacent pens 2present address: ecolink science, 107 kennedy street west, aurora, ontario, canada l4g 2l8 3r.r. #3, penetanguishene, ontario, canada l0k 1p0 4northwest region, regional operations division, ontario ministry of natural resources, 173 25th sideroad, rosslyn, ontario, canada p7k 0b9 81 (29.6 × 16.5 m) located within a mixed forest stand with little undergrowth and a partial canopy (50% in summer) of white pine (pinus strobus), white birch (betula papyrifera), trembling (populus tremuloides) and big tooth aspen (p. grandidentata). calves were weaned as described by addison et al. (1983) and from late october to the end of the experiment were fed ad libitum a ruminant ration containing 16% crude protein, 2.5% crude fat, and 6% crude fiber (united cooperative of ontario, mississauga, ontario, canada). husbandry of moose and experimental design for this study were as described in addison et al. (1994) with all animals assumed born on 15 may 1982. the 18 calves were divided into 3 treatment groups: moose with no winter ticks (n = 5; 2f:3m), moose infested with 21,000 larval winter ticks (n = 7; 3f:4m), and moose infested with 42,000 larval winter ticks (n = 6; 3f:3m). larval ticks were applied between mid-september and midoctober 1982, and all moose were euthanized at the end of the experiment (18–28 april 1983). the hair was dissolved and hides checked for ticks as described by addison et al. (1979). for months prior to the application of ticks, the study animals were attracted with food to a monitoring station where they stood quietly while we measured weight and took linear measurements. the tb was measured by inserting a standard, large animal mercury thermometer into the rectum. for 16 moose, tb was usually measured every 5–9 days from 24 november 1982 to 14 april 1983, except for a 2-week period from late january to mid-february 1983 (table 1). fewer data were available from the 2 other moose that were sacrificed prior to the completion of this study. the mean tb of individuals within and between sampling times was calculated using all 18 moose; however, data were missing for certain individuals on particular dates. all data for 3 moose with missing rectal temperatures for ≥3 of the 19 dates were removed from statistical analysis. further, because measurements were missing from 3 moose on one of the 19 dates, all were removed from the analysis for this date. after removing these data, each treatment group was comprised of 5 moose with measurements from 16 dates, for a total of 80 measurements per treatment. we tested for treatment effect (among groups), temporal effect, and an interaction effect between treatment and time using a two-factor anova with repeated measures of tb with the aov function in r (r core team 2013). results female and male calves respectively weighed 161 ± 8 and 178 ± 5 kg in midnovember, and 200 ± 17 and 218 ± 20 kg at the end of the experiment when 11 months old (addison et al. 1994). the 5 control moose harboured 0, 0, 4, 21, and 85 winter ticks at the conclusion of the experiment; the animal harbouring 85 ticks had limited (5%) hair loss (mclaughlin and addison 1986). in contrast, 1179–8290 ticks were recovered from the infested moose at the end of the experiment. hair loss was estimated at 23–44% in 8 of 10 infested animals, and 2 and 4% in the other 2 moderately infested moose (see mclaughlin and addison 1986). the mean tb (n = 321) was 38.2 ± 0.4 °c ranging from 36.8–40.7 °c. individual mean tb ranged from 38.0–38.3 °c with >99% of individual measurements from 36.8 – 39.4 °c (fig. 1). mean tb was not different among treatment groups (f2,12 = 0.207, p = 0.816), but did vary over time (f15,180 = 6.385, p < 0.001). there was a significant interaction effect between treatment and time (f30,180 = 1.561, p = 0.041) indicating that tb of treatment groups varied temporally. however, no discernible relationship existed as the mean tb of groups did not change in similar direction in all periods (fig. 2). 82 body temperatures of moose – addison et al. alces vol. 50, 2014 table 1. mean rectal body temperature (tb, °c) of standing captive calf moose exposed to 3 levels of winter tick loads (0, 21,000, 42,000 larvae) in fall 1982, algonquin provincial park, ontario; moose sample size is in parentheses. winter tick infestation level date 0 21,000 42,000 24 nov 1982 38.2 ± 0.2 (5) 38.3 ± 0.4 (6) 37.7 ± 0.5 (6) 30 nov 38.2 ± 0.1 (5) 38.0 ± 0.4 (7) 37.9 ± 0.5 (6) 5 dec 38.3 ± 0.2 (5) 38.0 ± 0.3 (6) 38.0 ± 0.6 (6) 12 dec 37.9 ± 0.4 (5) 37.8 ± 0.2 (7) 38.0 ± 0.3 (6) 19 dec 38.0 ± 0.3 (5) 38.4 ± 0.5 (6) 38.2 ± 0.7 (6) 26 dec 37.9 ± 0.1 (5) 38.0 ± 0.1 (7) 37.8 ± 0.1 (6) 3 jan 1983 38.5 ± 0.6 (4) 38.5 ± 1.0 (7) 38.3 ± 0.3 (6) 12 jan 38.4 ± 0.2 (5) 38.4 ± 0.3 (7) 38.4 ± 0.1 (6) 17 jan 37.7 ± 0.7 (4) 37.8 ± 0.6 (7) 38.0 ± 0.4 (6) 25 jan 38.4 ± 0.3 (5) 38.6 ± 0.1 (7) 38.4 ± 0.2 (6) 31 jan 37.9 ± 0.2 (5) 38.2 ± 0.5 (7) 38.1 ± 0.2 (6) 15 feb 38.1 ± 0.2 (5) 38.4 ± 0.1 (7) 38.3 ± 0.2 (6) 1 mar 38.1 ± 0.3 (5) 37.9 ± 0.3 (5) 38.0 ± 0.4 (5) 7 mar 38.1 ± 0.1 (5) 38.1 ± 0.3 (6) 38.3 ± 0.3 (5) 15 mar 38.3 ± 0.2 (5) 38.3 ± 0.4 (6) 38.6 ± 0.2 (5) 23 mar 38.4 ± 0.2 (5) 38.5 ± 0.3 (6) 38.6 ± 0.3 (5) 28 mar 38.3 ± 0.4 (5) 38.5 ± 0.2 (6) 38.5 ± 0.3 (5) 5 apr 38.3 ± 0.3 (5) 38.5 ± 0.4 (6) 38.5 ± 0.3 (5) 14 apr 38.5 ± 0.5 (5) 38.4 ± 0.5 (6) 38.4 ± 0.5 (5) 0 0.05 0.1 0.15 0.2 0.25 0.3 37 37.2 37.4 37.6 37.8 38 38.2 38.4 38.6 38.8 39 39.2 39.4 39.6 39.8 40 40.2 40.4 40.6 40.8 body temperature (°c) p ro po rt io na l f re qu en cy heavily infested moderately infested not infested figure 1. the frequency distribution of body temperature as measured in captive calf moose in ontario, 1982–1983. alces vol. 50, 2014 addison et al. – body temperatures of moose 83 discussion although one could postulate that hair loss is one factor influencing tb, the lack of difference in tb among the treatment groups was not surprising. the number of recovered ticks was relatively low in contrast with tick loads measured on heavily infested wild moose (samuel and barker 1979, samuel 2004). the number of larval winter ticks applied was an a priori estimate of the maximum numbers of ticks that would allow for the parasitic phases of the tick-moose cycle to be completed, while maintaining accept‐ able standards for the humane treatment of experimental animals, an objective that was achieved. for example, although re‐ duced pericardial and abdominal fat reservoirs occurred in the infested versus control moose (mclaughlin and addison 1986), we presume that all moose retained sufficient tissue reservoirs for adequate thermoregulation. further, given the wide range in volume of hair loss reported within a single treatment group (i.e., 2–24% in moderately infested moose; mclaughlin and addison 1986) and the limited number of moose per treatment group, it would be difficult to detect treatment differences. the negligible to limited seasonal variation in tb is consistent with previous reports of seasonal variation in tb in moose (franzmann et al. 1984) and wapiti (cervus elaphus) (parker and robbins 1984). on a cautionary note, the tbs measured in this study should not be considered representative of those of heavily infested wild moose with extensive hair loss. the captive moose received higher quality, more accessible food throughout winter compared to freeranging moose, and seldom experienced ambient temperatures considered thermally stressful (renecker and hudson 1986, addison and mclaughlin 2014). the significant interaction effect bet‐ ween treatment and time indicated that tb of treatment groups varied over time, but did so in different directions. environmental factors that might influence temporal differences remain unclear, but could include effects of handling during measurements as 36.0 36.5 37.0 37.5 38.0 38.5 39.0 39.5 11/5/82 11/25 12/15 1/4/83 1/24 2/13 3/5 3/25 4/14 5/4 b od y t em pe ra tu re ( °c ) heavy moderate none poly. (heavy) poly. (none) poly. (moderate) figure 2. the relationship between time and body temperature (rectal) of captive calf moose in 3 treatment groups of winter tick infestation: heavy (42,000), moderate (21,000), none. although treatment and time were statistically related, the temporal trend differed among treatment groups; ontario, 1982–1983. 84 body temperatures of moose – addison et al. alces vol. 50, 2014 higher tb occurs with increased excitability in immobilized moose (franzmann et al. 1984). however, we recognized no overt excitability in the study animals during measurements and the differences may simply reflect normal variation. the tbs measured in this study were lower than any reported from healthy moose, and likely reflect the psychical state of our calm tractable moose, or conversely, the more stressful condit‐ ions associated with measurements of freeranging moose. the upper end of the range of tb was consistent with data from prior studies (e.g., franzmann et al. 1984) and most similar to those of captive moose (38.0–39.7 °c) that were not immobilized (renecker and hudson 1986). seal et al. (1985) reported a mean tb of 38.6 °c for free-ranging moose immobilized from the ground as they approached mineral licks. in contrast, higher tb was reported for wild moose pursued and restrained (x̄ = 39.3 °c, range = 38.0– 40.4 °c), or pursued and immobilized (x̄ = 40.5 °c, range = 38.0–42.8 °c, roussel and patenaude 1975; x̄ = 39.1–39.7 °c, delvaux et al. 1999). most tbs of moose have been measured in adults and not young-of-the-year as reported here. differences in size and age likely have little if any influence on tb since in most ungulates tb varies little relative to body mass, and if variable, young animals generally have higher tb than adults (parker and robbins 1985). in summary, there was no evidence that presence of winter ticks as applied in this experiment had any direct influence on tb of moose. the tbs measured in our highly tractable animals were the lowest reported for healthy moose, and consistent with the view that level of excitability influences tb. importantly, they provide the baseline tb for resting moose that is important for metabolic modeling and comparison with tb measured via telemetry of free-ranging moose. acknowledgements we appreciate d. j. h. fraser for his coordination of many early aspects of this study. special thanks go to a. rynard, a. macmillan, m. jefferson, v. ewing, and d. bouchard for their steadfast assistance in collection of data and moose husbandry under adverse conditions. additional assistance was provided by c. pirie, m. a. mclaughlin, d. carlson, d. joachim and p. methner, and l. smith, k. paterson, k. long, a. jones, s. gadawski, s. fraser, d. fraser, and l. berejikian assisted in the earlier care of calves. we appreciate the assistance of c. d. macinnes and g. smith and staff for their administrative support. thanks to a. r. rodgers for advice with statistics. field work was conducted at the wildlife research station in algonquin park where r. keatley, p. c. smith and staff were of great help. thank you to m. lankester and anony‐ mous reviewers whose valuable suggestions were incorporated into the manuscript. references addison, e. m., f. j. johnson, and a. fyvie. 1979. dermacentor albipictus on moose (alces alces) in ontario. journal of wildlife diseases 15: 281–284. ———, and r. f. mclaughlin. 1993. seasonal variation and effects of winter ticks (dermacentor albipictus) on consumption of food by captive moose (alces alces) calves. alces 29: 219–224. ———, and ———. 2014. shivering by captive moose infested with winter ticks. alces 50: 87–92. ———, ———, and j. d. broadfoot. 1994. growth of moose calves (alces alces americana) infested and uninfested with winter ticks (dermacentor albipictus). canadian journal of zoology 72: 1469– 1476. alces vol. 50, 2014 addison et al. – body temperatures of moose 85 ———, ———, and d.j.h. fraser. 1983. raising moose calves in ontario. alces 18: 246–270. delvaux, h., r. courtois, l. breton, and r. patenaude. 1999. relative efficiency of succinylcholine, xylazine, and carfentanil mixtures to immobilize free-ranging moose. journal of wildlife diseases 35: 38–48. franzmann, a. w., c. c. schwartz, and d. c. johnson. 1984. baseline body temperatures, heart rates, and respiratory rates of moose in alaska. journal of wildlife diseases 20: 333–337. mclaughlin, r. f., and e. m. addison. 1986. tick (dermacentor albipictus)-induced winter hair-loss in captive moose (alces alces). journal of wildlife diseases 22: 502–510. parker, k. l., and c. t. robbins. 1984. thermoregulation in mule deer and elk. canadian journal of zoology 62: 1409–1422. ———, and ———. 1985. thermoregulation in ungulates. pages 161–182 in r. j. hudson and r. g. white, editors. bioenergetics of wild herbivores. crc press, boca raton, florida, usa. r core team. 2013. r: a language and environment for statistical computing. r foundation for statistical computing, vienna, austria. <http://www.r-project. org/>. renecker, l. a., and r. j. hudson. 1986. seasonal energy expenditures and thermoregulatory responses of moose. canadian journal of zoology 64: 322–327. roussel, y. e., and r. patenaude. 1975. some physiological effects of m99 etorphine on immobilized free-ranging moose. journal of wildlife management 39: 635–636. samuel, b. 2004. white as a ghost: winter ticks and moose. natural history series, volume 1. federation of alberta naturalists, edmonton, alberta, canada. samuel, w. m. 1991. grooming by moose (alces alces) infested with the winter tick, dermacentor albipictus (acari): a mechanism for premature loss of winter hair. canadian journal of zoology 69: 1255–1260. ———, and m. j. barker. 1979. the winter tick, dermacentor albipictus (packard, 1869) on moose alces alces (l.) of central alberta. proceedings of the north american moose conference and workshop 15: 303–348. seal, u. s., s. m. schmitt, and r. o. peterson. 1985. carfentanil and xylazine for immobilization of moose (alces alces) on isle royale. journal of wildlife diseases 21: 48–51. 86 body temperatures of moose – addison et al. alces vol. 50, 2014 http://www.r-project.org/ http://www.r-project.org/ body temperature of captive moose infested with winter ticks methods results discussion acknowledgements references alces20_129.pdf alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces22_443_moose&forestmgmt.pdf alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces26_86.pdf alces(23)_311bobcornellhonoured.pdf alces vol. 23, 1987 alces21_91.pdf alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces(25)_31.pdf alces22_361.pdf alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces(23)_181.pdf alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces24_118.pdf mass, morphology, and growth rates of moose in north dakota william f. jensen1, jason r. smith2, james j. maskey jr.3, james v. mckenzie (deceased)1, and roger e. johnson (retired)4 1north dakota game and fish department, 100 north bismarck expressway, bismarck, north dakota 58501 usa; 2north dakota game and fish department, 3320 east lakeside road, jamestown, north dakota 58401 usa; 3university of mary, 7500 university dr., bismarck, north dakota 58504 usa; 4north dakota game and fish department, 7928 45th street ne, devils lake, north dakota 58301 usa. abstract: this paper provides predictive formulas to estimate live weights of moose (alces alces andersoni) from hunter-harvested animals and evaluate growth rates of moose in north dakota, and reviews weight-related measurements among moose populations. from 1978–1990, morphometric data were collected on 224 hunter-killed moose harvested after the rut (10 november–12 december) in north dakota. body mass increased rapidly for both sexes from 0.5 years to 1.5 years-of-age. whole weight and total body length reached an asymptote for both sexes by 5.5 years; mean whole weight appeared to decline among older individuals. although field dressed weight was the best predictor of whole weight (r2 = 0.93; n = 154), total body length provided reasonably good estimates of whole weight (r2 = 0.76; n = 153). whole weight estimates based upon shoulder height (r2 = 0.33; n = 158) and hind-foot length (r2 = 0.46; n = 163) were less reliable. we also used morphometric variables to predict field dressed weight, carcass weight, and visceral weight. field dressed weight was the best predictor of antler width (r2 = 0.72; n = 108) and antler width was a good predictor of male age (r2 = 0.70; n = 119). when compared to other north american populations, average weights of moose harvested in north dakota tended to be higher in all age classes. additionally, the calf-to-yearling growth rate of female moose in north dakota was as high, or higher than in other populations. morphometric comparisons of free-ranging moose from various north american populations had much size overlap, with southern and eastern moose populations tending to have largest average adult body mass. sexual dimorphism of mature north dakota moose (> 4.5 years) was comparable to that in other populations. alces vol. 49: 1–15 (2013) key words: alces alces, body mass, morphometrics, sex, age, growth rates, north dakota. body weights and measurements of cervids provide insight on condition and health of local populations (clutton-brock et al. 1982, sauer 1984, verme and ullrey 1984, loudon 1987). they allow for the analysis of energetic requirements, energetic capability, and other metabolic parameters (schwartz et al. 1987), and how subspecies vary in size, shape, and rate of growth (bubenik 1998, geist 1998). comparative morphometric data on north american moose (alces alces) are limited (blood et al. 1967, timmerman 1972, schladweiler and stevens 1973, peterson 1974, franzmann et al. 1978, crichton 1980, adams and pekins 1995, lynch et al. 1995, broadfoot et al. 1996). additionally, measurement and definition of morphometric variables vary among studies making comparisons among populations problematic. difficulties in accessing hunter-killed moose, and the physical labor involved in handling these 1 animals have undoubtedly limited data collection. historically, moose in north dakota were restricted to the heavily forested areas of the turtle mountains, pembina hills, and the major tributaries of the red river. accounts of early traders in the area indicated that they were not as abundant as elk (cervus elaphus) or other big game species, and apparently disappeared from the state during the early 20th century (knue 1991). by the 1960s, moose had returned to north dakota and small numbers were occupying portions of their historic range. in 1977, the first modern moose hunting season allowed the harvest of 10 moose in cavalier, pembina, and walsh counties. the expansion of moose into the relatively accessible farmland of north dakota, coupled with the willingness of local farmers to assist hunters with loading and transporting animals with farm equipment, made it feasible for the north dakota game and fish department (ndgfd) to collect morphometric data. we provide analyses of 1) age and sex-specific weights and measurements, 2) measurement-weight relationships, 3) age and morphometric relationships, and 4) growth rates of hunter-killed moose from north dakota. our goal is to provide predictive formulas for estimating whole, field-dressed, carcass, and viscera weights of hunter-killed moose and to make comparisons with other north american and european populations. methods morphometric data were collected on 224 hunter-killed moose examined between 1978 and 1990. hunters were asked whenever possible to bring their moose in whole to a check station (i.e., prior to removal of viscera, hide, head, or legs). all moose were harvested between 10 november and 12 december after the rutting season; date of kill, sex, and legal descriptions (section, township, range) for all kill sites were recorded. the distribution of the animals examined was between 47.10 and 48.99° n and 97.14 and 100.41° w (fig. 1). moose hunting units m1c and m4 are comprised primarily of aspen (populus spp.) forests with intermingled cropland; the remainder of the harvest area was drift prairie with heavy conversion to cropland (fig. 1). for a more complete description of habitat see maskey (2008). weight was measured with local grain elevator scales with an accuracy of ±4.5 kg. recorded weights were whole weight (ww) which comprised a completely intact carcass except for loss of blood and tissue resulting from gunshot wounds; field dressed weight (fdw) which included carcass weight minus all thoracic and abdominal viscera; viscera weight (vw) which included all thoracic and abdominal organs and their contents including blood and the contents of the digestive system; and carcass weight (cw) which comprised the dressed carcass minus the head, hide, and legs below the hock joint. length was measured (nearest cm) before the moose was dressed-out following the methods of peterson (1974). with the carcass laid flat on its side, with head and spinal column supported on the same plane, total body length (tbl) was measured from tip of nose to tip of tail (point of last coccyx bone, excluding hair) by following the dorsal (spinous) processes of the vertebra. with the carcass laid flat on its side and the front leg positioned so that it was straight and perpendicular to the longitudinal axis of the body, hind-foot length (hfl) was measured from the calcaneum (heel bone of hock) to the tip of hoof, and shoulder height (sh) from the superior angle of the scapula (cartilaginous top of shoulder) to the distal tip of the front hoof. antler width (aw) was measured at the greatest spread between the tines and at a right angle to the longitudinal axis of the skull. prior to additional analyses, we 2 morphology of north dakota moose – jensen et al. alces vol. 49, 2013 used t-tests to determine whether any morphometric measurements differed between male and female moose of all ages. whole weight, fdw, cw, vw, and aw were compared to other morphometric measurements using simple linear regressions. all statistical analyses were conducted using r, version 2.11.1 (r development core team 2010). incisor eruption (peterson 1955) was used to identify young-of-the-year or calves (≤6 month-of-age) and front incisors were collected from moose ≥6 months of age for cementum annuli analysis (gasaway et al. 1978, haagenrud 1978). ageing was performed by matson's laboratory, milltown, mt, and by co-author r. johnson (ndgfd) after 1980. the relationship between each morphometric variable and age was examined with linear regression using square-roottransformed age as the dependent variable; to facilitate utility, we present backtransformed equations in results. growth rates for moose have been described as falling into 3 phases: 1) a selfaccelerated phase (schwartz 1998) of near exponential growth from birth to weaning (4–6 months old) allowing the calf to follow its mother over rough terrain and obstacles (geist 1998), 2) a second phase of rapid growth from calf to yearling (16–18 months old) allowing young moose to reach a body size (250–280 kg ww) that allows yearlings to confront predators (geist 1998), and 3) a self-inhibiting growth phase where seasonal peaks and troughs in body mass occur at different times for males and females (schwartz et al. 1987). results and discussion focus on these 3 growth phases. healthy north american moose calves have a mean birth weight ranging from fig. 1. locations of 224 moose harvested in north dakota, usa (1978–1990). each dot represents the site, to the nearest section, where ≥1 moose were harvested. alces vol. 49, 2013 jensen et al. – morphology of north dakota moose 3 12.6–18 kg for single calves (kellum 1941 in peterson 1955, franzmann et al. 1980, schwartz 1998) and 13.6 kg for twins (franzmann 1978). the mean weight for 1–3 day-old alaskan calves was 18.0 kg (n = 109; franzmann et al. 1980). average weights of captured calves < 2 weeks old in ontario averaged 15.7 kg (n = 8) for females and 17.3 kg (n = 10) for males (addison et al. 1994). the average weight of 43 captured neonate calves in alberta was 19.6 kg; however, some were captured as late as august (welch et al. 1985). lacking information on birth weights for north dakota moose, we used the range of 13–18 kg as the basis for calculating phase 1 growth rates. we determined percent change of growth rate during phase 1 by dividing calf weight at 0.5 years by the neonate weights of 13 and 18 kg, and we estimated the percent change during phase 2 by dividing the average yearling (1.5 years) weight by the average calf weight. we measured the level of sexual dimorphism for moose ≥4.5 years old. only these moose were included because they would be at or near their maximum size and this age range would best permit comparison to other studies. we calculated dimorphic ratios (male:female) for ww, fdw, cw, vw, tbl, hfl, and sh. we also used t-tests to examine whether these measures differed significantly between males and females ≥4.5 years old. results morphometric measurements sample sizes for ww, fdw, vw, and cw were obtained from 160, 166, 146, and 40 moose, respectively (tables 1 and 2). sample sizes for tbl, hfl, sh, and aw were obtained from 206, 196, 200, and 121 moose, respectively (tables 3 and 4). antler width appeared to plateau at 6.5 years and then decline in older males (table 4). morphometric relationships morphometric measurements were not significantly different for males and females of all ages. therefore, we combined sexes when conducting regression analyses for these variables. whole weight was best predicted by fdw (r2 = 0.93, n = 154), followed by tbl (r2 = 0.76, n = 153); fdw was best predicted by tbl (r2 = 0.70, n = 181) and cw (r2 = 0.65, n = 39). antler width was most highly correlated with fdw (r2 = 0.72, n = 108). all regression equations for predicting ww, fdw, cw, vw, and aw are provided in table 5. age was best predicted by ww (r2 = 0.71, n = 114), followed by fdw (r2 = 0.70, n = 166; table 6). antler width was also a reasonable estimator of male age (r2 = 0.70, n = 119; table 6). growth rates and patterns during phase 1, both post-rut male and female calves averaged 196 kg at about 7 months of age, representing a 989–1455% increase in body mass. the average growth rate would be 1% per day assuming a mean age of about 180 days, given a 1 december harvest date. the fdws of female and male calves were 61 and 72% of ww; fdw of yearling females and males were 73 and 70%, respectively. during growth phase 2, female and male calves (averaging 196 kg at 6–7 months) increased their ww over the next year by 69 and 65%, respectively. the fdw during phase 2 increased by 102 and 59.3% for females (n = 10) and males (n = 32), respectively (table 7). the fdws of female and male calves were 61 and 72% of ww; fdw of yearling females and males were 73 and 70%, respectively. whole weights, fdw, and tbl plateaued at 5.5 years for both males and females during the self-inhibiting growth phase (tables 1–4). 4 morphology of north dakota moose – jensen et al. alces vol. 49, 2013 table 1. age-weight relationship for female moose harvested in north dakota, usa (1978–1990). whole weight (kg) field dressed weight (kg) viscera weight (kg) carcass weight (kg) age n mean (± sd) range n mean (± sd) range n mean (± sd) range n mean (± sd) range 0.5 years 5 196.0 ± 21.6 176.9–226.8 4 119.1 ± 30.2 77.1–145.1 4 73.3 ± 21.1 54.4–99.8 2 127.7 ± 54.5 86.2–163.3 1.5 years 12 331.1 ± 44.1 272.2–444.5 10 241.1 ± 26.9 204.1–290.3 10 94.6 ± 25.1 68.0–154.2 2 195.0 ± 19.2 181.4–208.7 2.5 years 5 366.5 ± 20.9 331.1–385.6 6 255.8 ± 12.5 240.4–272.2 5 109.5 ± 17.0 84.8–127.0 4 197.3 ± 14.7 179.2–211.4 3.5 years 11 410.3 ± 50.9 349.3–526.2 12 285.4 ± 36.5 229.1–358.3 10 158.1 ± 27.6 95.3–172.4 1 281.2 281.2 4.5 years 2 437.7 ± 22.5 421.8–453.6 3 323.6 ± 42.1 294.8–371.9 2 138.3 ± 16.0 127.0–149.7 1 231.3 231.3 5.5 years 5 467.2 ± 39.9 417.3–512.6 5 330.7 ± 29.8 292.6–371.9 5 136.5 ± 30.5 90.7–167.8 1 220.0 220 6.5 years 2 437.7 ± 60.9 394.6–480.8 3 310.0 ± 15.9 294.8–326.6 2 127.0 ± 38.5 99.8–154.2 1 226.8 226.8 7.5 years 2 435.4 ± 25.7 417.3–453.6 2 310.7 ± 22.5 294.8–326.6 2 124.7 ± 3.2 122.5–127.0 1 239.5 239.5 8.5 years 3 444.5 ± 64.0 385.6–512.6 3 319.0 ± 41.2 281.2–362.9 3 125.5 ± 22.8 104.3–149.7 0 10.5 years 1 489.9 489.9 1 335.7 335.7 1 154.2 154.2 0 ≥ 1.5 years 43 397.7 ± 64.9 272.2–526.2 45 285.3 ± 43.2 204.1–371.9 40 127.5 ± 59.2 68.0–172.4 11 216.2 ± 29.1 179.2–281.2 table 2. weight categories (see methods) by age for male moose harvested in north dakota, usa (1978–1990). whole weight (kg) field dressed weight (kg) viscera weight (kg) carcass weight (kg) age n mean (± sd) range n mean (± sd) range n mean (± sd) range n mean (± sd) range 0.5 years 5 196.0 ± 34.3 140.6–226.8 4 141.7 ± 5.7 136.1–149.7 4 68.0 ± 12.3 49.9–77.1 0 1.5 years 27 323.6 ± 31.4 249.5–403.7 32 225.7 ± 24.5 181.4–290.3 27 96.3 ± 18.5 63.5–163.3 6 177.9 ± 24.4 147.4–207.3 2.5 years 35 402.3 ± 43.3 290.3–485.3 35 292.2 ± 26.7 217.7–349.3 30 119.1 ± 18.9 68.0–158.8 8 219.5 ± 37.9 158.8–281.2 3.5 years 22 446.3 ± 40.4 335.7–503.5 22 322.5 ± 30.1 254.0–367.4 20 121.8 ± 17.9 77.1–154.2 5 236.9 ± 32.2 195.0–270.8 4.5 years 8 444.1 ± 30.1 408.2–499.0 9 320.0 ± 36.6 272.2–381.0 8 123.6 ± 201 90.7–154.2 2 246.5 ± 4.8 243.1–249.9 5.5 years 7 496.6 ± 60.1 408.2–589.7 7 365.5 ± 22.0 340.2–408.2 6 141.7 ± 34.6 95.3–181.4 4 257.3 ± 26.3 235.0–295.3 6.5 years 7 479.5 ± 51.4 403.7–535.2 6 352.3 ± 33.5 303.9–394.6 6 121.0 ± 27.8 99.8–172.4 2 318.2 ± 88.8 255.4–381.0 7.5 years 0 1 349.3 349.3 0 0 9.5 years 1 453.6 453.6 1 335.7 335.7 1 117.9 117.9 0 ≥ 1.5 years 107 430.6 ± 69.2 249.5–589.7 113 311.7 ± 53.3 181.4–408.2 98 121.0 ± 24.1 63.5–181.4 27 245.5 ± 62.4 158.8–381.0 a l c e s v o l . 4 9 , 2 0 1 3 je n s e n e t a l . – m o r p h o l o g y o f n o r t h d a k o t a m o o s e 5 table 3. morphological measurements of female moose harvested in north dakota, usa (1978–1990). total body length (cm) hind foot length (cm) shoulder height (cm) age n mean (± sd) range n mean (± sd) range n mean (± sd) range 0.5 years 5 203.8 ± 12.2 190.0–219.7 5 71.0 ± 2.9 68.0–76.2 4 151.8 ± 9.7 143.0–164.0 1.5 years 13 248.5 ± 10.3 236.5–265.0 13 77.0 ± 2.6 71.1–81.9 13 175.5 ± 6.8 157.5–185.2 2.5 years 8 248.2 ± 10.4 232.0–264.7 6 77.9 ± 2.2 74.8–81.0 8 181.2 ± 4.7 177.3–184.9 3.5 years 15 257.5 ± 17.5 214.0–279.2 14 79.2 ± 3.4 74.1–86.1 15 188.2 ± 5.2 179.8–197.3 4.5 years 3 271.2 ± 7.7 266.1-280.0 3 79.0 ± 0.6 78.3–79.4 3 190.6 ± 6.7 182.9-194.7 5.5 years 5 273.2 ± 10.6 259.7–286.1 5 79.4 ± 2.4 77.4–82.9 5 193.4 ± 9.9 185.0–206.8 6.5 years 3 267.3 ± 5.0 263.6–272.9 3 80.7 ± 1.5 79.2–82.1 3 192.8 ± 0.4 192.4–193.1 7.5 years 2 263.1 ± 4.0 260.2–265.9 2 81.7 ± 1.2 80.8–82.5 2 190.9 ± 9.7 184.0–197.7 8.5 years 3 267.4 ± 12.1 255.2 279.4 3 77.9 ± 2.1 75.9–80.0 3 192.3 ± 7.2 188.4–200.6 10.5 years 1 261.1 261.1 1 80.3 80.3 1 189.1 189.1 ≥ 1.5 years 53 257.5 ± 14.8 214.0–286.1 50 78.6 ± 2.8 71.1–86.1 53 157.5 ± 206.8 157.5–206.8 table 4. morphological measurements of male moose harvested in north dakota, usa (1978–1990). total body length (cm) hind foot length (cm) shoulder height (cm) antler width (cm) age n mean (± sd) range n mean (± sd) range n mean (± sd) range n mean (± sd) range 0.5 years 6 201.4 ± 14.0 176.0–215.0 6 70.6 ± 2.5 66.0–73.5 6 153.0 ± 8.0 141.0–161.5 1 23.5 23.5 1.5 years 40 244.5 ± 11.9 218.0–272.3 38 77.7 ± 2.6 72.2–82.1 38 179.7 ± 7.1 167.5–194.5 28 69.9 ± 10.7 53.3–92.3 2.5 years 48 258.0 ± 12.9 227.0–288.3 46 80.0 ± 3.2 68.9–85.5 46 187.9 ± 9.4 162.0–203.5 40 89.1 ± 11.5 66.4–125.1 3.5 years 26 267.7 ± 13.9 240.0–295.3 26 81.0 ± 2.9 75.0–86.6 26 191.5 ± 8.4 176.0–206.3 26 100.1 ± 9.5 84.4–129.9 4.5 years 9 269.9 ± 9.4 257.5–282.8 7 80.1 ± 2.8 76.2–84.5 9 188.8 ± 6.3 181.0–200.5 9 107.1 ± 14.9 81.3–128.9 5.5 years 9 277.7 ± 12.9 258.0–295.9 9 82.7 ± 1.5 79.1–84.5 9 198.8 ± 6.4 189.0–205.8 9 122.8 ± 8.5 104.8–134.6 6.5 years 7 272.6 ± 12.8 255.9–289.4 7 79.2 ± 2.9 76.4–84.9 7 193.6 ± 13.1 178.5–219.7 7 128.6 ± 23.7 99.8–168.3 7.5 years 1 273.0 273.0 0 0 1 125.7 125.7 8.5 years 1 270.9 270.9 1 82.3 82.3 1 206.4 206.4 1 109.8 109.8 9.5 years 1 274.0 274.0 1 80.0 80.0 1 207.8 207.8 0 ≥ 1.5 years 142 259.0 ± 16.3 218.0–295.9 135 79.7 ± 3.1 68.9–86.6 137 187.7 ± 10.2 162.0–219.7 121 93.6 ± 21.0 53.3–168.3 6 m o r p h o l o g y o f n o r t h d a k o t a m o o s e – je n s e n e t a l . a l c e s v o l . 4 9 , 2 0 1 3 sexual dimorphism moose ≥4.5 years old had a sexual dimorphism ratio of 1.04 for ww (x = 485.8 kg for males [n = 15] and 452.1 kg for females [n = 15]), and 1.07 for fdw (x = 357.2 kg for males [n = 15] and 321.7 kg for females [n = 17]; table 8). these ratios were mid-range in comparison with other studies where weight dimorphism for ww and fdw ranged from 0.90–1.19 and 0.70–1.36, respectively (table 9). although measurements for male moose were larger for all but table 5. simple regression equations for weight-measurement relationships among moose from north dakota, usa (1978–1990). comparison n equation r2 whole wt. (ww) (kg) vs. field dressed wt.(fdw) (kg) 154 ww = (fdw×1.28) + 35.5 0.93 whole wt. (ww) (kg) vs. carcass wt. (cw) (kg) 39 ww = (cw×1.47) + 67.3 0.75 whole wt. (ww) (kg) vs. viscera wt. (vw) (kg) 156 ww = (vw×2.41) + 119.8 0.65 whole wt. (ww) (kg) vs. total body length (tbl) (cm) 153 ww = (tbl×0.36) − 520.2 0.76 whole wt. (ww) (kg) vs. shoulder height (sh) (cm) 158 ww = (sh×0.24) − 49.1 0.33 whole wt. (ww) (kg) vs. hind-foot length (hfl) (cm) 163 ww = (hfl×1.51) − 802.8 0.46 field-dressed wt. (fdw) (kg) vs. carcass wt. (cw) (kg) 39 fdw = (cw×0.84) + 104.4 0.65 field-dressed wt. (fdw) (kg) vs. viscera wt. (vw) (kg) 146 fdw = (vw×1.41) + 119.7 0.40 field-dressed wt. (fdw) (kg) vs. total body length (tbl) (cm) 181 fdw = (tbl×0.26) − 387.8 0.70 field-dressed wt. (fdw) (kg) vs. shoulder height (sh) (cm) 165 fdw = (sh×0.17) − 41.0 0.31 field-dressed wt. (fdw) (kg) vs. hind-foot length (hfl) (cm) 174 fdw = (hfl×0.11) + 196.4 0.03 carcass wt. (cw) (kg) vs. viscera wt. (vw) (kg) 32 cw = (vw×1.14) + 93.6 0.56 carcass wt. (cw) (kg) vs. total body length (tbl) (cm) 47 cw = (tbl×0.18) − 236.3 0.33 carcass wt. (cw) (kg) vs. shoulder height (sh) (cm) 40 cw = (sh×0.24) + 231.0 0.37 carcass wt. (cw) (kg) vs. hind-foot length (hfl) (cm) 46 cw = (hfl×0.013) + 207.8 0.001 viscera wt. (vw) (kg) vs. total body length (tbl) (cm) 156 vw = (tbl×0.09) + 0.011 0.40 viscera wt. (vw) (kg) vs. shoulder height (sh) (cm) 147 vw = (sh×0.05) + 22.7 0.14 viscera wt. (vw) (kg) vs. hind-foot length (hfl) (cm) 156 vw = (hfl×0.36) − 173.3 0.22 antler width (aw) (cm) vs. whole wt. (ww) (kg) 97 aw = (ww×2.57) − 137.2 0.65 antler width (aw) (cm) vs. field-dressed wt. (ww) (kg) 108 aw = (fdw×3.46) − 102.0 0.72 antler width (aw) (cm) vs. total body length (tbl) (cm) 129 aw = (tbl×0.81) − 1183.0 0.37 antler width (aw) (cm) vs. shoulder height (sh) (cm) 116 aw = (sh×0.39) + 198.8 0.13 antler width (aw) (cm) vs. hind-foot length (hfl) (cm) 125 aw = (hfl×.024) + 901.0 < 0.001 table 6. regression equations for age-measurement relationships among moose from north dakota, usa (1978–1990). comparison n equations r2 age vs. whole weight (ww) (kg) 114 age = ((ww×0.0043) – .066)2 0.71 age vs. field dressed wt. (fdw) (kg) 166 age = ((fdw×0.06) + 0.022)2 0.70 age vs. carcass wt. (cw) (kg) 39 age = ((cw×0.007) + 0.098)2 0.53 age vs. viscera wt. (vw) (kg) 145 age = ((vw×0.01) + 0.51)2 0.39 age vs. total body length (tbl) (cm) 203 age = ((tbl×0.0014) – 0.21)2 0.53 age vs. shoulder height (sh) (cm) 197 age = ((sh×0.0011) – 0.47)2 0.26 age vs. hind-foot length (hfl) (cm) 153 age = ((hfl×0.00035) + 1.31)2 0.014 age vs. antler width (aw) (cm) 119 age = ((aw×0.0016) + 0.26)2 0.70 alces vol. 49, 2013 jensen et al. – morphology of north dakota moose 7 table 7. comparisons of mean field dressed weight (fdw) of calves and percent change from calf (0.5 years) to yearling (1.5 years) age classes from 7 north american and 4 northern european moose populations. note: broadfoot et al. (1996) used moose of 11 months of age. females males location subspecies calf fdw (kg) [n] yearling fdw (kg) [n] change calf fdw (kg) [n] yearling fdw (kg) [n] change source north dakota a. a. andersoni 119.1 [4] 241.1 [10] 102.4% 141.7 [4] 225.7 [32] 59.3% this study vermont, new hampshire, and maine a. a. andersoni 108 [23] 216 [65] 100 % 112 [23] 199 [139] 77.7% adams and pekins (1995) quebec a. a. americana 108.6 [26] 192.6 [11] 77.3% 119.6 [19] 199.3 [24] 66.6% heyland (1964) in peterson (1974) quebec 108.6 [26] 203.4 [34] 87.3% 119.6 [19] 204.3 [51] 70.8% heyland (1966) in peterson (1974) ontario a. a. andersoni 156.5 [3] 230.9 [7] 47.5% 140.2 [7] 254.5 [19] 81.2% timmerman (1972) ontario a. a. andersoni 75.3 [1] 220.9 [2] 193.4% 115.2 [1] 240.4 [1] 108.7% simkin (1962) ontario a. a. andersoni 136.3 [3] 228.9 [5] broadfoot et al. (1996) manitoba a. a. andersoni 163.3 [1] 205.8 [4] crichton (1980), and pers. comm. alberta a. a. andersoni 93.9 [27] 161.9 [28] 72.4% 95.3 [21] 152.9 [34] 60.4% blood et al. (1967) montana a. a. shirasi 84.4 [14] 163.3 [15] 93.5% 99.3 [14] 170.6 [28] 71.8% schladweiler and stevens (1972) norway, southern a. a. alces 69.8 [74] 140.2 [115] 100.9% saether et al. (1996) norway, interior a. a. alces 68.5 [298] 139.8 [210] 104.1% saether et al. (1996) norway, alpine a. a. alces 63.2 [625] 125.1 [370] 97.9% saether et al. (1996) norway, northern a. a. alces 72.9 [7] 146.0 [122] 100.3% saether et al. (1996) 8 m o r p h o l o g y o f n o r t h d a k o t a m o o s e – je n s e n e t a l . a l c e s v o l . 4 9 , 2 0 1 3 vw and hfl, only fdw and cw differed significantly between sexes (table 8). discussion general observations field dressed weights were the best estimator of ww, although reliable estimates of ww were also obtained using tbl and sh (table 5). for all sexes the best estimate of age was ww followed by fdw, and aw was also a good estimator of male age (table 6). therefore, we suggest that collection of baseline data in local populations focus on fdw, tbl, aw, and incisor collection. although age can be reasonably predicted by aw for males, and ww and fdw for moose of both sexes, cementum annuli aging remains the most accurate method for determining age. further, cw and tbl appear to be fairly good predictors of fdw, while other measurements have limited use in estimating fdw (table 5). carcass weight and vw were poorly predicted by morphometric measurements (table 5). our results also indicate that ww of both sexes can range about ±20% of the mean within an age class. schwartz et al. (1987) reported overwinter weight loss can range from as little as 7% to as high as 23% of pre-rut body mass; thus, weights of individual moose in fallearly winter that are >20% below the local population average of an age class may indicate nutritional stress or other health concerns such as parasite infection. several authors have estimated the relationship between fdw and ww with varying results. in our study the slope of the regression between fdw and ww was 1.28 for both sexes combined (table 5) which was similar to that in studies with a combined relationship for both sexes. for example, peterson (1974) and crichton (1980) calculated slopes of 1.28 and 1.31 for both sexes. other authors have reported varying results for the relationship between ww and fdw. blood et al. (1967) estimated an overall carcass yield of 50% (1.50) for all ages, and similarly, schladweiler and stevens (1973) estimated ww:fdw ratios of 1.43 for adult females and 1.33 for adult males, and broadfoot et al. (1996) estimated 1.51 for females and 1.48 for males for 11-month old captive moose. other researchers have developed estimators of ww based on a variety of morphometric measurements. franzmann et al. (1978) developed equations for estimating ww from tbl, chest girth, hfl, and sh, and wallin et al. (1996) focused on chest circumference to estimate carcass body mass of moose in sweden after removal of head, skin, lower legs, kidneys, and viscera. because the methods and outcome of analyses have varied substantially among studies, we suggest that when comparing fdw and ww between populations, eliminate as many biases as possible (e.g., sex, age, and table 8. sexual dimorphism for 7 morphometric measurements and results of t-tests comparing these measurements between sexes for moose ≥4.5 years old from north dakota, usa (1978–1990). measurement sexual dimorphism t degrees of freedom p whole weight (ww) 1.04 1.3 34 0.21 field-dressed weight (fdw) 1.07 2.2 38 0.03 carcass weight (cw) 1.27 2.4 8 0.04 viscera weight (vw) 0.93 1.1 28 0.27 total body length (tbl) 1.02 1.5 40 0.13 hind-foot length (hfl) 1.0 0.11 29 0.91 shoulder height (sh) 1.16 1.5 28 0.14 alces vol. 49, 2013 jensen et al. – morphology of north dakota moose 9 table 9. comparisons of sexual dimorphism based on mean field dressed weight (fdw) and mean whole weight (ww) for adult females and males (≥4.5 years) from various moose populations in north america. * note: some references did not permit calculating weight using these age criteria. therefore, weights for quinn and aho (1989) were for mature animals aged as wear class vi (>6.5 years), schladweiler and stevens (1973) were for mature animals aged as wear class v (>5.5 years), blood et al. (1967) and franzmann et al. (1978) includes animals >3.5 years, geist (1998) included animals >4.5 years, and franzmann et al. (1987) were 5 year-old captive animals. females males females males location subspecies fdw (kg) [n] fdw (kg) [n] fdw sexual dimorphism ww (kg) [n] ww (kg) [n] ww sexual dimorphism source north dakota a. a. andersoni 321.7 [17] 343.2 [24] 1.07 452.1 [15] 471.3 [23] 1.04 this study quebec a. a. americana 267.3 [156] 348.0 [224] 1.30 heyland (1964, 1966) in peterson (1974) vermont, new hampshire, and maine a. a. americana 261.9 [76] 342.8 [251] 1.31 adams and pekins (1995) ontario a. a. andersoni 301 [5] 408 [10] 1.36 timmerman (1972) ontario a. a. andersoni 285.0 [3] 360.0 [2] 1.26 simkin (1962) ontario a. a. andersoni 461.0 [17] 496.0 [9] 1.07* quinn and aho (1989) manitoba a. a. andersoni 281.2 [7] 308.4 [13] 1.10 400.7 [3] 461.7 [8] 1.15 crichton (1980) alberta a. a. andersoni 205.0 [55] 220.0 [39] 1.07* 417.0 [6] 413.2 [3] 0.99* blood et al. (1967) alberta a. a. andersoni 413.0 [32] 456.0 [30] 1.10* geist (1998) alberta a. a. andersoni 412.4 [32] 461.1 [23] 1.12 lynch et al. (1995) montana a. a. shirasi 214.3 [29] 269.0 [35] 1.25* schaldweiler and stevens (1973) alaska a. a. gigas 400.5 [66] 454.6 [5] 1.14* franzmann et al. (1978) alaska a. a. gigas 499.0 [3] 594.0 [3] 1.19* franzmann et al. (1978) 1 0 m o r p h o l o g y o f n o r t h d a k o t a m o o s e – je n s e n e t a l . a l c e s v o l . 4 9 , 2 0 1 3 conversions to ww from fdw and morphometric measurements), and make only direct comparisons such as fdw and ww. growth rates and weights of north dakota moose our estimates of phase 1 growth rates were similar to that of schwartz (1998) who estimated a phase 1 accelerated growth of 1.3–1.6% per day for calves <165 days old (pre-rut). our rates were higher than those reported by addison et al. (1994) for captive calves with neonatal weights of 15.7 kg (n = 8) and 17.3 kg (n = 10) for females and males, respectively. at 187 days, or approximately 6 months old, their females averaged 149.2 kg (n = 6) and males 165.8 kg (n = 9), representing a 950 and 958% increase in mass, respectively. in short, during phase 1, north american moose display an impressive rate of growth of *1000% within a 6 month period. during growth phase 2, our female calves appeared to grow faster than male calves (tables 1 and 2), particularly when comparing fdw. this was also true for 5 of the 6 north american populations where similar comparisons could be made (table 7), suggesting that a larger proportion of body mass is being devoted to skeletal and muscle development at an earlier age in males. this additional skeletal and muscle mass may provide a selective advantage for male calves as they are often driven from their mother by courting males during the rut and must briefly survive alone. during the third (self-inhibiting) growth phase when body mass goes through seasonal peaks and troughs, the maximum weights for females and males occur midwinter and prerut, respectively (schwartz et al. 1987). body weight of north american female moose appears to plateau at 3–4 years (geist 1998), whereas weight of male moose plateaus at 5 years (peterson 1974). we found that both male and female weights peaked at 5.5 years (tables 1 and 2). while sample sizes were small, tbl also appeared to plateau at 5.5 years in north dakota (tables 3 and 4), which was similar as reported for alaskan moose (franzmann et al. 1978). comparative data are lacking, but these results suggest that north dakota moose with access to highly nutritious agricultural crops may maximize body weight earlier and continue to grow structurally longer than other moose populations. additional observations about moose growth rates in 6 of the 7 populations in which comparisons could be made, female calves weighed ≥5% less than male calves (table 7). the exception (timmermann 1972) where female calves weighted >10% more than males may be an artifact of small sample size (n = 3 female calves). overall, fdw of north american female and male calves averaged 101.4 kg (n = 69), and 112.0 kg (n = 61), respectively (table 7). estimates of ww from fdw using various correction factors are relatively common in the literature. however, actual ww of north american female and male moose calves in the literature are limited to blood et al. (1967), lynch et al. (1995), and this study. the average ww of female and male calves was 174 kg (n = 4) and 197 kg (n = 5) in alberta (blood et al. 1967), and 171 kg (n = 12) and 197 kg (n = 13) in ontario (lynch et al. 1995). we found that ww for female and male calves were identical, but male fdw appears to be higher. in short, the available data suggests that either maternal moose invest more resources in their male calves, or there is a selective advantage for male calves to develop greater muscle mass at an early age. during growth phase 2, 5 of 8 populations previously described (table 7) had heavier mean fdw for yearling males, but north dakota female yearlings grew at a alces vol. 49, 2013 jensen et al. – morphology of north dakota moose 11 faster rate than most. during this phase, female yearlings are more likely to remain in loose association with their mother during subsequent calving, while male yearlings disperse and are likely at relative disadvantage. overall, fdw of north american female and male yearlings averaged 202.2 kg (n = 155), and 199.1 kg (n = 313), respectively. based upon the fdw versus ww proportions for north dakota moose (tables 1 and 2), the mean ww of north american female and male moose yearlings would average 278 and 285 kg, respectively. during this growth phase, geist (1998) suggested that a solitary moose must be at least 250 kg to confront predators, basing his assumption on the size of adult ussuri or “dwarf” moose of manchuria (a. a. cameloides) (heptner and nasimovitch 1967, p. 72 in geist 1998). whole weights of north dakota moose and data on fdw we provide from other studies (table 9) are partially supportive of his assumption. sexual dimorphism it should be noted that ww and fdw in north dakota are subject to biases that may or may not be obvious. for example, weights of adult males from other studies sampled during pre-rut would likely result in a higher sexual dimorphic ratio because these animals should be in optimal physical condition. the low dimorphic ratio for ww and fdw in north dakota may relate directly to our data collection period during post-rut when males weigh less. higher sexual dimorphism of fdw of 1.3 was reported in vermont, new hampshire, and maine (adams and pekins 1995), quebec (peterson 1974), and ontario (timmermann 1972; table 9). the earliest starting dates for moose hunting seasons in new hampshire, maine, ontario, quebec, and vermont range from august–october (timmermann and buss 1995), whereas, moose in this study and in alberta (26 november-6 january; blood et al. 1967) were harvested post-rut. further, the sample size of ww in alberta was small (n = 6 females and 3 males) and included animals 3.5 years of age (blood et al. 1967). we assume that the value of 1.3 probably represents the pre-rut maximum sexual dimorphism, and ratios <1.1 probably reflect post-rut leaner male weights. other dimorphic ratios we report (tbl, hfl, and sh; table 8) were low and should not vary seasonally. subspecies comparisons sample sizes reported by simkin (1962) are too small (i.e., 1-2 individuals per category) for comparative purposes, but are reported here for completeness. north dakota calf and yearling moose were larger than in all other populations except in ontario (timmermann 1972; table 7). ontario moose were sampled during the first 2 weeks of an october hunting season from forestland long managed for pulpwood with an abundance of preferred and productive habitat and forage (timmermann 1972). comparisons were also made between mean fdw and ww of mature adult moose from north dakota with 9 other north american populations (table 9); however, age criteria used for determining adult or mature moose was not uniform for all populations. for example, moose from alberta (blood et al. 1967) and alaska (franzmann et al. 1978) included animals as young as 3.5 years, and females from north dakota were comparatively shorter in tbl than those from alaska (franzmann et al. 1978). however, our comparisons suggest that female moose from north dakota are markedly heavier than all other populations, with the exception of 5 year-old captive animals in alaska (franzmann et al. 1978) and animals sampled prior to the rutting season in ontario (quinn and aho 1989). 12 morphology of north dakota moose – jensen et al. alces vol. 49, 2013 these data also indicate that moose in north dakota grow as fast or faster during their initial years when compared to other north american and european moose populations (table 7). based upon tbl, moose from north dakota reached their maximum mass by 5.5 years. range expansion of moose from northern forested areas into the agriculturally dominated landscape of north dakota provides a unique opportunity to obtain and evaluate subspecies and regional morphometric variations. during the fall, moose in north dakota are frequently observed foraging on sunflowers and other agricultural crops. although food habits information in north dakota is limited to a single study (maskey 2008), seasonal use of agricultural foods was high, with row crops (primarily corn [zea mays]) composing 11 and 22% of the fall and winter diets (sheridan co., moose hunting unit m9; fig. 1). in the turtle mountains (bottineau co., moose hunting unit m4; fig. 1) where row crops are limited, alfalfa (medicago sativa) composed 13% of the summer and fall diet representing *90% of consumed forbs. this agricultural forage may help maximize growth rate and body mass of moose in north dakota faster than would occur in traditional forest habitat. management implications environmental conditions affect body mass (sand 1996, hjeljord and histol 1999, ericsson et al. 2002), and in turn, body mass influences reproductive potential (saether and haagenrud 1983, adams and pekins 1995). monitoring moose populations is difficult, particularly in forested environments; however, measuring body weight of yearling and adult moose has potential value in predicting the nutritional and reproductive status of moose populations (adams and pekins 1995). growth rates also have predictive value in estimating reproductive potential and habitat conditions. because measuring body weight of moose is often a difficult task, having reliable equations to predict ww from other morphological parameters provides important alternatives for measuring and monitoring moose populations. the predictive equations provided here will be useful to wildlife managers for estimating various weight parameters and age related to productivity, and may aid in law enforcement and immobilization protocols when body weight and age of moose are often necessary. acknowledgements we want to express our sincere thanks to the hunters and landowners for their cooperation during this study. additionally, we would like to thank the following department personnel for their assistance on check stations: s. allen, a. anderson, a. aufforth, b. bitterman, s. brashears, e. dawson, g. enyeart, t. ferderer, t. frank, j. gulke, a. harmoning, l. johnson, m. johnson, m. johnson, m. kanzelman, s. kohn, r. knapp, j. kobriger, a. kreil, g. link, b. lynott, m. mckenna, r. parsons, r. patterson, c. penner, h. pochant, c. pulver, g. rankin, b. renhowe, s. richards, r. rollings, j. samuelson, d. schmidt, j. schulz, b. stotts, r. sohn, l. tripp, and l. vetter. finally, we thank e. addison and 2 anonymous reviewers for providing helpful editorial comments. financial support was provided by the north dakota game and fish department federal aid project w-67-r. references adams, k. p., and p. j. pekins. 1995. growth patterns of new england moose: yearlings as indicators of population status. alces 31: 53–59. addison, e.m., r. f. mclaughlin, and j. d. broadfoot. 1994. growth of moose calves (alces alces americana) infected and uninfected with winter ticks alces vol. 49, 2013 jensen et al. – morphology of north dakota moose 13 (dermacentor albipictus). canadian journal of zoology 72: 1469–1476. blood, d. a., j. r. mcgillis, and a. lovass. 1967. weights and measurements of moose in elk island national park, alberta. canadian field-naturalist 81: 263–269. broadfoot, j. d., d. g. joachim, e. m. addison, and k. s. macdonald. 1996. weights and measurements of selected body parts, organs and long bones of 11-month-old moose. alces 32: 173–184. bubenik, a. b. 1998. evolution, taxonomy and morphology. pages 77–123 in a. w. franzmann and c. c. schwartz, editors. ecology and management of the north american moose. wildlife management institute, washington, d. c., usa. clutton-brock, t. h., f. e. guinness, and s. d. albon. 1982. red deer: behavior and ecology of two sexes. university of chicago press, chicago, illinois, usa. crichton, v. f. 1980. manitoba's second experimental moose hunt on hecla island. proceedings of the north american moose conference and workshop 16: 489–526. ericsson, g., j. b. ball, and k. danell. 2002. body mass of moose calves along an altitudinal gradient. journal of wildlife management 66: 91–97. franzmann, a. w. 1978. moose. pages 67-88 in j. l. schmidt and d. l. gilbert, editors. big game of north america: ecology and management. stackpole books, harrisburg, pennsylvania, usa. ———, w. b. ballard, c. c. schwartz, and t. h. spraker. 1980. physiologic and morphometric measurements in neonate moose and their cows in alaska. proceedings of the north american moose conference and workshop 16: 106–123. ———, r. e. leresche, r. a. rausch, and j. l. oldemeyer. 1978. alaskan moose measurements and weights and measurement-weight relationships. canadian journal of zoology 56: 298–306. gasaway, w. c., d. b. harkness, and r. a. rausch. 1978. accuracy of moose age determinations from incisor cementum layers. journal of wildlife management 42: 558–563. geist, v. 1998. deer of the world: their evolution, behavior, and ecology. stackpole books, mechanicsburg, pennsylvania, usa. haagenrud, h. 1978. layers in secondary dentine of incisors as age criteria in moose (alces alces). journal of mammalogy 59: 857–858. hjeljord, o., and t. histol. 1999. rangebody mass interactions of a northern ungulate a test of hypothesis. oecologia 119: 326–339. knue, j. 1991. big game in north dakota: a short history. north dakota game and fish department, bismarck, north dakota, usa. loudon, a. s. i. 1987. the influence of forest habitat structure on growth, body size, and reproduction in roe deer (capreolus capreolus l.). pages 559– 567 in c. m. wemmer, editor. biology and management of the cervidae. smithsonian institution press, washington d. c., usa. lynch, g. m., b. lajeunesse, j. willman, and e. s. telfer. 1995. moose weights and measurements from elk island national park, canada. alces 31: 199–207. maskey, jr., j. j. 2008. movements, resource selection, and risk analyses for parasitic disease in an expanding moose population in the northern great plains. ph. d. dissertation, university of north dakota, grand forks, north dakota, usa. peterson, r. l. 1955. north american moose. university of toronto press, toronto, canada. ———. 1974. a review of the general life history of moose. le naturaliste canadien 101: 9–21. 14 morphology of north dakota moose – jensen et al. alces vol. 49, 2013 quinn, n. w. s., and r. w. aho. 1989. whole weights of moose from algonquin park, ontario, canada. alces 25: 48–51. saether, b. e., r. anderson, o. hjeljord, and m. heim. 1996. ecological correlates of regional variation in life history of the moose alces alces. ecology 77: 1493–1500. ———, and h. haagenrud. 1983. life history of moose (alces alces): fecundity rates in relation to age and carcass weight. journal of mammalogy 64: 226–232. sand, h. 1996. life history patterns in female moose (alces alces): the relationship between age, body size, fecundity and environmental conditions. oecologia 106: 212–220. sauer, p. r. 1984. physical characteristics. pages 73-91 in l. k. halls, editor. white-tailed deer: ecology and management. stackpole books, harrisburg, pennsylvania, usa. schladweiler, p., and d. r. stevens. 1973. weights of moose in montana. journal of mammalogy 54: 772–775. schwartz, c. c. 1998. reproduction, natality and growth. pages 141–171 in a. w. franzmann and c. c. schwartz, editors. ecology and management of the north american moose. wildlife management institute, washington d.c., usa. ———, w. l. regelin, and a. w. franzmann. 1987. seasonal weight dynamics of moose. swedish wildlife research supplement 1: 301–310. simkin, d. w. 1962. weights of ontario moose. ontario fish and wildlife review 1: 10–12. timmermann, h. r. 1972. some observations of the moose hunt in the black sturgeon area of northwestern ontario. proceedings of the north american moose conference and workshop 8: 223–239. verme, l. j., and m. e. buss. 1995. the status and management of moose in north america early 1990s. alces 31: 1–14. ———, and d. e. ullrey. 1984. physiology and nutrition. pages 91–118 in l. k. halls, editor. white-tailed deer: ecology and management. stackpole books, harrisburg, pennsylvania, usa. wallin, k., g. cederlund, and a. pehrson. 1996. predicting body mass from chest circumference in moose alces alces. wildlife biology 2: 53–58. welch, d. a., m. l. drew, and w. m. samuel. 1985. techniques for rearing moose calves with resulting weight gains and survival. alces 21: 475–491. alces vol. 49, 2013 jensen et al. – morphology of north dakota moose 15 mass, morphology, and growth rates of moose in north dakota methods results morphometric measurements morphometric relationships growth rates and patterns sexual dimorphism discussion general observations growth rates and weights of north dakota moose additional observations about moose growth rates sexual dimorphism subspecies comparisons management implications acknowledgements references alces29_201.pdf alces29_169.pdf alces26_51.pdf pre-parturition movement patterns and birth site characteristics of moose in northeast minnesota amanda m. mcgraw1, juliann terry2, and ron moen1 1natural resources research institute, university of minnesota, 5013 miller trunk highway, duluth, minnesota 55811-1442; 2university of minnesota-duluth, swenson science building 207, 1035 kirby drive, duluth, mn 55812. abstract: habitat used immediately after parturition is important to survival of moose calves, though different habitat types may be functionally similar and thus contribute to the variability in habitat use reported in the literature. neonates are relatively immobile, which restricts movement of the cow-calf pair and makes both vulnerable to predation. the cow also requires adequate access to forage during the period when calf mobility is limited. we used fine-scale movement data to determine linear distance traveled to the birth site as well as habitat use by cow-calf pairs in northeast minnesota. all cows made long distance movements (x = 6 km) to the birth site where they localized in 1.72 ± 0.48 ha (95% kernel polygon) for approximately 7 ± 0.7 days. a mosaic of cover types that reflected availability across the landscape were used by the cow prior to localization at the birth site. birth site areas consisted of one cover type rather than the mosaic used before birth, and varied among cows, though bogs were used most often (40%). the small birth site area and use of bog habitat were likely a consequence of low calf mobility post-parturition. upon exiting the birth site, cow-calf pairs shifted toward use of mixed and young/regenerating forest which likely reflects the need and use for highly nutritious browse to meet the high energetic cost of lactation. alces vol. 50: 93–103 (2014) key words: alces alces, calving sites, minnesota, moose, parturition habitat. the time around parturition is critical to survival of offspring. the mother should select habitats that increase the survival of offspring and express behavior that reduces exposure of her and offspring to higher mortality risk. for species such as moose (alces alces), choices may be further restricted because the calf has limited mobility during the first weeks of life (altmann 1958, 1963). most moose give birth during a 19-day period in the month of may (sigouin et al. 1997). searches for calving sites typically take place after peak calving. opportunistic ground searches (addison et al. 1990, wilton and garner 1991), ground searches using vhf telemetry (bowyer et al. 1999, langley and pletscher 1994, leptich and gilbert 1986, scarpitti et al. 2007), and searches from aircraft for vhf collared cow-calf pairs (bailey and bangs 1980, mcgraw et al. 2011) have all been used to locate calving sites. cows will make a longer distance move followed by localization at calving, indicating that monitoring of daily locations can help identify when calving occurs (testa et al. 2000). for logistical reasons, most descriptions of pre-parturition movement patterns and birth site characterization have relied on relatively few locations in each parturition event. single daily locations are typically obtained in vhf telemetry studies to determine if a cow has localized or given birth (testa et al. 2000). ground searches for maternal beds accurately describe birth locations, but do not fully delineate the entire area used around the birth location during the postparturition period when cow-calf movements 93 are limited. aerial and ground vhf searches more accurately describe portions of the post-parturition area used by calf-cow pairs than the actual birth site. the limitations of single or few observations of cow-calf pairs during parturition may contribute to the high variation in vegetative cover and vegetation density, visibility, and proximity to water that has made describing generalized calving site characteristics difficult (addison et al. 1990, poole et al. 2007). variability of calving habitats across regions is probably also a function of available habitat types within a study area. moose are reported to birth on hill tops in quebec and ontario (addison et al. 1990, wilton and garner 1991, chekchak et al. 1998), and some swim to islands where available (addison et al. 1993). undisturbed lowland areas dominated by cedar and near water were important for calving in maine (leptich and gilbert 1986), and in new hampshire moose used mature, mixed, and coniferous forests, perhaps because open water and islands were rare (scarpitti et al. 2007). post-parturition areas had a higher than expected bog component in minnesota, though results were variable among cows (mcgraw et al. 2011). some cow moose select birth sites that provide hiding cover but do not necessarily have the highest quality or quantity of forage available (leptich and gilbert 1986, langley and pletscher 1994, bowyer et al. 1999). this is often interpreted as a trade-off between avoiding predators and meeting nutritional requirements (bowyer et al. 1999), and may be important to consider as a factor influencing calving site selection in minnesota where black bear (ursus americanus) and wolves (canis lupus) occur. some cows in british columbia calved in areas with lower forage availability that also had lower predation risk, while others calved in areas with higher forage availability and presumably higher predation risk (poole et al. 2007). understanding habitat and space use behavior during parturition may be especially important in northeast minnesota because the moose population is declining and recruitment rates in recent years are the lowest on record (delgiudice 2013). characteristics of the birth site and post-parturition areas have not been studied in detail in minnesota, in large part because post-parturition locations have been obtained from vhf telemetry flights that occur after peak calv‐ ing, and provide only one location within 2–4 weeks of calving (mcgraw et al. 2011). gps-collared moose can be used to locate calving sites by identifying a longer movement followed by localization (poole et al. 2007). our objective was to identify movement patterns indicative of calving by using location data recorded at 20 min intervals from gps-collared moose. we then evaluated fine-scale movements and habitat use of the cow while the calf had limited mobility. additional objectives were to characterize size, cover type composition, and length of time spent at the birth site by cow-calf pairs. the fine-scale gps locations allowed us to re-examine our previous data from vhf-collared moose (mcgraw et al. 2011) and compare past results with a more precise and robust dataset that can more accurately describe habitat use and movement patterns of moose during parturition. study area both the vhfand gps-collared moose studies occurred in approximately the same 3,700 km2 area of northeast minnesota (47°30′n, 91°20′w; fig. 1). land ownership is mostly public (∼82%) and includes portions of the superior national forest as well as state, county, and tribal lands. a significant portion of private ownership exists as blocks of industrial forest land. 94 moose parturition patterns – mcgraw et al. alces vol. 50, 2014 a boreal forest mix is the matrix from which moose in northeast minnesota choose a calving location. the region is part of the northern superior uplands (minnesota department of natural resources [mndnr] 2010) and is transitional from northern hardwoods in the south to canadian boreal forests in the north (pastor and mladenoff 1992). important habitat types in the home ranges of moose are young mixed conifer and deciduous forests, including aspen (populus tremuloides), paper birch (betula papyrifera), and balsam fir (abies balsamea). early successional forests (11–30 years post-disturbance) are used because forage is within reach of moose (kelsall et al. 1977). summer ranges consist largely of black spruce (picea mariana) lowlands as well as uplands and cut over areas dominated by paper birch, aspen, and balsam fir (peek et al. 1976). in early summer, moose generally use upland, lowland, and plantation areas in proportion to their occurrence (peek et al. 1976). northeast minnesota has a continental climate with severe winters and warm summers. precipitation usually occurs as snow from december–march. methods adult cow moose were darted from helicopters and fitted with gps collars (lotek fig. 1. study area in northeast minnesota where black dots indicate birth sites (n = 20) of gps collared moose, 2012. alces vol. 50, 2014 mcgraw et al. – moose parturition patterns 95 wireless, inc., newmarket, ontario, canada) in january and february 2011 (mccann et al. 2014). blood samples were taken and blood progesterone levels were used as an indication of pregnancy at capture. gps collars recorded locations every 20 min for 2 years. locations, movement, and habitat use of collared females were analyzed for patterns indicative of parturition beginning 1 may as previously defined in a vhf study (lenarz et al. 2011, mcgraw et al. 2011). the area occupied by the cow following its initial localization in may was considered the birth site because cows birth shortly after the initial localization (testa et al. 2000, poole et al. 2007). some cows remained in the immediate vicinity of the birth site, and others moved a short distance from the birth site and localized again (i.e., a secondary localization event). the area and time spent at both sites, as well as the total postparturition area were calculated. pre-parturition movement patterns we monitored pre-parturition movements using 20-min gps location data for each cow (n = 52) during the month of may. the short time scale between gps locations allowed us to calculate the distance moved each day throughout parturition. the length and duration of the movement was recorded from the last location in a cluster of foraging paths and bed sites to the initial localization at the birth site following the linear path. straight-line distances from the last feeding bout to the birth site were also calculated for comparison with past vhf studies. initial localizations at birth sites were identified by cows occupying smaller areas for longer durations and with less variation in location than foraging or bedding locations. in 2012 births were verified using helicopter searches to visually observe cow-calf pairs at the birth site shortly after parturition. vhf flight simulations we used calving dates and locations from gps-collared moose to better describe the vhf data set used previously to des‐ cribe post-parturition habitat in minnesota (mcgraw et al. 2011). we simulated postparturition locations using gps data to estimate the distance that cow-calf pairs in the vhf data set were from the birth site based on calving dates from gps-collared moose in 2012. this was necessary because calves located in the vhf data set could have been up to 4 weeks old. we also compared habitat characteristics at birth sites of gpscollared moose to the habitat characteristics of the simulated vhf data set, and to habitat characteristics of the actual vhf data set. all 12 flight dates from the 2004–2008 vhf study (mcgraw et al. 2011) were randomly assigned to each cow in the current study to simulate the location of the cow at 1200 hr using excel 2010 (microsoft corporation, redmond, washington, usa). the random assignment of flight dates was repeated to estimate the distance from the actual birth site to the simulated vhf post-parturition location. we then calculated the straight line distances from the gps birth site to the simulated vhf post-parturition site. birth site and post-parturition area birth sites (n = 20) were identified by viewing location data in googleearth (google inc. 2013 (version 7.1.1.1888, mountain view, california, usa) to locate clusters of points following long movements in may. dates and times of entry into and exit from the birth sites were identified visually. entrance into the birth site was defined as the first point in a cluster. the cow was still considered to be in the birth site when making short movements and re-localizing. the cow left the birth site when she moved and did not localize again or made a large movement from the last cluster of points at the 96 moose parturition patterns – mcgraw et al. alces vol. 50, 2014 birth site. birth site areas were calculated as 50% and 95% kernel polygons using all locations occurring between birth site entry and exit dates (fig. 2). we calculated 50% and 95% kernel polygons in the geospatial modelling environment (beyer 2012) using the plug-in estimator and a 10 m output cell size. we used the land use land cover (lulc) habitat classification system to determine cover type composition of birth sites (mcgraw et al. 2011). the lulc raster data set was derived from landsat thematic mapper (tm) images at a 30 m resolution (mndnr 2007). source imagery dates ranged from june 1995 to june 1996. the lulc classification system defined 16 cover types in northeast minnesota with >95% accuracy. more than 90% of the study area consisted of 6 terrestrial cover types: mixed, coniferous and deciduous forests, wet bog, marshes and fens, and regenerating forests (moen et al. 2011). we calculated cover type composition within 50% and 95% kernel birth site polygons using arcmap 10.1 (esri, redlands, california). cover type composition in these polygons was compared to cover type outputs of post-parturition areas identified during the vhf study (mcgraw et al. 2011) with anova. we used statistix (version 9.0; analytical software, boca raton, florida) and excel 2010. significance level for all tests was set at p = 0.05. unless otherwise noted, means are presented throughout as x � se. results all gps-collared cows with high progesterone levels (4.98 ± 0.3 ng/ml) at capture showed localization behavior indicative of calving in may. most cows that localized (46 of 52) made a long distance movement of 6 km ± 0.8 km (range = 1–33 km; fig. 2) over 17 ± 1.7 h (range = 2–57 h) before stopping at the birth site (fig. 2). linear path distances calculated using 20 min gps locations were twice as long as straight line distances from the beginning of the long distance movement to the birth site. the straight-line distance from the start of the long-distance movement to the calving site was 3 ± 0.5 km (range = 0.3–23 km). cows with low progesterone levels (0.33 ± 0.2 ng/ml) were not pregnant. all cows with low progesterone levels did not make a long distance movement and did not localize. cows remained localized at the birth site for 4 ± 0.4 days (range = 1–15 days). of the 52 cows that initially localized, 32 moved 133 ± 17.9 m (mode = 80 m; range = 30– 460 m) before localizing again for an additional 3 ± 0.6 days (range = 0–16 days). the remaining 20 cows did not move away from the site where they first localized. in total, cow-calf pairs spent 7 ± 0.7 days (range = 1–18 days) in the post-parturition area. the 50% kernel areas for 20 cows were 0.42 ± 0.06 ha and the 95% kernels were 1.72 ± 0.48 ha in the post-parturition period (fig. 3). the cover type trends indicating selection by cows were consistent, though not statistically different from post-parturition areas calculated in the previous vhf study (fig. 4). the proportion of bog cover type continued to increase as polygon size around birth sites fig. 2. distance moved by cows (n = 52) before localization at parturition. distances were measured as a linear path from locations collected every 20 min by gpscollared moose. alces vol. 50, 2014 mcgraw et al. – moose parturition patterns 97 decreased (34 ± 11% of 50% kernel polygons), while the amounts of mixed forest (24 ± 9% of 50% kernel polygons) and regenerating young forests (6 ± 5% of 50% kernel polygons) declined. primary and secondary localization sites for each cow, defined using 50% kernel polygons, were within a single cover type (table 1). bog habitats (35%) were used more than other available cover types by gps-collared cows, followed by mixed forest (25%) and conifer (20%). composition of cover types used 5 days prior to the long distance movement to birth site was variable among cows, but included a higher diversity of cover types than after localizing at the birth site (fig. 5). the trend in cover type use indicated a general movement away from variable use of mixed forests and young and regenerating stands to use of one cover type, more often bogs. 95% kernels50% kernelsgps locations cow 1 cow 0 100 200 400 meters 2 fig. 3. example of birth site areas as defined by 50% (black lines) and 95% (gray lines) kernel polygons for 2 cows. each point indicates the location of each cow at the birth site at 20 min intervals. fig. 4. change in cover type composition as the area surrounding the known cow/calf locations (ppa) and random locations were incrementally reduced from 100 to 5 ha (mcgraw et al. 2011), compared to composition of kernel birth site areas as defined using gps collar location data collected at 20 min intervals throughout the calving period. 98 moose parturition patterns – mcgraw et al. alces vol. 50, 2014 the average calving date during the gps study was 14 may (mode = 17 may; range = 3–27 may), with 70% of births occurring between 9 and 20 may. post-parturition locations from the vhf study were obtained by observing cow-calf pairs from helicopters after peak calving, between 21 may and 5 june each year (mcgraw et al. 2011). simulation based on calving dates of gpscollared cows indicated that mean calf age was 12 days (95% ci: 9.9–13.4 days) when cow-calf pairs were located during the vhf study. cow-calf locations for gps-collared cows corresponding to the simulated locations from the vhf study were 2 km (95% ci: 0.2–3.8 km) from actual birth sites. discussion in the current study in which actual birth sites were identified (n = 20), the bog cover type was used in higher proportion than its availability. however, there was still considerable variability in cover type composition among birth sites, with mixed forest used less than expected if birth site selection was random, whereas coniferous and deciduous forests were used about in proportion to availability. in the 5 days before long distance movement to birth sites, cows used a wide variety of cover types. upon movement to the birth site, cows tended to move to a specific cover type in which they localized to give birth and remained for 7 days. these data are consistent with past studies (addison et al. 1990, langley and pletscher 1994, chekchak et al. 1998, bowyer et al. 1999, scarpitti et al. 2007) demonstrating considerable variation in habitat types used table 1. proportion of birth sites (n = 20) in each cover type, based on the area inside 50% kernel birth site polygons. cover type birth site (%) home range (95% fixed kernel) northeast minnesota (%) deciduous forest 10 5 9 mixed forest 25 39 40 bogs 40 21 13 coniferous forest 20 17 23 regenerating 5 16 7 other 0 2 8 fig. 5. proportion of cover types used during periods beginning 5 days prior to the long distance movement (pre-ldm), during the long distance movement (ldm), localization at the birth site (birth), and 5 days post localization at the birth site (post-birth). alces vol. 50, 2014 mcgraw et al. – moose parturition patterns 99 as calving sites. this implies that many habitat types are functionally similar in terms of what aspects are required to successfully rear calves. as a result, specific protective measures of calving sites would be difficult to implement given the wide variety of suit‐ able habitat available. while calves are relatively immobile after birth, bog cover types were used by nearly half of cows for birth sites in minnesota. bogs likely provide hiding cover for calves as well as some foraging opportunities and access to water for cows while their movements are restricted by the calf. as the cow-calf pairs moved away from the birth site, variability of cover type composition increased, with more time spent in mixed forests as well as young and regenerating forests. movement to foraging habitat shortly after spring green up, when browse species have the highest nutritional content, likely allows cows to meet the energetic demands of lactation. despite the lag time between parturition and when cow-calf pairs were located during vhf telemetry flights, cover type composition was consistent with those observed with gps data. while the cow-calf pair moves away from the birth site after about one week, some cover types may remain important and sought out 3–4 weeks postparturition. as the post-parturition area surrounding vhf telemetry locations of cows with 12 day old calves was decreased, the proportion of bog cover type increased and mixed forest types decreased (mcgraw et al. 2011). straight-line distances for cows were half as long (3 km) as the movement lengths measured by following the linear path (6 km). using 20 min gps location data allowed us to refine movement calculations, and as a result, we identified that a greater proportion of moose in this study (88%) made long distance pre-parturition movements than reported in the literature (20%; bowyer et al. 1999). previous studies have measured straight-line distances from 2 discrete observations. while long distance movements observed in northeast minnesota are similar in length to those reported in central alaska (7.3 ± 2.3 km; bowyer et al. 1999), they are twice the length of those reported in south central alaska (4 km; testa et al. 2000). these differences are likely due to the method of measurement. in this study, duration and length of the pre-parturition movement were calculated by measuring the linear path of the cow with 20 min location data, whereas past vhf studies located cows once (testa et al. 2000) or twice daily (bowyer et al. 1999) during the calving period. more frequent locations will influence the measured distance moved per day during the long distance pre-parturition movement. calves are most vulnerable during the first weeks of life and affect cow movement until the calf is more mobile. cows remain in visual or vocal distance during the initial postparturition days when calves are less mobile (cederlund et al. 1987, van ballenberghe and ballard 2007). the duration of time spent in the birth area was ∼7 days which is consistent with the amount of time female moose were protective of birth sites at a research facility in russia (bogomolova and kurochkin 2002); however, it is shorter than the 3–4 weeks reported elsewhere (altmann 1963, bowyer 1999). this could be a result of the limitations of vhf technologies or reflect different definitions of the birth site area, ranging from the point of parturition to the larger area of restricted cow-calf movement in the weeks following birth. the size of the birth site area used during the first post-partum week tended to be small and located in a single cover type; however, this is partly a function of the coverage resolution (30 × 30 m cell size). variability among cows in terms of cover type selection for birth sites could also be an anti-predator 100 moose parturition patterns – mcgraw et al. alces vol. 50, 2014 strategy (bowyer et al. 1999). black bears (ursus americanus) and wolves (canis lupus) occur throughout moose habitat in northeast minnesota and their possible effect on birth site selection strategy should be considered, as restriction of the birth site area could be a function of predator avoidance. a cow remaining localized in a small area while its calf is relatively immobile reduces the area in which a predator may encounter the pair (bowyer et al. 1999). eventual movement from the birth site after ~7 days may be a function of depleted forage and continued predator avoidance. though the birth site area may be small, it stands to reason that the longer the pair remains at the birth site, the more likely they are to encounter a predator. cow-calf pairs that moved short distances within a few days post-partum remained in the same cover type when they localized again. these short movements could also reflect low availability of forage or disturbance (bowyer et al. 1999). the temporal scale at which we were able to monitor gps-collared cows is much finer than was previously possible, enabling us to more accurately define the spatial extent of birth site areas. it is possible that past observations using ground searches and/or telemetry overlooked small movements to secondary birth sites, resulting in calculation of smaller birth site and post-parturition areas. aerial birth site searches should be interpreted with caution if they are flown only once or twice a year after peak calving. simulated locations of cows that represented the 2004–2008 vhf flights occurred 12 days after localization of cows at the birth site, and after most cow-calf pairs would have left the area (mcgraw et al. 2011). the identification of birth site characteristics from the simulated vhf flights would have been 2 km from the actual post-parturition habitat when the calf is more mobile. the use of gps collars to collect locations every 20 minutes vastly improved our ability to identify pre-parturition movement patterns and allowed us to more accurately define the spatial extent of calving areas. future research using gps technology to record fine scale movement patterns is needed to determine when cows with calves resume normal activity levels after parturition. defining and identifying this will lead to more accurate description of post-parturition habitat use and cow-calf movements. acknowledgements funding for this work was provided by the environment and natural resources trust fund of minnesota, the university of minnesota duluth, and the natural resources research institute. partial summer support for a. mcgraw and j. terry was provided by the integrated biosciences graduate program, university of minnesota duluth. this is contribution number 566 from the center for water and the environment at the natural resources research institute, university of minnesota duluth. references addison, e. m., r. f. mclaughlin, d. j. h. fraser, and m. e. buss. 1993. observations of preand post-partum behaviour of moose in central ontario. alces 29: 27–33. ———, j. d. smith, r. f. mclaughlin, d. j. h. fraser, and d. g. joachim. 1990. calving sites of moose in central ontario. alces 26: 142–153. altmann, m. 1958. social integrations of the moose calf. animal behavior 6: 155– 159. ———. 1963. naturalistic studies of ma‐ ternal care in moose and elk. pages 233–253 in h. l. rheingold, editor. maternal behavior in mammals. wiley & sons, new york, new york, usa. bailey, t. n., and e. e. bangs. 1980. moose calving areas and use on the kenai alces vol. 50, 2014 mcgraw et al. – moose parturition patterns 101 national moose range, alaska. proceedings of the north american moose conference and workshop 16: 289–313. beyer, h. l. 2012. geospatial modelling environment (version 0.7.2.1). (software). url: <http://www.spatialecology. com/gme>. bogomolova, e. m., and y. a. kurochkin. 2002. parturition activity of moose. alces supplement 2: 27–31. bowyer, r. t., v. van ballenberghe, j. g. kie, and j. a. k. maier. 1999. birth-site selection by alaskan moose: maternal strategies for coping with a risky environment. journal of mammalogy 80: 1070–1083. cederlund, g., f. sandegren, and k. larsson. 1987. summer movements of female moose and dispersal of their offspring. journal of wildlife management 51: 342–352. chekchak, t., r. courtois, j-p. ouellet, l. breton, and s. st-onge. 1998. characteristics of moose (alces alces) calving sites. canadian journal of zoology 76: 1663–1670. delgiudice, g. d. 2013. 2013 aerial moose survey. minnesota department of natural resources, st. paul, minnesota, usa. <http://files.dnr.state.mn.us/recrea tion/hunting/moose/moose_survey_2011. pdf> (accessed march 2013). kelsall, j. p., e. s. telfer, and t. d. wright. 1977. the effects of fire on the ecology of the boreal forest, with particular reference to the canadian north: a review and selected bibliography. occasional paper #323. canadian wildlife service, ottawa, canada. langley, m. a., and d. h. pletscher. 1994. calving areas of moose in northwest‐ ern montana and southeastern british columbia. alces 30: 127–135. leptich, d. j., and j. r. gilbert. 1986. characteristics of moose calving sites in northern maine as determined by multivariate analysis: a preliminary investigation. alces 22: 69–81. lenarz, m. s. 2011. 2011 aerial moose survey. minnesota department of natural resources, st. paul, minnesota, usa. <http://files.dnr.state.mn.us/recrea tion/hunting/moose/moose_survey_2011. pdf> (accessed march 2011). mccann, n. p., r. a. moen, and s. k. windels. 2014. identifying thermal refugia for a cold-adapted mammal facing climate change. oecologia (in review). mcgraw, a. m., r. a. moen, and m. schrage. 2011. characteristics of postparturition areas of moose in northeast minnesota. alces 47: 113–124. minnesota department of natural resources (mndnr). 2007. landsat based land use-land cover. <http:// deli.dnr.state.mn.us/metadata.html?id=l 250000120604> (accessed april 2013). ———. 2010. ecological classification system: laurentian mixed forest prov‐ ince. <http://www.dnr.state.mn.us/ecs/ 212/index.html> (accessed february 2011). moen, r. a., m. e. nelson, and a. edwards. 2011. radiotelemetry locations, home ranges, and aerial surveys in minnesota. alces 47: 101–112. pastor, j., and d. j. mladenoff. 1992. the southern boreal-northern hardwood forest border. pages 216–240 in h. h. shugart, r. leemans, and g. b. bonan, editors. a systems analysis of the global boreal forest. cambridge univer‐ sity press, cambridge, united kingdom. peek, j. m., d. l. urich, and r. t. mackie. 1976. moose habitat selection and relationships to forest management in northeastern minnesota. wildlife monographs 48: 3–65. poole, k. g., r. serrouya, and k. stuartsmith. 2007. moose calving strategies in interior montane ecosystems. journal of mammology 88: 139–150. scarpitti, d. l., p. j. pekins, and a. r. musante. 2007. characteristics of neonatal moose habitat in northern new hampshire. alces 43: 29–38. 102 moose parturition patterns – mcgraw et al. alces vol. 50, 2014 http://www.spatialecology.com/gme http://www.spatialecology.com/gme http://files.dnr.state.mn.us/recreation/hunting/moose/moose_survey_2011.pdf http://files.dnr.state.mn.us/recreation/hunting/moose/moose_survey_2011.pdf http://files.dnr.state.mn.us/recreation/hunting/moose/moose_survey_2011.pdf http://files.dnr.state.mn.us/recreation/hunting/moose/moose_survey_2011.pdf http://files.dnr.state.mn.us/recreation/hunting/moose/moose_survey_2011.pdf http://files.dnr.state.mn.us/recreation/hunting/moose/moose_survey_2011.pdf http://deli.dnr.state.mn.us/metadata.html?id=l250000120604 http://deli.dnr.state.mn.us/metadata.html?id=l250000120604 http://deli.dnr.state.mn.us/metadata.html?id=l250000120604 http://www.dnr.state.mn.us/ecs/212/index.html http://www.dnr.state.mn.us/ecs/212/index.html sigouin, d., j-p. ouellet, and r. courtois. 1997. geographical variation in the mating and calving periods of moose. alces 33: 85–95. testa, j. w., e. f. becker, and g. r. lee. 2000. movements of female moose in relation to birth and death of calves. alces. 36: 155–162. van ballenberghe, v., and w. b. ballard. 2007. population dynamics. pages 223–246 in a. w. franzmann and c. c. schwartz, editors. ecology and management of the north american moose, 2nd edition. university press of colorado, boulder, colorado, usa. wilton, m. l., and d. l. garner. 1991. preliminary findings regarding elevation as a major factor in moose calving site selection in south central ontario, canada. alces 27: 111–117. alces vol. 50, 2014 mcgraw et al. – moose parturition patterns 103 pre-arturition movement patterns and birth site characteristics of moose in northeast minnesota study area methods pre-rturition movement patterns vhf flight simulations birth site and post-arturition area results discussion acknowledgements references alces21_359.pdf alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces27_220.pdf alces(23)_49.pdf alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces24_69.pdf alces26_91.pdf alces24_7.pdf alces(23)_preface.pdf alces vol. 23, 1987 alces26_154.pdf alces21_103.pdf alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces27_1.pdf alces20_161.pdf alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces29_213.pdf alces(25)_36.pdf variation in metatarsal morphology among subgroups of north american moose (alces alces) william j. silvia1, rolf o. peterson2, john a. vucetich2, william f. silvia1, and alexander w. silvia1 1department of animal and food sciences, university of kentucky, lexington, kentucky, 40546-0215; 2school of forest resources and environmental science, michigan technological university, houghton, michigan, 49931 abstract: the objectives of this study were to characterize variation in dimensional data from the metatarsus of 4 different subpopulations of north american moose (alces alces) that are known to differ in stature, and to determine if specific metatarsal width measurements (proximal, middle, distal) can be used to accurately predict metatarsal length in these subpopulations. we found that subpopulations differ in the dimensions of their metatarsal bones. alaskan moose (a. a. gigas) are significantly larger in the length and width of the metatarsus than non-alaskan moose. moose from isle royale have significantly shorter metatarsal bones than the other groups which is associated with a proportional reduction in the middle metatarsal width; the ratio of middle width:length was similar across groups in contrast to the proximal: and distal width:length ratios. these dimensions were not reduced proportionally in isle royale specimens as these ratios were greater in the isle royale moose than in other groups. predictive equations for estimating metatarsal length from each of the 3 width measurements were developed. the length could be predicted accurately from each of the width measurements if separate predictive equations were developed for specimens collected from isle royale versus the other subgroups. these data indicate that considerable variation exists in the dimensions of a single bone, the metatarsus, in subgroups of the same species. valid predictive equations developed using data sets from one subgroup may not provide accurate predictions when applied to other subgroups of the same species. alces vol. 50: 159–170 (2014) key words: alces, metatarsus, moose, morphology, variation estimating the body size of individuals is an important part of any population assessment. direct measures (e.g., shoulder height, heart girth, body weight) of large species are often difficult to obtain in the field, and estimates of body size are often made from extrapolations of other body parts. foot length is correlated with live or carcass weight in many ungulate species (bandy et al. 1956, mcewan and wood 1966, roseberry and klimstra 1975, martin et al. 2013) including moose (alces alces) (franzmann et al. 1978, lynch et al. 1995, jensen et al. 2013). for ungulates, both living and recently deceased, this is most often measured along the plantar surface from the calcaneal protuberance to the tip of the longest toe. for animal remains that are collected after significant decomposition, it may be more convenient and consistent to measure the length of the metatarsus itself, commonly referred to as the cannon bone. the length of the metatarsus is correlated with body size across mammalian species (mcmahon 1975, alexander et al. 1979). for example, the length of the metatarsus is correlated with body weight and growth rate in cattle (coble et al. 1971b), and length and width william j. silvia, department of animal and food sciences, 409 wp garrigus bldg, university of kentucky, lexington, kentucky 40546-0215 159 of the metatarsus were smaller in female than male cattle (coble et al. 1971a), a clear indication of sexual dimorphism. the length of the metatarsus is an excellent indicator of fetal age in sheep (santucci et al. 1993), the length of the metatarsus in growing lambs is directly related to maternal nutrition during gestation (pálsson and vergés 1952), and the heritability of metatarsal dimensions is relatively high (coble et al. 1971a). the length of the metatarsus has been used as an indirect measure of body size in moose (alces alces; peterson 1977). in the field, it is quite common to find metatarsal bones from moose that have been broken or damaged in such a way that an accurate length cannot be determined. however, portions of the metatarsus are often intact permitting accurate measurement of the width at some point along the length of the bone. recognizing that metatarsal dimensions are of great utility in field research with moose and that there is considerable size variation among subpopulations of moose, our first objective was to characterize the length and 3 specific width measurements of metatarsal bones collected from 4 groups of moose: 1) isle royale national park (subspecies undetermined, either a. a. americana or a. a. andersoni), 2) extant alaskan moose (subspecies a. a. gigas), 3) fossilized alaskan moose (subspecies undetermined), and 4) mainland, excluding alaska (includes subspecies a. a. americana, a. a. andersoni, a. a. shiras). our second objective was to determine if specific metatarsal widths (proximal, middle, distal) can be used to accurately predict metatarsal length of north american moose, and to determine if the relationships between length and specific widths vary among the 4 subgroups. methods quantitative measurements of metatarsal morphology of adult north american moose were made on 4 subgroups. the first subgroup consisted of 420 moose from isle royale national park (48°06’ n, 88°30’ w; peterson 1977) located in lake superior approximately 30 km from the ontario, canada coastline. the precise origin of moose on isle royale is unknown, but the founding animals could be either a. a. americana or a. a. andersoni subspecies; however, isle royale moose are morphologically different from both subspecies (peterson et al. 2011). these metatarsal specimens are currently housed at michigan technological university’s (mtu) ford center in alberta, michigan. the second group of specimens was from 170 modern alaskan moose and included specimens housed at 1) the museum of the north, university of alaska, fairbanks, alaska (collected from denali national park; 63°20’ n, 150°30’ w; n = 65), 2) the mtu ford center (collected in the kenai national wildlife refuge [knwr] at 60° 20’ n, 150°30’ w; n = 95), 3) the american museum of natural history (amnh), new york, new york (collected throughout alaska; n = 6), and 4) the field museum of natural history (fmnh), chicago, illinois (collected throughout alaska; n = 3). the third group of 49 metatarsal bones was fossil material from the late pleistocene age that was collected from several sites 10–35 km north of fairbanks, alaska (65° n, 147°40’ w) (frick 1930, wilkerson 1932) and was part of the frick collection at the amnh (n = 49); these are presumed from the subspecies a. a. gigas. the fourth set of 34 specimens, referred hereafter as mainland moose, was collected from a variety of sites in canada and the united states (excluding alaska) and included subspecies a. a. americana, a. a. andersoni, and a. a. shiras. these specimens are housed at the 1) mtu ford center (collected by the michigan and minnesota departments of natural resources (n = 20), 2) the amnh (n = 7), 3) the fmnh (n = 1), 4) brown 160 metatarsal dimensions in alces – silva et al. alces vol. 50, 2014 university (n = 1), 5) harvard university (n = 2), and 6) the university of kentucky (n = 3). all specimens were from moose either killed by hunters or vehicular collisions. specimens collected in isle royale national park, denali national park, or the knwr were obtained from animals that died of natural causes. on isle royale, the majority resulted from predation by wolves (peterson 1977); as a result, animals that were more susceptible to predation (due to age, injury, disease) may be overrepresented. the sex of specimens was determined from examination of soft tissue (when present) and morphological characteristics of the associated skull (when present). a general age was determined by the size of the remains and the complement of deciduous and permanent teeth (peterson et al. 1983). when necessary, tissue from the metatarsal bones was removed manually with a knife and/or by prolonged immersion in hot water (>80 °c). quantitative measurements of the cannon bone were made using 2 sizes of manual vernier calipers. the length was measured using a 24-inch, cen-tech aluminum caliper (harbor freight tools inc., camarillo, california, usa) that was modified by adding a vertical fence to each side, extending the height to approximately 2.2 cm (fig. 1). the width of each metatarsus was measured at the proximal end, midpoint, and distal end with a standard 5-inch manual caliper (helios, germany; fig. 2a). the width at the proximal end was measured at the widest point, typically within 1 cm of the end (fig. 2b). the width at the distal end was also measured at the widest point, but the precise location varied; in some, it was very close to the end at the lateral and medial edges of the corresponding articular condyles, and in others it was proximal to the condyles, in the approximate location of the epiphyseal plate (fig. 2c). these measurements were used to calculate width:length ratios for each width (i.e., proximal, middle, and distal). the condition of the epiphyseal plate was classified as either unfused or fused. the unfused classification included specimens in which the 2 portions of the metatarsus were separable, and specimens in which the 2 portions were not separable but a distinct suture was clearly visible (fig. 3). specimens were classified as coming from adults only if the distal epiphyseal growth plate was no longer visible. the effects of subgroup and/or sex on quantitative measurements (length and width of the metatarsus, width:length ratio) were evaluated with analysis of variance using the glm procedure of sas (1985). the relationships between the length of the cannon bone and the 3 width measurements were evaluated with linear regression using the reg procedure of sas (1985). the accuracy of the regression equations in predicting metatarsal length from width measurements was evaluated with paired t-test using the means procedure of sas (1985). fig. 1. the modified vernier caliper used to measure the length of the cannon bone. note that vertical fences were added to each of the ‘jaws’ of the caliper to extend the height. alces vol. 50, 2014 silva et al. – metatarsal dimensions in alces 161 results metatarsal length and width the length of the metatarsus was different for each of the subgroups (p < 0.01; fig. 4a). the fossil metatarsal bones from alaskan moose were the longest, followed by those of modern alaskan moose, mainland moose, and lastly isle royale moose. the width of the metatarsus at the proximal end was greater in the 2 alaskan subgroups than in the other subgroups (p < 0.01; fig. 4b); the alaskan subgroups did not differ (p = 0.10), nor did the non-alaskan subgroups (p = 0.64). the ratio of proximal metatarsal width:metatarsal length was different among groups (p < 0.01; fig. 4c). among all subgroups the ratio was highest fig. 2. dorsal view of the cannon bone with points of measurement indicated. panel a shows the measurement of length (dashed line) and proximal, middle, and distal widths (vertical arrows); 2 possible points for measure of the distal width are indicated. panel b shows a detailed view of the proximal end indicating the point of measurement more precisely. panel c shows a more detailed view of the distal end and the 2 possible points for measurement. fig. 3. the dorsal view of the distal end of the cannon bone from an adult (panel a) and a juvenile animal (panel b). although not separable, the epiphyseal plate is clearly visible on the juvenile specimen (arrow). 162 metatarsal dimensions in alces – silva et al. alces vol. 50, 2014 in specimens from isle royale (p < 0.05); the ratio among the other 3 subgroups did not differ (p > 0.70). to further examine the relationship between proximal width and length, the effect of width and subgroup on metatarsal length was determined (n = 285). proximal width had a significant effect (p < 0.01), but subgroup did not (p = 0.06). there was an interaction between proximal width and subgroup on metatarsal length (p = 0.01). since the ratio width:length appeared to be different for isle royale moose compared to the other subgroups, a second analysis was conducted without isle royale moose. again, the effect of proximal width was evident (p < 0.01), but not subgroup (p = 0.27) or the interaction term of proximal width and subgroup (p = 0.20), implying that the alaskan and mainland subgroups are similar and a different relationship exists for isle royale moose. middle (n = 226) and distal width measurements (n = 224) were not available from the fossil specimens; therefore, comparisons could only be made among the modern subgroups. the width of the metatarsal at the midpoint was greater in alaskan moose than those from non-alaskan subgroups (p < 0.01; fig. 5a); this width was similar in isle royale and non-alaskan subgroups (p > 0.30). the ratio middle width: length was not different among subgroups (p = 0.44; fig. 5b). as with the proximal metatarsal width, the distal metatarsal width was greater in the alaskan subgroup than non-alaskan subgroups (p < 0.01; fig. 5c). the ratio distal metatarsal width:metatarsal length also differed among subgroups (p < 0.01; fig. 5d). the distal width:length ratio was greater for isle royale moose than the other groups (p < 0.01); the other groups did not differ (p = 0.12). the effect of sex on metatarsal dimensions was analyzed with all specimens in which sex could be determined, which excluded the fossil subgroup. the length of the metatarsus was greater in males than females (p < 0.01; fig. 6a). as in the first analysis, the length of the metatarsus differed among subgroups (p < 0.01), and metatarsal length was longer in males than females in all subgroups. a significant interaction fig. 4. the effect of subgroup (isle royale [isro], mainland [non-isro from the lower 48 contiguous united states and canada], alaska, and fossil alaska) on metatarsal length (a), proximal metatarsal width (b), and the ratio of proximal metatarsal width to metatarsal length (c). bars with different letter superscripts are different (p < 0.05). alces vol. 50, 2014 silva et al. – metatarsal dimensions in alces 163 between sex and subgroup was also found (p < 0.01). this interaction was strongest in alaskan moose that had the longest metatarsal length and largest difference between males and females. the effect of sex on the relationship between each of the 3 width measurements and metatarsal length was examined fig. 5. the effect of subgroup (isle royale [isro], mainland [non-isro from the lower 48 contiguous united states and canada] and alaska) on middle metatarsal width (a), the ratio of proximal metatarsal width to metatarsal length (b), distal metatarsal width (c), and the ratio of distal metatarsal width to metatarsal length (d). bars with different letter superscripts are different (p < 0.05). fig. 6. the effect of sex and subgroup on metatarsal dimensions: length of the metatarsus (panel a) in which effects of sex, subgroup, and their interaction were observed (p < 0.01). proximal, middle, and distal width of the metatarsus is shown in panels b, c, d. effects of sex and subgroup on all 3 width measurements were observed (p < 0.01) but not of their interaction (p ≥ 0.09). 164 metatarsal dimensions in alces – silva et al. alces vol. 50, 2014 separately in the isle royale and non-isle royale subgroups. in both cases, the effect of proximal width was evident (p < 0.01; table 1). sex had no effect on length that was not already accounted for by proximal width (p > 0.21). the interaction term of sex with proximal width on metatarsal length was also not significant (p > 0.20) in either the isle royale or non-isle royale subgroups. similarly, there were no effects of sex or the interaction of sex and width on the relationship between middle or distal width on the length of the metatarsal (table 1). predictive equations for metatarsal length based on widths a quantitative description of the relationship between proximal width and metatarsal length was investigated with linear regression. separate regression analyses were conducted for the isle royale and non-isle royale subgroups using the following simple model: metatarsal length ¼ m � proximal metatarsal width þ b þ e ð1þ where: m = slope, b = y-intercept, and e = error term. comparison of the estimates of slope and y-intercept for the two groups (isle royale versus non-isle royale) indicated substantial difference (table 2, fig. 7). these relationships explained a high percentage of the variation in metatarsal length for isle royale (r2 = 0.47) and non-isle royale specimens (r2 = 0.66) (table 2). the accuracy of the regression lines in predicting metatarsal length from proximal width was evaluated by comparing measured lengths to estimated lengths from specimens not used to derive the regression equations; specimens from both groups were included in this test. the length of metatarsal bones from both groups was more accurately predicted using the separate regression equations derived from the respective data sets (table 3). the same analytical procedures were used to examine the relationships between middle width and metatarsal length, and distal width and metatarsal length; middle and distal widths were not available from fossil alaskan moose. there was no effect of subgroup or the width by subgroup interaction term, indicating consistency across all subgroups. subsequently, regression analysis was used and prediction equations developed with the combined subgroup data (fig. 8, 9, table 4). again, width measurements accounted for a large percentage of the variation in metatarsal length (r2 = 0.55 and 0.53 for middle and distal widths, respectively). table 1. the effect of sex on the relationship between the width of the metatarsal at 3 points of measurement (proximal, middle, and distal) and the length of the metatarsal in specimens from isle royale and non-isle royale locations including alaska (modern and fossil), canada, and the 48 contiguous united states (excluding isle royale). isle royale non-isle royale proximal width n 102 120 width p < 0.01 p < 0.01 sex p > 0.21 p > 0.78 sex × width interaction p > 0.20 p > 0.83 middle width n 65 121 width p < 0.01 p < 0.01 sex p > 0.89 p > 0.79 sex × width interaction p > 0.83 p > 0.84 distal width n 65 120 width p < 0.01 p < 0.01 sex p > 0.82 p > 0.89 sex × width interaction p > 0.86 p > 0.83 alces vol. 50, 2014 silva et al. – metatarsal dimensions in alces 165 as with proximal width measurements, a reasonably accurate estimate of metatarsal length was obtained from either middle or distal width (table 5). as expected, the length of the isle royale specimens tended to be overestimated. although not justified based on the initial analysis, a more accurate estimate of metatarsal length was developed using separate equations derived from the isle royale and non-isle royale data (table 6, fig. 8, 9). accuracy was substantially improved for the isle royale and alaskan subgroups (table 7), but not mainland moose that was a small heterogeneous group representing 3 subspecies. table 3. comparison of measured and predicted metatarsal lengths (mm) in isle royale and non-isle royale moose using separate predictive equations developed from proximal metatarsal lengths (mm). non-isle royale moose include modern and fossil specimens from alaska, and modern specimens from the 48 contiguous united states (excluding isle royale) and canada. isle royale alaska mainland sample size 6 6 6 ave. proximal metatarsal width 53.6 57.7 51.6 ave. metatarsal length 384.5 414.3 389.8 predicted metatarsal length from proximal metatarsal width (isle royale) 388.2 398.9 382.7 difference between, range, and probability that true and predicted lengths differ (isle royale) −3.7 −13.6−9.8 15.5 5.4−33.4 7.1 −0.4−18.8 p = 0.40 p = 0.02 p = 0.04 predicted metatarsal length from proximal metatarsal width (non-isle royale) 403.7 420.2 395.5 difference between, range, and probability that true and predicted lengths differ (non-isle royale) −19.2 −28.0−5.4 −5.9 −14.7−9.0 −5.6 −16.7−2.1 p = 0.01 p = 0.15 p = 0.08 table 2. the regression parameters describing the different relationship between proximal metatarsal width and metatarsal length in specimens from isle royale compared to other populations in canada and the united states including alaska. isle royale non-isle royale sample size 110 159 significance level p < 0.01 p < 0.01 adjusted r2 0.47 0.66 slope (se) 2.67 (0.27) 4.09 (0.23) y-intercept (se) 245 (14) 185 (13) fig. 7. scatter plot depicting the relationship between the proximal metatarsal width and metatarsal length for the 4 subgroups (isle royale [isro], mainland [non-isro from the lower 48 contiguous united states and canada], alaska, and fossil alaska). regression lines for the isle royale (dotted) and non-isle royale (solid line) groups are shown. 166 metatarsal dimensions in alces – silva et al. alces vol. 50, 2014 discussion the measurements of metatarsal length and width indicated that the alaskan subgroups are larger in relative size. the isle royale subgroup is different from the other subgroups with shorter metatarsal length and correspondingly larger proximal: and distal width:length ratios. the length of the metatarsus was shorter in isle royale moose than the other subgroups and may reflect the trend for large herbivores to experience a reduction in size when isolated on small islands (peterson et al. 2011), which conforms to the ‘island rule’ (van valen 1973, lomolino 2005). given this hypothesis and the short history of isle royale moose, these data demonstrate the remarkable speed at which this phenomenon can occur. the metatarsal length:width ratios also provide insight into the biological mechanism by which reduction in metatarsal size occurred on isle royale. long bones, including metatarsals, initially form in 3 parts, the proximal epiphysis (proximal articular surface), diaphysis (shaft), and distal epiphysis. growth ceases when the cartilaginous epiphyseal plates separating these portions ossify. it appears that the reduced size in isle royale specimens is limited to the diaphysis with both the length and width of the diaphysis affected proportionally. the widths at the proximal and distal epiphyses do not appear to be reduced, particularly when compared to the mainland group. thus, the shortening effect appears to be mediated solely through the diaphysis and this isolated effect may facilitate the identification of specific genes mediating such evolutionary action. the length of the metatarsus could be predicted accurately from each of the width measurements, particularly if separate predictive equations were developed for specimens from isle royale versus other subgroups. the greatest deviation between predicted and actual metatarsal length was only 4.3% using the specific equations; refinements to these predictive equations are presumably possible. for example, the distal width measurement was taken either at the distal fig. 8. scatter plot depicting the relationship between middle metatarsal width and metatarsal length for 3 subgroups (isle royale [isro], mainland [non-isro from the lower 48 contiguous united states and canada], and alaska). regression lines derived from isro specimens (dotted line), non-isro specimens (solid line), and for all specimens combined (dashed line) are shown. fig. 9. scatter plot depicting the relationship between distal metatarsal width and metatarsal length for 3subgroups (isle royale [isro], mainland [non-isro from the lower 48 contiguous united states and canada], and alaska). regression lines derived from isro specimens (dotted line), non-isro specimens (solid line), and for all specimens combined (dashed line) are shown. alces vol. 50, 2014 silva et al. – metatarsal dimensions in alces 167 epiphysis or at the distal articular condyle, whichever was wider; however, a more accurate equation might be developed with a single, consistent measurement. predictive equations based on middle and distal widths for the non-isle royale subgroups improved the accuracy of prediction for the alaskan, but not mainland group, possibly reflecting the potential heterogeneity within the mainland group. it may indicate that separate equations need to be developed for subpopulations within. finally, possible differences in the method of sample collection among data sets should be considered. the majority of mainland specimens were collected by hunters or the result of vehicular accidents, whereas specimens from isle royale, kenai national wildlife refuge, and denali national park were collected from moose presumably dying of natural causes. animals that were particularly susceptible to predation may be overrepresented in these groups. a more robust sample size reflecting consistent sampling and population variation would presumably improve the relationships presented in this paper. acknowledgements the authors would like to thank the field workers who collected metatarsal specimens on isle royale and contributed to the collection at michigan technological university. we are indebted to the officials at the united states department of the interior, national table 4. the regression parameters describing the relationship between middle metatarsal width (mm) and metatarsal length (mm), and distal metatarsal width (mm) and metatarsal length. common equations were developed using specimens from all subgroups. middle metatarsal distal metatarsal sample size 210 208 significance level p < 0.01 p < 0.01 adjusted r2 0.55 0.53 slope (se) 4.81 (0.30) 3.38 (0.22) y-intercept (se) 242 (10) 172 (15) table 5. comparison of measured metatarsal length (mm) and predicted metatarsal length (mm) in isle royale and non-isle royale moose based on middle and distal metatarsal widths (mm). common predictive equations were derived using specimens from all subgroups. isle royale alaska mainland sample size 6 6 6 ave. metatarsal length 384.5 414.3 389.8 ave. middle width of metatarsal 32.5 35.2 31.8 predicted metatarsal length from middle metatarsal width 397.8 411.2 394.7 difference between, range, and probability that true and predicted length differ (mm) −13.3 −30.0−6.2 3.2 −5.2−15.9 −4.8 −9.3−0.1 p = 0.08 p = 0.34 p = 0.01 ave. distal width of metatarsal 66.9 70.9 64.3 predicted metatarsal length from distal metatarsal width 398.3 411.7 389.7 difference between true length and predicted length (mm) and range (below) −13.8 −23.3−0.0 2.7 −13.6−11.9 0.2 −11.7−12.0 probability that the true length and the predicted length are different p = 0.04 p = 0.53 p = 0.96 168 metatarsal dimensions in alces – silva et al. alces vol. 50, 2014 parks service, and isle royale national park for granting access to the park and permitting collection of specimens. we would also like to thank mr. p. poore for the modification of the vernier calipers used to measure the length of the cannon bone. this research was supported in part by the kentucky agricultural experiment station and is published with the approval of the director (publication number 14-07-014). references alexander, r. m., a. s. jayes, g. m. o. maloiy, and e. m. wathuta. 1979. allometry of the limb bones of mammals from shrews (sorex) to elephant (loxodonta). journal of zoology 189: 305–314. bandy, p. j., i. m. cowan, w. d. kitts, and a. j. wood. 1956. a method for the assessment of the nutritional status of wild table 7. comparison of measured and predicted metatarsal lengths (mm) from middle and distal metatarsal widths (mm) in isle royale moose with those from non-isle royale moose. the alaska and mainland equations were derived from measurements from non-isle royale moose that included modern and fossil specimens from alaska, and modern specimens from the 48 contiguous united states (excluding isle royale) and canada. isle royale alaska mainland sample size 6 6 6 ave. metatarsal length 384.5 414.3 389.8 ave. middle width of metatarsal 32.5 35.2 31.8 predicted metatarsal length from middle metatarsal width 388.3 414.5 400.5 difference, range, and probability that true and predicted lengths differ −3.8 −16.2−13.7 −0.2 −7.3−14.5 −10.7 −15.5−5.5 p = 0.50 p = 0.96 p < 0.01 ave. distal width of metatarsal 66.9 70.9 64.3 predicted metatarsal length from distal metatarsal width 388.0 415.5 396.5 difference, range, and probability that true and predicted lengths differ −3.5 −13.5−11.0 −1.2 −16.6−10.4 −6.7 −16.8−6.4 p = 0.47 p = 0.78 p = 0.10 table 6. the regression parameters describing the relationship between proximal metatarsal width (mm) and metatarsal length (mm) in specimens from isle royale compared to non-isle royale moose. non-isle royale moose included modern and fossil specimens from alaska, and modern specimens from the 48 contiguous united states (excluding isle royale) and canada. middle metatarsal width distal metatarsal width isle royale non-isle royale isle royale non-isle royale sample size 71 139 71 137 significance p < 0.01 p < 0.01 p < 0.01 p < 0.01 adjusted r2 0.41 0.54 0.50 0.54 slope (se) 3.30 (0.47) 4.07 (0.32) 2.38 (.29) 2.91 (0.23) y-intercept (se) 280 (15) 270 (11) 228 (19) 209 (16) alces vol. 50, 2014 silva et al. – metatarsal dimensions in alces 169 ungulates. canadian journal of zoology 34: 48–52. coble, d. s., l. l. wilson, j. p. hitchcock, h. varela-alvarez, and m. j. simpson. 1971a. sire, sex and laterality effects on bovine metacarpal and metatarsal characters. growth 35: 65–77. ——— , ———, m. j. simpson, h. varelaalvarez, j. p. hitchcock, j. h. ziegler, j. d. sink, and j. l. watkins. 1971b. relation of bovine metacarpal and metatarsal characters to growth and carcass characters. growth 35: 79–89. franzmann, a. w., r. e. leresche, r. a. rausch, and l. l. oldemeyer. 1978. alaskan moose measurements and weights and measurement–weight relationships. canadian journal of zoology 56: 298–306. frick, c. 1930. alaska’s frozen fauna. natural history (the journal of the american museum of natural history) 30: 71–80. jensen, w. f., j. r. smith, j. j. maskey jr., j. v. mckenzie, and r. e. johnson. 2013. mass, morphology and growth rates of moose in north dakota. alces 49: 1–15. lomolino, m. v. 2005. body size evolution in insular vertebrates: generality of the island rule. journal of biogeography 32: 1683–1699. lynch, g. m., b. lajuenesse, j. willman, and e. s. telfer. 1995. moose weights and measurements from elk island national park, canada. alces 31: 199–207. martin, j. g. a., m. festa-bianchet, s. d. cote, and d. t. blumstein. 2013. detecting between-individual differences in hind-foot length in populations of wild mammals. canadian journal of zoology 91: 118–123. mcewan, e. h., and a. j. wood. 1966. growth and development of the barren ground caribou. i. heart girth, hind foot length and body weight relationships. canadian journal of zoology 44: 401–411. mcmahon, t. a. 1975. allometry and biomechanics: limb bones in adult ungulates. american naturalist 109: 547–563. palsson, h., and j. b. verges. 1952. effect of the plane of nutrition on growth and the development of carcass quality in lambs. part i. the effects of high and low planes of nutrition at different ages. journal of agricultural science 42: 1–92. peterson, r. o. 1977: wolf ecology and prey relationships on isle royale. u.s. national park service scientific monograph series 11. u.s. government printing office, washington d. c., usa. ——— , c. c. schwartz, and w. b. ballard. 1983. eruption patterns of selected teeth in three north american moose populations. journal of wildlife management 47: 884–888. ———, j. a. vucetich, d. beyer, m. schrage, and j. raikkonen. 2011. phenotypic variation in moose: the island rule and the moose of isle royale. alces 47: 125–133. roseberry, j. l., and w. d. klimstra. 1975. some morphological characteristics of the crab orchard deer herd. journal of wildlife management 39: 48–58. sas. 1985. user’s guide: statistics. sas institute inc., cary, north carolina, usa. santucci, v. l., j. a. kuller, a. f. battelli, s. a. laifer, and d. i. edelstone. 1993. fetal metatarsal length: an accurate predictor of gestational age and weight in the ovine fetus. gynecologic and obstetric investigation 35: 76–79. van valen, l. 1973. a new evolutionary law. evolutionary theory 1: 1–33. wilkerson, a. s. 1932. some frozen deposits in the goldfields of interior alaska: a study of the pleistocene deposits of alaska. american museum novitates 525: 1–22. 170 metatarsal dimensions in alces – silva et al. alces vol. 50, 2014 variation in metatarsal morphology among subgroups of north american moose (alces alces) methods results metatarsal length and width predictive equations for metatarsal length based on widths discussion acknowledgements references alces22_377.pdf alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 temporal assessment of physical characteristics and reproductive status of moose in new hampshire daniel h. bergeron1,2, peter j. pekins 1 , and kristine rines2 1department of natural resources and the environment, university of new hampshire, durham, new hampshire 03824, usa; 2new hampshire fish and game department, new hampton, new hampshire 03256, usa. abstract: biological data collected from harvested moose (alces alces) were analyzed to assess whether temporal change has occurred in the physical and reproductive condition of moose from 1988–2009 in new hampshire. measurements included age and field-dressed body weight of both sexes, number of corpora lutea (cl) and ovulation rate of females, and antler beam diameter (abd) and antler spread of males. similar data were obtained from maine and vermont for comparative analysis. the only significant changes (p <0.05) occurred in the yearling age class: mean body weight of both sexes, number of cl, and abd all declined in new hampshire. the current ovulation rate (∼20%) and mean body weight (<200 kg) of yearling females in new hampshire and vermont were considered low. the declines measured in yearlings, yet relative stability in adults, are consistent with the presumption that winter ticks (dermacentor albipictus) impact the productivity of moose populations through reduced calf survival and growth and fecundity of yearlings. density-dependent factors related to habitat change are also discussed given the recent, rapid expansion of moose in the 3 states. continued monitoring of physical parameters and productivity of harvested moose, particularly the yearling cohort, is warranted to better assess the relationships among winter ticks, habitat quality, and moose populations. alces vol. 49: 39–48 (2013) key words: alces alces, body weight, moose, new england, physical characteristics, reproductive status age-specific body weight is directly related to the health and production of male and female moose (alces alces) (schwartz and hundertmark 1993), and onset of ovulation in yearlings (saether and heim 1993). antler measurements that are used routinely to estimate the health of white-tailed deer (odocoileus virginianus) populations are also used to gauge population status of moose (e.g., child et al. 2010); antler size in moose is influenced by many factors including nutritional status and health (bubenik 1997). in new hampshire, age, antler spread, antler beam diameter (abd), number of points, corpora lutea (cl) count, and field-dressed body weight of hunterharvested moose have been measured since 1988. adams and pekins (1995) found differences in body weight and number of cl in yearling cow moose relative to other age classes, but no difference within age classes from regions with different moose density. they concluded that yearling moose were useful for estimating herd health due to their substantial weight gain, change in antler characteristics, and onset of ovulation in this age class. because their data were from a relatively new and expanding moose population in the 1980–1990s, they encouraged future analyses to assess both temporal and regional trends. musante et al. (2010) found that the ovulation rate and cl count of yearling moose in new hampshire declined from 39 1988–1998 to 1999–2004, yet were un‐ changed in adults. in a comprehensive study including habitat use (scarpitti 2006) and age-specific mortality rates, they concluded that epizootics of winter ticks (dermacentor albipictus) caused periodic, annual high mortality in calves and lower fecundity in yearlings. given the relationships between certain physical characteristics and nutritional status of a moose population, periodic analysis of physical and reproductive data should reveal trends and change in the relative condition of the moose population in new hampshire. in this study we assessed temporal trends in physical characteristics and relative nutritional and reproductive status of moose in new hampshire from 1988– 2009, a period that encompassed previous studies (i.e., adams and pekins 1995, musante et al. 2010) and 5 additional years. further, we analyzed similar data from neighboring states maine and vermont to produce a regional assessment. methods study area we used data collected by new hampshire fish and game department (nhfg) personnel at mandatory harvest check stations. moose/data were from 3 northern regions that differed in moose population density (nhfg 2009) (fig. 1); the 3 regions from highest to lowest density were connecticut (ct) lakes (0.83 moose/km2), north (0.61 moose/km2), and white mountains (0.26 moose/km2), respectively (k. rines, unpubl. data, 2009). elevation in the study area ranges from ∼120–1900 m, average snow depth ranges from 0–60 cm, and ambient temperature ranges from ∼−30 to 30° c (noaa 1971–2000). the ct lakes and north regions were dominated by commercial hardwood species including sugar (acer saccharum) and red maple (a. rubrum), yellow birch (betula alleghaniensis), and american beech (fagus grandifolia). red spruce (picea rubens) and balsam fir (abies balsamea) tend to be the dominant species at higher elevations (>760 m) and in cold, wet lowland sites (degraaf et al. 1992). these regions are predominately forested and the majority of the land is privately owned and commercially harvested using various silvicultural techniques (degraaf et al. 1992); they contain ∼10% wetlands and open water, and are interspersed with trails and logging roads. the ct lakes region is hilly with few high mountains, while the north is characterized by higher forested terrain. the white mountains region contains the white mountain national forest which covers 304,050 ha and is ∼97% forested. it contains the highest elevations in new hampshire and is dominated by beech, sugar maple, and yellow birch; other common species include white ash (fraxinus americana), red maple, red spruce, and eastern hemlock (tsuga canadensis). timber harvest in this region is at smaller scale than the other regions, with maximum clear-cut size of ∼10–12 ha (degraff et al. 1992, sperduto and nichols 2004). white-tailed deer are sympatric with moose throughout the study area, and at low-moderate density (<4/km2). field measurements physical measurements of harvested moose in 1988–2009 were divided into 3 time periods (1988–1998, 1999–2004, and 2005–2009) and analyzed by region. measurements included age and fielddressed body weight for both sexes, number of cl, abd, antler spread, and number of points. a micrometer was used to measure abd on one antler at 2 perpendicular sites 2.54 cm above the pedicle; the average diameter was recorded. antler spread was the maximum distance measured between any 2 points, and an antler point was ≥2.54 cm long. ovaries were collected and stored in denatured ethyl-alcohol and sectioned later to 40 physical and reproductive status in nh – bergeron et al. alces vol. 49, 2013 visually count the number of cl (cheatum 1949). age was determined by cementum annuli counts from a lower incisor (sergeant and pimlott 1959). a subset of similar data was obtained from maine and vermont; maine data included only field-dressed body weight of cows and vermont data were from 1993–2009. data analysis new hampshire data were analyzed initially by time period and sample region, and combined statewide for comparison with maine and vermont data. analysis of variance (anova) was used to test for age-specific differences in physical parameters; age classes were 0.5, 1.5, 2.5, 3.5, 4.5, 5.5, and ≥6.5 years. a shapiro-wilk test was used to test if the data were normally distributed and a bartlett test was used to check for homogeneity of variance (zar 1999). pairwise comparisons were made with the tukey test. analyses were performed with systat v. 13. significance for all tests was assigned a priori at α = 0.05. results the analysis included measurements from >3000 and 1500 male moose, and >1500, 1300, and 2500 female moose in new hampshire, vermont, and maine, respectively. in new hampshire, sample size was >10 in the middle age classes (1.5–3.5 years) in all regions in any given time period; sample size was >20 in all age classes/time periods for state comparisons. fig. 1. location of 3 study regions with different moose density (high-low) used to evaluate temporal trends in physical and reproductive status of moose in northern new hampshire, 1988–2009. alces vol. 49, 2013 bergeron et al. – physical and reproductive status in nh 41 females statewide means for body weight and cl counts for all age classes are presented in table 1. in new hampshire the only significant differences between time periods in any region occurred in the yearling age class; albeit, body weight declined in most age classes in successive periods (table 1). body weight of yearlings declined significantly (∼25 kg) from 1988–1998 to 2005– 2009 in all regions (fig. 2): ct lakes (p = 0.033), north (p = 0.000), white mountains (p = 0.003). the number of cl in yearlings also declined from 1988–1998 to 2005–2009 in all regions (fig. 3): ct lakes (43%, p = 0.009), north (68%, p = 0.000), white mountains (76%, p = 0.003). the cl count was ∼0.20 across all regions in 2005–2009, declining from 0.60–0.80 since 1988–1998. the ovulation rate in yearling cows declined from 56 to 21% from 1988–1998 to 2005–2009 in new hampshire, and from 36 to 16% in vermont. the average body weight of yearling cows with 0 cl was 199 kg in new hampshire and 198 kg in vermont. yearling body weight declined 6% in vermont (11 kg, p = 0.001) from 1999–2004 to 2005–2009 (fig. 2), and cl counts, though not different, also declined to <0.20 (fig. 3). the cl count was lower in vermont than new hampshire in 1988– 1998 (45%, p = 0.030) and 1999–2004 (38%, p = 0.030) (fig. 3); there was no difference in 2005–2009, albeit all counts were historical lows. yearling body weight in new hampshire and vermont was not different. body weight of maine year‐ lings increased 3% from 1988–1998 to 1999–2004 (p = 0.012) (fig. 2). body weight was 6% lower in maine than new hampshire in 1988–1998 (p = 0.000), but 7% higher in 2005–2009 (p = 0.000). mean body weight of maine yearlings increased 9% from 1988–1998 to 2005–2009, and only maine had a statewide mean >200 kg table 1. statewide means (± sd) of field-dressed body weight and number of corpora lutea of harvested female moose in 3 consecutive time periods in new hampshire, 1988–2009. the only significant differences (p <0.05) occurred in the 1.5 year (yearling) age class (*); all parameters declined from 1988– 1998 to 2005–2009. age 1988–1998 1999–2004 2005–2009 body weight (kg) 0.5 110 ± 25 (74) 105 ± 20 (51) 107 ± 22 (45) 1.5 211 ± 31 (175) 203 ± 27 (206) 190 ± 29 (165)* 2.5 258 ± 34 (167) 250 ± 29 (132) 238 ± 31 (117) 3.5 255 ± 35 (102) 246 ± 29 (85) 258 ± 31 (87) 4.5 268 ± 34 (55) 263 ± 35 (68) 247 ± 43 (60) 5.5 261 ± 32 (46) 260 ± 32 (48) 246 ± 29 (40) ≥6.5 258 ± 36 (106) 263 ± 31 (133) 257 ± 36 (131) # corpora lutea 1.5 0.65 ± 0.65 (187) 0.42 ± 0.52 (200) 0.21 ± 0.42 (169)* 2.5 1.26 ± 0.66 (174) 1.09 ± 0.53 (142) 0.98 ± 0.48 (127) 3.5 1.29 ± 0.62 (102) 1.17 ± 0.60 (90) 1.08 ± 0.54 (91) 4.5 1.53 ± 0.65 (62) 1.26 ± 0.56 (72) 0.98 ± 0.61 (62) 5.5 1.37 ± 0.61 (46) 1.30 ± 0.63 (54) 1.13 ± 0.67 (48) ≥6.5 1.46 ± 0.73 (108) 1.31 ± 0.60 (151) 1.13 ± 0.66 (142) 42 physical and reproductive status in nh – bergeron et al. alces vol. 49, 2013 fig. 2. mean (± se) field-dressed body weight (kg) of harvested yearling female moose in 3 sample regions of new hampshire (1988–2009), and statewide means in new hampshire, vermont, and maine. body weight declined (p <0.05) in new hampshire and vermont from 1988–1998 to 2005–2009; conversely, body weight increased in maine. for reference, yearlings with body weight <200 kg are considered non-reproductive. 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 ct lakes north white mt new hampshire vermont c o rp o ra l u te a sample area 1988-1998 1999-2004 2005-2009 time period fig. 3. mean (± se) number of corpora lutea (cl) in harvested yearling female moose in 3 sample regions of new hampshire (1988–2009), and statewide means in new hampshire and vermont. number of cl declined (p <0.05) in new hampshire from 1988–1998 to 2005–2009; although not different, the decline in vermont was ∼50%. alces vol. 49, 2013 bergeron et al. – physical and reproductive status in nh 43 from 1999–2009. the proportion of yearlings >200 kg in new hampshire, vermont, and maine was 44, 32, and 62%, respectively, in 2005–2009. males statewide means for body weight, abd, and antler spread are presented in table 2. in new hampshire the only significant differences between time periods in any region occurred in the yearling age class; albeit, all characteristics in table 2 declined in most age classes in successive periods. yearling body weight declined 28, 16, and 30 kg from 1988–1998 to 2005–2009 in the ct lakes (12%, p = 0.000), north (7%, p = 0.011), and white mountain (14%, p = 0.000) regions, respectively (table 3). yearling abd declined 11% in the ct lakes (p = 0.023) and 9% in the white mountains (p = 0.014) regions (table 3) from 1988– 1998 to 2005–2009. yearling antler spread declined 13, 11, and 15% from 1988–1998 to 2005-2009 in the ct lakes (p = 0.034), north (p = 0.026), and white mountains (p = 0.001) regions, respectively (table 3). as in new hampshire, vermont yearlings declined in each physical characteristic except abd; body weight declined 9% (p = 0.003) (table 3) and antler spread 7% table 2. statewide means (± sd) of field-dressed body weight, antler beam diameter (abd), and antler spread of harvested bull moose in 3 consecutive time periods in new hampshire, 1988–2009. the only significant differences (p <0.05) occurred in the 1.5 year (yearling) age class (*); all parameters declined from 1988–1998 to 2005–2009. age 1988–1998 1999–2004 2005–2009 body weight (kg) 0.5 119 ± 23 (67) 114 ± 26 (42) 115 ± 25 (46) 1.5 222 ± 39 (377) 206 ± 24 (235) 201 ± 29 (184)* 2.5 271 ± 42 (361) 262 ± 28 (246) 253 ± 30 (219) 3.5 311 ± 36 (229) 294 ± 30 (251) 284 ± 33 (214) 4.5 335 ± 40 (174) 317 ± 32 (172) 312 ± 34 (150) 5.5 350 ± 37 (108) 331 ± 32 (96) 319 ± 37 (93) ≥6.5 352 ± 37 (180) 344 ± 32 (243) 335 ± 36 (218) abd (mm) 1.5 36 ± 9 (415) 34 ± 7 (262) 34 ± 6 (199)* 2.5 45 ± 7 (391) 44 ± 5 (275) 42 ± 5 (251) 3.5 49 ± 6 (258) 47 ± 5 (291) 46 ± 4 (243) 4.5 54 ± 7 (191) 51 ± 6 (195) 50 ± 6 (162) 5.5 56 ± 8 (124) 54 ± 6 (114) 54 ± 5 (106) ≥6.5 60 ± 6 (214) 59 ± 7 (271) 58 ± 6 (236) antler spread (cm) 1.5 66 ± 11 (372) 60 ± 12 (247) 59 ± 11 (191)* 2.5 90 ± 11 (363) 85 ± 12 (275) 81 ± 11 (247) 3.5 107 ± 15 (246) 98 ± 14 (289) 96 ± 15 (242) 4.5 120 ± 16 (183) 112 ± 16 (191) 109 ± 16 (157) 5.5 126 ± 16 (114) 121 ± 12 (121) 118 ± 16 (106) ≥6.5 133 ± 15 (197) 131 ± 16 (269) 128 ± 15 (232) 44 physical and reproductive status in nh – bergeron et al. alces vol. 49, 2013 (p = 0.049) from 1988–1998 to 2005–2009 (table 3). there was no difference in body weight between new hampshire and vermont yearlings; antler spread was greater in new hampshire than vermont in 1988–1998 (9%, p = 0.031) and 2005–2009 (5%, p = 0.028), and abd was 6% larger in vermont than new hampshire in 1999–2004 (p = 0.033). discussion prior research in new hampshire (musante 2006, musante et al. 2010) indicated that new hampshire's moose population was effectively stable due to low annual growth rate (estimates = 0.95-1.07). population stability occurs despite the belief that habitat quality is high (scarpitti et al. 2005, scarpitti 2006) and adult productivity and survival are also high (musante et al. 2010). the population is presumably most influenced by winter ticks that cause periodic, high mortality of calves and reduced productivity in yearling cows (musante et al. 2010). our data indicate that body weight and cl count of yearling females have continued to decline through 2005–2009 to about 190 kg and 0.20 cl (table 1), respectively; ovulation rates of yearlings in north america average 49% (range = 0–100%, schwartz 2007). conversely, the ovulation rate of adults was not low in new hampshire or vermont (most age classes >90%, table 1); however, the cl count of adults was in decline in all age classes across the study period (table 1). yearling females <200 kg are considered non-reproductive (adams and pekins 1995), and not coincidently, mean body weight of table 3. means (± sd) of field-dressed body weight, antler beam diameter (abd), and antler spread of harvested 1.5 year-old bull moose in 3 consecutive time periods in 3 regions of new hampshire, 1988– 2009. significant declines (p <0.05) of all parameters occurred in all regions of new hampshire from 1988–1998 to 2005–2009, except abd in the north. body weight and antler spread declined (p <0.05) in vermont from 1988–1998 to 2005–2009. 1988–1998 1999–2004 2005–2009 body weight (kg) ct lakes 232 ± 43 (80) 209 ± 26 (44) 204 ± 34 (43) north 223 ± 30 (119) 212 ± 23 (80) 207 ± 27 (61) white mt. 222 ± 44 (102) 194 ± 22 (38) 192 ± 23 (36) new hampshire 222 ± 39 (377) 206 ± 25 (235) 201 ± 29 (184) vermont 216 ± 27 (58) 202 ± 28 (127) 196 ± 27 (247) abd (mm) ct lakes 38 ± 10 (85) 34 ± 6 (47) 34 ± 6 (47) north 35 ± 7 (134) 34 ± 7 (99) 34 ± 5 (65) white mt. 37 ± 9 (113) 33 ± 7 (44) 34 ± 7 (40) new hampshire 36 ± 9 (415) 34 ± 7 (262) 34 ± 6 (199) vermont 34 ± 6 (59) 36 ± 6 (128) 34 ± 7 (258) antler spread (cm) ct lakes 68 ± 21 (76) 60 ± 14 (44) 59 ± 10 (44) north 64 ± 16 (123) 59 ± 12 (96) 57 ± 11 (63) white mt. 69 ± 21 (99) 57 ± 11 (40) 59 ± 13 (39) new hampshire 66 ± 11 (372) 60 ± 12 (247) 59 ± 11 (191) vermont 60 ± 11 (54) 60 ± 12 (118) 56 ± 12 (247) alces vol. 49, 2013 bergeron et al. – physical and reproductive status in nh 45 cows with 0 cl was 199 kg in new hampshire (1988–2009) and 198 kg in vermont (1993–2009). productivity from the yearling age class in new hampshire and vermont is expectedly low based on ovulation rates ≤20% that are considerably lower (30–50%) than those measured prior to 2000. the mean cl count in new hampshire (0.22) and vermont (0.16) was equal to half the proportion of yearlings >200 kg (44 and 32%, respectively); assuming this relationship, the mean cl in maine is probably >0.30, as 62% of yearlings were >200 kg. several factors including habitat quality, weather, and disease/parasites contribute to declining trends in physical parameters of a moose population, the latter 2 typically of short-term impact. however, musante et al. (2010) believed that moose in new hampshire were mostly influenced by the annual impact, and particularly epizootics, of winter ticks. mortality of their radio-collared moose was mostly due to winter kill/parasites (41%) associated with winter tick infestations; mortality due to hunting, road-kill, poaching, predation, and weather was not considered major during the 4-year study. further, habitat was considered adequate because field-dressed weights, reproductive data, and survival of adults were not low or declining, or representative of a habitatlimited population. although our analysis identified no statistical decline in physical characteristics or ovulation rates of adults, body weight of males and females and age-specific cl counts trended downward across the ∼20-year period (tables 1 and 2). calves are most severely impacted by winter tick infestations and some mortality is likely an annual event; however, even surviving calves presumably experience lower body weight and reduced fecundity as yearlings (samuel 2004, 2007, musante et al. 2010). the declining trend in yearling condition in new hampshire and vermont from 1988–2009 suggests that average tick loads might impact moose populations through reduced fitness and fecundity of yearlings. although the field-dressed body weight of yearling cows in maine has been stable at 205 kg since 1999, it is less than the peak weight in new hampshire in 1988–1998 (217 kg, fig. 2). as a region, it is evident that productivity of yearling cows is low with cl counts probably <40% even in maine based on comparative data from new hampshire and vermont (fig. 2 and 3). new hampshire's moose population was still expanding in 1988–1998, and their physical characteristics may have peaked during this period of high resource availability related to extensive forest harvesting in the 1980s (see bontaites and gustafson 1993). their gradual decline since 1988 may reflect the combined influences of saturation of available habitat, reduced availability of preferred habitat, and gradual decline in habitat quality due to subsequent forest maturation. further, concern exists about forest regeneration in the face of dense populations in northern areas of all 3 states, and isolated examples exist (see bergeron et al. 2011); that these populations may express selflimiting impacts on habitat quality, hence fecundity, is possible. however, the steep decline in yearling body weight and that the yearling ovulation rate is well below the north american average suggests that other contributing factors exist, particularly given the relative stability of measurements in adult moose. in fact, winter ticks cause age-specific impacts because calves have higher, relative tick numbers than adults, and severe hairloss is evident on calves even in low/average tick years (samuel and barker 1979, samuel 2004, sine et al. 2009, bergeron 2011). the lack of a local epizootic of winter tick since 2002 and the declining trend in yearling physical characteristics supports the hypothesis that annual winter tick numbers affect population dynamics through reduced growth and 46 physical and reproductive status in nh – bergeron et al. alces vol. 49, 2013 fecundity of yearling moose (i.e., surviving calves). recent warmer and shorter winters that maximize spring survival and autumn questing of winter ticks presumably enhance this relationship by causing an increase in annual tick numbers, and likely increase the probability of an epizootic that produces substantial calf mortality; anecdotal reports from all 3 states suggest that a local epizootic in combination with deep snow caused high calf and yearling mortality in winter 2010–2011. the relative influences of habitat, population density, weather, and parasites on the population dynamics of moose is difficult to ascertain, and likely varies temporally. collection of long-term data sets of tick numbers and physical parameters of harvested moose in concert with annual, spring hair-loss surveys would better document the relationships between winter tick and population dynamics of moose in new hampshire. acknowledgements funding for this research was provided by the nhfg and data were available because of the dedication of nhfg biologists at harvest check stations. we are grateful to the many students from the university of new hampshire who helped at check stations through the years. we thank l. kantar of the maine department of inland fisheries and wildlife, and c. alexander of the vermont fish and wildlife department for providing comparative data. n. fortin assisted with tables and figures, and early versions of the paper. references adams, k. p., and p. j. pekins. 1995. growth patterns of new england moose: yearlings as indicators of population status. alces 31: 53–59. bergeron, d. h. 2011. assessing relationships of moose populations, winter ticks, and forest regeneration in northern new hampshire. m.s. thesis, university of new hampshire, durham, new hampshire, usa. ———, p. j. pekins, h. f. jones, and w. b. leak. 2011. moose browsing and forest regeneration: a case study in northern new hampshire. alces 47: 39–51. bontaites, k., and k. a. gustafson. 1993. the history of moose and moose management in new hampshire. alces 29: 163–167. bubenik, a. b. 1997. evolution, taxonomy, and morphology. pages 77-123 in a. w. franzmann and c. c. schwartz, editors. ecology and management of the north american moose. smithsonian institution press, washington, d.c., usa. cheatum, e. l. 1949. the use of corpus lutea for determining ovulation incidence and variation in the fertility of white-tailed deer. cornell veterinarian 39: 282–291. child, k., d. a. aitken, and r. v. rea. 2010. morphometry of moose antlers in central british columbia. alces 46: 123–134. degraaf, r. m., m. yamisaki, w. b. leak, and j. w. lanier. 1992. new england wildlife: management of forested habitats. general technical report ne-144, usda forest service, northeast experiment station, radnor, pennsylvania, usa. musante, a. r. 2006. characteristics and dynamics of a moose population in northern new hampshire. m.s. thesis, university of new hampshire, durham, new hampshire, usa. ———, p. j. pekins, and d. l. scarpitti. 2010. characteristics and dynamics of a regional moose alces alces population in the northeastern united states. wildlife biology 16: 185–204. new hampshire fish and game department (nhfg). 2009. wildlife harvest summary. new hampshire fish and game department, concord, new hampshire, usa. saether, b., and m. heim. 1993. ecological correlates of individual variation in age alces vol. 49, 2013 bergeron et al. – physical and reproductive status in nh 47 at maturity in female moose (alces alces): the effects of environmental variability. journal of animal ecology 62: 482–489. samuel, w. m. 2004. white as a ghost: winter ticks and moose. natural history series, volume 1. federation of alberta naturalists, edmonton, alberta, canada. ———. 2007. factors affecting epizootics of winter ticks and mortality of moose. alces 43: 39–48. ———, and m. barker. 1979. the winter tick, dermacentor albipictus (packard, 1869) on moose alces alces (l.) of central alberta. proceedings of the north american moose conference and workshop 15: 303–348. scarpitti, d. 2006. seasonal home range, habitat use, and neonatal habitat characteristics of cow moose in northern new hampshire. m.s. thesis, university of new hampshire, durham, new hampshire, usa. ———, c. habeck, a. r. musante, and p. j. pekins. 2005. integrating habitat use and population dynamics of moose in northern new hampshire. alces 41: 25–35. schwartz, c. c. 2007. reproduction, natality, and growth. pages 141–171 in a. w. franzmann and c. c. schwartz, editors. ecology and management of the north american moose. smith‐ sonian institution press, washington, d.c., usa. ———, and k. j. hundertmark. 1993. reproductive characteristics of alaskan moose. journal of wildlife management 57: 454–468. sergeant, d. e., and d. h. pimlott. 1959. age determination in moose from sectioned incisor teeth. journal of wildlife management 23: 315–321. sine, m., k. morris, and d. knupp. 2009. assessment of a line transect field method to determine winter tick abundance on moose. alces 45: 143–146. sperduto, d. d., and w. f. nichols. 2004. natural communities of new hampshire. new hampshire natural heritage bureau, concord, new hampshire, usa. zar, j. h. 1999. biostatistical analysis, 4th edition. prentice hall, inc., englewood cliffs, new jersey, usa. 48 physical and reproductive status in nh – bergeron et al. alces vol. 49, 2013 temporal assessment of physical characteristics and reproductive status of moose in new hampshire methods study area field measurements data analysis results females males discussion acknowledgements references alces(23)_195.pdf alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 opinions about moose and moose management at the southern extent of moose range in connecticut andrew m. labonte1, howard j. kilpatrick1, and john s. barclay2 1connecticut department of energy and environmental protection, wildlife division, 391 route 32, north franklin, connecticut 06254; 2wildlife conservation research center, university of connecticut, 1376 storrs road, unit 4087, storrs, connecticut, usa. 06269. abstract: increasing moose (alces alces) populations in the northeastern united states present new challenges for wildlife managers who must balance beneficial and adverse aspects of moose populations. it is important that managers understand stakeholder attitudes and values about moose and incorporate such into outreach and management programs. the objective of this research was to assess landowner and hunter perceptions about status, management, and concerns associated with a small moose population in connecticut. the majority of landowners and hunters correctly believed that <100 moose existed in connecticut, half believed that the population was increasing but had no opinion about appropriate size, and few had ever observed a moose in connecticut or been involved in a moose-vehicle accident (mva). landowner support for viewing areas was high and moose hunting low unless mvas increased; support for hunting moose was high among hunters. if human-moose conflicts increase, principally mvas, we expect reduced public support for the resident moose population. proactive education and management are suggested to reduce human-moose conflicts, mvas, and increase acceptance of hunting as a possible population management tool. alces vol. 49: 83–98 (2013) key words: alces alces, connecticut, moose, human dimensions, survey. moose (alces alces) populations have increased throughout northern new england over the past 30 years presenting management challenges to balance their beneficial and potentially adverse aspects (wattles and destefano 2011). moose provide intrinsic economic value to both consumptive and non-consumptive users (schwartz and bartley 1991), with watching and hunting as major revenue generators (wolfe 1987, timmermann and rodgers 2005), especially in northern new england. populations reaching levels sufficient for recreational opportunities may result in higher levels of adverse consequences in the form of moose-vehicle accidents (mvas) and ecological damage (mirick 1999, timmermann and rodgers 2005), although such conflicts can also occur in small populations and suburban areas (mcdonald et al. 2012). assessing attitudes of various stakeholder groups toward a wildlife species is useful to understand societal support and opposition about current and potential management decisions (bath and enck 2003), and importantly, incorporating stakeholder attitudes into outreach and management programs (teel et al. 2002). natural resource agencies increasingly emphasize stakeholder participation in decision-making (lauber and knuth 1997) and management of human-wildlife interactions (ericsson 2003) to implement plans (flanigan 1987, hartig and thomas 1988, pinkerton 1991, landre and knuth 1993), strengthen public relationships (landre and knuth 1993), and reduce conflict (erickson 1979, twight and patternson 1979, nelkin 1984, blahna and yontsshepherd 1989). 83 relative to other big game species, human dimensions (hd) research with moose was initially limited in north america (wolfe 1987). an evaluation of articles from 1974-2001 in alces indicated that the majority of early hd research pertained to hunting of moose or mvas, with less attention to public values and attitudes towards moose (ericsson 2003). in the northeastern united states, states have used hd research to evaluate public opinion about their initial moose management programs in vermont (alexander 1993), new hampshire (donnelly and vaske 1995), and new york (lauber and knuth 1997, 1999). relative to nonconsumptive recreation in new hampshire, silverberg et al. (2001) measured knowledge level, attitudes, and motivation of wildlife viewers at a moose viewing site. and recently, hd research was used to evaluate public opinions about moose to provide effective educational strategies to reduce human-moose conflicts, principally mvas, in prince george, british columbia (mcdonald et al. 2012). however, similar hd information is non-existent in connecticut, the southern extent of moose range in new england. although the potential for moose populations to continue expanding in connecticut is unclear, developing management strategies and programs that are both effective and acceptable to the public is important relative to managing human-moose conflicts. our objective was to survey landowners and hunters about the status, management, and associated concerns with the moose population in connecticut. study area and background connecticut (12,548 km2) was the fourth most densely populated area (3,500,000 people, 278 people/km2) in the united states at the time of this research (connecticut economic resource center 2006, 2010). located in southern new england, it is bounded on the south by long island sound, and by the states of rhode island to the east, massachusetts to the north, and new york to the west. connecticut is about half forested (55.6%), 20% developed or barren, 16.7% turf, grass, or agricultural field, 4.4% wetlands (non-forested, forested, and tidal), and 3.2% water (hochholzer 2010). historic accounts suggest that moose existed in connecticut prior to the 18th century (trumbull 1797, deforest 1964); however, goodwin (1935) noted that at the beginning of the 18th century there was no record of moose in connecticut. further, the lack of archaeological deposits of moose suggests that they likely existed in low numbers, if at all (n. bellantoni, connecticut state archeologist, pers. commun.). a few reports of transient moose occurred between 1916 and 1956 (connecticut wildlife 2000), and on 18 september 1956, the board of fisheries and game (currently the department of energy and environmental protection, deep) passed an emergency regulation that gave full protection to moose in connecticut. sporadic reports of moose occurred until the early 1990s (kilpatrick et al. 2003), and in 1992 the deep began documenting all credible sightings and mvas. in 1996 a question was added to the annual deer hunter questionnaire asking them to report any moose observation during the deer (odocoileus virginianus) hunting season. in 1998, the wildlife division of deep adopted a directive (deep2431-d1) that outlined procedures for responding to problem moose situations that included hazing, capture and relocation, and euthanasia. since 2000, observations of cows with calves confirmed the establishment of a small resident population (kilpatrick et al. 2003). an empirical model based on public sightings of moose reported to the deep conservatively estimated the population at ∼64 in 2004 (labonte and kilpatrick 2006) with ∼75 present at the 84 opinions about moose – labonte et al. alces vol. 49, 2013 time of this survey in 2008 (labonte 2011). despite low moose numbers, connecticut was experiencing 2–4 mvas annually (deep, unpublished data) and deep staff were exploring options to implement a moose management strategy to address increasing mvas. however, it was unknown if the general public or hunting community would support a management strategy that included moose hunting given the minimal population. understanding public and hunter opinions about moose and moose management is essential for developing an effective moose management plan in connecticut. methods based on the distribution of moose sightings by the public (kilpatrick et al. 2003), hunters (labonte et al. 2008), and reported mvas (deep, unpublished data) (fig. 1), northern connecticut was selected as the study area for the landowner survey. based on geographic features and an assessment of human population densities, towns in northern connecticut were delineated into 3 groups for the landowner survey (fig. 1) and were used for landscape level comparisons. towns were grouped as central (n = 13), eastern (n = 16), and western (n = 20) (table 1, fig. 1). landowner survey a database containing the names and addresses of landowners in the 49 study towns was obtained from municipal town offices. we set a sampling rule to include private landowners and removed all identifiable outliers (e.g., limited liability companies, corporations, companies, schools, churches, trustees, towns). we deleted duplicate landowner records (i.e., multiple ownerships) to compile a list of landowners with an equal likelihood of being randomly selected and receiving a single survey. we calculated minimum sample sizes required for each landscape based on a stratified random sampling approach (scheaffer et al. 1996). a mail survey was chosen because it can include complex questions, access geographically dispersed groups, and recipients can reply at their convenience with low potential for social desirability bias (decker et al. 2001). we used a 3-wave survey with a variation of the repeated mailing technique (dillman 1978). surveys were mailed to randomly selected landowners stratified among the 3 landscapes (eastern, central, western) in january 2008; 2 follow-up surveys were mailed to non-respondents about 4–8 weeks apart. after 3 attempts to contact landowners by mail, we contacted a sub-sample of nonrespondents by telephone to assess nonresponse bias. we used likert-scale questions in the surveys (likert-scale numbers indicated by each response were used to calculate mean response scores) to assess beliefs and experiences about wildlife (5-point scale), concerns about moose, support for hunting (5point scale), and acceptability of situations involving moose (6-point scale). there were 3 general types of questions with 3 response categories: 1) landowner beliefs and experiences (agree, neutral, disagree), 2) landowner opinions about management (support, neutral, oppose), and 3) landowner concerns (acceptable, not acceptable/no action, not acceptable/action). the study protocol and survey were reviewed and approved by the connecticut wildlife division, the northeast wildlife damage management cooperative, and the institutional review board (irb), office of research compliance at the university of connecticut; the irb chair deemed the survey exempt from irb status. surveys were conducted in accordance with federal guidelines in which minors (<18 years of age) were excluded, results were not identifiable to individuals, and surveys involved no risk to individuals. alces vol. 49, 2013 labonte et al. – opinions about moose 85 fig. 1. the study area was in the state of connecticut located in the northeastern portion of the united states. landowner surveys were conducted in 2008 in the northern portions of connecticut (shaded areas) where most moose-vehicle accidents (•) occurred, while hunter surveys were conducted at town halls ( ) located throughout the state. 86 opinions about moose – labonte et al. alces vol. 49, 2013 hunter survey we selected 31 of 169 (18%) town clerks to distribute the survey to any resident or non-resident hunter purchasing a connecticut firearms hunting license or combination hunting/fishing license; towns and sampling period were selected based on the highest volume of hunting license sales in 2004. surveys were distributed during 3 sampling periods (january, april, and october 2008) which were chosen to obtain a representative sample of each hunter group; many hunters purchase a license to pursue game during a specific season and the timing coincided with peak issuance. packets containing an instruction letter, return envelope, and specific number of surveys were mailed to the town clerks before each sample period; the number of surveys per town was based upon the respective volume of 2004 license sales. town clerks were instructed to provide a survey to every other individual that purchased a hunting or combination hunting/ fishing license; upon completion, they collected the survey and mailed all after each sampling period. we generated questions to evaluate hunting activity, participation in outdoor-related activities, and perceptions and opinions about connecticut's moose population. we used a 5-point likert-scale question to assess support for hunting and grouped responses into 3 categories: support, neutral, or oppose. the review and approval of the study protocol and survey were identical to the landowner survey. analysis we treated ordinal-level (likert scale) data as interval-level data as previous studies have validated the use of such data in survey research (nunnally and bernstein 1994, zinn and andelt 1999, daley et al. 2004). we calculated levene's test (p < 0.05) for equality of variances and the kolmogorov-smirnov test of normality; based on these results we used the kruskal-wallis test (p < 0.05) for all analysis at the landscape level, and the mann-whitney u test (p < 0.05) for comparisons between landowners and hunters. pearson chi-square tests (p < 0.05) were used to examine nominal level variables and compare responses between respondents and non-respondents. all analyses were conducted using systat 12.0 (systat 2007). results respondent demographics landowner survey — surveys were returned from 622 of 2,023 landowners (35.7% eastern, 31.3% central, 37.9% western); proportionally, 66% from the first, 20% from the second, and 14% from the final mailing. there was no difference among landscapes in gender (χ2 = 3.44, p < 0.178) and age of respondents (χ2 = 0.410, p < 0.999); 56.4% were male and the mean age of all respondents was 54.4 (sd = 14.7) years. after 3 attempts by mail, we contacted 51 non-respondents by telephone to assess non-response bias for specific questions. hunting was not commonly allowed in any landscape but was higher in the western (16%, χ2 = 13.6, p < 0.001) and eastern (14.7%, χ2 = 20.3, p < 0.001) than the central landscape (3%). table 1. human densities and landscape level (eastern, central, western) characteristics in connecticut, 2008. location eastern central western number of towns 16 13 20 population 79/km2 185/km2 71/km2 % forest 65.4 29.8 67.9 % commercial/ residential 14.2 43.2 11.7 % turf/agriculture 12.4 21.1 12.6 % wetlands 4.6 2.8 3.8 % water 2.3 1.8 3.3 % other 1.1 1.3 0.7 alces vol. 49, 2013 labonte et al. – opinions about moose 87 hunter survey — surveys were completed by 446 of 485 hunters (91.9%) and due to this high response rate, we did not assess non-response bias. gender of hunters was primarily male (97.6%) and the mean age was 48.1 (sd = 12.5) years; the majority had harvested deer (65.2%) and a few bear (7.0%) and moose (3.6%). the majority (>60%) would participate in nonconsumptive moose recreation (watching, photography) and half (50.8%) would hunt moose (table 2). landowner beliefs and experiences with wildlife most landowners believed that wildlife and management were important, and the mean response scores were similar across all landscape levels except for huntingrelated questions. in general, the majority of landowners were not unsupportive of hunting, but 30-60% were neutral/or disagreed with some aspect of hunting (table 3). knowledge about moose landowner survey — landowners were asked to identify the moose from 3 sketches depicting a deer, moose, and bear. responses were combined as no differences existed among landscapes (χ2 = 1.562, p = 0.458); most (90.3%) correctly selected the image of the moose with the remainder selecting the deer. respondent and nonrespondent opinions about the number of moose existing in connecticut were not different (χ2 = 2.316, p = 0.128) and no adjustments were made. all responses were combined because no differences (χ2 = 4.315, p = 0.634) among landscapes existed in perceptions about how many moose exist in connecticut. the majority (64%) correctly estimated that there were <100 moose in connecticut, and >90% estimated <500 moose (table 2). landowner-hunter comparisons — a similar proportion of landowners (63.9%) and hunters (67.4%) believed that <100 moose existed in connecticut (χ2 = 1.31, p = 0.253) (table 2). more hunters (27.7%) than landowners (18.5%) believed that <10 moose existed in connecticut (χ2 = 11.9, p = 0.001); although both were <10%, conversely, more landowners than hunters believed that >500 moose existed (χ2 = 8.6, p = 0.003). the primary source of information influencing opinions about the size of the moose population was from other table 2. landowner and hunter opinions about the moose population in connecticut, usa, 2008. lower case n refers to # of respondents. percent of respondents survey question landowner hunter number of moose (n) 590 408 0 3.0 6.9 <10a 18.5 27.7 <100a 63.9 67.4 100–499 28.0 29.0 >500 8.0 3.5 status of moose population (n) 606 430 increasing 51.8 67.6 decreasing 7.8 <1.0 stable 10.0 11.6 no opinion 30.4 20.0 opinion of moose population (n) 604 427 too high 3.0 3.9 too low 25.9 40.6 just right 15.7 19.2 no opinion 54.9 36.1 activities would participate in (n) 626 404 watching moose 62.1 33.8 photographing moose 50.7 27.5 hunting moose 10.7 50.8 other 2.0 1.0 none 20.0 19.0 aincludes all respondents who indicated 0 or <10. 88 opinions about moose – labonte et al. alces vol. 49, 2013 table 3. landowner beliefs and experiences about wildlife in connecticut, usa, 2008. beliefs and experiences about wildlife % responsea agree neutral disagree no opinion mean response scoresb c e w c e w c e w c e w c e w hc pc n i notice birds and wildlife around me daily 98 99 96 1 0 2 1 1 1 0 0 1 1.65 1.77 1.70 4.60 0.101 626 observing and learning about wildlife is important to me 88 92 89 10 7 7 2 1 3 0 0 1 1.34 1.47 1.40 3.12 0.210 624 hunting animals for any purpose should not be permitted 19 12 22 22 16 12 58 71 65 2 2 1 −0.54 −0.89 −0.68 7.66 0.022 623 it is important to manage some wild animal populations 84 86 86 9 5 9 6 8 4 1 1 1 1.08 1.16 1.18 3.09 0.214 622 wild animal populations should be managed for the benefit of all people 68 69 74 16 16 13 14 13 13 1 3 1 0.78 0.84 0.81 0.38 0.826 620 participation in hunting helps people appreciate wildlife and natural processes 36 53 44 23 23 15 36 22 37 4 3 4 −0.01 0.40 0.03 8.62 0.013 623 if wildlife populations are abundant, it is ok to use them as a natural renewable resource 53 65 55 22 15 24 19 17 17 6 3 5 0.45 0.71 0.49 5.13 0.077 613 regulated hunting is an acceptable use of a natural resource 65 76 69 15 9 12 16 13 17 4 2 3 0.63 0.94 0.70 9.88 0.007 621 c = central, e = eastern, w = western. alikert scale ranged from −2 (“strongly disagree”) to 2 (“strongly agree”). to evaluate percentages, responses were truncated into “agree, neutral, disagree.” bnot included in analysis are the number of respondents who choose “no opinion.” ch and p values for kruskal-wallis test statistic comparison between eastern, central, and western groups. a l c e s v o l . 4 9 , 2 0 1 3 l a b o n t e e t a l . – o p in io n s a b o u t m o o s e 8 9 sources (33.1%) for landowners and personal experience (37%) for hunters. opinions about moose landowner survey — respondent and non-respondent opinions about the status of connecticut's moose population (χ2 = 5.997, p = 0.112) and the number of moose in connecticut (χ2 = 6.374, p = 0.095) were not different and no adjustments were made. no difference among landscapes (χ2 = 0.835, p = 0.659) existed between the proportion believing that the moose population was either increasing (about half) or decreasing, or that believed it was too high (∼3%) or too low (∼25%) (χ2 = 2.71, p = 0.257); therefore, responses were combined (table 2). likewise, no landscape differences existed (χ2 = 2.68, p = 0.262) and responses were combined for the proportion of landowners (∼70%) who would support designating viewing areas for moose watching. landowner-hunter comparisons — about half (51.8%) of landowners and 2/3 of hunters believed that connecticut's moose population was increasing, but few (3 and 4%, respectively) believed the population was too high (table 2). more hunters (68%) than landowners (52%) believed that the population was increasing, and fewer that it was decreasing (χ2 = 33.1, p <0.001). there was no difference (χ2 = 0.559, p = 0.455) in the proportion of landowners and hunters who believed that the population was too low or too high; although, measurably more hunters thought the population was too low (40.6 vs. 25.9%; table 2). from a list of 3 potential moose-related activities if moose were common in connecticut, landowners favored watching and photography (62.1 and 50.7%), and hunters favored hunting and watching (50.8 and 33.8%). the proportion of landowners and hunters who would participate in watching (χ2 = 60.8, p < 0.001), photographing (χ2 = 41.9, p < 0.001), or hunting moose (χ2 = 247.6, p < 0.001) was different. the proportion of landowners and hunters who would not participate in any moose activity was similar (∼20%; χ2 = 0.057, p < 0.811) (table 2). interactions with moose landowner survey — a minority (15%) of landowners observed moose in 29 towns and differences existed among landscapes (χ2 = 14.3, p = 0.001). twice as many landowners observed moose in western (27.0%) than central (12.0%, χ2 = 13.6, p < 0.001) and eastern landscapes (12.6%, χ2 = 6.07, p = 0.014) which were not different (χ2 = 0.031, p = 0.860) (table 4). an additional 51 landowners reported moose tracks or other sign with the same landscape differences (χ2 = 13.3, p = 0.001); more moose tracks and sign were observed in western (21.8%) than in central (7.8%, χ2 = 13.2, p < 0.001) and eastern (10.0%, χ2 = 3.99, p = 0.046) landscapes which were not different (χ2 = 0.464, p = 0.496) (table 4). only landowners in western landscapes had been in a mva (n = 4) in connecticut. although the rate of mva experiences in any landscape was low over all (<5%), landscape differences existed (χ2 = 8.29, p = 0.016) (table 4). landowners in western landscapes (4.9%) were in more mvas than those in central landscapes (<1.0%, χ2 = 7.45, p = 0.006); there was no difference between western and eastern (1.0%, χ2 = 2.71, p = 0.100) or eastern and central (χ2 = 0.001, p = 0.979) landscapes (table 4). hunter survey — moose were observed by 20% of hunters (n = 91) in 36 towns, with 71 others observing tracks or scat in 14 towns where sightings occurred, as well as in 13 other towns. landowner concerns about moose landowner concerns were not different among landscapes regarding health, safety, 90 opinions about moose – labonte et al. alces vol. 49, 2013 or damage-related issues (h = 0.059–2.115, 0.742 >p >0.347), and were combined for analysis (table 5). the majority were only concerned about mva, with <20% very concerned (table 5). moose population management landowner survey — responses were combined because mean scores were not different among landscapes (h = 1.44–5.59, 0.487 > p > 0.061) for any scenario regarding moose population management (table 6). a minority (31%) of landowners supported hunting as a method to control moose populations in connecticut based on their current level of concern; their support was highest if hunting was carefully regulated and controlled by the state, or if the moose population and number of mvas were increasing (both 54%). conversely, the vast majority of hunters (83-88%) supported hunting under all scenarios (table 6). the proportion of landowners and hunters who supported hunting was different “if it was carefully regulated and controlled by the state” (u = 53,194, χ2 = 211.53, table 4. landowner interactions with moose in connecticut, usa, 2006–2007. moose-human interactions % response yes no c e w c e w χ2 pa observed moose 12.0 12.6 27.0 88.0 87.4 73.0 14.30 0.001 in yard 1.5 4.6 5.3 98.5 95.4 94.7 5.56 0.062 outside yard 3.8 3.4 13.7 96.2 96.6 86.3 13.55 0.001 crossing road 5.8 2.3 13.7 94.2 97.7 86.3 9.88 0.007 other 3.5 5.7 5.3 96.5 94.3 94.7 1.26 0.531 observed tracks/scat 7.8 10.0 21.8 92.2 90.0 78.2 13.30 0.001 moose-vehicle accident 1.0 (0.0b) 1.0 (0.0b) 4.9 (4.0b) 99.0 99.0 95.1 8.29 0.016 e = eastern (n = 87), c = central (n = 343), w = western (n = 95). aχ2 and p values for pearson chi-square comparison between eastern, central, and western groups. bmva reported just in connecticut. table 5. landowner concerns about moose interactions in connecticut, usa, 2008. concerns about moose % response mean response scoresa hb pb no concern some concern very concerned no opinion encountering a moose 67.4 24.9 4.0 3.7 1.47 1.263 0.532 the cost of residential property damage caused by moose 57.2 30.9 4.9 7.1 1.61 2.115 0.347 being injured in a motor vehicle accident that involves a moose 28.0 50.7 18.6 2.8 2.33 1.385 0.500 potential problems that moose may cause to the ecosystem 52.5 31.3 4.9 11.3 1.66 0.596 0.742 overall current level of concern related to moose 57.3 34.6 3.4 4.7 1.58 0.662 0.718 alikert scale ranged from 1 (“not concerned”) to 4 (“very concerned”). to evaluate percentages, “slightly concerned” and “somewhat concerned” responses were truncated into “some concern.” bh and p values for kruskal-wallis test statistic comparison between eastern, central, and western groups. alces vol. 49, 2013 labonte et al. – opinions about moose 91 table 6. landowner and hunter opinions about managing moose populations using hunting in connecticut, usa, 2008. concerns about moose % response mean response scoresa support neutral oppose hb pb uc pc χ2 land hunt land hunt land hunt land hunt land land based on your current level of concern? 31 na 29 na 40 na −0.23 2.05 0.35 if your level of concern increases? 47 na 25 na 29 na 0.20 3.98 0.13 if hunting were carefully regulated and controlled by the state? 54 88 22 6 24 6 0.34 1.41 2.82 0.24 53,194 0.00 211.5 if you knew that the moose population would be maintained at its current level? 41 83 30 8 29 9 0.09 1.23 2.69 0.26 49,524 0.00 206.2 if you knew that hunting is currently allowed in other new england states? 41 na 30 na 29 na 0.10 5.59 0.06 if you knew the likelihood of a human fatality was greater d? 54 85 26 8 21 7 0.44 1.37 1.44 0.48 18,731 0.00 268.0 alikert scale ranged from −2 (“strongly oppose”) to 2 (“strongly support”). to evaluate percentages, “strongly support” and “support” were truncated into “support,” and “oppose” and “strongly oppose” were truncated into “oppose.” bh and p values for kruskal-wallis test statistic comparison between eastern, central, and western groups. cu and p values for mann-whitney u test between landowners and hunters. dif you knew the likelihood of a human fatality was greater for a moose-vehicle accident than a deer-vehicle accident and that the moose population and number of moosevehicle accidents were increasing in connecticut? na = not asked on survey. 9 2 o p in io n s a b o u t m o o s e – l a b o n t e e t a l . a l c e s v o l . 4 9 , 2 0 1 3 p < 0.001), “if they knew that the moose population would be maintained at its current level” (u = 49,524, χ2 = 206.22, p < 0.001), and “if the moose population and number of mvas was increasing in connecticut” (u = 18,731, χ2 = 268.01, p < 0.001) (table 6). the range of responses was evenly distributed (13–18% per response, n = 6 responses) for those either primarily supporting or opposing hunting moose in connecticut (table 7). in general, those supporting hunting of moose wanted to either hunt moose or linked human-moose conflicts with need for hunting. conversely, those opposed to hunting moose were either unsupportive of hunting or believed that the population/conflict rate was too low (table 7). landowner opinions about roadside sightings and moose-vehicle accidents no differences existed at the landscape level in opinions about roadside sightings (h = 3.7–5.8, 0.15 > p > 0.054), mvas (h = 0.61–2.8, 0.23 > p > 0.73), or fatalities resulting from a mva (h = 2.2 – 3.0, p > 0.22). the proportion of landowners who deemed “it not acceptable and some action should be taken” increased substantially in all categories if the overall problem of mvas rose (table 8). discussion although few landowners hunted or permitted hunting on their property, observing and learning about wildlife was important to most landowners and they were supportive of designating viewing areas for moose. hunting activity, beliefs, and experiences with wildlife if hunting was involved, and direct interactions with moose and mvas were influenced by landscape. but, knowledge, opinions about moose and moose management, and concerns about moose were similar across landscapes despite landscape differences in moose experiences, albeit experiences were low (<20%) in all landscapes. we found that landowner and hunter knowledge about moose abundance was limited, as in massachusetts 20 years ago (vecellio et al. 1993). a small number (<50) of landowners and hunters combined believed no moose existed in connecticut. the main source of information about moose for landowners was from non-deep sources, table 7. landowner responses regarding reasons why they primarily supported or opposed hunting to control moose populations in connecticut, usa, 2008. primarily supported hunting n % respondents regulated hunting is a legitimate method to control moose population growth 306 18.1 moose threaten human safety 254 15.1 deep officials are well trained to handle problems associated with moose 252 14.9 moose population is too high or may become too high 244 14.5 moose cause damage to crops or property 244 14.5 want the opportunity to hunt moose 222 13.2 don't know 101 6.0 other 63 3.7 primarily opposed to hunting moose are not a threat to human safety at their current level 211 16.3 moose do not cause enough damage to warrant management 205 15.8 moose population is too low and does not warrant management 198 15.3 do not support hunters killing moose 190 14.6 disagree with hunting 181 14.0 do not support deep killing moose 176 13.6 do not know 85 6.6 other 51 3.9 alces vol. 49, 2013 labonte et al. – opinions about moose 93 table 8. landowner opinions about roadside sightings and moose-vehicle accidents in connecticut, usa, 2008. % response acceptable not acceptable/no action not acceptable/ action mean response scoresa concerns about moose c e w c e w c e w c e w hb pb a moose is on or near a busy highway occasionally 35.6 39.2 23.7 13.9 13.4 19.6 50.5 47.4 56.7 3.31 3.29 3.60 3.742 0.154 moose are frequently on or near busy highways 14.6 19.8 10.2 10.9 14.6 9.2 74.5 65.6 80.6 4.13 4.01 4.36 5.837 0.054 1 moose-vehicle collision occurs each year statewide 38.1 34.4 31.6 21.5 36.5 34.7 40.4 29.2 33.7 3.16 3.05 3.11 0.618 0.734 2-5 moose-vehicle collisions occur each year statewide 26.5 26.6 20.4 15.5 18.1 18.4 58.0 55.3 61.2 3.80 3.78 3.93 1.009 0.604 6-10 moose-vehicle collisions occur each year statewide 18.1 21.3 10.5 15.4 7.9 14.7 66.5 70.8 74.7 4.14 4.26 4.40 2.878 0.237 >10 moose-vehicle collisions occur each year statewide 13.2 17.8 9.4 12.4 7.8 8.3 74.4 74.4 82.3 4.39 4.49 4.68 2.746 0.253 a human fatality results from a motorist hitting a moose in connecticut 16.7 20.0 10.5 21.0 24.4 23.2 62.4 55.6 66.3 4.08 3.82 4.23 2.964 0.227 2-5 human fatalities result from a motorist hitting a moose in connecticut 10.8 13.3 6.3 14.2 10.0 10.4 75.0 76.7 83.3 4.52 4.56 4.69 3.069 0.216 e = eastern, c = central, w = western. alikert scale was 1 (“acceptable”), 2 (“not acceptable/no management action taken”), 3 (“not acceptable/action should be taken”). bh and p values for kruskal-wallis test statistic comparison between eastern, central, and western groups. 9 4 o p in io n s a b o u t m o o s e – l a b o n t e e t a l . a l c e s v o l . 4 9 , 2 0 1 3 whereas hunters were most influenced by personal experience and deep communications. it is not surprising that ∼25% of landowners and hunters believed <10 moose existed (table 2), because few ever observe a moose in connecticut. many landowners and hunters had no opinion about the moose population status (20–30%) or how many moose should exist in connecticut (35–55%) (table 2). the low response rate probably reflects their lack of experience, familiarity, and interest in moose. riley and decker (2000) also found a large portion of people lacked opinions about cougars in montana, presumably for the same reasons. they suggested that lack of opinion may indicate 1) a lack of general concern in the everyday lives of residents, 2) stakeholder perceptions that managers do not listen to stakeholder concerns, or 3) distrust in delegation of decision-making to managers. overall, the majority of landowners had few concerns about moose except mvas. less than half supported using hunting as a method to control moose populations in connecticut based on their current knowledge of population levels, as opposed to hunters who were strongly supportive. more than half of landowners were supportive of moose hunting if it was carefully regulated and controlled by the state. although all forms of hunting are controlled by state fish and wildlife agencies, kilpatrick et al. (2007) found that landowners often are unaware of regulations or requirements that govern wildlife resources and expressed increased support for certain regulations or requirements although they already existed. predictably, if the number of roadside sightings, mva, or the number of related human fatalities increased, the proportion finding such unacceptable also increased. given that the first reported mva in connecticut occurred in 1995 and the annual rate remains low (2.3 mva per year), it is not surprising that residents are minimally concerned about moose. overall, few landowners (<1%) had ever been involved in a mva in connecticut. if the frequency of moose sightings along roads increases substantially, support for controlling moose populations will presumably increase regardless of the number of mvas or human fatalities. about 50% believed that a moose near a busy highway was unacceptable requiring action, and 58% believed action was required at 2–5 mva per year, the current reported mva rate (table 7). a similar situation occurred with elk (cervus elaphus) in urban areas of flagstaff, arizona (lee and miller 2003), where most respondents were concerned about being in an automobile accident involving an elk or seeing one along a roadside. the collective ability for humans to accept the presence and consequences of any wildlife species will eventually define the wildlife acceptance capacity (wac) for that species (decker and purdy 1988). in anchorage, alaska where moose populations exceed habitat carrying capacity (wac is either higher or lower), only half of residents supported moose hunting (whittaker et al. 2001). in british columbia, mcdonald et al. (2012) found that most respondents suggested reducing attractants or relocating moose to alleviate moosehuman conflicts, presumably over hunting, however sample size was small (n <100). acceptance of hunting among certain stakeholders may be influenced more by basic beliefs about hunting which are based on fundamental value orientations toward use or protection of wildlife (fulton et al. 1996, zinn et al. 1998). in connecticut, because moose are of such low numbers and few residents have any direct experience with moose, an associated wac is probably not measurable or is exceedingly high. we expect a reduction in wac if moosehuman conflicts increase measurably and alces vol. 49, 2013 labonte et al. – opinions about moose 95 advocate for a proactive management strategy that would increase public education about moose, mvas, and the potential role of hunting to help protect human safety. educational efforts should improve public awareness through posted warnings about local moose on department of transportation variable message boards (vmbs), erecting moose-crossing signs in appropriate areas, and meeting with stakeholder groups. the effectiveness of vmbs and signs to reduce mvas is unknown, but they should alert drivers otherwise unaware about moose in connecticut. a multi-faceted strategy should increase public awareness and education about moose in connecticut and aid in developing a long-term moose management program beyond simply minimizing mvas. acknowledgements we thank g. chasko, p. curtis, j. enck, m. gregonis, m. ortega, d. may, and r. ricard for reviewing surveys and drafts of this manuscript, and j. brooks, t. goodie, p. lewis, t. muni, and a. ocampo for assisting with data collection. this project was supported by the university of connecticut, college of agriculture and natural resources, department of natural resources and the environment, wildlife conservation research center, the northeast wildlife damage management cooperative, and the connecticut department of energy and environmental protection, wildlife division, federal aid in wildlife restoration project 49-35. references alexander, c. e. 1993. the status and management of moose in vermont. alces 29: 187–195. bath, a. j., and j. w. enck. 2003. wildlifehuman interactions in national parks in canada and the usa. national park service research review 4: 1–32. blahna, d. j., and s. yonts-shephard. 1989. public involvement in resource planning: toward bridging the gap between policy and implementation. society and natural resources 2: 209–227. connecticut economic resource center. 2006. state-by-state analysis. rocky hill, connecticut, usa. <http://www. cerc.com/content/resources.asp> (accessed december 2010). ———. 2010. connecticut cerc state profile. rocky hill, connecticut, usa. <http://www.cerc.com/townprofiles/ customer-images/connecticut.pdf> (accessed december 2010). connecticut wildlife. 2000. wildlife management through the century. connecticut wildlife 20(6): 14–17. daley, s. s., d. t. cobb, p. t. bromley, and c. e. sorenson. 2004. landowner attitudes regarding wildlife management on private land in north carolina. wildlife society bulletin 32: 209–219. decker, d. j., t. l. brown, and w. f. siemer. 2001. human dimensions of wildlife management in north america. the wildlife society, bethesda, maryland, usa. ———, and k. g. purdy. 1988. toward a concept of wildlife acceptance capacity in wildlife management. wildlife society bulletin 16: 523–527. deforest, j. w. 1964. history of the indians of connecticut from the earliest known period to 1850. archon books, hamden, connecticut, usa. dillman, d. a. 1978. mail and telephone surveys: the total design method. wiley –interscience, new york, new york, usa. donnelly, m. p., and j. j. vaske. 1995. predicting attitudes towards a proposed moose hunt. society and natural resources 8: 307–319. erickson, p. a. 1979. the role of the public. pages 309–321 in p. a. erickson, editor. environmental impact assessment: 96 opinions about moose – labonte et al. alces vol. 49, 2013 http://www.cerc.com/content/resources.asp http://www.cerc.com/content/resources.asp http://www.cerc.com/townprofiles/customer-images/connecticut.pdf http://www.cerc.com/townprofiles/customer-images/connecticut.pdf principles and applications. academic press. new york, new york, usa. ericsson, g. 2003. of moose and man: the past, the present, and the future of human dimensions in moose research. alces 39: 11–26. flanigan, f. h. 1987. public involvement in the chesapeake bay, usa program. pages 512–517 in s. k. majumdar, l. w. hall, and h. m. austin, editors. contaminant problems and management of living chesapeake bay resources. pennsylvania academy of science. easton, pennsylvania, usa. fulton, d., m. j. manfredo, and j. lipscomb. 1996. wildlife value orientations: a conceptual and measurement approach. human dimensions of wildlife 1: 24–47. goodwin, g. 1935. the mammals of connecticut. state geological and natural history survey, bulletin 53. hartford, connecticut, usa. hartig, j. h., and r. l. thomas. 1988. development of plans to restore degraded areas in the great lakes. environmental management 12: 327–347. hochholzer, h. 2010. connecticut's forest resource assessment and strategy; building a better tomorrow for connecticut's forests today. department of environmental protection, bureau of natural resources, forestry division, hartford, connecticut, usa. kilpatrick, h. j., r. riggs, a. m. labonte, and d. celotto. 2003. history and status of moose in connecticut 2002. bureau of natural resources, wildlife division, department of environmental protection, hartford, connecticut, usa. ———, a. m. labonte, and j. s. barclay. 2007. acceptance of deer management strategies by suburban landowners and bowhunters. journal of wildlife management 71: 2095–2101. labonte, a.m. 2011. an assessment of moose (alces alces american) and moose management in connecticut. ms thesis, university of connecticut, storrs, connecticut, usa. ———, and h. j. kilpatrick. 2006. moose (alces alces) in connecticut: establishment, expansion, and future expectations. proceedings of the 63rd annual northeast fish and wildlife conference. groton, connecticut, usa. ———, ———, and w. reid. 2008. 2007 connecticut deer summary report. department of environmental protection, bureau of natural resources, wildlife division, hartford, connecticut, usa. landre, b. k., and b. a. knuth. 1993. the role of agency goals and local context in great lakes water resources public involvement programs. environmental management 17: 153–165. lauber, t. b., and b. a. knuth. 1997. fairness in moose management decisionmaking: the citizen's perspective. wildlife society bulletin 25: 776–787. ———, and ———. 1999. measuring fairness in citizen participation: a case study of moose management. society and natural resources 11: 19–37. lee, m. e., and r. miller. 2003. managing elk in the wildland-urban interface: attitudes of flagstaff, arizona residents. wildlife society bulletin 31: 185–191. mcdonald, a. m. h., r. v. rea, and g. hesse. 2012. perceptions of moosehuman conflicts in an urban environment. alces 48: 123–130. mirick, p. g. 1999. back in the company of moose. massachusetts wildlife 49: 17–25. nelkin, d. 1984. approaches to participation in natural resource decisions. pages 175–182 in c. w. churchman, a. h. rosenthal, and s. h. smith, editors. natural resource administration. westview press, boulder, colorado, usa. nunnally, j. c., and i. h. bernstein. 1994. psyclandtric theory. third edition. mcgraw-hill, new york, new york, usa. alces vol. 49, 2013 labonte et al. – opinions about moose 97 pinkerton, e. 1991. locally based water quality planning contributions to fish habitat protection. canadian journal of fish and aquatic science 48: 1326–1333. riley, s. j., and d. j. decker. 2000. wildlife stakeholder acceptance capacity for cougars in montana. wildlife society bulletin 28: 931–939. scheaffer, r. l., w. mendenhall iii, and l. ott. 1996. elementary survey sampling. fifth edition. duxbury press, wadsworth publishing company inc, raritan, new jersey, usa. schwartz, c. c., and b. bartley. 1991. reducing incidental moose mortality: considerations for management. alces 27: 227–231. silverberg, j. k., p. j. pekins, and r. a. robertson. 2001. impacts of wildlife viewing at dixville notch wildlife viewing area. proceedings of the 2001 northeast recreation research symposium, bolton landing, new york, usa. systat software. 2007. systat 12.0 statistics iii. systat software, inc., san jose, california, usa. teel, t. l., r. s. krannich, and r. h. schmidt. 2002. utah stakeholders' attitudes toward selected cougar and black bear management practices. wildlife society bulletin 30: 2–15. timmermann, h. r., and a. r. rodgers. 2005. moose: competing and complementary values. alces: 85–120. trumbull, b. d. 1797. a complete history of connecticut, civil and ecclesiastical, from the emigration of its first planters from england, in 1630 to 1713. hudson and goodwin, hartford, connecticut, usa. twight, b. w., and j. j. patterson. 1979. conflict and public involvement: measuring consensus. journal of forestry 77: 771–776. vecellio, g. m., r. d. deblinger, and j. e. cardoza. 1993. status and management of moose in massachusetts. alces 29: 1–7. wattles, d. w., and s. destefano. 2011. status and management of moose in the northeastern united states. alces 47: 53–68. whittaker, d., m. j. manfredo, p. j. fix, r. sinnott, s. miller, and j. j. vaske. 2001. understanding beliefs and attitudes about an urban wildlife hunt near anchorage, alaska. wildlife society bulletin 29: 1114–1124. wolfe, m. l. 1987. an overview of the socioeconomics of moose in north america. swedish wildlife research supplement 1: 659–675. zinn, h. c., and w. andelt. 1999. attitudes of fort collins, colorado residents toward prairie dogs. wildlife society bulletin 27: 1098–1106. ———, m. j. manfredo, j. j. vaske, and k. wittmann. 1998. using normative beliefs to determine the acceptability of wildlife management actions. society and natural resources 11: 649–662. 98 opinions about moose – labonte et al. alces vol. 49, 2013 opinions about moose and moose management at the southern extent of moose range in connecticut study area and background methods landowner survey hunter survey analysis results respondent demographics landowner beliefs and experiences with wildlife knowledge about moose opinions about moose interactions with moose landowner concerns about moose moose population management landowner opinions about roadside sightings and mooseicle accidents discussion acknowledgements references alces29_175.pdf alces27_226erratumcropped.pdf alces28_59.pdf alces24_164.pdf alces21_393.pdf alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 shivering by captive moose infested with winter ticks edward m. addison1,2 and robert f. mclaughlin1,3 1wildlife research and development section, ontario ministry of natural resources, 300 water street, 3rd floor north, peterborough, ontario, canada k9j 8m5. abstract: occurence and rate of shivering were measured to assess thermoregulatory responses of captive moose (alces alces) infested with winter ticks (dermacentor albipictus). shivering was observed on 47 occasions in 5 of 8 infested moose calves from october to april; in contrast, 4 moose calves not infested with winter ticks did not shiver under identical weather conditions. only 5 shivering bouts occurred from october to march, all on a single day. the other 42 shivering bouts occurred in april with bouts lasting 1–103 min. during the april bouts, ambient temperature was 1– 4 °c (42 of 42), maximum wind speed was ≤12 km/h (38 of 42), and it was raining (30 of 42). shivering was associated with 23–44% hair loss in april, but not during cold weather in mid-winter despite 5–10% hair loss in march. maintaining stable core body temperature during late winter-early spring could compromise the energetic balance of wild free-ranging moose with extensive hair loss and abundant ticks, in conditions equivalent to or worse than measured in this study. alces vol. 50: 87–92 (2014) key words: alces alces, dermacentor albipictus, behavior, moose, winter ticks, shivering, thermoregulation. there are numerous studies and much conjecture as to how winter ticks (dermacentor albipictus) adversely affect moose (alces alces). effects on growth, food habits, blood composition, and bioenergetics have been studied in experimentally infested captive moose (mclaughlin and addison 1986, glines and samuel 1989, welch et al. 1990, addison and mclaughlin 1993, addison et al. 1994, addison et al. 1998). samuel (2004) and musante et al. (2007) modeled the potential blood loss from winter ticks and concluded that the elevated energy expenditure and protein imbalance associated with blood loss is deleterious to survival of moose calves with heavy infestations. mclaughlin and addison (1986) speculated that heat loss through winter tick-induced disruption of the hair coat in late winter-early spring may cause accelerated loss of body fat as measured in well fed, captive moose. in contrast, welch et al. (1990) suggested that hair loss may impose only nominal thermoregulatory costs on free-ranging moose and may possibly facilitate dissipation of excess heat in warm, early spring weather. shivering in moose is of interest because it is an involuntary form of thermogenesis that usually occurs when low ambient temperatures (ta) require an increase in metabolic rate to maintain core body temperature (tb) (i.e., the lower critical temperature, tlc) (hohtola 2004). shivering at or near the tlc has been observed in other cervids (parker and robbins 1984). our objective was to determine under what conditions shivering occurs in captive moose calves infested with winter ticks. this information will increase understanding of the metabolic impacts during late winter and early spring weather when moribund and dead moose with severe winter tick 2present address: ecolink science, 107 kennedy street west, aurora, ontario, canada l4g 2l8 3present address: r.r. #3, penetanguishene, ontario, canada l0k1p0 87 infestations are most frequently observed in the wild. methods the experiments were conducted in algonquin provincial park, ontario (45° 30′ n, 78° 35′ w) where 13 of 18 calves were captured at <2 weeks of age in may 1982; 5 calves were from other areas in central and northeastern ontario (addison and mclaughlin 1993). male and female calves were paired in each of 6 adjacent pens (29.6 × 16.5 m) located within a mixed forest stand with little undergrowth and a partial canopy (50% cover in summer) of white pine (pinus strobus), white birch (betula papyrifera), trembling (populus tremuloides) and big tooth aspen (p. grandidentata). calves were weaned as described by addison et al. (1983) and from late october to the end of the experiment were fed ad libitum a ruminant ration containing 16% crude protein, 2.5% crude fat, and 6% crude fiber (united cooperative of ontario, mississauga, ontario, canada). three treatment groups of moose were established: 4 controls without winter ticks, 4 infested with 21,000 winter tick larvae (moderate), and 4 infested with 42,000 larvae (heavy). where possible, siblings were assigned to different treatment groups. control moose were sprayed with acaricide (dursban m., dow chemical of canada ltd., sarnia, ontario, canada) twice in november and powdered with rotenone in december, january, and february to control accidental infestation by larval ticks. all moose were euthanized at the end of the experiment when their hair was dissolved and hides were checked for remaining ticks (addison et al. 1979). moose were weighed on a calibrated platform scale (canadian scale company model 7045) approximately every 7 days from 16–325 days of age (addison et al. 1994). these same 12 moose were used in concurrent studies of hair loss that was measured every 4–10 days from 23 january – 15 april 1983 by measuring the area of hair loss and depth of hair removed (mclaughlin and addison 1986). shivering and other behaviors were recorded simultaneously from 3 observation booths positioned above the back of each of a pair of 6 adjacent pens. three observers recorded observations of 2 moose in a single pen during a 2.5 h period, after which a second group of 3 observers replaced them. these 2 groups then alternated in successive 2.5 h periods throughout the day; all observations were in daylight hours. the 6 moose observed simultaneously included 2 from each treatment group; subsequently, the remaining 6 moose (2 per treatment) were observed the following day and this process continued until 27–30 h of observations were recorded monthly (october – april) for each moose. behaviour was recorded in 60, 1-min intervals during 1 h observation periods; 2403 h of observation occurred from mid-october 1982 to mid-april 1983. the ta, wind speed, and precipitation were recorded at the beginning of each observation period, and those data were attributed to each observation min in that period. results female and male calves weighed 128 ± 6 and 144 ± 15 kg when 4½ months old in early october and 200 ± 17 and 218 ± 20 kg, respectively, at the end of the experiment at 11 months of age (addison et al. 1994). the 4 control moose harbored 0, 4, 21, and 85 winter ticks at the end of the experiment, in contrast to 1179–8290 ticks recovered from the infested moose. fortytwo of 47 shivering bouts were observed in 2 individual moose in the moderate treatment group; 29 bouts were by a single moose (m4) (mclaughlin and addison 1986). the ta was relatively mild on most observation days particularly in october, 88 shivering in tick infested moose – addison and mclaughlin alces vol. 50, 2014 november, march, and april. in december – february (the coldest months) there were 59, 18, and 5 h of observations when ta was −10 to −19, −20 to −29, and −30 to −32 °c, respectively (table 1). no wind was measured in 123 h of observation, winds 20– 40 km/h occurred most frequently in october – december, and there were 14 h of observations in february when wind was >20 km/h (table 1). precipitation during observation periods was lowest in february (table 1). hours of precipitation were generally similar in september – january and april, with most april precipitation as rain. no shivering was observed from october – january and in march. one infested moose had 5 short shivering bouts within 35 min on a single day in february; ta was −10 to −11 °c with wind speed of 12 km/h during 1 bout, and 28 km/h during the other 4 bouts. these bouts averaged 5 min, ranging 1–10 min in length. the other 42 bouts occurred in april and averaged 17.8 min, ranging from 1–103 min in length. the ta was 1–4 °c, maximum wind speed was ≤12 km/h (38 of 42), and it rained during 30 of the 42 april bouts. shivering occurred when moose were standing (n = 25) or recumbent (n = 20), and when both standing and recumbent (n = 2). other activities during shivering bouts included recumbent with head rested on the ground (n = 9), walking (n = 9), ruminating (n = 8), feeding (n = 4), defecating (n = 4), drinking or eating snow (n = 2), urinating (n = 2), grooming (n = 2), rubbing against other objects (n = 1), trotting (n = 1), and galloping (n = 1). discussion shivering in moose during relatively warm and often wet april weather is of special interest because it seems inconsistent with results of prior studies. the insulative properties of moose hair are as high or higher table 1. weather characteristics during monthly observations of captive calf moose infested with winter ticks, algonquin provincial park, ontario, 1982–1983. values represent hours of observation during specific conditions; total monthly hours per category may differ due to rounding. the proportional monthly hours of rain are in parentheses. oct nov dec jan feb mar apr ambient temperature (°c) >0 64 12 10 0 0 33 50 0 to −9 1 48 26 42 19 26 10 −10 to −19 0 0 19 15 25 0 0 −20 to −29 0 0 5 3 10 0 0 −30 to −32 0 0 0 0 5 0 0 wind speed (km/h) 0 7 13 7 28 22 21 25 1–9 5 8 5 5 2 16 3 10–19 24 22 28 25 21 16 21 20–29 24 19 20 2 14 1 9 30–40 1 3 0 0 0 0 0 precipitation rain 15 (56) 10 (27) 10 (28) 7 (19) 0 13 (57) 26 (84) snow 17 27 26 29 7 10 5 alces vol. 50, 2014 addison and mclaughlin – shivering in tick infested moose 89 than that of any arctic mammal (scholander et al. 1950), hence cold weather under most conditions would not cause thermoregulatory stress to moose. increased metabolic rate indicative of moose reaching their winter tlc was not observed from −25 to −30 °c (renecker and hudson 1986), and captive, well fed moose calves were almost as cold tolerant as adults (renecker et al. 1978). absence of shivering in december, january, and most of february, despite very cold weather, was consistent with metabolic measurements and predictions (renecker and hudson 1986). in contrast, certain of our moose had extensive hair loss in april and shivering occurred in some but not all tick-infested moose during relatively mild temperatures in april. insulative properties of hair are positively correlated with the depth of hair for many arctic mammals and cattle (scholander et al. 1950, bennett 1964). however, wet pelts and skin substantially increase heat loss (scholander et al. 1950, holmes 1981) and wind reduces the insulative quality of hair coats (scholander et al. 1950, parker and robbins 1985). hair loss of our moose was generally ∼5% at the end of february, but for most infested moose increased to 23–44% by mid-april. most shivering was by a single moose with 31% hair loss during rainy and moderately windy conditions in april (mclaughlin and addison 1986). distinct individual variation in shivering was evident as 62 and 17% of shivering bouts were by 2 moose (moderate). it may seem anomalous that only 2 moose in the moderate group shivered the most, whereas only limited shivering was observed in the heavy treatment group. it is possible that the quantity of larval ticks applied in the 2 treatments was not different biologically. for example, hair loss was not substantially different between the moderate (23–31%) and heavy (28–44%) treatment groups. further, hair loss in the moderate group was highly variable (2–24%) at the conclusion of the experiment (mclaughlin and addison 1986). moose also displayed variation in grooming and rubbing behaviors within each of the treatment groups. little behavioral response was evident in our tick-infested moose in alberta during a relatively warm winter and spring (welch et al. 1990). only 1 trial occurred below −10 °c, and none below −25 °c, temperatures above the tlc of moose (renecker and hudson 1986). it has been suggested that hair loss might have a thermoregulatory advantage by dissipating excess heat during warm spring weather (welch et al. 1990). unfortunately, variables affecting thermo‐ regulation such as precipitation, wind, and radiant energy are seldom documented (parker and robbins 1985), and lack of standardization in techniques and variable diets are cause for caution when comparing bio‐ energetic studies of cervids (renecker and hudson 1986). in our study, the duration and intensity of wind and precipitation were generally similar in all months except for less precipitation in february (table 1). the only obvious weather differences between april and the prior 4 months included more precipitation as rain in april and warmer ta in march and april (table 1). there was also an exponential increase in hair loss (mclaughlin and addison 1986) and rapid growth by adult winter ticks in april (addison and mclaughlin 1988), hence higher blood loss as estimated in previous studies (see samuel 2004, musante et al. 2007). of note is that our experimental moose had >20,000 fewer ticks and much less hair loss at the end of the experiment than measured, on average, on free-ranging moose succumbing to heavy winter tick infestation in alberta (samuel and barker 1979, samuel 2004). the greater attrition of fat reservoirs in infested than non-infested moose, despite high quality food, presumably occurred over a period of 1–2 months (mclaughlin 90 shivering in tick infested moose – addison and mclaughlin alces vol. 50, 2014 and addison 1986). it is unlikely that the loss of body condition in infested moose could be attributed only to energy demands of maintaining core tb associated with hair loss. models developed by samuel (2004) and musante et al. (2007) indicate that the energy cost associated with replacing direct blood loss has a substantial metabolic impact during the 2–3 week peak period of female adult engorgement. these energy costs and higher thermoregulatory costs in march and april likely contributed to the reduced body condition of the infested calves as described by mclaughlin and addison (1986). it is also possible that due to depleted and less oxygenated red blood cells, there is less general motor activity and reduced movement, foraging, and production of body heat in heavily infested free-ranging moose. most moose experimentally infested with up to 42,000 larval winter ticks had limited hair loss and did not shiver during cold weather in mid-winter or even with substantial hair loss in spring. when shivering did occur it was for extended periods (hours) in spring, at or above freezing temperatures when moose were wet. the combination of rainy weather and severe hair loss from heavy tick infestation likely has more thermoregulatory impact than recognized previously. calves with severe hair loss and reduced body fat due to heavy winter tick infestation likely have an elevated tlc in late winter-early spring that could further compromise their survival. acknowledgements we appreciate d. j. h. fraser for his coordination of many early aspects of this study. special thanks go to a. rynard, a. macmillan, m. jefferson, v. ewing, and d. bouchard for their steadfast assistance in collection of data and for moose husbandry under adverse conditions. additional assistance was provided by c. pirie, m. a. mclaughlin, d. carlson, d. joachim and p. methner. l. smith, k. paterson, k. long, a. jones, s. gadawski, s. fraser, d. fraser, and l. berejikian assisted in the earlier care of calves. we appreciate the assistance of c. d. macinnes and g. smith and staff for their administrative support. field work was conducted at the wildlife research station in algonquin park where r. keatley, p. c. smith, and staff were of great help. thank you to r. moen, b. samuel, and anonymous reviewers for providing excellent suggestions for improvements to earlier drafts of the manuscript. references addison, e. m., f. j. johnson, and a. fyvie. 1979. dermacentor albipictus on moose (alces alces) in ontario. journal of wildlife diseases 15: 281–284. ———, and r. f. mclaughlin. 1988. growth and development of winter tick, dermacentor albipictus, on moose (alces alces). the journal of parasitology 74: 670–678. ———, and ———. 1993. seasonal variation and effects of winter ticks (dermacentor albipictus) on consumption of food by captive moose (alces alces) calves. alces 29: 219–224. ———, ———, and j. d. broadfoot. 1994. growth of moose calves (alces alces americana) infested and uninfested with winter ticks (dermacentor albipictus). canadian journal of zoology 72: 1469–1476. ———, ———, and ———. 1998. effects of winter tick (dermacentor albipictus) on blood characteristics of captive moose (alces alces). alces 34: 189–199. ———, ———, and d. j. h. fraser. 1983. raising moose calves in ontario. alces 18: 246–270. bennett, j. w. 1964. thermal insulation of cattle coats. proceedings of the australian society of animal production 5: 160–166. alces vol. 50, 2014 addison and mclaughlin – shivering in tick infested moose 91 glines, m. v., and w. m. samuel. 1989. effect of dermacentor albipictus (acari: ixodidae) on blood composition, weight gain, and hair coat of moose, alces alces. experimental and applied acarology 6: 197–213. hohtola, e. 2004. shivering thermogenesis in birds and mammals. pages 241–252 in b. n. barnes and h. v. carey, editors. life in the cold: evolution, mechanisms, adaptation, application. twelfth international hibernation symposium. biological papers of the university of alaska, number 27. fairbanks, alaska, usa. holmes, c. w. 1981. a note on the protection provided by the hair coat or fleece of the animal against the thermal effects of simulated rain. animal production 32: 225–226. mclaughlin, r. f., and e. m. addison. 1986. tick (dermacentor albipictus)-induced winter hair-loss in captive moose (alces alces). journal of wildlife diseases 22: 502–510. musante, a. r., p. j. pekins, and d. l. scarpitti. 2007. metabolic impacts of winter tick infestations on calf moose. alces 43: 101–110. parker, k. l., and c. t. robbins. 1984. thermoregulation in mule deer and elk. canadian journal of zoology 62: 1409–1422. ———, and ———. 1985. thermoregulation in ungulates. pages 161–182 in r. j. hudson and r. g. white, editors. bioenergetics of wild herbivores. crc press, boca raton, florida, usa. renecker, l. a., and r. j. hudson. 1986. seasonal energy expenditures and thermoregulatory responses of moose. canadian journal of zoology 64: 322–327. ———, ———, k. christophersen, and c. lewis. 1978. effect of posture, feeding, low temperature, and wind on energy expenditures of moose calves. proceedings of the north american moose conference and workshop 14: 126–140. samuel, b. 2004. white as a ghost: winter ticks and moose. natural history series, volume 1, federation of alberta naturalists, edmonton, alberta, canada. samuel, w. m., and m. j. barker. 1979. the winter tick, dermacentor albipictus (packard, 1869) on moose alces alces (l.), of central alberta. proceedings of the north american moose conference and workshop 15: 303–348. scholander, p. f., v. walters, r. hock, and l. irving. 1950. body insulation of some arctic and tropical mammals and birds. the biological bulletin 99: 225–236. welch, d. a., w. m. samuel, and r. j. hudson. 1990. bioenergetic consequences of alopecia induced by dermacentor albipictus (acari: ixodidae) on moose. journal of medical entomology 27: 656–660. 92 shivering in tick infested moose – addison and mclaughlin alces vol. 50, 2014 shivering by captive moose infested with winter ticks methods results discussion acknowledgements references alces24_78.pdf alces26_163conferenceworkshop.pdf population genetic structure of moose (alces alces) of south-central alaska robert e. wilson1,2, thomas j. mcdonough3, perry s. barboza1, sandra l. talbot4, and sean d. farley5 1university of alaska fairbanks, institute of arctic biology, fairbanks, alaska 99775; 3alaska department of fish and game, 3298 douglas st., homer, alaska 99603; 4u. s. geological survey, alaska science center, 4210 university drive, anchorage, alaska 99508; 5alaska department of fish and game, 333 raspberry road, anchorage, alaska 99518 abstract: the location of a population can influence its genetic structure and diversity by impacting the degree of isolation and connectivity to other populations. populations at range margins are often thought to have less genetic variation and increased genetic structure, and a reduction in genetic diversity can have negative impacts on the health of a population. we explored the genetic diversity and connectivity between 3 peripheral populations of moose (alces alces) with differing potential for connectivity to other areas within interior alaska. populations on the kenai peninsula and from the anchorage region were found to be significantly differentiated (fst = 0.071, p < 0.0001) with lower levels of genetic diversity observed within the kenai population. bayesian analyses employing assignment methodologies uncovered little evidence of contemporary gene flow between anchorage and kenai, suggesting regional isolation. although gene flow outside the peninsula is restricted, high levels of gene flow were detected within the kenai that is explained by male-biased dispersal. furthermore, gene flow estimates differed across time scales on the kenai peninsula which may have been influenced by demographic fluctuations correlated, at least in part, with habitat change. alces vol. 51: 71–86 (2015) key words: alaska, genetic diversity, gene flow, moose, population genetic structure the pattern of geographical variation in genetic diversity and divergence is dictated by the interaction of genetic drift, gene flow, and natural selection (eckert et al. 2008), and these evolutionary processes can be influenced by the location of a population within the species’ geographic range (briggs 1996, wisely et al. 2004, howes and lougheed 2008). at the local and regional scales, the relative location of a population can strongly impact patterns of dispersal and degree of isolation influenced by both historical and contemporary events (vucetich and waite 2003, eckert et al. 2008), ultimately determining the level of genetic structure and diversity. genetic diversity is lowest at the range margins and highest at the center of a species distribution (yamashita and polis 1995, schwartz et al. 2003, eckert et al. 2008, howes and lougheed 2008). marginal populations are more likely to be isolated, occur in patchy habitats, and may reflect recent colonization. peripheral populations are less likely to receive immigrants whereas the core populations typically occupy prime habitat and experience greater levels of gene flow (hoffmann and blows 1994, brown et al. 1995, wisely et al. 2004, miller et al. 2010, schrey et al. 2011). evolutionary theory suggests that the reduction of genetic diversity within peripheral populations can impede adaptation to 2present address: u. s. geological survey, alaska science center, 4210 university drive, anchorage, alaska 99508 71 changing environmental conditions (bradshaw 1991, hoffmann and parsons 1991, hoffmann and blows 1994, blows and hoffmann 2005). such adaptation is largely determined by the availability of additive genetic variation in heritable traits with fitness consequences. several studies have shown that even small changes in genetic variation can have large effects on population fitness (frankham 1995, reed and frankham 2003) including juvenile survival (coulson et al. 1999, mainguy et al. 2009, silva et al. 2009), antler growth (von hardenberg et al. 2007), and parasite resistance (coltman et al. 1999) within ungulates. however, adaptability is also affected by effective dispersal which can have either positive or negative effects on the population depending on the rate of gene flow and the strength of selection acting on the local population (garcía-romos and kirkpatrick 1997, akerman and bürger 2014, bourne et al. 2014, frankham 2015). thus, examining conditions (habitat, genetic diversity, gene flow rates, and life history) under which peripheral populations exist can aid in understanding the processes that maintain geographical ranges, predicting the consequences of climate change (parmesan and yohe 2003, root et al. 2003, hampe and petit 2005), and conserving populations at range margins (howes and lougheed 2008). the kenai peninsula is a peripheral region situated in south-central alaska that was separated from the mainland by a narrow (16 km wide) isthmus at the end of the last ice age. due to its diverse landscape, biodiversity in this region is unusually high at this latitude (morton et al. 2009), and the moose (alces alces) is one of the most recognizable and socio-economically important species. moose populations on the kenai peninsula are characterized by fluctuations in population size, peaking after the occurrence of forest fires that promote optimal forage habitat (oldemeyer et al. 1977). while moose populations on the kenai have fluctuated between 5,000–8,000 animals over the past several decades (t. j. mcdonough, unpublished data), these fluctuations have not been uniform across moose management units on the peninsula. while population size in game management unit (gmu) 15c in southwest kenai has increased, numbers in gmu 15a (northwest) have declined drastically, ∼40% in the last 20 years as quality forage has diminished since the last major fire in 1969. relative isolation from neighboring regions with a strong history of fluctuations in population size might lead to reduced genetic variability on the kenai peninsula which could ultimately be detrimental to the long-term health of moose in this region. using microsatellite loci, we compared levels of genetic variation and gene flow in 2 areas within the kenai peninsula and the anchorage area. these 3 areas are situated on the periphery of overall moose distribution in alaska but differ in levels of potential connectivity to the core area of interior alaska. first, we investigated the connectivity between gmus on the kenai peninsula that have been affected by a long history of land alteration and demographic changes. second, we predicted that 2 sites within the disjunct kenai peninsula region, where opportunities for genetic exchange may be more limiting than anchorage, would exhibit relatively lower genetic diversity. methods sample collection a total of 163 moose were sampled from 3 populations in south-central alaska (fig. 1). ear-plugs and blood were taken from 33 collared female moose in 2008–2010 and 2012 from the city of anchorage and adjacent eagle river (called anchorage hereafter). in addition, muscle tissue was taken from 32 hunter-killed moose (16 female, 15 male, and 1 unknown) during the winter of 72 population genetic structure – wilson et al. alces vol. 51, 2015 2011–2012. in spring 2012, blood was taken from radio-collared female moose from gmu 15a (n = 49; 3,367 km2) and gmu 15c (n = 49; 3,030 km2) on the kenai peninsula, the borders of which are approximately 20 km apart. anchorage samples are archived at the molecular ecology laboratory, u.s. geological survey, anchorage, alaska, and kenai peninsula samples at the alaska department of fish and game, homer, alaska. all animal capturing and genetic sampling were conducted under division of wildlife conservation acuc approval (# 2012–2007, 2013–2021, and 90–05) and under the university of alaska fairbanks iacuc approval (# 14885 and 182744). molecular techniques genomic dnawas extracted from blood and tissue samples using a “salting out” procedure described by medrano et al. (1990), with modifications described in sonsthagen et al. (2004). genomic dna concentrations were quantified using fluorometry and diluted to 50 ng ml–1 working solutions. individuals were initially screened at 17 microsatellite loci. thirteen autosomal loci were found to be polymorphic of which 9 with dinucleotide repeat motifs were selected for further analysis that were polymorphic in all populations: bl42, bm888, bm203, bm2830 (bishop et al. 1994), nvhrt21, nvhrt22 (røed and midthjell 1998), rt1, rt5, and rt30 (wilson et al. 1997). polymerase chain reaction (pcr) amplification and electrophoresis followed protocols described in roffler et al. (2012). ten percent of the samples were amplified and genotyped in duplicate for the 9 microsatellite loci for quality control. analysis of genetic diversity and population genetic subdivision we calculated allelic richness, inbreeding coefficient (fis), observed and expected heterozygosities, and tested for deviations from hardy-weinberg equilibrium (hwe) and linkage disequilibrium (ld) for each microsatellite locus and population in fstat ver. 2.9.3 (goudet 1995). the degree of genetic subdivision among moose populations was assessed by calculating overall and pairwise fst and rst, adjusting for multiple comparisons using bonferroni correction (α = 0.05) in arlequin v3.5.1.3 (excoffier and lischer 2010). because the upper possible fst value for a set of microsatellite loci is usually <1.0 (hedrick 2005), we used recodedata, version 1.0 (meirmans 2006) to calculate the uppermost limit of fst for our data set. we also used the bayesian-clustering program stucture 2.2.3 (pritchard et al. 2000) to determine the level of population structure in the autosomal microsatellite data set. we performed 2 sets of analyses to look at structure within south central alaska: 1) between anchorage and kenai peninsula and 2) within the kenai peninsula (gmu 15a and 15c). structure assigns individuals to populations maximizing hardy-weinberg equilibrium and minimizing linkage disequilibrium. the analysis was conducted for 1– 10 populations (k) using an admixture fig. 1. sampling areas for three moose populations in south-central alaska: anchorage, game management unit (gmu) 15a (northwest kenai peninsula), and gmu15c (southwest kenai peninsula). alces vol. 51, 2015 wilson et al. – population genetic structure 73 model with 100,000 burn-in iterations and 1,000,000 markov chain monte carlo (mcmc) iterations without providing a priori information on the geographic origin of the individuals; the analyses were repeated 10 times for each k to ensure consistency across runs. we used the ▵k method of evanno et al. (2005) and evaluation of the estimate of the posterior probability of the data given k, ln p(d), to determine the most likely number of groups at the uppermost level of population structure. for the kenai peninsula analysis we used the locprior which is able to detect weak signals of population structure in datasets not detectable under standard models (hubisz et al. 2009). we determined if location was informative by the value of r, which parameterizes the amount of information contained by the location of the samples. values of r > 1 indicates either there is no population structure or that structure is independent of locality. gene flow we estimated gene flow between moose populations using 2 methodologies: migrate v3.2.16 (beerli and felsenstein 1999, 2001) and bayesass 3.0 (wilson and rannala 2003). these programs differ in the underlying model used to estimate gene flow. migrate uses a steady-state twoisland coalescent model of population differentiation which incorporates parameters scaled to the mutation rate (µ), the effective population size parameter θ (4neµ), and the migration rate m (m/µ) between populations. bayesass uses an assignment methodology which does not incorporate genealogy or assume that populations are in hardy-weinberg equilibrium (wilson and rannala 2003). thus, estimates of migration rate can be interpreted differently and at different temporal scales. bayesass reflects gene flow over the last several generations (referred to as contemporary gene flow hereafter) whereas migrate gene flow estimates are averaged over the past n generations, where n equals the number of generations the populations have been at mutation-drift equilibrium (beerli and felsenstein 1999, 2001). it is generally agreed that microsatellite mutation rates are several orders of magnitude higher than mutation rates of dna sequences (mitochondrial or nuclear; schlötterer 2000, ellegren 2004). thus, microsatellite markers can reflect recent (within the last 10,000 years) and almost contemporaneous events, but increases in homoplasy associated with microsatellites reduce their ability to capture older demographic events (hartl and clark 2007, hughes 2010). therefore, migrate analyses are referred to as estimating recent gene flow. migrate was run with a full migration model; θ (4neµ, composite measure of effective population size and mutation rate) and all pairwise migration parameters were estimated individually from the data. gene flow was estimated using maximum likelihood search parameters; 10 short chains (5000 trees used out of 1,500,000 sampled), 10 long chains (15,000 trees used out 5,250,000 sampled), and 5 static heated chains (1, 1.33, 2.0, 4.0, and 1,000,000; swapping interval = 1). full models were run 10 times to ensure the convergence of parameter estimates. bayesass was initially run with the default delta values for allelic frequency (p), migration rate (m), and inbreeding (f). subsequent runs incorporated different delta values to ensure that acceptance rate for proposed changes was between 20–40% for each parameter to maximize log likelihood values and ensure the most accurate estimates (wilson and rannala 2003). final delta values used were ▵p = 0.5 (27% acceptance rate), ▵m = 0.2 (27%), and ▵f = 0.85 (31%). we performed 10 independent runs (10 million iterations, 1 million burn-in, 74 population genetic structure – wilson et al. alces vol. 51, 2015 and sampling frequency of 1000) and 2 additional longer runs (50 million iterations, 5 million burn-in) with different random seeds to ensure convergence and consistency across runs. convergence was also assessed by examining the trace file in program tracer v1.5 to ensure proper mixing of parameters (rambaut and drummond 2007). population demography to estimate the effective population size (ne) for each gmu on the kenai peninsula and anchorage area, we used the approximate bayesian computation method (beaumont et al. 2002) implemented in the program onesamp 1.2 (tallmon et al. 2008). we used a lower prior of 100 for all populations and a maximum prior that reflected the current census size (1,000 for anchorage, 2,000 for gmu 15a, and 3,000 for gmu 15c). similar values were obtained for larger maximum possible effective population sizes. lastly, we used bottleneck which compares the number of alleles and gene diversity at polymorphic loci under the infinite allele model (iam; maruyama and fuerst 1985), stepwise mutation model (smm; ohta and kimura 1973), and twophase model of mutation (tpm; di rienzo et al. 1994; parameters: 79% ssm, variance 9; piry et al. 1999, garza and williamson 2001). one thousand simulations were performed for each population and parameters were changed among 5 runs to evaluate the robustness of results. significance was assessed using a wilcoxon sign-rank test which determines if the average of standardized differences between observed and expected heterozygosities is significantly different from zero (cornuet and luikart 1996). significant heterozygote deficiency relative to the number of alleles indicates recent population growth, whereas heterozygote excess relative to the number of alleles indicates a recent population bottleneck (cornuet and luikart 1996). bottleneck compares heterozygote deficiency and excess relative to genetic diversity, not to hardyweinberg equilibrium expectation (cornuet and luikart 1996). results genetic diversity and population subdivision multilocus genotypes were collected from 163 individuals and each individual had a unique genotype. the number of alleles per locus observed ranged from 3.4– 4.7 per population with an overall estimate of 5.1 (tables 1, 2). the observed heterozygosity ranged from 43–55% with an overall mean heterozygosity of 49%. the kenai peninsula exhibited a 19% lower allelic richness (20% in gmu 15a and 25% in gmu 15c) compared to the anchorage area, and 3x more private alleles were observed in the anchorage region (table 1). in addition, the observed (ho) and expected (he) heterozygosity was significantly lower (all p-values < 0.0001) in the kenai peninsula (ho by 18%, he by 16% expected), in gmu 15c (15%, 14%), and in gmu 15a (22%, 20%). on average, individuals on the kenai showed a greater level of homozygosity (kenai: 4.94 loci [sd = 1.46] vs. anchorage: 3.98 loci [sd = 1.51]; t = 4.02, p < 0.0001). the inbreeding coefficient (fis) did not differ significantly from zero in any population (table 1). all loci and populations were in hwe and linkage equilibrium. significant genetic structure was observed at the 9 microsatellite loci between anchorage and the two gmus on the kenai peninsula (table 3). no significant difference was found within the kenai peninsula. the upper limit of the fst for our microsatellite data set was 0.499. therefore, the overall fst of 0.071 accounted for 14.2% of the maximum possible level of genetic structure and 19% for the pairwise alces vol. 51, 2015 wilson et al. – population genetic structure 75 comparisons between anchorage and kenai peninsula gmus. structure uncovered genetic partitioning within south central moose populations, supporting a two-population model (δk = 188.3, average ln p(d) = −2758.7). most individuals from anchorage were assigned to one genetic cluster (87.7%), whereas individuals from kenai gmu 15a and 15c were assigned to a second cluster with high probability, 93.6 and 92.6%, respectively (fig. 2). seven anchorage individuals were assigned to the anchorage cluster with <60% certainty, conversely, only a single kenai individual was assigned to the kenai cluster with <60% certainty. genetic partitioning was not observed within kenai peninsula, as including capture location (lociprior) was not informative (r > 9). gene flow restricted gene flow over the past several generations was observed under the bayesass model between anchorage and kenai peninsula, with 96.8% (93.3–100%) of the anchorage population comprised of a non-migrant origin (fig. 3). within the kenai peninsula, there was a signal of a northern direction of contemporary gene flow from gmu 15c into 15a (proportion of individuals with migrant origin: 27.8% in 15a vs. 6.9% in 15c); although the 95% confidence intervals do overlap (fig. 3). asymmetrical recent gene flow as estimated by migrate was observed among sampled populations. the directionality of gene flow was from kenai peninsula into anchorage (fig. 3). the number of migrants per generation (nem) ranged from 2.56 (gmu 15a; 1.97–3.29) and 2.78 (gmu 15c; 2.20–3.48) into anchorage and 0.99 (0.74–1.32) and 1.08 (0.78–1.47) into the kenai gmu 15a and 15c, respectively. within kenai there was a signal of asymmetrical gene flow from gmu 15a into 15c (3.3 migrants/generation; fig. 3). population demography the estimated effective size using the bayesian computation method for the anchorage region was 74.3 (95% ci: 67.6– 83.0). gmus 15a and 15c on the kenai peninsula had lower estimated effective sizes with non-overlapping confidence intervals with anchorage (table 1). the bottleneck analysis showed no evidence of significant table 1. estimates of genetic diversity of the moose sampled from three locales in south-central alaska, including: average number of alleles, allelic richness (ar), observed and expected heterozygosities (ho/he), inbreeding coefficient (fis), effective population size (ne) estimated in onesamp and sample size (n) calculated from nine microsatellite loci. allelic richness is based on smallest sample size of 65 for anchorage and overall kenai. within kenai peninsula (gmu 15a and gmu 15c) based on sample size of 49. kenai peninsula anchorage gmu 15a gmu 15c overall kenai no. alleles 4.67 3.67 3.44 4.00 no. private alleles 10 1 2 3 ar 4.59 3.67 3.44 3.78 ho (sd) / 0.55 (0.02)/ 0.43 (0.02)/ 0.47 (0.02)/ 0.45 (0.02)/ he (sd) 0.56 (0.06) 0.45 (0.07) 0.48 (0.05) 0.47 (0.06) fis 0.007 0.056 0.031 0.043 ne 74.3 (67.6–83.0) 47.9 (43.2–56.8) 36.8 (33.3–43.9) 145.5 (123.3–255.9) n 65 49 49 98 76 population genetic structure – wilson et al. alces vol. 51, 2015 heterozygosity excess or deficit under the smm or tpm. however, there was evidence of a recent population decline (heterozygote excess) based on the infinite allele model (iam) for kenai gmu 15c. discussion climatic and glaciation history has played a major role in shaping the evolutionary history of many taxa in south-central alaska. it was not until approximately 7,000 years before present that the kenai peninsula became distinct and relatively isolated from the mainland by a 16 km wide mountainous isthmus (pielou 1991, muhs et al. 2001). this isolation has fostered genetically or morphologically distinct populations for a variety of taxa (e.g., wolverine table 2. estimates of observed and expected heterozygosity, number of alleles per locus for nine autosomal nuclear microsatellite loci assayed in three moose populations in south-central alaska. all loci were in hardy-weinberg equilibrium. sample size is in parentheses; ho = heterozygosity observed, he = heterozygosity expected, and na = number of alleles. kenai peninsula locus anchorage (65) gmu 15a (49) gmu 15c (49) overall kenai (98) all populations (165) nvhrt22 ho/he 0.69/0.76 0.49/0.54 0.57/0.53 0.53/0.53 0.60/0.68 na 6 5 4 6 6 nvhrt21 ho/he 0.49/0.50 0.55/0.46 0.39/0.45 0.47/0.45 0.48/0.48 na 5 3 2 3 5 rt1 ho/he 0.49/0.46 0.27/0.29 0.39/0.38 0.31/0.34 0.38/0.40 na 2 2 2 2 2 rt5 ho/he 0.54/0.52 0.18/0.21 0.25/0.32 0.21/0.26 0.34/0.40 na 4 3 3 3 4 rt30 ho/he 0.55/0.58 0.69/0.67 0.74/0.72 0.71/0.70 0.65/0.66 na 5 4 4 4 5 bm203 ho/he 0.20/0.20 0.37/0.41 0.51/0.50 0.44/0.46 0.34/0.38 na 5 3 4 4 6 bm2830 ho/he 0.46/0.49 0.37/0.43 0.41/0.41 0.39/0.42 0.42/0.45 na 2 2 2 2 2 bm888 ho/he 0.63/0.65 0.22/0.26 0.20/0.27 0.21/0.27 0.38/0.46 na 4 4 3 4 4 bl42 ho/he 0.91/0.84 0.71/0.81 0.80/0.75 0.76/0.79 0.82/0.83 na 9 6 7 8 12 overall loci ho/he 0.55/0.56 0.43/0.45 0.47/0.48 0.45/0.47 0.49/0.53 na 4.67 3.67 3.44 4.00 5.11 table 3. pairwise and overall values of fst and rst calculated from nine microsatellite loci. significant values after bonferroni correction (p < 0.0001) are marked with an asterisk. fst rst anchorage – kenai gmu 15a 0.094* 0.014 – kenai gmu 15c 0.092* 0.028 kenai gmu 15a – kenai gmu 15c 0.001 0.000 overall 0.071* 0.016 alces vol. 51, 2015 wilson et al. – population genetic structure 77 [gulo gulo], tomasik and cook 2005; american marten [ursus americanus], robinson et al. 2007; song sparrow [melospiza melodia], patten and pruett 2009). the moose populations residing on the kenai are no exception. using a multi-locus approach, we observed that moose on the kenai were genetically distinct from those in the 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 pr o b ab ili ty o f a ss ig n m en t anchorage kenai gmu 15a kenai gmu 15c fig. 2. structure analysis showing posterior probability of assignment of individuals to each (k = 2) genetic cluster. white bar represents the estimated probability of assignment to cluster one and grey bar is the estimated probability of assignment to cluster two. 0.99 (0.74-1.32) 2.56 (1.97-3.29) 3.29 (2.63-4.07) 1.18 (0.89-1.54) a b 2.78 (2.20-3.48) 1.08 (0.78-1.47) 0.02 (0.00-0.05) 0.01 (0.00-0.03) 0.28 (0.13-0.43) 0.07 (0.00-0.25) 0.01 (0.00-0.02) 0.01 (0.00-0.04) fig. 3. estimates of (a) recent (number of migrants per generation, nem) estimated in migrate and (b) contemporary (proportion of individuals with migrant origin, m) calculated in bayesass for moose populations in south-central alaska as calculated from nine microsatellite loci with relative magnitude indicated by width of arrow; 95% confidence intervals are in parentheses. 78 population genetic structure – wilson et al. alces vol. 51, 2015 mainland anchorage population and exhibited significantly lower levels of genetic diversity at microsatellite loci. loss of genetic diversity between peninsula and mainland populations residing in areas with barriers that limit dispersal (e.g., peninsulas and islands) across the landscape are expected to have lower genetic variation (gaines et al. 1997). our results were consistent with gaines et al. (1997) prediction: moose occupying the kenai peninsula had significantly reduced genetic diversity (∼18%) compared to the nearest mainland population in anchorage. a reduction of genetic variability has also been reported for other alaskan moose populations (hundertmark 2009, schmidt et al. 2009) as well as other mammals on the kenai peninsula (e.g., canada lynx [lynx canadensis], schwartz et al. 2003). the loss of genetic variation in peripheral populations may be due to numerous factors such as limited number of connections to other populations or smaller population size (schwartz et al. 2003). cook inlet waters, mountains, and a highway and railways may represent formidable dispersal barriers for moose between these regions. although kenai peninsula and anchorage are in close geographic proximity (straight line distance over land is ∼105 km), the costs of dispersal over the rugged terrain and highways or swimming across the inlet are likely high. in agreement with limited effective dispersal, we found restricted contemporary gene flow between kenai peninsula and mainland anchorage populations with confidence intervals suggesting there has been limited genetic exchange over the past several generations. telemetry studies of the sampled females in this study showed that individuals remained in the same general area throughout the year (farley et al. 2012, t. j. mcdonough, unpublished data), further suggesting a low likelihood of long-distance dispersal between these two regions. however, connectivity could be mediated through a contact zone north of the isthmus located at portage valley that is used by black bears (ursus americanus) (robinson et al. 2007). the isthmus is not an absolute/strong barrier as movements of radio-collared moose occur across the isthmus; this movement was restricted within intermountain valleys that spanned both sides of the isthmus (t. lohuis, alaska department of fish and game, unpublished data). furthermore, structure analysis estimated a low probability assignment to a genetic cluster for ∼12% of the individuals in anchorage, suggestive of genetic exchange that has occurred during or after population divergence, with higher rate going into the anchorage area based on the migrate analysis. this northward direction of gene flow is also found in other peninsular populations (schmidt et al. 2009) and may reflect post-colonization gene flow rates. further study of moose in areas between anchorage and kenai peninsula might identify if a contact zone exists for moose at the isthmus as seen in other mammals, or if these regions are truly isolated as indicated by the contemporary gene flow analysis. relationships within the peninsula unlike the potential strong barriers to dispersal between the peninsula and mainland populations, there are relatively few natural barriers to movement in the western part of the peninsula, and gene flow estimates suggest that there is ongoing genetic exchange. the directionality of gene flow across the western kenai peninsula has not remained constant over time. differences in directionality across time scales may be attributed to the fluctuating nature of moose population dynamics that is correlated at least in part with habitat change, in particular alces vol. 51, 2015 wilson et al. – population genetic structure 79 in gmu 15a where population size fluctuates with major fire events (oldemeyer et al. 1977, schwartz and franzmann 1989, loranger et al. 1991). the habitat in gmu 15a has changed drastically over the last century after major fires in 1947 and 1969 transformed previously low quality habitat to ideal foraging habitat, which subsequently declined to the current condition (oldemeyer et al. 1977, schwartz and franzmann 1989). if periodic population increase has been sufficiently frequent throughout the history of moose in this area, and dispersal is influenced by population density and habitat quality, we might expect the directionality of gene flow to change over time with more moose dispersing from areas of high productivity into areas of lower density or less preferred habitat as competition for resources increases. indeed, contemporary gene flow estimated in bayesass indicates gene flow from a higher density area (gmu 15c) with better quality habitat into an area characterized by poor habitat conditions and lower density area (15a). moose populations on the kenai peninsula have also fluctuated in size partially due to human activities (land development and forest fires), with changes in habitat potentially affecting fertility and survival of young (klein 1970, franzmann and arneson 1973, schwartz and franzmann 1989, testa and adams 1989). while moose populations initially respond positively to wildfires through the emergence of optimal habitat, populations eventually decline as the habitat changes to late succession (non-optimal forage) vegetation. during the 20 years following the last major fire in gmu 15a (1969), the population has declined by approximately 40%. current and previous assessment of calf survival from this area has identified low calf survival (franzmann et al. 1980, t. j. mcdonough, unpublished data). such a drastic decline in population size coupled with low productivity can negatively impact genetic diversity of a population; this may partially explain the significantly low genetic diversity on the kenai peninsula. a reduction in genetic diversity can lower viability and fecundity (falconer 1981, ralls et al. 1983, frankham 1995, crnokrak and roff 1999), and at the extreme can lead to inbreeding depression; decreased viability and fecundity occur currently on the kenai peninsula (franzmann and schwartz 1985, adf&g 2013, unpublished data). whether lower reproductive rates (twinning rates and calf survival) on the kenai peninsula are correlated solely with genetic variability or are influenced in addition, or solely by environmental factors, is an area for future investigation. conservation implications although the effects of inbreeding depression can diminish over time (lynch 1977), a general loss of genetic diversity can be detrimental over evolutionary time as it may lower the ability of populations to respond to environmental stressors such as novel predators, parasites, or climatic conditions (lacy 1987, quattro and vrijenhoek 1989, leberg 1993). following the recommendations of frankham et al. (2014), all 3 populations fall below the minimum effective population size of 1,000 required to maintain long-term viability. in addition, the gmus on the kenai peninsula, when considered separately, have an effective population size lower than both recent (> 100; frankham et al. 2014) and earlier (> 50; franklin 1980, soulé 1980) recommendations to avoid inbreeding depression. indeed, the kenai peninsula does have a higher inbreeding coefficient (although not significantly different from zero) and higher levels of homozygosity. however, when considering the kenai peninsula as a single population, the effective population exceeds 100 but remains below the threshold for long-term viability. 80 population genetic structure – wilson et al. alces vol. 51, 2015 neutral loci are commonly used to infer evolutionary history of populations and make inferences about overall variation (see howes and lougheed 2008), but it is still unclear whether the trends in putatively neutral loci are reflective of quantitative-trait variation found in genes for physiological, morphological, or life history traits that are likely important for the adaptive potential of populations (merilä and crnokrak 2001, reed and frankham 2001, eckert et al. 2008). therefore, a conclusion that reduced genetic diversity observed at neutral microsatellite markers reflects reduction of diversity in the genome overall is premature. our results showing significant population structure and limited connectivity to outside populations for the kenai peninsula provide a working hypothesis for the potential effects on genetic diversity, which can be tested by assaying both selectively neutral and functional diversity. such studies can provide greater resolution on the processes responsible for the distribution of genetic diversity among moose populations within southcentral alaska. acknowledgements funding was provided by joint base elmendorf-richardson, under the guidance of h. griese, c. garner, d. battle, and r. graham, and u. s. geological survey. extensive field assistance was provided by the military conservation agent program on elmendorf air base. j. crouse, j. selinger, and pilots j. decreft, t. levanger, and j. fieldman, and the kenai national wildlife refuge provided field assistance on the kenai peninsula. z. grauvogel, k. sage, and s. sonsthagen provided laboratory and manuscript advice, while k. hundertmark provided advice on marker selection and g. roffler and e. solomon provided help with maps. any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the u.s. government. references akerman, a., and r. bürger. 2014. the consequences of gene flow for local adaptation and differentiation: a two-locus deme model. journal of mathematical biology 68: 1135–1198. beaumont, m. a., w. zhang, and d. j. balding. 2002. approximate bayesian computation in population genetics. genetics 162: 2025–2035. beerli, p., and j. felsenstein. 1999. maximum-likelihood estimation of migration rates and effective population numbers in two populations using a coalescent approach. genetics 152: 763–773. ———, and ———. 2001. maximum likelihood estimation of a migration matrix and effective population sizes in n subpopulations by using a coalescent approach. proceedings of the national academy of sciences usa 98: 4563–4568. bishop, m. d., s. m. kappes, j. w. keele, r. t. stone, s. l. f. sunden, g. a. hawkins, s. s. toldo, r. fries, m. d. grosz, j. yoo, and c. w. beattie. 1994. a genetic linkage map for cattle. genetics 136: 619–639. blows, m. w., and a. a. hoffmann. 2005. a reassessment of genetic limits to evolutionary change. ecology 86: 1371–1384. bourne, e. c., g. bocedi, j. m. j. travis, r. j. pakeman, r. w. brooker, and k. schiffers. 2014. between migration load and evolutionary rescue: dispersal, adaptation and the response of spatially structured populations to environmental change. proceedings of royal society b 281: 20132795. bradshaw, a. d. 1991. the croonian lecture, 1991: genostasis and the limits to evolution. philosophical transactions of the royal society b: biological sciences 333: 289–305. alces vol. 51, 2015 wilson et al. – population genetic structure 81 briggs, j. c. 1996. biogeography and punctuated equilibrium. biogeographica 72: 151–156. brown, j. h., d. w. mehlman, and g. c. stevens. 1995. spatial variation in abundance. ecology 76: 2028–2043. coltman, d. w., j. g. pilkington, j. a. smith, and j. m. pemberton. 1999. parasite-mediated selection against inbred soay sheep in a free-living, island population. evolution 53: 1259–1267. coulson, t., s. albon, j. slate, and j. pemberton. 1999. microsatellite loci reveal sex-dependent responses to inbreeding and outbreeding in red deer calves. evolution 53: 1951–1960. cornuet, j. m., and g. luikart. 1996. description and power analysis of two tests for detecting recent population bottlenecks from allele frequency data. genetics 144: 2001–2014. crnokrak, p., and d. a. roff. 1999. inbreeding depression in the wild. heredity 83: 260–270. di rienzo, a., a. c. peterson, j. c. garza, a. m. valdes, m. slatkin, and n. b. freimer. 1994. mutational processes of simple-sequence repeat loci in human populations. proceedings of the national academy of sciences usa 91: 3166–3170. eckert,c. g., k. e. samis, and s.c. lougheed. 2008. genetic variation across species’ geographical ranges: the central-marginal hypothesis and beyond. molecular ecology 17: 1170–1188. ellegren, h. 2004. microsatellites: simple sequences with complex evolution. nature reviews genetics 5: 435–445. evanno, g., s. regnaut, and j. goudet. 2005. detecting the number of clusters of individuals using the software structure: a simulation study. molecular ecology 14: 2611–2620. excoffier, l., and h. e. l. lischer. 2010. arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under linux and windows. molecular ecology resources 10: 564–567. falconer, d. s. 1981. an introduction to quantitative genetics. longman, london, england. farley, s., p. barboza, h. griese, and c. gardner. 2012. characterization of moose movement patterns, movement of black bears in relation to anthropogenic food sources, and wolf distribution and movement on jber lands, of elmendorf afb and fort richardson ap. alaska department of fish and game report, juneau, alaska, usa. franklin, i. r. 1980. evolutionary change in small populations. pages 135–149 in m. e. soulé and b. a. wilcox, editors. conservation biology: an evolutionaryecological perspective. sinauer associates, sunderland, massachusetts, usa. frankham, r. 1995. inbreeding and conservation: a threshold effect. conservation biology 9: 792–799. ———. 2015. genetic rescues of small inbred populations: meta-analysis reveals large and consistent benefits of gene flow. molecular ecology doi: http:// 10.1111/mec.13139. ———, c. j. a. bradshaw, and b. w. brook. 2014. genetics in conservation management: revised recommendations for the 50/500 rules, red list criteria and population viability analyses. biological conservation 170: 56–63. franzmann, a.w., and p. d. arneson. 1973. moose research center studies. alaska department of fish and game report, soldotna, alaska, usa. ———, and c. c. schwartz. 1985. moose twinning rates: a possible population condition assessment. journal of wildlife management 49: 394–396. ———, ———, and r. o. peterson. 1980. moose calf mortality in summer on the kenai peninsula, alaska. journal of wildlife management 44: 764–768. gaines, m. s., j. e. diffendorfer, r. h. tamarin, and t. s. whittam. 1997. the 82 population genetic structure – wilson et al. alces vol. 51, 2015 effects of habitat fragmentation on the genetic structure of small mammal populations. journal of heredity 88: 294–304. garcía-ramos, g., and m. kirkpatrick. 1997. genetic models of adaptation and gene flow in peripheral populations. evolution 51: 21–28. garza, j. c., and e. g. williamson. 2001. detection of reduction in population size using data from microsatellite loci. molecular ecology 10: 305–318. goudet, j. 1995. fstat (version 1.2): a computer program to calculate f-statistics. journal of heredity 86: 485–486. hampe, a., and r. j. petit. 2005. conserving biodiversity under climate change: the rear edge matters. ecology letters 8: 461–467. hartl, d. l., and a. g. clark. 2007. principle of population genetics, 4th edition. sinauer associates, sunderland, massachusetts, usa. hedrick, p. w. 2005. a standardized genetic differentiation measure. evolution 59: 1633–1638. hoffmann, a. a., and m. w. blows. 1994. species borders: ecological and evolutionary perspectives. trends in ecology and evolution 9: 223–227. ———, and p. a. parsons. 1991. evolutionary genetics and environmental stress. oxford university press, oxford, england. howes, b. j., and s. c. lougheed. 2008. genetic diversity across the range of temperate lizard. journal of biogeography 35: 1269–1278. hubisz, m. j., d. falush, m. stephens, and j. k. pritchard. 2009. inferring weak population structure with the assistance of sample group information. molecular ecology resources 9: 1322–1332. hughes, a. l. 2010. reduced microsatellite heterozygosity in island endemics supports the role of long-term effective population size in avian microsatellite diversity. genetica 138: 1271–1276. hundertmark, k. 2009. reduced genetic diversity in two introduced and isolated moose populations in alaska. alces 45: 137–142. klein, d. r. 1970. food selection by north american deer and their response to over-utilization of preferred plant species. pages 25–46 in a. watson, editor. animal populations in relation to their food sources. british ecological society symposium 10. blackwell, oxford, england. lacy, r. c. 1987. loss of genetic diversity from managed populations: interacting effects of drift, mutation, immigration, selection, and population subdivision. conservation biology 1: 143–158. leberg, p. 1993. strategies for population reintroduction: effects of genetic variability on population growth and size. conservation biology 7: 194–199. loranger, a. j., t. n. bailey, and w. w. larned. 1991. effects of forest succession after fire in moose wintering habitats on the kenai peninsula, alaska. alces 27: 100–109. lynch, c. b. 1977. inbreeding effects upon animals derived from a wild population of mus musculus. evolution 31: 526–537. mainguy, j., s. d. côté, and d. w. coltman. 2009. multilocus heterozygosity, parental relatedness and individual fitness components in a wild mountain goat, oreamnus americanus population. molecular ecology 18: 2297–2306. maruyama, t., and p. a. fuerst. 1985. population bottlenecks and non-equilibrium models in population genetics. ii. number of alleles in a small population that was formed by a recent bottleneck. genetics 111: 675–689. medrano, j. f., e. aasen, and l. sharrow. 1990. dna extraction from nucleated red blood cells. biotechniques 8: 43. meirmans, p. g. 2006. using the amova framework to estimate a standardized alces vol. 51, 2015 wilson et al. – population genetic structure 83 genetic differentiation measure. evolution 60: 2399–2402. merilä, j., and p. crnokrak. 2001. comparison of genetic differentiation at marker loci and quantitative traits. journal of evolutionary biology 14: 892–903. miller, m. j., e. bermingham, j. klicka, p. escalante, and k. winker. 2010. neotropical birds show a humped distribution of within-population genetic diversity along a latitudinal transect. ecological letters 13: 576–586. morton, j. m., m. bower, e. berg, d. magness, and t. eskelin. 2009. long term ecological monitoring program on the kenai national wildlife refuge, alaska: an fia adjunct inventory. pages 1–17 in w. mcwilliams, g. moisen, and r. czapiewski, compilers. proceedings of the forest inventory and analysis (fia)symposium 2008. rmrsp-56cd. usda forest service, rocky mountain research station, fort collins, colorado, usa. muhs, d., t. a. ager, and j. e. beget. 2001. vegetation and paleoclimate of the last interglacial period, central alaska. quaternary science reviews 20: 41–61. ohta, t., and m. kimura. 1973. a model of mutation appropriate to estimate the number of electrophoretically detectable alleles in a finite population. genetic research 22: 201–204. oldemeyer, j. l., a. w. franzmann, a. l. brundage, p. d. arneson, and a. flynn. 1977. browse quality and the kenai moose population. journal of wildlife management 41: 533–542. parmesan, c., and g. yohe. 2003. a globally coherent fingerprint of climate change impacts across natural systems. nature 421: 37–42. patten, m. a., and c. l. pruett. 2009. the song sparrow, melospiza melodia, as a ring species: patterns of geographic variation, a revision of subspecies, and implications for speciation. systematics and biodiversity 7: 33–62. pielou, e. c. 1991. after the ice age: the return of life to glaciated north america. university of chicago press, chicago, illinois, usa. piry, s., g. luikart, and j. m. cornuet. 1999. bottleneck: a computer program for detecting recent reductions in the effective population size using allele frequency data. journal of heredity 90: 502–503. pritchard, j. k., m. stephens, and p. donnelly. 2000. inference of population structure using multilocus genotype data. genetics 155: 945–959. quattro, j. m., and r. c. vrijenhoek. 1989. fitness differences among remnant populations of the sonoran topminnow, poeciliopsis occidentalis. science 245: 976–978. ralls, k., j. ballou, and r. l. brownell jr. 1983. genetic diversity in california sea otters: theoretical considerations and management implications. biological conservation 25: 209–232. rambaut, a., and a. j. drummond. 2007. tracer v1.4. < http://beast.bio.ed.ac.uk/ tracer> (accessed august 2013). reed, d. h., and r. frankham. 2001. how closely correlated are molecular and quantitative measures of genetic variation? a meta-analysis. evolution 55: 1095–1103. ———, and ———. 2003. correlation between fitness and genetic diversity. conservation biology 17: 230–237. robinson, s. j., l. p. waits, and i. d. martin. 2007. evaluating population structure of black bears on the kenai peninsula using mitochondrial and nuclear dna analyses. journal of mammalogy 88: 1288–1299. røed, k. h., and l. midthjell. 1998. microsatellites in reindeer, rangifer tarandus, and their use in other cervids. molecular ecology 7: 1773–1776. roffler, g. h., l. g. adams, s. l. talbot, g. k. sage, and b. w. dale. 2012. range overlap and individual movements 84 population genetic structure – wilson et al. alces vol. 51, 2015 http://beast.bio.ed.ac.uk/tracer http://beast.bio.ed.ac.uk/tracer during breeding season influence genetic relationships of caribou herds in southcentral alaska. journal of mammalogy 95: 1318–1330. root, t. l., j. t. price, k. r. hall, s. h. schneider, c. rosenzweig, and j. a. pounds. 2003. fingerprints of global warming on wild animals and plants. nature 421: 57–60. schlötterer, c. 2000. evolutionary dynamics of microsatellite dna. chromosoma 109: 365–371. schmidt, j. i., k. j. hundertmark, r. t. bowyer, and k. g. mccracken. 2009. population structure and genetic diversity of moose in alaska. journal of heredity 100: 170–180. schrey, a. w., m. grispo, m. awad, m. b. cook, e. d. mccoy, h. r. mushinsky, t. albaryrak, s. bensch, t. burke, l. k. butler, r. dor, h. b. fokidis, h. jensen, t. imboma, m. m. kesslerrios, a. marzal, i. r. k. stewart, h. westerdahl, d. f. westneat, p. zehtindjiev, and l. b. martin. 2011. broad-scale latitudinal patterns of genetic diversity among native european and introduced house sparrow (passer domesticus) populations. molecular ecology 20: 1133–1143. schwartz, c. c., and a. w. franzmann. 1989. bears, wolves, moose and forest succession, some management considerations on the kenai peninsula, alaska. alces 25: 1–10. schwartz, m. k., l. s. mills, y. ortega, l. f. ruggiero, and f. w. allendorf. 2003. landscape location effects genetic variation of canada lynx (lynx canadensis). molecular ecology 12: 1807–1816. silva, a. d., j.-m. gaillard, n. g. yoccoz, a. j. m. hewison, m. galan, t. coulson, d. allaine, l. vial, d. delorme, g. van laere, f. klein, and g. luikart. 2009. heterozygosity-fitness correlations revealed by neutral and candidate gene markers in roe deer from a long-term study. evolution 63: 403–417. sonsthagen, s. a., s. l. talbot, and c. m. white. 2004. gene flow and genetic characterization of northern goshawks breeding in utah. condor 106: 826–836. soulé, m. e. 1980. thresholds for survival: maintaining fitness and evolutionary potential. pages 151–169 in m. e. soulé and b. a. wilcox, editors. conservation biology: an evolutionary-ecological perspective. sinauer associates, sunderland, massachusetts, usa. tallmon, d. a., a. koyuk, g. h. luikart, and m. a. beaumont. 2008. onesamp: a program to estimate effective population size using approximate bayesian computation. molecular ecology resources 8: 299–301. testa, j. w., and g. p. adams. 1989. body condition and adjustments to reproductive effort in female moose (alces alces). journal of mammalogy 79: 1345–1354. tomasik, e., and j. a. cook. 2005. mitochondrial phylogeography and conservation genetics of wolverine (gulo gulo) of northwestern north america. journal of mammalogy 86: 386–396. von hardenberg, a., b. bassano, m. festabianchet, g. luikart, p. lanfranchi, and d. coltman. 2007. age-dependent genetic effects on a secondary sexual trait in male alpine ibex, capra ibex. molecular ecology 16: 1969–1980. vucetich, j. a., and t. a. waite. 2003. spatial patterns of demography and genetic processes across the species’ range: null hypotheses for landscape conservation genetics. conservation genetics 4: 639–645. wilson, g. a., and b. rannala. 2003. bayesian inference of recent migration rates using multilocus genotypes. genetics 163: 1177–1191. ———, c. strobeck, l. wu, and j. w. coffin. 1997. characterization of microsatellite loci in caribou rangifer tarandus, and their use in other artiodactyls. molecular ecology 6: 697–699. alces vol. 51, 2015 wilson et al. – population genetic structure 85 wisely, s. m., s. w. buskirk, g. a. russell, k. b. aubry, and w. j. zielinksi. 2004. genetic diversity and structure of the fisher (martes pennanti) in a peninsular and peripheral metapopulation. journal of mammaology 85: 640–648. yamashita, t., and g. a. polis. 1995. a test of the central-marginal model using sand scorpion populations (paruroctonus mesaensis, vaejovidae). journal of arachnology 23: 60–64. 86 population genetic structure – wilson et al. alces vol. 51, 2015 population genetic structure of moose (alces alces) of south-entral alaska methods sample collection molecular techniques analysis of genetic diversity and population genetic subdivision gene flow population demography results genetic diversity and population subdivision gene flow population demography discussion loss of genetic diversity between peninsula and mainland relationships within the peninsula conservation implications acknowledgements references moose antler morphology and asymmetry on isle royale national park kenneth j. mills1,3 and rolf o. peterson2 1department of biological sciences, michigan technological university, houghton, michigan, usa 49931; 2school of forestry and wood products, michigan technological university, houghton, michigan, usa 49931. abstract: isle royale national park, an island archipelago in lake superior, supports moose at higher density (1–4/km2) relative to most other north american sites. we compared antler size and asymmetry measurements from isle royale moose that died of natural causes to measurements available for other regional moose populations in published literature. we used these comparisons to test predictions that antlers of isle royale moose would be smaller and more asymmetric that other regional populations due to the high population density and the resulting ecological conditions on isle royale. moose on isle royale follow the same patterns of antler development as elsewhere, reaching maximum size at 7–8 years of age with slight declines after age 10–12. however, these moose develop antlers that are much smaller than all measured north american subpopulations. antler size was most comparable to moose from scandinavia where moose exist at comparably high population density. boone and crockett score, which is commonly used to compare antler size, performed poorly at ranking individuals with large antlers suggesting that more biologically relevant measures such as antler volume should be considered for comparisons of antler size. pedicle constriction was found to be a reliable indicator of senescence among old bulls. antler asymmetry was negatively related to antler size and was more extreme than asymmetry measured in alaskan moose. moose age had no detectable effect on the degree of antler asymmetry. in general, bull moose on isle royale develop smaller, more asymmetric antlers than other north american subpopulations which exist at lower density, consistent with the hypothesis that these qualities are related to nutrient limitation caused by high population density. results, however, may also reflect genetic differences and artifacts of sampling. alces vol. 49: 17–28 (2013) key words: alces alces, antler, asymmetry, development, isle royale national park, moose. moose (alces alces) develop large ant‐ lers during a relatively short growing period, requiring an intake of nutrients and expenditure of energy above that required for maintenance of basal functions (stewart et al. 2000). the ability to acquire and allocate resources necessary for antler development is influenced by factors such as age, body size, nutrition, genetics, and population and environmental conditions (sæther and haagenrud 1985, clutton-brock and albon 1989, markusson and folstad 1997, stewart et al. 2000, strickland and demarais 2000, bowyer et al. 2001, schmidt et al. 2001). as secondary structures in sexually dimorphic cervids, antlers have significance in sexual selection and are correlated with social dominance and mating success (cluttonbrock and albon 1989, bartoš 1990, solberg and sæther 1994, pélabon and joly 2000, stewart et al. 2000). these developmental, morphological, and sociobehavioral attributes allow antlers to be useful parameters in ecological research. 3present address: wyoming game and fish department, po box 850, pinedale, wy, 82941. kenneth.mills@wyo.gov 17 antler size typically increases until bulls reach maximum body and antler size between the ages of 5 and 10 years (stewart et al. 2000, bowyer et al. 2001). after age 10, antler size tends to decline (sæther and haagenrud 1985, bubenik 1990, bubenik 1998, stewart et al. 2000, bowyer et al. 2001), and simultaneously there is increasing evidence of physical senescence (hindelang and peterson 1994). age and body mass, then, both influence energetic investment in antler development (scribner and smith 1990). antler development patterns of isle royale moose that die of wolf predation and other natural causes will reflect overall nutritional condition as well as the culling influence of mortality factors. also, the large number of relatively old moose in the population (peterson 1977) should illuminate the poorly understood influence of senescence on antler development (bubenik 1998). asymmetry, defined as random deviations from perfect bilateral symmetry, is present to varying degrees in all bilateral morphological traits (palmer and strobeck 1986, bubenik 1990, bowyer et al. 2001). antlers are bilateral secondary structures and, therefore, portray differential degrees of asymmetry which depend on developmental stability, environmental quality, and individual fitness (e.g., nutritional status, inbreeding, injury, parasite load, age) and thus may be useful for comparisons between individuals and populations (palmer and strobeck 1986, clutton-brock and albon 1989, solberg and sæther 1994, alados et al. 1995, folstad et al. 1996, møller et al. 1996, markusson and folstad 1997, pélabon and van breukelen 1998, pélabon and joly 2000, bowyer et al. 2001, schmidt et al. 2001). antler asymmetry has an inverse relationship with antler size for many cervid species, which may be indicative of relative individual fitness regardless of age (markusson and folstad 1997, pélabon and van breukelen 1998, bowyer et al. 2001, ditchkoff et al. 2001). population wide stressors, such as reduced nutrition, may also manifest themselves through patterns in antler asymmetry and thus measures of antler asymmetry at broader scales may also be useful for comparisons between populations. reduced predator species diversity has allowed moose population density to reach uncommonly high levels on isle royale national park compared to most other north american subpopulations (peterson 1995, karns 1998, peterson et al. 2003), where a relative shortage of nutrition could reduce individual fitness and limit the ability of bull moose to allocate excess energy toward antler development (brown 1990). nutritional restriction due to high density may also manifest itself in the degree of antler asymmetry at the scale of the individual and the population (pélabon and van breukelen 1998, pélabon and joly 2000, bowyer et al. 2001). likewise, wolf predation and starvation are the only significant sources of mortality for moose on isle royale (peterson 1977, peterson 1999), so age structure and thus antler characteristics likely differ from other populations where antler morphology has been studied (gasaway et al. 1987, nygrén 2000, stewart et al. 2000, bowyer et al. 2001). therefore, antler characteristics may provide a basis for comparing condition and nutritional status of moose at isle royale and other geographic sites (bowyer et al. 2001). herein we assess antler size relative to age and antler asymmetry relative to age and antler size for bull moose collected on isle royale national park. we predict that patterns of antler development and asymmetry will follow similar general patterns measured for other north american populations. however, we also expect that antlers for moose on isle royale will be smaller and more asymmetric than other north american populations due to the nutritional 18 moose antler morphology – mills and peterson alces vol. 49, 2013 restriction caused by high population density (see also peterson et al. 2011). study area moose have existed on isle royale (544 km2) for the past century and in the last half-century they have been cropped by an unmanipulated population of gray wolves (canis lupus). both species have been protected since the establishment of isle royale national park in 1940 (mech 1966). wolf and moose populations have been counted each year since 1959. both predator and prey exist at relatively high density, with moose fluctuating from about 500 (1/km2) to over 2,000 (4/km2) animals during 1959– 2002, with a mean of 2.03 ± 0.11/km2 (se; range = 0.92–4.45/km2) during that period (peterson 1999, r. peterson, unpublished data). population densities for moose in other regions of north america are generally below 1/km2 (karns 1998). likewise, moose populations located on the nearest mainland in southwest ontario and northeast minnesota, the likely source for moose on isle royale, generally range from 0.20–0.40/km2 (mech 1966, karns 1998, ontario ministry of natural resources, unpublished data). methods skulls of male moose with polished antlers were collected during field studies at isle royale during 1970–2001. ages of moose were estimated from counts of annular cementum lines. antler size was measured in accordance with the boone and crockett club (b&c) scoring system (boone and crockett club 2011, gasaway et al. 1987). a net dry score for each set of antlers, tallied in inches, was calculated as follows: [spread + (2 � smallest palm length) + (2 � smallest palm width) + (2 � smallest beam circumference) + (2 � least number of points)] (see boone and crockett club 2011 for details on scoring methods). the remaining measurements were recorded in centimeters (gasaway et al. 1987). the largest diameter of both left and right pedicles on each skull was measured to study how this skull character varies with age. some pedicles showed an apparent constriction at the point where the antler joins the pedicle, which has not been described previously in the scientific literature (fig. 1). therefore, both constricted and unconstricted pedicle measurements were taken for these individuals in order to quantify this morphological trait. the constricted measurement was taken at the area of greatest constriction just before the antler base, while the unconstricted measurement was taken directly medial to the constricted area. scoring systems such as b&c may have limitations that affect the results of comparative studies (gasaway et al. fig. 1. constriction of the pedicle (outlined in white) just medial to the base of the antler was evident for many antlered bulls collected from isle royale national park. alces vol. 49, 2013 mills and peterson – moose antler morphology 19 1987, bubenik 1998). therefore, we also determined antler volume to directly measure antler size using water displacement. prior to measurement, each antler was saturated in water until all air pockets were filled prior to measurement. in order to measure the accuracy of this technique, we determined volume for 10 antlers, 3 times each. each individual measurement for each antler was compared to the mean of the 3 measurements for that antler to determine the error of each measurement. finally, the total mean error of the 30 measurements was calculated to confirm that the error was within acceptable limits (i.e., < 5%). we then compared two of the most used measures of antler size, b&c score and spread (boone and crockett club 2011), to the respective total volume measurement for each individual to determine the degree to which these scores accurately estimate antler size using exponential regression. second-order polynomial equations were fitted to data relating antler character size to moose age to evaluate variation in antler size with age and age-related growth of antlers compared to that of alaskan moose as measured by bowyer et al. (2001). a dunnett's test (zar 1999) was used to determine if the mean maximum sizes for the 20 largest isle royale moose for both b&c score and spread were smaller than the same measurements from multiple subpopulations of north american moose, as determined by gasaway et al. (1987), and moose from finland as determined by nygrén (2000). we also plotted comparative growth curves for isle royale moose, selected north american subpopulations, and a swedish subpopulation of moose as adapted from gasaway et al. (1987). growth curves were determined by using 3-year running averages except for the oldest and youngest age classes, which are presented as actual means. we pooled individuals in the 14 year age class and older for the isle royale subpopulation. relative antler asymmetry was determined by taking the difference between the large and small side of each measured antler parameter for each individual (i.e., palm width, palm length, beam circumference, number of points, pedicle diameter, and volume) divided by the respective large side for each measured antler parameter for that individual (e.g., [large palm width – small palm width] ÷ large palm width = relative asymmetry of the palm width for that individual moose). we then assessed the relationship between relative asymmetry and moose age using linear regression. we also used linear regression to measure the relationship between relative asymmetry and the mean size of the respective antler parameter. we used a one-sample t-test to compare the mean relative asymmetry for palm width, palm length, beam circumference, and number of points for isle royale moose to the mean relative asymmetry of the respective measures for alaskan moose as determined by bowyer et al. (2001). we tested whether asymmetry was fluctuating or directional for each lateral antler character using a wilcoxon signedrank test (see palmer and strobeck 1986, zar 1999, pélabon and joly 2000, bowyer et al. 2001). results the total number of skulls in the sample was 106, but not all parameters could be measured for some specimens because of weathering prior to collection. antlers for isle royale moose were smaller than alaskan subpopulations in palm width, palm length, beam circumference, number of points and spread (fig. 2, 3). for b&c score and spread, isle royale moose were smaller than all other north american subpopulations measured (all p <0.05; table 1). antler spread from isle royale 20 moose antler morphology – mills and peterson alces vol. 49, 2013 isle royale y = –0.2219× 2 + 4.6907× –3.5515 r2 = 0.2962 p<0.001 y = –0.6256×2 + 13.112× –12.102 r2 = 0.3778 p<0.001 y = –1.0131×2 + 20.658× +9.3598 r2 = 0.3638 p<0.001 y = –24.816×2 + 513.7× –419.44 r2 = 0.278 p<0.001 y = –0.0876×2 + 1.8086× +7.3762 r2 = 0.3642 p<0.001 y = –0.043×2 + 0.897× + 1.1549 r2 = 0.194 p<0.001 isle royale alaska p a lm w id th ( cm ) 45 40 35 30 25 20 15 10 5 0 25 20 15 10 5 0 4500 4000 3500 3000 2500 2000 1500 1000 500 0 p a lm le n g th ( cm ) b & c s co re ( in ) v o lu m e ( m l ) b e a m c ir cu m fe re n ce ( cm ) 110 100 90 80 70 60 50 40 30 20 10 0 160 140 120 100 80 60 40 20 0 age (years) 0 2 4 6 8 10 12 14 age (years) 0 2 4 6 8 10 12 14 n u m b e r o f p o in ts 12 11 10 9 8 7 6 5 4 3 2 1 0 fig. 2. regression analyses of antler characteristics in relation to age of bull moose collected from isle royale national park. raw data was used to generate a second order polynomial regression equation for isle royale moose. regression lines for alaskan moose were obtained from bowyer et al. (2001). sample sizes for the isle royale sample are as follows: palm width, n = 68; number of points, n = 74; palm length, n = 67; beam circumference, n = 91; b&c score, n = 64; volume, n = 68. alces vol. 49, 2013 mills and peterson – moose antler morphology 21 moose was also smaller than the palmate antler category from finland (∣q∣ = 3.1696, p <0.05), and was marginally different from the non-palmate antler category (∣q∣ = 1.9245, p ≈ 0.05; table 1). isle royale moose also appear to have maximum antler spread similar to that of moose from sweden, although raw data were not available for the swedish subpopulation (fig. 3). for moose at isle royale, maximum antler size is reached between the ages of 7 and 8 years for all measured parameters, except for b&c score, which reached its maximum at 6 years (fig. 2, 3). generally, a slight decrease in size occurred after 10–12 years of age, with incipient physi‐ cal senescence (fig. 2). this was evident by the malformed or misshapen antlers of several senescent individuals (see bubenik 1998). the volume measurement technique was determined to be accurate to within a mean of 1.9 ± 0.3% (range = 0.2–5.5%). age-related change in antler volume was similar to other size measurements, reaching a maximum at age 7, then decreasing more slightly after age 10 (fig. 2). the relationship between b&c score and total volume (left + right) was exponential and variable for individuals with high b&c scores (fig. 4a). antler spread also was exponentially related to total volume and was more variable as spread increased (fig. 4b). pedicle diameter portrayed the same antler development pattern as other parameters, reaching maximum size at 8 years (fig. 5a). however, it did not appear to decline as an indication of senescence as other parameters did. pedicle constriction was present in some moose as early as 7 years and increased with age to a maximum at 16–18 years (fig. 5b). the degree of relative asymmetry was not related to moose age for any bilateral antler parameter (all p > 0.458), but was negatively related to antler size for most bilateral antler categories including volume (f = 0.27, p = 0.002; fig. 6), palm width (f = 1.61, p = 0.000), beam circumference (f = 10.82, p = 0.001), and number of points (f = 0.74, p = 0.000). relative asymmetry had no relationship with antler size for palm length (f = 0.07, p = 0.799) or pedicle diameter (f = 0.15, p = 0.697) the degree of relative asymmetry for isle royale moose was much larger than in alaskan moose for palm length, palm width, and beam circumference but was not different for number of points (table 2). wilcoxon signed-rank tests showed that left and right antler sides were not different for palm length, palm width, beam circumference, number of points, volume, or pedicle diameter (z = 0.061, p = 0.952; z = 1.056, p = 0.291; z = 0.002, p = 0.998; z = −0.836, p = 0.403; z = 0.679, p = 0.497; z = 0.808, p = 0.419, respectively). fig. 3. comparative growth curves for selected north american subpopulations and a swedish subpopulation of moose as adapted from gasaway et al. (1987). curves are plotted by using 3-year running averages except for the oldest and youngest age classes, which are actual means. for the isle royale national park subpopulation (n = 76), individuals in the 14 year age class and older are pooled. 22 moose antler morphology – mills and peterson alces vol. 49, 2013 discussion population density for moose on isle royale, where there is predation only by gray wolves, is an order of magnitude higher than most other areas of north america (peterson 1999), but comparable to many moose ranges in scandinavia (0.8–1.8/km2; cederlund and markgren 1987, hörnberg 2001). isle royale moose, to a greater extent than other moose populations, are also subjected to strong selection by wolf predation, and are thereby more naturally regulated than other hunted populations. these two ecological characteristics make interpopulation comparisons involving moose at isle royale particularly compelling. however, it is necessary to address this difference in terms of sample selection when comparing datasets collected from individuals subjected to natural mortality and those collected from hunter-killed individuals. neither sample is randomly selected; in the case of isle royale, individuals were collected after death from natural causes, and so probably include proportionately higher numbers of individuals in poor condition and/or older age classes. with other datasets, individuals were measured table 1. antler spread and boone and crockett score for the 20 largest moose from selected regions of north america and finland. (data adapted from gasaway et al. 1987 and nygrén 2000). spread (cm) boone and crockett score subspecies/region mean se max. mean:max. mean se max. n gigas alaska1 182.6 2.64 207 0.88 247.1 0.71 255 20 gigas x andersoni yukon and northwest territories1 170.2 2.18 191.8 0.89 232.9 1.57 247.3 20 gigas x andersoni northern british columbia2 154.7 2.64 172.7 0.90 215.7 0.91 229.1 20 andersoni western canada (except north british columbia) and minnesota2 154.7 2.41 178 0.87 217.3 1.27 226.9 20 andersoni x americana ontario2 151.6 2.79 181.6 0.83 201.3 1.35 211.6 20 americana eastern canada and maine2 154.4 2.49 181.9 0.85 202.9 2.73 238.6 19 shirasi western usa3 133.9 2.69 151.9 0.88 188.2 1.73 205.5 20 andersoni isle royale2 107.0 3.18 129.4 0.83 133.4 2.04 151.7 20 alces finland palmate 114.8 0.46 149 0.77 511 nonpalmate 111.9 0.86 139 0.81 1considered alaska-yukon moose by boone and crockett club. 2considered canadian moose by boone and crockett club. 3considered shiras moose by boone and crockett club. alces vol. 49, 2013 mills and peterson – moose antler morphology 23 following hunter harvest, which would introduce biases based on hunter selection (e.g., hunter selection for larger than average bulls, antler size restrictions imposed by wildlife management agencies). the mean:maximum ratios presented in table 1 suggest that the regional datasets are likely similar, and therefore comparable. it is likely that the true maximum antler size realized by isle royale moose is larger than that presented herein, but cast antlers that are significantly larger than the largest represented in this dataset are rarely found during fieldwork on the island (r. peterson, michigan technological university, unpublished data). this evidence suggests that these datasets have at least acceptable levels of comparability, but comparisons should still be considered with a y = 369.46e0.0212x r 2 = 0.7808 p < 0.001 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140 160 boone & crockett score (in) t o ta l v o lu m e ( m l ) b y = 0.0019x3.196 r 2= 0.758 p < 0.001 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 spread (cm) t o ta l v o lu m e ( m l ) fig. 4. regression analyses of the relationship between total antler volume and (a) boone & crockett score (n = 65) or (b) spread (n = 65) for antlered bull moose collected from isle royale national park. a y = 0.0173x2 + 0.0085x – 0.0498 r 2 = 0.2987 p < 0.001 0 2 4 6 8 10 age (years) m e a n p e d ic le c o n st ri ct io n ( m m ) b y = –0.2584x2 + 6.6782x + 25.188 r 2 = 0.6205 p < 0.001 0 20 40 60 80 100 120 0 2 4 6 8 10 12 14 16 18 0 2 4 6 8 10 12 14 age (years) l a rg e st p e d ic le d ia m e te r (m m ) fig. 5. regression analysis for (a) pedicle constriction (n = 91) and (b) pedicle diameter (n = 95) in relation to age of antlered bull moose collected from isle royale national park. y = –5e-05x + 0.2754 r 2= 0.1403 p = 0.002 –0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 1000 2000 3000 4000 5000 6000 7000 8000 mean volume (ml) r e la tiv e a sy m m e tr y fig. 6. linear regression between relative asymmetry of left and right antler sides against mean antler volume for bull moose collected from isle royale national park (n = 68). 24 moose antler morphology – mills and peterson alces vol. 49, 2013 caution due to the potential biases caused by differences in sampling methodologies (e.g., sample sizes, sampling duration, sample collection protocols). moose present on isle royale develop smaller antlers than all other reported sub‐ populations in north america, and their antlers are similar to or smaller than two subpopulations reported for scandinavia. however, antler development through age follows much the same patterns as other populations, reaching a maximum size after 7 to 8 years, which is maintained until senescence at around age 12 (gauthier and larsen 1985, bowyer et al. 2001). the fact that isle royale moose appear to have a restricted ability to produce larger antlers should be a function of ecological conditions on the island, with nutrient limitation induced by high population density being the most fundamental difference between this island population and those in mainland areas. this is demonstrated when comparing antler size of isle royale moose to antler measurements collected from moose in southwest ontario. the maximum antler spread and b&c score for isle royale was 22.2 cm or 49.6 in smaller than the mean of the 19 largest moose measured from the mainland ontario population (table 1). this analysis suggests a significant reduction in antler size in the century following moose colonization on the island, with the primary difference between these groups being population density (karns 1998, peterson 1999, r. o. peterson, unpublished data, ontario ministry of natural resources, unpublished data). most comparative antler studies use composite scores of linear measures such as the b&c scoring system or simply antler spread. these scores are easy to calculate, but may have significant limitations (gasaway et al. 1987, bubenik 1998). we determined that b&c score and spread do not accurately rank large antlered individuals, in many cases ranking larger individuals below smaller individuals (figs. 4a and b). the b&c score, therefore, may have limited usefulness when comparing antler size between similar populations, especially when comparing primarily large antlered bulls. volume, on the other hand, should be a more accurate measure of antler size because it is directly related to energetic investment during antler development. this suggests that researchers should consider biologically relevant morphological metrics such as volume when conducting comparative studies on antlers. moose numbers on isle royale are naturally regulated with no human interference, which allows individual moose the opportunity to reach ages when signs of senescence would be expected. in most cases, the second order polynomials used in regression estimated reductions in size for older individuals. despite this, most measured antler parameters had only slight reductions in antler size for post-prime age individuals, table 2. comparison of mean relative asymmetry (ra; large small/large) for antler characters from 1,501 harvested alaskan moose and antlered bull moose collected from isle royale national park. data for antler characters from alaska were obtained from bowyer et al. (2001) and gasaway et al. (1987). alaska isle royale antler character ra se ra se n t p palm width 0.10 0.002 0.20 0.033 67 3.04 0.003 palm length 0.07 0.002 0.16 0.034 66 2.60 0.011 beam circumference 0.03 0.001 0.06 0.011 100 2.46 0.016 # points 0.19 0.005 0.20 0.026 74 0.44 0.660 alces vol. 49, 2013 mills and peterson – moose antler morphology 25 which was consistent for moose measured in alaska (bowyer et al. 2001). however, there was a small proportion of old and senescent individuals that developed small and drastically asymmetric antlers (see bubenik 1998). pedicle constriction may be a better indicator of declining reproductive vigor in older individuals. pedicle constriction was observed in both large and small antlered individuals as well as individuals with normal and abnormal antler morphology. constriction was first apparent in some bulls that were 7 years of age, the same age that antlers begin to reach their maximum, mature size, and it increased with age, though not all older individuals had measurable restrictions. a. b. bubenik (pers. commun.) suggested that pedicle constriction resulted from testosterone insufficiency, which may begin well after sexual maturity and increase with reproductive senescence. antler asymmetry for moose on isle royale was fluctuating and was most pronounced among moose with small antlers at the extremes of age and development. although some older, senescent individuals developed very small and asymmetric antlers (see bubenik 1998), overall there was little evidence to suggest that age has any governing effect on antler asymmetry. therefore, antler asymmetry should be a valid indicator of individual fitness and condition regardless of age, with the individuals in the best condition developing the largest and most symmetric antlers. likewise, asymmetry may also provide a basis for comparisons of fitness and condition between populations. in this case, isle royale moose portrayed greater degrees of relative asymmetry than alaskan subpopulations, the only subpopulation for which asymmetry measurements were available (bowyer et al. 2001). high levels of antler asymmetry population-wide, as measured for moose from isle royale, may reflect more nutrient limitation and developmental instability. in general, bull moose on isle royale develop smaller, more asymmetric antlers than other north american subpopulations, even those within the same geographic region, suggesting that these qualities are the result of nutrient limitation caused by high population density (peterson et al. 2011). these findings are consistent with the evidence of slight dwarfism associated with high population density and lack of selection by wolf predation during the first half of the 20th century (peterson et al. 2011). however, this correlative study did not quantify or eliminate other potential contributing factors, such as genetic founder effects and effects of sampling methodology. this study also supports the contention that antlers are useful indicators for both individual and population condition (e.g., markusson and folstad 1997, pélabon and van breukelen 1998, strickland and demarais 2000, schmidt et al. 2001), although future research should attempt to specifically evaluate fitness in relation to measures of antler size and asymmetry for moose. acknowledgements we thank p. dewitt, e. parker, c. peterson, and d. suhonen for assistance in antler measurements, and financial assistance from the u.s. national science foundation (deb-0918247), the u.s. national park service (co-op agreement no. j631005n0040003), and rop, the robbins chair at michigan technological university in sustainable management of the environment. references alados, c. l., j. escós, and j. m. emlen. 1995. fluctuating asymmetry and fractal dimension of the sagittal suture as indicators of inbreeding depression in dama and dorcas gazelles. canadian journal of zoology 73: 1967–1974. bartoš, l. 1990. social status and antler development. pages 442–459 in 26 moose antler morphology – mills and peterson alces vol. 49, 2013 g. a. bubenik and a. b. bubenik, editors. horns, pronghorns, and antlers: evolution, morphology, physiology and social significance. springer-verlag, new york, new york, usa. boone and crockett club. 2011. http:// www.boone-crockett.org/pdf/sc_moose. pdf <accessed june 2012>. bowyer, r. t., k. m. stewart, j. g. kie, and w. c. gasaway. 2001. fluctuating asymmetry in antlers of alaskan moose: size matters. journal of mammalogy 82: 814–824. brown, r. d. 1990. nutrition and antler development. pages 426–441 in g. a. bubenik and a. b. bubenik, editors. horns, pronghorns, and antlers: evolution, morphology, physiology and social significance. springer-verlag, new york, new york, usa. bubenik, a. b. 1990. epigenetical, morphological, physiological, and behavioral aspects of evolution of horns, pronghorns, and antlers. pages 3–113 in g. a. bubenik and a. b. bubenik, editors. horns, pronghorns, and antlers: evolution, morphology, physiology and social significance. springer-verlag, new york, new york, usa. ———. 1998. evolution, taxonomy and morphophysiology. pages 77–124 in a. w. franzmann and c. c. schwartz, editors. ecology and management of the north american moose, 1st edition. smithsonian institution press, washington, d. c., usa. cederlund, g., and g. markgren. 1987. the development of the swedish moose population, 1970–1983. swedish wildlife research supplement. 1: 55–61. clutton-brock, t. h., and s. d. albon. 1989. red deer in the highlands. bsp professional books, oxford, england. ditchkoff, s. s., r. l. lochmiller, r. e. masters, w. r. starry, and d. m. leslie jr. 2001. does fluctuating asymmetry of antlers in white-tailed deer (odocoileus virginianus) follow patterns predicted for sexually selected traits. proceedings of the royal society of london, b 268: 891–898. folstad, i., p. arneberg, and a. j. karter. 1996. antlers and parasites. oecologia 105: 556–558. gasaway, w. c., d. j. preston, d. j. reed, and d. d. roby. 1987. comparative antler morphology and size of north american moose. swedish wildlife research supplement 1: 311–325. gauthier, d. a., and d. g. larsen. 1985. geographical variation in moose antler characteristics, yukon. alces 21: 91–101. hindelang, m., and r. o. peterson. 2000. skeletal integrity in moose at isle royale national park: bone mineral density and osteopathology related to senescence. alces 36: 61–68. hörnberg, s. 2001. changes in population density of moose (alces alces) and damage to forests in sweden. forest ecology and management 149: 141–151. karns, p. d. 1998. population distribution, density and trends. pages 125–140 in a. w. franzmann and c. c. schwartz, editors. ecology and management of the north american moose, 1st edition. smithsonian institution press, washington, d. c., usa. markusson, e., and i. folstad. 1997. reindeer antlers: visual indicators of individual quality? oecologia 110: 501–507. mech, l. d. 1966. the wolves of isle royale. united states national park service fauna series number 7. united states government printing office, washington, d. c., usa. møller, a. p., j. j. cuervo, j. j. soler, and c. zamora-muñoz. 1996. horn asymmetry and fitness in gemsbok, oryx g. gazella. behavioral ecology 7: 247–253. nygrén, k. 2000. directional asymmetry in moose. alces 36: 147–154. palmer, a. r., and c. strobeck. 1986. fluctuating asymmetry: measurement, analysis, pattern. annual review of ecology and systematics 17: 391–421. alces vol. 49, 2013 mills and peterson – moose antler morphology 27 http://www.boone-crockett.org/pdf/sc_moose.pdf http://www.boone-crockett.org/pdf/sc_moose.pdf http://www.boone-crockett.org/pdf/sc_moose.pdf pélabon, c., and l. van breukelen. 1998. asymmetry in antler size in roe deer (capreolus capreolus): an index of individual and population fitness. oecologia 116: 1–8. ———, and p. joly. 2000. what, if anything, does visual asymmetry in fallow deer antlers reveal? animal behaviour 59: 193–199. peterson, r. o. 1977. wolf ecology and prey relationships on isle royale. national park service scientific monograph series no. 11. united states government printing office, washington, d. c., usa. ———. 1995. the wolves of isle royale: a broken balance. willow creek press, minocqua, wisconsin, usa. ———. 1999. wolf-moose interaction on isle royale: the end of natural regulation? ecological applications 9: 10–16. ———, d. beyer, m. schrage, and j. räikkönen. 2011. phenotypic variation in moose: the island rule and the moose of isle royale. alces 47: 125–133. ———, j. a. vucetich, r. e. page, and a. chouinard. 2003. temporal and spatial aspects of predator-prey dynamics. alces 39: 215–232. sæther, b.-e., and h. haagenrud. 1985. geographical variation in the antlers of norwegian moose in relation to age and size. journal of wildlife management 49: 983–986. schmidt, k. t., a. stien, s. d. albon, and f. e. guinness. 2001. antler length of yearling red deer is determined by population density, weather and early lifehistory. oecologia 127: 191–197. scribner, k. t., and m. h. smith. 1990. genetic variability and antler development. pages 460–473 in g. a. bubenik and a. b. bubenik, editors. horns, pronghorns, and antlers: evolution, morphology, physiology and social significance. springer-verlag, new york, new york, usa. solberg, e. j., and b.-e. sæther. 1994. male traits as life-history variables: annual variation in body mass and antler size in moose (alces alces). journal of mammology 75: 1069–1079. stewart, k. m., r. t. bowyer, j. g. kie, and w. c. gasaway. 2000. antler size relative to body mass in moose: tradeoffs associated with reproduction. alces 36: 77–83. strickland, b. k., and s. demarais. 2000. age and regional differences in antlers and mass of white-tailed deer. journal of wildlife management 64: 903–911. zar, j. h. 1999. biostatistical analysis. 4th edition. prentice hall, upper saddle river, new jersey, usa. 28 moose antler morphology – mills and peterson alces vol. 49, 2013 moose antler morphology and asymmetry on isle royale national park study area methods results discussion acknowledgements references alces27_8.pdf alces22_69.pdf alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces29_219.pdf alces22_245.pdf alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces(25)_48.pdf alces20_187.pdf alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces24_10.pdf alces vol. 45, 2009 heikkilä & tuominen moose in conserved forests in finland 49 the influence of moose on tree species composition in liesjärvi national park in southern finland risto heikkilä1 and marita tuominen2 1finnish forest research institute, vantaa research unit, p.o. box 18, fi-01301 vantaa, finland 2university of helsinki, faculty of forest ecology, p.o. box 27, fi-00014 helsinki, finland abstract: intensive forest management has promoted a rapid increase in finland’s moose (alces alces) population since the 1970s. the main objective of this study was to determine the role of moose browsing in modifying natural processes of protected forests that are influenced by high moose populations in adjacent managed forests. this study occurred in liesjärvi national park located in the mid-boreal vegetation region of finland. forest stands were sampled with line-plot sampling (50 m² plots at 100 m distances) in the older (oa; 1956) and newer (na; 2005) parts of the park. we found that long-term selective browsing in oa retarded the development of young stands in favor of norway spruce (picea abies) and low-growing broadleaf species. browsing in recent years was relatively intensive in na where young regeneration areas still existed from previous forest management. the most intensive browsing occurred on 18.6 % of trees in na and 3.1 % in oa; young palatable tree species were taller in na than oa. also, in oa the density of preferred aspen (populus tremula) and rowan (sorbus aucuparia) trees was relatively low in the height class that produces the dominant tree canopy. despite short-term intensive browsing, na appeared better able to recover to a natural forest state. fecal pellet groups associated with young scots pine (pinus sylvestris) and browsing of birch (betula spp.) and aspen indicated the importance and role of forage quantity and quality on winter range of moose. the amount of consumed new twig biomass was 20-fold greater in na compared to oa, indicating a difference in the size of the moose population and presumably habitat quality between the areas. the effect of browsing on different tree species was measured at the stand level in oa in an area restored with prescribed burning 11 years previous. comparative measurements in two exclosures and adjacent open areas indicated that regeneration in the burned area was browsed intensively and growth of young trees was retarded, except spruce. the major impacts of browsing on aspen and rowan identify the need for new approaches to maintain forest diversity. a crucial issue will be the contradiction between preferred and sustained high moose harvests and the desire for natural forest diversity in conservation areas. alces vol. 45: 49-58 (2009) key words: alces alces, browsing, tree species, diversity, moose, forest conservation. moose (alces alces) herbivory plays an essential role in the dynamics of natural forests (risenhoover and maass 1987, pastor and naiman 1992, persson et al. 2000). in intensively managed boreal forests of finland, natural or unmanaged ecosystems are maintained in relatively small conservation areas where natural processes are the sole disturbance factor. even during a long unmanaged history, the characteristics of conserved forests are probably affected to some degree by outside influences. for example, forest management over large areas has favored moose populations by creating abundant regeneration similarly as natural disturbance processes such as fire (linder et al. 1997, lavsund et al. 2003). continuous forest management more effectively maintains moose populations than the sporadic occurrence of forest fires occurring irregularly in natural forests. consequently, the moose density in natural forests adjacent to managed forests is often higher than expected relative to normal regeneration rate and turnover of a natural forest. moose in conserved forests in finland heikkilä & tuominen alces vol. 45, 2009 50 the impacts on vegetation by moose, expressed through selective browsing (bergström and hjeljord 1987), have become acute in the case of certain tree species like balsam fir (abies balsamea) and aspen (populus spp.) in north america (brandner et al. 1990, kay 1997) and fennoscandia (kouki et al. 2004). attention has been paid to the question of population overabundance and generally to the role of moose as a disturbance factor in managed forests (gill 1992, edenius et. al 2002). in the long term, high-density moose populations can damage forest habitats in the absence of predation or human control (mclaren et al. 2004). heikkilä et al. (2003) suggested that browsing can cause considerable change in the early successional habitats of managed and natural forests in finland. nature conservation preserves are often located in close proximity to managed forests that are occupied by moose. depending on when preserves are established, they may reflect characteristics of previous forest management and fluctuations in the moose population. restoration of managed forests for conservation purposes often requires treatment strategies that allow natural processes (e.g., burning). such work has occurred in both older and more recently conserved areas in finland. the effects of the traditional, annual concentration of moose on their winter range (lavsund 1987, andersen 1991) may be important in this respect, because young conserved areas may still retain the characteristics of managed forests preferred by moose. further, intensive browsing may occur in older conserved areas from moose seeking sanctuary from nearby hunted areas. it has been suggested that some hunting activity should be permitted in conserved forest areas in finland (ympäristöministeriö 2006). in this study we analyzed the effects of moose browsing on a conserved forest area currently being restored in liesjärvi national park in southern finland. it was hypothesized that moose browsing may alter tree species composition and forest diversity. we questioned whether possible changes were unnatural and conflicted with goals to maintain this natural conservation area. study area the study area was the liesjärvi national park established in southern finland in 1956 (fig. 1). part of the original 660 ha area had been under forest management resulting in middle-aged and old closed forest; no inspection was made to classify the structure of the forest afterward. without disturbance factors, norway spruce (picea abies) became dominant over less shade tolerant broadleaved species and scots pine (pinus sylvestris). shaded undergrowth trees were common with silver birch (betula pubescens) and downy birch (betula pubescens) more dominant fig. 1. the location of liesjärvi national park in southern finland. the geobotanical vegetation zones are after kalliola (1973: 1 = hemiboreal zone, 2 = southern boreal zone, 3 = middle boreal zone, and 4 = northern boreal zone. alces vol. 45, 2009 heikkilä & tuominen moose in conserved forests in finland 51 than aspen and rowan (sorbus aucuparia). new regeneration occurred mostly as shaded undergrowth in small natural gaps and at forest edges. in 1993 an area of about 3 ha was burned to restore a dry forest site dominated by scots pine. in 2005 the park was enlarged to 2,200 ha by adding a new area where intensive forest management had not occurred for about 20 years. despite initial restoration measures, the new area (na) was less natural than the original, older area (oa) relative to forest age structure, although it had denser young stages and an increased proportion of moist sites. the characteristics of a managed forest were still prominent in the forest stands of different age classes. the average density of moose in the area was 3.1-4.0 moose/1,000 ha (ruusila et al. 2006). hunter surveys (information from game management association of tammela) indicated that the local winter density of moose after hunting was 4.6 moose/1,000 ha (range 4.3-5.1) in 2000-2005. methods forest inventory data, browsing intensity, and pellet groups were measured in both the oa and na portions of the park in 2005. we established 50 m2 circular plots on parallel transects with a distance of 100 m between transects and plots. because sampling in the oa was intended to be relatively intensive, the plot size was reduced to 20 m2 to measure all trees within a plot; trees >6 m high and fecal pellet groups were measured in the 50 m2 plots. the height of all tree species ≤6 m was measured to an accuracy of +10 cm. the diameter at breast height (1.3 m) was measured on trees >6 m high. to assess forest structure, the abundance of tree species was calculated in 4 height categories: ≤ 2 m, 2.14 m, 4.1-6 m, and >6 m. five tree condition categories were described: healthy, lightly damaged (no effect on growth), moderately damaged (growth loss obvious), seriously damaged (retarded growth and development), and dead. browsing intensity, including shoot and bark damage, was scaled in 4 categories according to a visual estimate of lost biomass: 0-25%, 26-50%, 51-75%, and >75%. the most intensively utilized trees were classified as bushy deformed. fresh moose browsing points on side shoots were counted to estimate consumed biomass according to the bite diameter and weight calculations for tree species (heikkilä and härkönen 1993); new and old main stem breakages were counted. old and new bites and breakages were distinguished according to the point and color of browsing. new (lying on the forest litter) and old fecal pellet groups (at least 20 pellets) were counted to compare presence and habitat use of moose in the study areas (neff 1968, härkönen and heikkilä 1999). we also used 2 exclosures built in the oa 10 years earlier where a 3 ha area was restored by burning. data were collected both inside and outside of 9, 20 m² sample plots in the 25 m x 25 m exclosures. height of tree species was measured according to the landscape-level inventory. statistical analyses were performed using sas program version 9.3.1 (sas institute, 2005), and comparisons between areas were made with a non-parametric mann-whitney u-test. spearman rank correlation at the sample plot level was used to analyze habitat selection according to fecal pellet groups and characteristics of the forests. height differences between exclosures and open areas were analyzed with one-way anova (spss package). results norway spruce and downy birch dominated the tree species composition in oa, and spruce, scots pine, silver birch, and other broadleaf species were abundant in na. more willow, rowan, and pine occurred in na (table 1). the availability of palatable trees (excluding spruce) ≤6 m in height was 447 trees/ha moose in conserved forests in finland heikkilä & tuominen alces vol. 45, 2009 52 greater in oa than in na. the mean height of trees ≤6 m was greater in na than in oa for all species except grey alder (alnus incana) (table 2). the number of trees ≤4 m was 9060/ha in oa and 6060/ha in na; about one third was spruce in both areas. the total tree density in na was twice that in oa for trees in the 2.1-4 m height category (1633/ha vs. 826/ha); pines and broadleaf trees accounted for 76% of trees in na and 62% in the more spruce-dominated oa. the proportion of trees ≤2 m was considerably higher in oa than in na (88% vs. 70%), and their density in oa was twice that in na. moose browsing occurred on 25% of trees in oa, whereas 42% of trees were browsed in na (fig. 2). intensive browsing (>75%) occurred on 18.6% of trees in na and 3.1% in oa. browsing was 30% on scots pine in both areas. aspen, rowan, and willows were intensively browsed in na. bushy deformed trees totaled 224 trees/ha in oa, and included pine, aspen, downy birch, rowan, and willows. in na only 48 trees/ha were bushy deformed of which >90% were willows and the rest rowans. the mean bite diameter was 2.76 oa na u-value p-value pine 904 ± 162 1550 ± 146 6.7290 <0.0001 spruce 4655 ± 948 2334 ± 230 -1.8487 0.0645 silver birch 298 ± 70 343 ± 64 3.5385 0.0004 downy birch 2441 ± 626 1898 ± 238 2.1439 0.0320 aspen 159 ± 39 217 ± 55 1.3537 0.1758 willows 255 ± 165 441 ± 106 4.5307 <0.0001 rowan 928 ± 168 1490 ± 211 4.2995 <0.0001 juniper 57 ± 52 23 ± 14 1.4411 0.1495 grey alder 46 ± 23 188 ± 65 3.6821 0.0002 other spp 23 ± 14 table 1. mean density (± se) of tree species (trees/ha) in liesjärvi national park, southern finland, 2005. the park was established in 1956 and consisted of two areas with different histories of forest management and moose population density; the original area was designated oa (old area), and a new area added in 2005 was designated na (new area). oa na u-value p-value pine 109.8 ± 5.0 193.0 ± 5.6 -11.6159 <0.0001 spruce 91.7 ± 2.8 133.3 ± 3.0 19.6262 <0.0001 silver birch 188.8 ± 13.2 242.2 ± 8.4 -5.0225 <0.0001 downy birch 118.1 ± 2.8 213.4 ± 4.1 20.6175 <0.0001 aspen 76.4 ± 8.0 151.6 ± 7.6 -7.4651 <0.0001 willows 63.7 ± 2.3 157.6 ± 5.2 -14.2084 <0.0001 rowan 76.7 ± 3.6 141.2 ± 2.6 -18.0767 <0.0001 juniper 58.1 ± 2.9 161.6 ± 13.6 4.0430 <0.0001 grey alder 237.6 ± 35.7 181.3 ± 12.5 0.9246 0.3552 other spp 65.0 ± 13.0 table 2. the mean (± se) height (cm) of tree species ≤6 m high in liesjärvi national park, southern finland, 2005. the park was established in 1956 and consisted of two areas with different histories of forest management and moose population density; the original area was designated oa (old area), and a new area added in 2005 was designated na (new area). alces vol. 45, 2009 heikkilä & tuominen moose in conserved forests in finland 53 mm (± 0.19 se) in na and 2.03 mm (± 0.21 se) in oa. the lowest diameter occurred on downy birch (1.64 mm ± 0.11 se) and the highest on pine (3.52 mm ± 0.11 se). thicker than average bites were taken from rowan and pine in both areas, and from aspen and alder in na. the estimated consumption of new twig biomass was 2.44 kg/ha in na and 0.09 kg/ha in oa. the number of old and new stem breakages was higher in oa than na (table 3). the stem damage/tree was consistently higher in oa than na for all tree species except juniper (juniperus communis). rowan had the highest number of breakages per tree in both oa (9.3) and na (0.9). silver birch, willows, and aspen were also broken more than once in oa. in general, the ≤6 m high trees in oa were less seriously affected than those in na. the proportion of dead trees was 16% in na and 6.5% in oa. of species preferred by moose, 65% of aspen, rowan, and willows were affected in both oa and na; dead trees were 7% of the total in oa and 30% in na. the oa na u-value p-value new twig browsing/ha 909.0 ± 48.9 155.0 ± 9.5 2.8777 0.0040 old stem breakage/ha 5294.0 ± 256.2 2111.0 ± 159.0 -24.5731 <0.0001 new stem breakage/ha 67.0 ± 4.0 50.0 ± 2.0 5.6691 <0.0001 stem breakage/tree 2.9 ± 0.1 0.4 ± 0.0 12.5000 0.0110 pellet groups/ha 18.6 ± 6.3 80.3 ± 20.5 4.0442 <0.0001 table 3. damage associated with moose browsing and total number of fecal pellet groups in liesjärvi national park, southern finland, 2005. the park was established in 1956 and consisted of 2 areas with different histories of forest management and moose population density; the original area was designated oa (old area), and a new area added in 2005 was designated na (new area). fig. 2. proportions of 4 categories of moose browsing intensity on tree species in liesjärvi national park in southern finland. the park was established in 1956; the original area was designated oa (old area), and a new area added in 2005 was designated na (new area). moose in conserved forests in finland heikkilä & tuominen alces vol. 45, 2009 54 proportion of live aspen was 75% in oa (126 aspens/ha) and 62% in na (114 aspens/ha). most live trees 4.1-6.0 m tall (313/ha in oa and 260/ha in na) were spruce and downy birch (79%) in oa, and downy birch and pine (67%) in na. all aspen (2.1/ha) were dead and willows were absent in oa. in na 5.9 aspens/ha were live of which 4.4/ha were injured; no dead aspens were found. in oa 16.3 rowans/ha were healthy or injured, whereas 12.0 rowans/ha were injured and 3 rowans/ ha were dead in na. grey alder was either injured or dead in both oa and na. conifers were the dominant trees >6 m tall in both oa (81%) and na (70%). aspen, rowan, and willows accounted for 2% of trees >6 m in oa and 3.5% in na. in oa 13% of the 48 aspens/ha were dead; 4% of 33 aspens/ha were dead and 22% affected by bark stripping in na. the maximum stem diameter of aspen was ≤20 cm and that of rowan and willows ≤15 cm. the diameter of the largest conifers was 46-50 cm in oa and 36-40 cm in na. the abundance of fecal pellet groups indicated that moose used na more than oa during the previous winter (80.3 groups/ha ± 20.5 se vs. 18.6 ± groups/ha ± 6.3 se, mann-whitney u = 4.0442, p <0.0001). in oa the number of fecal pellet groups correlated positively with the number of pines <2.5 m tall (r = 0.21, p = 0.004), the number of stem breakages of silver birch (r = 0.318, p <0.0001) and aspen (r = 0.168, p = 0.023), and the total number of moose-affected pines and downy birches (r = 0.465, p <0.0001). in na the number of pellet groups correlated positively with the number of pines <2.5 m tall (r = 0.198, p = 0.02). the fresh browsing of pine correlated with pellet groups (r = 0.332, p <0.0001); fresh willow browsing was nearly significant (p = 0.08). the number of pellet groups increased with the number of stem breakages on pine (r = 0.210, p = 0.016) and silver birch (r = 0.216, p = 0.011), and with rowans that were seriously damaged (r = 0.179, p = 0.037) or dead (r = 0.250, p = 0.003). measurements within the exclosures indicated that moose reduced (p <0.05) the growth (height) of pine, silver birch, and aspen during the 11 year period post-burn (fig. 3). downy birch was uncommon and spruce was not browsed. aspen was 1 m higher in the exclosures, and rowan did not occur outside the exclosure. willows were absent inside the exclosures. silver birch and pine combined accounted for about 80% of the total sapling density per ha, about 7,500 inside and 10,500 outside the exclosures. discussion liesjärvi national park in finland provides an excellent example of how previous and current forest and moose management can influence natural forest succession in a conservation area. the management history of the original portion of the park (oa) was different from that added later (na), both in forest management and moose population density. the forests in oa developed with minimal disturbance in the 1950-1970s when the moose population density was relatively low thereby allowing tree species typical of moose forage to grow. the subsequent rapid increase in moose population density and the forest management history in na (added 0 50 100 150 200 250 pinus sylvestris picea abies betula pendula betula pubescens populus tremula inside outside height, cm *** *** *** fig. 3. height of tree species inside and outside two exclosures in a forest restoration treatment area in oa, 10 years post-burn, liesjärvi national park, southern finland. alces vol. 45, 2009 heikkilä & tuominen moose in conserved forests in finland 55 in 2005) created the structural differences evident in forests in oa and na. the longterm impact from browsing that occurred after the 1970s retarded the development of young trees in oa, and created a low-growing community of palatable tree species (table 2; risenhoover and maass 1987, abaturov and smirnov 2002). the recent intensive browsing in na (>75 % browsing of 18.6 % of trees vs. 3.1% in oa) is reflected in the high proportion of injured and dead trees. in this short time na experienced more damage than oa (table 2, fig. 2). however, the availability of palatable trees in na was greater than in oa, especially in the 2.1-4.0 m height class that contains relatively abundant moose forage (parker and morton 1978, heikkilä and härkönen 1998). in the absence of past disturbances other than browsing, the oa forest developed towards spruce dominance (table 1). this trend gradually reduced habitat value for moose that prefer highly available deciduous forage (saether and andersen 1990, ball et al. 2000) that is associated with high moose population densities. however, an abundant moose population affects tree species diversity in ways that are difficult to predict (mclaren et al. 2004). because the local management goal is to maintain moose populations at levels that provide widespread hunting opportunity, understanding browsing impacts by high populations is critical. our data indicate that browsing pressure within a conservation area needs to be related to management goals in surrounding managed forests to best ascertain the effects on biodiversity within conservation areas. the high pellet group density in na indicated that a concentrated winter population of moose may reduce and alter diversity of tree species (andersen 1991). generally, the pellet group density increased with browsing intensity which is consistent with earlier findings (heikkilä and härkönen 1993). in both oa and na the number of pellet groups correlated positively with the density of young pine that is palatable winter browse (lundberg et al. 1990). the positive correlation between pellet groups and stem breakage on aspen in oa indicated higher and more frequent use by moose than in conserved forests in koli national park in eastern finland where moose used aspen much less (härkönen et al. 2008). gap dynamics may keep aspen inaccessible to moose and completely avoid browsing damage (syrjänen et al. 1994, cumming et al. 2000, edenius et al. 2002). local moose population density may also influence whether aspen escapes browsing damage. the risk of reduced biodiversity is recognized in managed forests located in highdensity moose winter range (heikkilä and härkönen 1993). most concern is for rowan and especially aspen, one of the most threatened tree species (kouki et al. 2004). although rowan was relatively abundant in liesjärvi national park, only a few individuals grew beyond the browsing height of moose. because rowan does not resist intensive browsing well, it commonly remains low-growing (saether 1990). rowan disperses widely even from a few seed producing trees, whereas aspen reproduces mainly from suckering (zackrisson 1985). aspen was much less abundant than rowan in both oa and na, and only a few were >2 m-≤ 6 m high. no live aspen was found in the 4.1-6.0 m height class in oa indicating a lack of recruitment of dominant trees and risk for future forest diversity. there are several ways of enhancing the diversity of intensively browsed tree species beyond population management of large ungulates, for example, as in aspen communities of conserved forests in north america (suzuki et al. 1999, kaye et al. 2005, romme et al. 2005). the strategy of aspen to resist ungulate browsing presupposes favorable conditions for regeneration. small-scale restoration by burning at the stand level was attempted in liesjärvi national park to compensate for moose in conserved forests in finland heikkilä & tuominen alces vol. 45, 2009 56 the lack of natural disturbance. however, no young aspens in the burned area grew beyond moose browsing after 10 years. rowan, preferred by moose year-round, was also absent due to intensive browsing, and even low-density moose populations can influence tree species composition. our results suggest that a small-scale restoration needs to be supplemented with protection against moose browsing; partial fencing might ensure aspen regeneration (mclaren et al. 2004, edenius and ericsson 2007). one crucial issue is the conflict between a sustained high harvest desired by hunters and the need for natural high forest diversity. diversity may vary widely in conserved areas (kouki et al. 2004); for example, härkönen et al. (2008) reported that browsed aspens recover well at low moose density. however, moose populations often impact development of conserved forests by retarding primary succession and causing conifer dominance (davidson 1993). further, the absence of natural disturbance results in a closed spruce-dominated boreal forest (linder et al. 1997). high density moose populations require large-scale disturbances to create preferred regeneration habitat to maintain balance between browsing pressure and forage availability. our data indicate that the continuous selective browsing pressure by moose in liesjärvi national park gradually reduced forage diversity and availability. it is clear that potentially threatened species and the composition and availability of all forage trees need to be addressed in management plans because management practices to date have not prevented a critical drop in forest health and diversity in liesjärvi national park. a cooperative decision-making process among adjacent landowners and moose managers is needed to help establish and maintain natural development in recently conserved forest areas. acknowledgements we would like to thank mr. jorma sillanpää for the help in carrying out field work, and 2 anonymous referees for their valuable comments on the manuscript. references abaturov, b. d., and k. a. smirnov. 2002. effects of moose population density on development of forest stands in central european russia. alces supplement 2: 1-5. andersen, r. 1991. habitat deterioration and the migratory behavior of moose (alces alces l.) in norway. journal of applied ecology 28: 102-108. ball, j. p., k. danell, and p. sunesson. 2000. response of a herbivore community to increased food quality and quantity: an experiment with nitrogen fertilizer in a boreal forest. journal of applied ecology 37: 247-255. bergström, r., and o. hjeljord. 1987. moose and vegetation interactions in northwestern europe and poland. swedish wildlife research supplement 1: 213-228. brandner, t. a., r. o. peterson, and k. l. risenhoover. 1990. balsam fir on isle royale: effects of moose herbivory and population density. ecology 71: 155164. cumming, s. g., f. k. a. schmiegelow, and p. j. burton. 2000. gap dynamics in boreal aspen stands: is the forest older than we think? ecological applications 10: 744-759. davidson, d. w. 1993. the effects of herbivory and granivory on terrestrial plant succession. oikos 68: 23-35. edenius, l., m. bergman, g. ericsson, and k. danell. 2002. the role of moose as a disturbance factor in managed boreal forests. silva fennica 36: 57-67. _____, and g. ericsson. 2007. aspen demographics in relation to spatial context and ungulate browsing: implications for conservation and forest management. biology and conservation 135: 293-301. alces vol. 45, 2009 heikkilä & tuominen moose in conserved forests in finland 57 _____, _____, and p. näslund. 2002. selectivity by moose vs. spatial distribution of aspen: a natural experiment. ecography 25: 289-294. gill, r. m. a. 1992. a review of damage by mammals in north temperate forests. 3. impact on trees and forests. forestry 65: 363-388. heikkilä, r., and s. härkönen. 1993. moose (alces alces l.) browsing in young scots pine stands in relation to the characteristics of their winter habitats. silva fennica 27: 127–143. _____, and _____. 1998. the effects of salt stones on moose browsing in managed forests in finland. alces 34: 435-444. _____, p. hokkanen, m. kooiman, n. ayguney, and c. bassoulet. 2003. the impact of moose browsing on tree species composition in finland. alces 39: 203-213. härkönen, s., k. eerikäinen, r. lähteenmäki, and r. heikkilä. 2008. does moose browsing threaten european aspen regeneration in koli national park, finland? alces 44: 31-40. _____, and r. heikkilä. 1999. use of pellet group counts in determining density and habitat use of moose alces alces in finland. wildlife biology 5: 233-239. kalliola, r. 1973. suomen kasvimaantiede. (the finnish botanical geography) werner söderström osakeyhtiö, porvoo, finland. (in finnish). kay, c. e. 1997. is aspen doomed? journal of forestry 95 (5): 4-11. kaye, m. w., d. binkley, and t. j. stohlgren. 2005. effects of conifers and elk browsing on quaking aspen forests in the central rocky mountains, usa. ecological applications 15: 1284-1295. kouki, j., k. arnold, and p. martikainen. 2004. long-term persistence of aspen – a key host for many threatened species – is endangered in old-growth conservation areas in finland. journal of nature conservation 12: 41-52. lavsund, s. 1987. moose relationships to forestry in finland, norway and sweden. swedish wildlife research supplement 1: 229-244. _____, t. nygren, and e. j. solberg. 2003. status of moose populations and challenges to moose management in fennoscandia. alces 39: 109-130. linder, p., b. elfving, and o. zackrisson. 1997. stand structure and successional trends in virgin boreal forest reserves in sweden. forest ecology and management 98: 17-33. lundberg, p., m. åström, and k. danell. 1990. an experimental test of frequencydependent food selection: winter browsing by moose. holarctic ecology 13: 177-182. mclaren, b. e., b. a. roberts, n. djanchékar, and k. p. lewis. 2004. effects of overabundant moose on the newfoundland landscape. alces 40: 45-59. neff, d. j. 1968. the pellet-group count technique for big game trend, census, and distribution: a review. journal of wildlife management 32: 597-614. parker, g. r., and l. d. morton. 1978. the estimation of winter forage and its use by moose on clearcuts in northcentral newfoundland. journal of range management 31: 300-304. pastor, j., and r. j. naiman. 1992. selec-selective foraging and ecosystem processes in boreal forests. american naturalist 139: 690-705. persson, i.-l., k. danell, and r. bergström. 2000. disturbance by large herbivores in boreal forests with special reference to moose. annales zoologici fennici 37: 251-263. risenhoover, k. l., and s. a. maass. 1987. the influence of moose on the composition and structure of isle royal forests. canadian journal of forest research 17: 357-364. moose in conserved forests in finland heikkilä & tuominen alces vol. 45, 2009 58 romme, w. h., m. g. turner, g. a. tuskan, and r. a. reed. 2005. establishment, persistence, and growth of aspen (populus tremuloides) seedlings in yellowstone national park. ecology 86: 404-418. ruusila, v., m. pesonen, r. tykkyläinen, a. karhapää, and m. wallén. 2006. hirvikannan koko ja vasatuotto vuonna 2005. (moose population size and calf production in 2005). riistantutkimuksen tiedote 211. 7 s. (in finnish). saether, b.-e. 1990. the impact of differ-the impact of different growth patterns on the utilization of tree species by a generalist herbivore, the moose alces alces: implications for optimal foraging theory. behavioural mechanisms of food selection. nato asi series, volume g 20: 323-340. _____, and r. andersen. 1990. resource limitation in a generalist herbivore, the moose alces alces: ecological constraints on behavioural decisions. canadian jour-canadian journal of zoology 68: 993-999. suzuki, k., h. suzuki, d. binkley, and t. j. stohlgren. 1999. aspen regeneration in the colorado front range: differences at local and landscape scales. landscape ecology 14: 231-237. syrjänen, k., r. kalliola, a. puolasmaa, and j. mattsson. 1994. landscape structure and forest dynamics in subcontinental russian european taiga. annales zoo-ian european taiga. annales zoo-annales zoologici fennici 31: 19-34. ympäristöministeriö (the ministry of environment). 2006. metsästys eteläisen suomen kansallispuistoissa. (hunting in the national parks of southern finland). ympäristöministeriön asettaman työryhmän raportteja 10/2006. (in finnish). zackrisson, o. 1985. some evolutionary aspects of the life history characteristics of broadleaved tree species found in the boreal forest. pages 17-36 in b. hägglund and g. peterson, editors. broadleaves in boreal silviculture an obstacle or an asset? swedish university of agricultural sciences, department of silviculture. report 14. alces vol. 45, 2009 sipko – reintroduction of large herbivores in russia 35 status of reintroductions of three large herbivores in russia taras p. sipko institute of ecology and evolution ras, leninskii pr. 33, 119071 moscow, russia. abstract: reintroductions of muskoxen (ovibus moschatus), european bison (bison bonasus), and moose (alces alces) have occurred recently in russia. although the process of capturing and moving muskoxen was problematic in remote areas, the reintroduction of animals from canada and the usa successfully restored this extirpated species, and the current population in northern russia serves as a source for further transplants. european bison populations were stagnant and suffered from inbreeding in russia prior to reintroduction of captive animals from throughout europe. the population in orlovskoye polesie national park has experienced population growth with improved genetic potential. of concern is that reintroductions in other areas of russia were unsuccessful and the global population of european bison is not improving. moose from the penzhina river area in russia were successfully reintroduced to the kamchatka peninsula where they were absent for >400 years. the population is growing and dispersing across the peninsula from the transplant sites, and is among the largest physically in eurasia. alces vol. 45: 35-42 (2009) key words: alces alces, bison bonasus, population, reintroduction, restoration, ovibus moschatus, russia. the primary goal of reintroducing large herbivores in russia is to restore biological diversity in northern ecosystems. a secondary goal is to provide a dependable and renewable food supply for residents of northern russia. in this paper i summarize reintroduction efforts with muskoxen (ovibus moschatus), european bison (bison bonasus), and moose (alces alces) that were undertaken for different ecological reasons and circumstances. prior to the reintroduction efforts muskoxen were extirpated, the resident population of european bison was stagnant and suffered from inbreeding associated with a small founder population, and moose, although increasing in certain areas of russia, were absent for centuries from the proposed reintroduction area. shorter and earlier descriptions of these efforts can be found in "re-introduction news" a newsletter of the iucn (sipko et al. 2006, sipko and gruzdev 2006, sipko and mizin 2006). reintroduction of muskoxen in northern russia background and approach a considerable part of russia’s landmass borders the arctic ocean and has severe climatic conditions including long periods of cold temperatures. remains of muskoxen discovered on the taimyr peninsula were 2000-4000 years old indicating that they inhabited the region within relatively recent geological time (vereshagin and barishnikov 1985). previous research indicated that this region was capable of supporting >2 million muskoxen without damage to the fragile northern ecosystem. reintroduction of muskoxen into suitable habitat was expected to have a positive impact on the ecological community because of increased utilization of vegetation resulting in faster turnover of energy at all trophic levels. because yakushkin (1998) found that immature male muskoxen dispersed up to 800 km from reintroduction of large herbivores in russia – sipko alces vol. 45, 2009 36 their natal area, the plan called for herds to be reintroduced within 600-700 km of each other to encourage genetic interchange. sites along the shoreline of the arctic ocean were selected for the first reintroduction that occurred in 1974 when 10 animals were delivered from canada (banks island) to the eastern part of the taimyr peninsula. it was successful and eventually muskoxen spread north, east, and south (putorana plato) of the taimyr peninsula. the population was estimated at 2,500 in 2002 (sipko et al. 2003), and was nearly 4,000 by 2005. a second reintroduction on vrangel island in 1975 used 20 muskoxen obtained from the usa (nunivak island, alaska). population growth was slow because mortality was high in the initial acclimatization period. by 2003, the population was 750 animals (gruzdev and sipko 2003), and recent estimates indicate that the population has stabilized at 800-850. additional reintroductions were achieved by relocating muskoxen from the vrangel island and taimyr peinsula populations (table 1). the objective was to capture and relocate muskoxen that were 0.3-3.5 years old, however, most captured animals were 0.5 years old. vrangel island is a nature preserve and only vehicles with low-pressure tires and snowmobiles are allowed. once located, muskoxen were surrounded by people with dogs and selected animals were chemically immobilized with a dart gun. sedated animals were isolated from the herd and placed into containers for transportation to a holding enclosure. after the capture quota was met, a helicopter transported them to another enclosure on the mainland where they were placed in individual containers, loaded onto an airplane, and transported to the reintroduction site or temporary holding enclosure with local transportation equipment. the methods of capture, handling, and containment on taimyr island were the same as at vrangel island. a helicopter was used to locate and deliver muskoxen either to a temporary holding enclosure or directly to the reintroduction site. they were usually kept in the temporary enclosure for a period before release. discussion the initial reintroduction of muskoxen to northern russia in the 1970s proved to be successful, and set the stage for further reintroductions in other parts of russia (sipko et al. 2007). in addition to those captured for subsequent reintroductions, 81 muskoxen were also captured for zoological parks and domeslocation region year number 1st breeding 2008 population east taimyr peninsula krasnoyarsk 1974, 1975 30 1975 ~6500 wrangel island chukotka 1975 20 1977 ~ 800 bulun yakutia 1996 24 1999 >300 anabar yakutia 1997, 2000 41 2000 >150 begichev island yakutia 2001, 2002 25 2003 >50 allaikhov yakutai 2000 11 2004 64 taas-yrach yakutai 2001-2003 18 2004 12 tamma yakutai 2002, 2003 22 2004 0 polar ural yamal 1997, 1998, 2001, 2003 63 1999 108 kolima magadan 2004 22 none 20 total 284 >8000 table 1. a summary of the location, history, and status of reintroduced herds of muskoxen in northern russia. alces vol. 45, 2009 sipko – reintroduction of large herbivores in russia 37 tication experiments. regular surveillance of muskoxen is hampered by the remoteness of the reintroduction areas. surveys conducted in yakutia in 2005 indicated that 347 muskoxen were in 4 reintroduction areas, a >3x population increase in 10 years. the fastest growth rate occurred in the allaikhov herd where 18 calves were produced in 3 years. the muskoxen population in the bulun region split into 2 nearly equal sized herds; one herd dispersed 120 km west to the delta of the lena river where it resides currently. reintroductions of muskoxen in northern russia are problematic. difficult working conditions, remote locations, and numerous animal transfers with different modes of transportation meant that animals needed skilled animal care specialists to accompany them. the overall mortality rate was 10-15% during the process of capture and containment prior to release. muskoxen from the taimyr peninsula adapted well to relocation sites in central siberia, and those from vrangel island established viable herds in eastern russia. presumably, the introduction of muskoxen from 2 separate populations will have a positive impact on genetic variability that influences survival, productivity, and stability of new herds. a program for additional introductions is underway, and plans are in development to introduce muskoxen to the mountain ranges of northern asia. reintroduction of european bison in central russia background and approach the global population of european bison has not expanded in the past 15 years, remaining at approximately 3,000. small herds scattered in free-roaming populations and captive-rearing facilities typically consisted of 5-7 ancestors from the 12 original animals that were founders of all contemporary bison (belousova 1993). these circumstances caused high inbreeding coefficients in the lowland populations (44%) and moderate coefficients (26%) in the lowland-caucasian line (olech 1998). recent studies indicated that inbreeding occurred over a much longer period of time, thus, the actual inbreeding coefficient is probably much higher. as a result, phenotypic expressions of inbreeding depression are evident in certain populations (sipko 2002). it is estimated that an effective population size of 500 individuals is required to preserve genetic polymorphism that enables a population to adapt and evolve in a constantly changing environment to prevent extinction (soule and wilcox 1980). an effective population must have an adequate sex ratio and sufficient mature and sexually active animals that comprise 25-35% of the population. thus, establishing a population of 1,500-2,000 should ensure long-term viability and survival of a species. however, at least 2 geographically isolated populations are deemed necessary to reduce the risk of disease or unforeseen events that might decimate a population. russia has sufficient geographical area of suitable habitat to accommodate multiple distinct populations of european bison. two areas of appropriate size and ecological conditions were selected for the reintroductions; importantly, they also offered protection as designated wildlife reserves. one area was in the european broadleaf forest of the bryansk-oryol-kaluga region in the european (i.e., central) part of russia. it was a large, contiguous forest tract extending from the boundary of ukraine along the desna river (black sea basin) northeast to the oka river (caspian sea basin), and had few natural or artificial barriers to impede bison movement; railways and highways bisected the region in only one location. the forest was 30-50 km wide and stretched >400 km. the eastern section was known as the oka defense line of the state of moscow during the 14th-17th centuries. these types of frontier forests acted as a line of defense reintroduction of large herbivores in russia – sipko alces vol. 45, 2009 38 until the middle of the 18th century and were strictly protected from harvesting and use by people. as a result, this tract of contiguous broadleaf forest was one of only a few areas that remained largely intact in its natural state, and is a protected natural area. designated sections that contributed to this forest tract included 1) the desnaynskostarogutsky national park (ukraine) with 162 km2 area situated on the boundary of ukraine and russia, 2) the bryansky les biosphere reserve with 1230 km2 in the bryansk region of russia, and 3) the adjacent protected areas of the oryol and kaluga regions in the north including orlovskoye polesie national park (777 km2), kaluzhskie zaseki state nature reserve (185 km2), and ugra national park and biosphere reserve (986 km2). the second area used to reintroduce bison was the ust-kubenskoye hunting facility and surrounding region located about 400 km north of moscow in the vologda region (vologodskai oblast) on the 590 n latitude parallel. the landscape is a series of raised terrace plains at 110-200 m elevation within the severnai dvina river drainage; 64% was forest dominated by conifers (55%, mostly abies spp.). the ustkubenskoye hunting facility is 260 km2 with the russkii sever national park (1664 km2) at the western boundary. the southern border begins at the shore of lake kubenskoe, and the northern and western boundaries adjoin 8,000 km2 of federal forest lands. this area was considered optimal because european bison have survived there long-term without human support. further, resident animals have shown evidence of twinning that is uncommon in bison, suggesting high habitat quality. logging has produced large areas of secondgrowth forest with a shrub-layer suitable for foraging, and most agricultural areas were abandoned and these regenerating lands also provide rich food resources for bison. bison used in the reintroduction were from captive breeding centers in russia and west europe (table 2) to potentially enhance their future genetic viability. their gene pools were quite distinct because bison from russia and west europe have been isolated for almost 100 years. bison were transported in group or individual containers to temporary enclosures and released 1-2 months later. location region year number 2008 population comments cherga mountains altay 1982-1984 12 34 orlovskoye pollesye national park (opnp) oryol 1996-2001, 2006 75 143 successful kaluzhskie zaseki state nature reserve kaluga 2001 0 na dispersal from opnp petrovskoe hunting facility kaluga 2,007 9 9 ust-kubenskoye hunting facility vologda 1991, 1994 5 24 vilikoozerskoe hunting facility vladimir 1989, 1994, 2002, 2004, 2007 25 15 4 bison present at 2nd transplant muromskij sanctuary vladimir 2001-2004 13 16 sknjatinskoe hunting facility tver 1986, 1991 33 3 unsuccessful branskij les state nature reserve bryansk 1999-2000 11 0 unsuccessful total 183 >225 table 2. a summary of the location, history, and status of reintroduced herds of european bison in northern russia. alces vol. 45, 2009 sipko – reintroduction of large herbivores in russia 39 discussion our method of transporting bison over long distances in individual containers proved successful. based on our experience, using large containers containing several animals, as done with cattle, was problematic and should be avoided if possible. bison often injured each other during group transport, resulting in high injury and mortality rates during the reintroduction effort. the bison population in the orlovskoye polesie national park has the greatest genetic potential compared to other bison groups in the world (table 3). the population is growing rapidly with 20 calves born in 2005. animals have dispersed to areas adjacent to the park and regularly appear in the kaluzhskie zaseki state nature reserve. it appears that 3 separate herds have formed from the original population. additional releases into these areas are needed in order to quickly establish optimally sized populations. the region situated between the volga and oka rivers contains a large bison population. however, the region has much industry and transport and communication lines and this lowland area has little coniferous forest. the vilikoozerskoe hunting facility, muromskijj sanctuary, and the sknjatinskoe hunting facility located here have insufficient area for further expansion of the population. slow population growth and numerous mortalities are evident, and further reintroductions are not considered worthwhile. the release of bison into the bryansky les state nature reserve proved to be unsuccessful. long migration patterns and poaching in the russia-ukraine border areas resulted in their demise. there is need to supplement the bison population in the ust-kubenskoye hunting facility. also, the introduction of captive bison from the netherlands into the bukovina population in east carpathian had limited success. the introduced males were unable to compete during the rut with local bulls native to the rugged mountain conditions. it is concerning that european bison numbers are not increasing worldwide. the european bison pedigree book (2002) noted that overall growth was weak with 172 births and 112 deaths overall. arguably, there are insufficient animals to successfully establish new viable populations (sipko and kazmin 2004). a new reintroduction effort in yakutia will focus on a captive breeding and release program. in 2006, 30 wood bison were donated and transported from alberta, canada and relocated in a fenced enclosure 120 km from the city of yakutsk. successful births occurred in 2008 (6) and 2009 (7); 4 of the original animals have died. the long term plan is to establish a free-ranging population in yakutia through gradual release of young bison. reintroduction of moose to the kamchatka peninsula background and approach the kamchatka peninsula has been occupied and developed by russians since the 17th century with no evidence of moose country source number russia prioksko-terrasnyjj zapovednik 21 okskijj zapovednik 24 zoo rostov na donu 1 zoo st. peterburg 1 belarus belovezhskaja pushha 2 netherlands natuurpark lelystad 14 germany springe 6zoo dortmund tierpark chemnitz switzerland tierpark dahlholzly 4 wildpark langenberg finland zoo helsinki 1 belgium han-sur lesse 1 total 75 table 3. origin of european bison transplanted in the orlovskoye pollesye national park, russia. reintroduction of large herbivores in russia – sipko alces vol. 45, 2009 40 inhabiting this region during that time. the wildlife of kamchatka is low in diversity when compared to the mainland and both lynx (felis lynx) and squirrel (sciurus vulgaris) appeared only in the 20th century (valentsev and mosolov 2004). however, archeological evidence indicates that moose were present during the 11th-16th centuries in southern and eastern areas of the peninsula (vereschsgin and nikolaev 1979); this information lead to the interest of reintroducing moose on the kamchatka peninsula. moose have continuously inhabited northeastern mainland russia, but populations have been low in recent centuries. growth of these populations has been documented only since the mid-20th century. the human population has been localized in small settlements leaving vast tracts of land without hunting or poaching pressure. further, the whole region was involved in a wolf extermination (poisoning) program. as a consequence, the moose population in the mountain taiga sector of the penzhina river basin expanded to 2000 animals by 1974 (fil 1975), thus was considered a suitable donor population for the reintroduction. after explorating the peninsula, the kamchatka river valley was considered most suitable for moose. to the east, west, and south were mountain ranges that protected this area from deep snow, and food resources in the valley appeared adequate for moose. although the southern kamchatka region had considerable food resources for moose, deep snow and high risk of poaching were considered problematic. snow depth frequently reaches 120 cm but is not uniform throughout the region because of the local effect of wind. because many rivers do not freeze due to frequent thaw cycles and volcanic heat sources, the southern area of the kamchatka peninsula was considered to have highest potential as moose habitat. the reintroduction occurred in 2 stages, each with unique characteristics. the first stage was in 1977-1988 when 63 moose were captured along the tributaries of the penzhina and belaya rivers (pavlov 1999) and subsequently moved to the kamchatka river valley in the central part of the peninsula. captures were selective for 9-month old calves that were caught with the aid of a helicopter; they were pushed toward the treeless part of the bottomland and immobilized by darting from the helicopter. they were transported by helicopter to an enclosure where they remained 5-7 days, after which they were immobilized and placed in individual containers and transported by helicopter to the release site; transport took approximately 9 h. moose were kept in an enclosure at the release site for 15 days to allow them to recover and acclimatize to the local environment; they were released immediately if snow depth exceeded 60 cm. the first birth was documented in 1979. the second stage of the reintroduction occurred in march and april, 2004-2005 (table 4) when 26 moose were captured, transported, and released in the southern part of the kamchatka peninsula. most were 11 months old; 4 animals were 2 years of age. captures occurred in the kamchatka river basin where tall forest cover reduced the effectiveness of the helicopter. when moose were detected, the helicopter drove them towards openings in the forest where snowmobiles were used to overtake the animals. they were immobilized, hobbled, and transported by sled to an individual holding container where they remained until fully recovered (1-3 h), then loaded into the helicopter and transported to the reintroduction site; transportation time was 5-6 h. they were immediately released in the golygina and udochka river valleys. most moose were released on thermal terrain that was snow-free during winter because of volcanic heat. a cow moose with accompanying calf was observed in autumn 2005. discussion a stable moose population appears to be established and a few animals have even alces vol. 45, 2009 sipko – reintroduction of large herbivores in russia 41 dispersed to the western coast of kamchatka. a census conducted in the central kamchatka region in 2004 estimated the population at 1698-1775 animals (sipko et al. 2004). deep snow exceeding 150 cm in winter 2004-2005 caused considerable concern about winter mortality of released moose. however, they overwintered successfully with only a single mortality to a bear (ursus arctos) the following spring. this successful reintroduction of moose is important for developing tourism and hunting on the kamchatka peninsula. steady population growth has been realized in the southern area of the peninsula where twinning predominates. physical measurements of 32 moose indicate that individuals are realizing larger growth than their donor population from central kamchatka. it is presumed that there is optimal forage production and availability on the volcanic substrate, and that moose are using accessible marine grass (seaweed). however, deep snow is of concern relative to selective and heavy browsing pressure during winter. moose inhabiting the kamchatka peninsula are among the largest of all eurasian specimens, and this seems characteristic of the local population. body weight of large bull moose ranges from 600-750 kg, outside spread of antlers are 161.5-181 cm (n = 6; the largest bull was killed in 2002 in ust-kamchatskyi district), and 11 month-old calves weigh 220-325 kg (n = 10). the large size of these animals posed problems during immobilization, handling, and transport. increased drug volumes created considerable difficulty during recovery from drug-induced shock, and their large size caused handling problems during translocations by helicopter. moose appeared to be more sensitive to capture and transportation problems than muskoxen and european bison; much time and effort was spent in recovering them to a normal physical state. one important factor lending success to the second stage of the reintroduction was that the moose had improved resistance to capture and handling stress. in previous efforts capture mortality often exceeded 50%, whereas mortality was insignificant in the second stage of the reintroduction. acknowledgements my most sincere thanks are extended to v. g. tikhonov and s. s. egorov for the source year release area number of moose 2008 population total female penjinskii: palmatkina, essoveem, chichill, and belaya (white) rivers 1977 milkovskii 4 3 1978 milkovskii 9 5 1979 milkovskii 12 5 1980 milkovskii 12 4 1981 milkovskii 26 14 1982 milkovskii and elizovskii ~2000 chukotka: anadyr river valley 1988 smirnihovskii: sakhalin island region 10 7 10-15 kamchatka peninsula: milkovskii area 2004 ust’-bolsheretskii 11 6 2005 ust’-bolsheretskii 15 7 >45 table 4. a summary of the location, history, and status of moose reintroductions on kamchatka peninsula, russia. reintroduction of large herbivores in russia – sipko alces vol. 45, 2009 42 pleasure of the many days of joint work during the reintroduction of muskoxen and bison in yakutia. references belousova, i. p. 1993. influence of inbreeding on viability of european bison in russian breeding centers. pages 29-43 in k voprosu o vozmozhnosti sokhraneniya zubra v rossii. onti pnts ran, push-onti pnts ran, pushchino, russia. (in russian with english summary). fil, v. i. 1975. penszinskii the moose. journal of hunting and hunting economy 3: 12-13. (in russian). gruzdev, a. r., and t. p. sipko. 2003. productivity and demography of muskoxen on vrangel island. rangifer report 11: 30. olech, w. 1998. the inbreeding of european bison population and its influence on viability. 49th eaap meeting, warsaw, poland, august 24-27. abstract only. pavlov, m. p. 1999. acclimatization of the hunting-trade in animals and birds in the ussr. volume 3. moscow, russia. (in russian). sipko, t. p. 2002. zubr (bison bonasus l.). pages 386-405 in a population and genetic analysis: questions of a modern hunting economy. centrohotkontrol, moscow, russia. (in russian with english summary). _____, v. i. fil, and a. r. gruzdev. 2004. moving moose in kamchatka. the siberian zoological conference. september 15-22, 2004, novosibirsk, russia. abstract only. (in russian). _____, _____, and _____. 2006. re-introduction of moose into kamchatka, russia, re-introduction news. iucn/ssc. 25: 26-27. _____, and a. r. gruzdev. 2006. reintroduction of muskoxen in northern russia. re-introduction news. iucn/ ssc. 25: 25-26. _____, _____, and k. n. babashkin. 2003. demography and productivity of muskoxen in taimyr. rangifer report 7: 40-41. _____, _____, v. g. tikhonov, and s. s. egorov. 2007. capturing and reintroduction of muskoxen in the north russia. rangifer report 11: 32. _____, and v. d. kazmin. 2004. modern problems of european bison protection and their solution in russia. pages 123128 in proceedings of european bison conservation. mammals research institute pas, centre of excellence bioter, bialowieza, poland. _____, and i. a. mizin. 2006. re-introduction of european bison in central russia. re-introduction news iucn/ssc. 25: 27-28. soule, m. e., and b. a. wilcox. 1980. conservation biology: an evolutionaryecological perspective. sinauer associates publishers, inc., sunderland, massachusetts, usa. valentsev, a. s., and v. i. mosolov. 2004. lynx in kamchatka peninsula. proceedings of kamchatka branch of pacific institute of geography, far eastern division, russian academy of sciences. petropavlovskkamchatskii pechatnyi dvor publishing house. 5: 10-27. (in russian). vereshchagin, n. k., and g. f. barishnikov. 1985. extinction of mammas in northern eurasia. pages 3-38 in mammals of northern eurasia. l.: zoological institute an, ussr. (in russian). _____, and a. i. nikolaev. 1979. animals hunted by neolithic tribes on the shores of kamchatka. bulletin (moip) moscow communities verifiers nature. branch biology. 84: 40-44. (in russian). yakushkin, g. d. 1998. muskoxen in taimyr. north russian academy of agrarian sciences, novosibirsk, russia. (in russian). alces22_395.pdf alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces27_127.pdf alces(23)_221.pdf alces vol. 23, 1987 rodgersar text box alces vol. 23, 1987 alces vol. 23, 1987 alces(25)_172.pdf alces29_181.pdf alces24_178.pdf alces vol. 45, 2009 sipko and kholodova – eurasian moose population 25 fragmentation of eurasian moose populations during periods of population depression taras p. sipko and marina v. kholodova institute of ecology and evolution, russian academy of sciences, 33 leninsky prospect, moscow 119071, russia abstract: changes in the distribution of eurasian moose (alces alces) populations during the pleistocene and holocene eras were analyzed from historical and contemporary literature. we focused on how range boundaries varied, suitable habitat was fragmented, and how local and regional populations were isolated, especially during periods of population depression. we discuss how the occurrence and duration of isolation of local populations likely influenced the genetic structure of eurasian moose. we question the geographic division of certain subspecies, and suggest that our analysis be used to reinterpret and revise genetic structure of eurasian moose populations. alces vol. 45: 25-34 (2009) key words: alces alces, eurasia, fragmentation, genetic structure, history, isolation, moose, population depression, range, subspecies. the eurasian moose (alces alces) popula-population returned to most of its original range during the 20th century. historically it experienced numerous range reductions and fragmentations that were followed by restoration and dispersal into new areas. recent research (danilkin 1999) suggests that species differentiation in moose could be greater than believed previously. in order to best understand and interpret the genetic structure of eurasian moose (i.e., its phylogeography and polymorphism), it is necessary to identify dispersal centres where genetic diversity was presumably highest. assuming that mitochondria are inherited maternally and females disperse shorter distance than males, we expect that the geographic distribution of mitochondrial dna haplotypes was fairly stable and should reflect past migration routes of moose. we collected and analyzed extensive archaeological and paleontological data to trace moose range during periods of substantial population depression. we regard our results as preliminary because these data are incomplete, and distribution of moose populations varied in space and time and was not always documented accurately. distribution and taxonomoy subspecies of moose in eurasia are considered to have no distinct differences; variable morphological characteristics relative to geographic location represent a cline (markov and danilkin 1996, danilkin 1999). moose constitute one macropopulation in europe (rozhkov et al. 2002, davydov et al. 2004) with european moose (alces alces alces) inhabiting europe, altai, and western siberia up to the yenisei river (danilkin 1999). in summer and winter the yenisei river is a stem of a large river system rich in valleys that are preferred moose habitat with no hindrance to moose migration. we suspect that the subspecies boundary lies either westward along the ob and yenisei rivers divide, or eastward along the divide of the yenisei and lena river basins; further research is required to better delineate this boundary. caucasian moose (a. a. caucasicus) in the caucasus area were eliminated by the beginning of the 20th century. repopulation of the north caucasus region by this subspecies in the 20th century indicates that caucasian and european moose habitats were probably well eurasian moose population – sipko and kholodova alces vol. 45, 2009 26 connected in the past, and calls into question whether to identify caucasian moose as a separate subspecies (danilkin 1999). the subspecies status seems more appropriate for the moose population in the westernmost part of europe. yakut moose (a. a. pfizenmayeri) inhabit the area to the east of the yenisei river up to the stanovoi ridge, khakassia (tikhonov 1990), northern mongolia to the south, and the tchersky or verkhoyansky ridges to the northeast. kolyma moose (a. a. buturlini) are distributed in northeast siberia to the east of the tchersky ridge (zheleznov 1990). we think that the boundary of this subspecies runs along an arc formed by the verkhoyansky ridge and subtar-khayata ridge; further genetic research is needed to delineate this boundary. ussuri moose (a. a. cameloides) are restricted to the southwestern part of siberia and the amur basin; the distribution of this subspecies requires further study. it’s highly possible that moose populations inhabiting the area east of baikal lake (ditsevich 1990) and further east in steppe valleys of the selenga and orkhon basins, and in northeast china also belong to this subspecies. the northern boundary runs along the stanovoy ridge (kutcherenko 1975) and the lena river divide (danilkin 1999). biological parameters in russia moose live mainly in forest habitats but also occupy the forest steppe and forest tundra. moose were documented in tundra on the lena river estuary along the coastline of the arctic sea (nasimovich 1955), and moose remnants were found on new siberian islands in the arctic sea (vereshchagin 1967). they were also documented in the desert near aral lake, >500 km from their original ecotopes (heptner and nasimowitsch 1974). moose have also been documented in the alpine zone to 2500 m elevation (semenov-tjan-shansky 1948, heptner et al. 1961). although typically sedentary, certain populations migrate and individuals may disperse long distances. it took them several decades to disperse >1500 km to reach the caucasus foothills (yasan 1966), but they easily cross waterways 10-15 km wide (timofeeva 1974). their size, mobility, and relatively high reproductive rate aid them in repopulating vacated areas. moose in the pleistocene era the earliest evidence of moose remains dates back to the mid-pleistocene era (vereshchagin 1967) that had no less than 3 periods of glaciation (gerasimov and markov 1939) that dramatically altered landscapes and moose range (fig. 1). according to vereshchagin (1967), there is scarce evidence of moose in the ice age, whereas moose remains are considerable in the holocene era. active morphogenetic processes apparently occurred in the late pleistocene era (boeskorov 2001). moose distribution during the ice age can be described only in general terms. because the greatest part of europe was covered by glacial sheets and lakes, moose probably inhabited a rather narrow territory between the ocean and glacial lakes adjoining the alpine and nordic ice sheets. this was possibly a connecting link between the moose population in southwest europe overgrown with forests very much like those of modern scandinavia (woillard 1979), and that inhabiting vast areas of the south ural mountains where forest refugia existed (panov 1999). the european moose population in this area seems to be the most abundant and genetically diverse, and as the glacial sheet retreated northward it covered eastern europe. much later, about 10,000 years ago when the southern part of scandinavia was free of ice, european moose with red deer (cervus elaphus) and wisent (bison bonasus) found their way from west europe to fennoscandia using denmark as a land bridge (filonov 1983), and later dispersed from the east through the karelian isthmus. in west siberia when alces vol. 45, 2009 sipko and kholodova – eurasian moose population 27 the last glacial period was at its utmost, the glacier blocked the yenisei and ob rivers forming mansi lake, a large reservoir twice as large as the black sea; surplus water ran via the turgai channel to the caspian sea (groswald 1983). during this period western and eastern siberian moose appeared to be totally isolated, which explains the chromosomal differences between the eastern and western populations. the altai-sayan mountain region had no solid glacial sheet during the ice age (gerasimov and markov 1939), although a chain of mountain glaciers factored into the isolation of northern and southern moose populations in this area. there was no solid glacial sheet in east siberia during the pleistocene (groswald 1998, 1999) and moose occupied all suitable areas. southeast siberia is limited by lake baikal in the west and by the stanovoi highlands, then the stanovoi range, and part of the dzhugdzhur range in the east. although moose currently cross these mountain systems, these mountains were covered with vast glaciers in the pleistocene era (preobrazhenskiy 1960, groswald 1984) and were impassable for some period isolating the so-called ussuri moose from the rest of the population. northeastern asia is a huge amphitheatre sloping towards the arctic ocean that is characterized by strong orographic contrasts; though subdued mountains prevail, they are joined with highlands and plains. the verkhoyansk mountains present an orographic barrier of the area in the west. to the south of the verkhoyansk range, the sette-daban and the yudom range stretch divided by the yudom-mai highlands, and further along the okhotsk sea coastline lies the dzhugdzhur range. the tchersky range stretches 1800 fig. 1. distribution of ice sheets, mountain glaciers, and ice-dammed lakes in eurasia during the midpleistocene era (according to groswald 1984) that influenced the distribution and range of moose. major seas include the aral (as), black (bs), and caspian (cs). glacial sheets include the chukchi (ch), east siberian (es), karskii (ka), ohotskii (oh), and scandinavian (sc). mountain glaciers include the altai (as), baikal (ba), central asian (ca), tibetan (ti), and verkhoyansk (ve). other features include lake mansijskoe (ml), amur river (am), and the turgaiskii trench (tu). eurasian moose population – sipko and kholodova alces vol. 45, 2009 28 km northwest to the east of the verkhoyansk mountains. glaciers developed in the mountains to various extant and during glacial maximums reached highland valleys which had several glacial and interglacial periods (groswald and kotljakov 1989). highlands occupy the inner part of the area, and lowlands lie along the coast and narrow stretches penetrate between the mountains to the south. these valleys formed refugia where isolated moose populations survived. four moose populations that formed during the ice age are documented in the area (safronov 2008). further, during the sartan ice age there was a forest refugium in the middle reach of the anadyr (kozhevnikov and zheleznovchokotskij 1995) where a moose population most closely related to alaskan moose might have existed. range recession in eurasia moose were distributed across most of europe during the early holocene era (heptner et al. 1961, vereshchagin 1967). later the range retreated eastward; the last moose was killed in saxony in 1777 and in galicia in west ukraine in 1769 (gebel 1879). by the end of the 18th century moose were eliminated in belovezhskaya pusha (sablina 1955). in the beginning of the 20th century moose were still in east prussia (now kaliningrad region, russia; obermeier 1913) but were never encountered after world war i. thus, the european moose was preserved only in russia and the nordic countries by the mid20th century. european part of russia during the period of utmost population depression, the southern boundary of moose range retreated 450-1000 km northward, the northern boundary 500-600 km southward (danilkin 1999), and the range was fragmented (fig. 2). the northern boundary corresponded to the northern extent of the forest zone and reached 65° n in the ural mountains (sokolov 1959). the southern boundary coincided with the latitudinal flow of major rivers such as the volga, kama, and belaya rivers (filonov 1983). figure 2 depicts the location of 20 isolated moose populations; in the following text each population is described both temporally and geographically with an accompanying number [#] identified on figure 2. the main area was divided roughly into western [1] and eastern parts [2] along the vologda-arkhangelsk railroad and the white sea-baltic canal; there is no information regarding the duration of this fragmentation. the western part was characterized by irregular moose distribution of variable configuration and included the leningrad, pskov, novgorod, and tver regions, the western part of the vologda region, the northern part of the smolensk region, and north of byelorussia (serzhanin 1961). there is good reason to believe that a small breeding population survived in the area of the pripyat and pinsk [3] marshes (serzhanin 1961, galaka 1964). a small population of moose also survived in the bryansk forests [4] along the left bank area of the desna river (fedosov and nikitin 1951) and moose were also documented in the sumy region (galaka 1964). moose also survived extirpation in the meshcherskaya lowland [5], the boggy interfluve of the volga and oka rivers (severtsev 1854, kulagin 1932). one other isolated population existed in the area between the tsna and sura rivers [6] in the mordovia, penza, and tambov regions (filonov 1983). scandinavian peninsula current norwegian and swedish moose populations are abundant, yet in the beginning of the 19th century only small, isolated groups survived in the southwest of the scandinavian peninsula [7] (markgren1974, danell and bergström 2008). the same situation occurred in finland where moose disappeared by the mid-19th century (markgren1974, nygrén et al. 2008). single animals migrated gradually alces vol. 45, 2009 sipko and kholodova – eurasian moose population 29 from adjoining regions of south karelia [8] where the moose population was not abundant (vereshchagin and rusakov 1979). data on the kola peninsula [9] is rather contradictory. some authors (semenov-tjan-shanskij 1948, kirikov 1966) state that the species disappeared from the area and later repopulated it from adjoining territories. others (vereshchagin and rusakov 1979) quote data supporting that a moose population survived on the peninsula. west siberia the moose population between the ural mountains and the yenisei river in west siberia [10] was at its minimum by the beginning of the 1920s. the northern boundary arched southward reaching 63° n, and the range was a strip of land about 450 km wide (laptev 1958), with a 250-500 km wide gap in the region of the ob river dividing the area into eastern and western parts (laptev 1958, yurlov 1965). it is suggested that migration could occur between these areas, but insufficient evidence exists to support or refute this idea. this gap existed for an unknown period, but we believe it lasted no less than 100 years. south siberia the moose population of the altai-sayan mountains was at its minimum at the end of the 19th century (sobansky 1975). moose were also eliminated in the adjoining regions lying to the north and northwest in the kemerovo region in kuznetsk alatau (sobansky 1992), in the south altai, and the adjoining kazakhstan regions (sludsky 1953). moose were considered fully extirpated in altai (filonov 1983), nevertheless, a small population may have possibly survived in the upper reaches of the abakan, biya [12] (dmitriev 1938), katun, and tchuya rivers [11] (sobansky 1975). repopulation of the altai region resulted from migration from both the east and west (filonov 1983). moose was never abundant in the sayan fig. 2. the southern boundary of moose range retreated 450-1000 km northward and the northern boundary 500-600 km southward during the period of severest population depression, this figure identifies the location of 20 isolated moose populations during this period; each population is described both temporally and geographically in the text. eurasian moose population – sipko and kholodova alces vol. 45, 2009 30 region. in the beginning of the 20th century they disappeared from the khakas-minusinsk basin (skalon et al. 1941), as well as from the west sayan where only single animals were encountered migrating from the west. moose survived only in the east sayan, east of tuva, and in the eastern part of the tannu-olu range [13]; the latter were connected with a mongolian population (yanushevich 1952). the mongolian population occupied the area south of the altai-sayan region and was limited to mountainous taiga regions. moving west to east, moose left in the mongolian altai were encountered only at the beginning of the eastern part of the tanu-ola mountains. further east moose occupied an area around habsugul lake, the khangai mountains up to the khabgai-nuru range in the south, and the upper reaches of the onon, kerulen, and tola rivers and other rivers in the khantae uplands (bannikov 1954). east siberia this region was a northern siberian upland between the yenisei and lena rivers where the distribution of moose has changed considerably (michurin and mironenko 1967). during the period of lowest population at the end of the 19th-early 20th centuries, the range boundary crossed the yenisei river at 590 30’ n and went northeast crossing the podkamennaya tunguska river. it followed the verkhnyaya tunguska and podkamennaya tunguska divide to 1000 e, where it turned north to the upper reach of the kotui river basin, went along the kotui to 700 n, and then east to the mid-reach of the anabar river [14] (naumov 1934, heptner et al. 1961). according to middendorf (1869), in the mid-19th century moose were occasionally documented along the banks of the nizhnyaya tunguska river. however, maak (1887), whose expedition in 1854–55 visited the vilyuy river starting not far from the nizhnyaya tunguska river that flows east to the lena river, reported that moose were absent in the area and local people had no knowledge of them. moose were also absent in central yakutia (tavrovsky et al. 1971). we suggest that northern [14] and southern [15] populations in this region were isolated for an extended period. southeast siberia moose range has changed little in the region of the trans-baikal and amur basin [16]. the southern boundary along the sea of japan retreated northward to 440 45’ n (kaplanov 1948), and moose abandoned the ussuri and part of the amur bottomlands [18] (zhitkov 1914, rakov 1965). more apparent changes occurred in southwestern china where at the beginning of the 20th century moose were distributed to the north of the chita-kharbinvladivostock railroad (oshanin 1934, zhen zuoxin 1956). they inhabited the great and little khingan mountains and were in direct contact with moose on the opposite bank, often crossing the amur river (rakov 1964). thus, this chinese population and the adjacent russian population should be considered the same. one other population survived in east manchuria at the interfluve of the ussuri and sungari rivers [17] (abramov 1949); it was isolated long enough to develop different antler morphology than moose in the sikhote alin (rakov 1965). further, in an east [18]-west [16] direction along the chita–khabarovsk transect, average body weight increases stepwise to the west boundary of the khabarovsk territory (rozhkov et al. 2001) where the little khingan meets the bureinsky range. it is possible that these mountain ranges divide genetically and taxonomically different moose populations. northeast siberia moose disappeared before the beginning of the 18th century in the kamchatka peninsula (vereshchagin and nikolaev 1979) and sakhalin island (kozyrev 1960, alekseev 1974) alces vol. 45, 2009 sipko and kholodova – eurasian moose population 31 where none were documented by the first russian explorers. in the 1820s moose were rare in the whole region (mensbir 1878) and were not documented in the koryak district (fil and demyanyuk 1972). the range boundary along the pacific coast retreated westward to the kolyma tributary of the omolon river [19] (zheleznov 1990, sipko et al. 2004). moose did not inhabit the central axial zone of the mountain ranges of the verkhoyansk uplands (tavrovsky et al. 1971), though some survived in mountain valleys [20] forming 4 isolated populations (safronov 2009). repopulation of the kamchatka peninsula occurred from recent transplants (sipko et al. 2004, 2008). conclusion moose range in russia underwent repeated changes during the pleistocene and holocene eras in response to climatic and landscape changes. this complicated geographic history caused periods of isolation for particular local moose populations. some isolated populations were depressed severely and probably passed through a genetic bottleneck that influenced their local genetic structure. because certain populations demonstrate distinct morphological and physical differences, we suggest that the taxonomic and phylogenetic structure of moose range in eurasia is more complex than considered previously. further genetic studies of distinct populations would help elucidate such genetic relationships and differences. references abramov, k. g. 1949. the distribution, ecology, and economy of moose in priamurie. the nature branch of biology 54 (1): 1728. moscow, russia. (in russian). alekseev, a. i. 1974. the amur expedition, 1849-1855. publishing house of nauk academy, leningrad, russia. (in russian). bannikov, a. g. 1954. the mammals of the mongolian people’s republic. publishing house of the academy of sciences of ussr, moscow, russia. (in russian). boeskorov, g. g. 2001. systematization and origin of modern moose. siberian branch of the russian academy of science, novosibirsk, russia. (in russian). danell, k., and r. bergström. 2008. the history of moose management in sweden an 800 year perspective. 6th international moose symposium, yakutsk, russia, 1420 august, 2008. abstract only. danilkin, a. a. 1999. the deer (cervidae), mammals of russia and adjacent regions. moscow, russia. (in russian). davydov, a. v., o. p. piskunov, a. v. pronjaev, and j. i. rozhkov. 2004. spatial differentiation moose of eurasian moose by results of an estimation of the hunting trophies. bulletin of game biologists 1: 36-41. moscow, russia. (in russian). ditsevich, v. n. 1990. systematization and morphology of east siberian moose. 3rd international moose symposium. syktyvkar, russia, 27 august-5 september. abstract only. dmitriev, v. v. 1938. ungulates of the altaisky nature reserve and surroundings (east altai and west sayan). the altay reserve 1: 171-262. (in russian). fedosov, l. v., and k. n. nikitin. 1951. pages 6-8 in fauna of the bryansk area. bryansk publishing house, bryansk, russia. (in russian). fil, v. i., and v. p. demjanjuk. 1972. distribution and number of moose in the kamchatka area. russian scientific research institute of hunting and fur farming 34: 32-36. kirov, russia. (in russian). filonov, k. p. 1983. moose. lesnaya promyshlennost, moscow, russia. (in russian). galaka, b. a. 1964. expansion of moose range in ukraine. pages 35-43 in biology and economy of the moose. rosselhozizdat, moscow, russia. (in russian). gebel, g. 1879. influence of the person and its culture on an empire of animals. nature eurasian moose population – sipko and kholodova alces vol. 45, 2009 32 and hunting 1: 95-117. moscow, russia. (in russian). gerasimov, i. p., and k. k. markov. 1939. the glacial age in territory of the ussr. works of institute of geography, vol. 33, academy of sciences, moscow– leningrad, russia. (in russian). groswald, m. g. 1983. integumentary glaciers of continental shelves. science, moscow, russia. (in russian). _____. 1984. the euro-asian glacial cover. pages 130-131 in glaciology the dictionary gidrometeoizdat. leningrad, russia. (in russian). _____. 1998. materials glaciology. 84: 121129. moscow, russia. (in russian). _____. 1999. euro-asian hydrosphere accidents and a congelation of arctic regions. moscow, russia. (in russian). _____, and v. m. kotljakov. 1989. great glacial systems and drain of northern eurasia and its significance for inter-regional correlations (quaternary period). paleogeography and litologia 5-13. kishinev, shtiincza, ussr. (in russian). heptner, w. g., and a. a. nasimowitsch. 1974. der elch (the moose). ziesmanverlag, wittenburg lutherstadt, germany. (in german). _____, a. a. nasimovich, and a. g. bannikov. 1961. mammals of the soviet union. artiodactyls and non-artiodactyls. moscow, russia. (in russian). kaplanov, l. g. 1948. tiger, red deer, and moose. otdelenia zoologia no. 14 (29). moscow, russia. (in russian). kirikov, s. v. 1966. hunting animals, natural environment, and man. nauka, moscow, russia. (in russian). kozhevnikov, j. p., and n. k. zheleznovchukotskij. 1995. beringia: history and evolution. moscow, russia. (in russian). kozyrev, r. v. 1960. the ancient past of sakhalin. yuzhno-sakhalinsk, russia. (in russian). kulagin, n. m. 1932. moose of the ussr. publishing house of nauk academy, leningrad, russia. (in russian). kutcherenko, s. p. 1975. the distribution, number, and use of ussurijskogo moose. pages 108-109 in ungulate fauna of the ussr. мoscow, russia. (in russian). laptev, i. p. 1958. mammals of a taiga zone of western siberia. tomsk university publishing house, tomsk, russia. (in russian). maak, r. k. 1887. viljujsky district of the yakut area. st. petersburg, russia in 1994 edition: viljujsky district of the yakut area. yana publishing house, moscow, russia. (in russian). markgren, g. 1974. the moose in fennoscandia. canadian naturalist 101: 185-194. markov, g. g., and a. a. danilkin. 1996. kraniometricheskaja in eurasia (the characteristics of moose in eurasia). russian academy of sciences, biological series p: 244-247. moscow, russia. (in russian). menzbir, m. 1878. characteristics of distribution of animals. nature and hunting 10: 32-34. moscow, russia. (in russian). michurin, l. n., and o. n. mironenko. 1967. about moose in the putorano mountains. pages 72-75 in biology and economy of the moose. rosselhozizdat, moscow, russia. (in russian). middendorf, a. 1869. travel in northern and eastern siberia by siberian fauna. 2(5) st. petersburg, russia. (in russian). nasimovich, a. a. 1955. the role of snow cover in the lives of ungulates in the ussr. publishing house of the academy of sciences of ussr, moscow, russia. (in russian). naumov, н. п. 1934. mammals of the tungus district. work of the polar commission. publishing house of the academy of sciences of ussr, moscow, russia. (in russian). nygrén, k., p. danilov, and t. nugren. 2008. alces vol. 45, 2009 sipko and kholodova – eurasian moose population 33 the dynamics and management of north european moose. 6th international moose symposium. yakutsk, russia, 14-20 august, 2008. abstract only. obermeier, g. 1913. the prehistoric person. translated from german, p.j. schmidt. edition of brockhaus and efron, st. petersburg, russia. (in russian). oshanin, a. 1934. trade and sports hunting from northern manchuria. collection of works 7: 12-24. moscow, russia. (in russian). panov, n. k. 1999. history of development of vegetation of a mountain part of the southern urals mountains in the late pleistocene and holocene, on paleontological data. pages 144-158 in n. g., smirnova, editor. historical ecology of animals in the southern ural mountains. sverdlovsk, russia. (in russian). preobrazhenskiy, v. s. 1960. kodarskiy glacial region (trans-baikal). ninth scetion of igy program (glaciology). no. 4. publishing house of the academy of sciences of ussr, moscow, russia. (in russian). rakov, n. v. 1964. geography and ecology of moose populations in amur-ussurijskij kraj. pages 44-100 in biology and economy of the moose. rosselhozizdat, moscow, russia. (in russian). _____. 1965. distribution and ecology of moose in amur-ussurijskij kraj. pages 28-65 in biology and economy of the moose. rosselhozizdat, moscow, russia. (in russian). rozhkov, j. i., a. v. pronjaev, o. v. piskunov, n. e. ovsjukova, a. v. davydov, and l. v. rozhkova. 2001. moose. the population-biological analysis of license information. hunting russian animals. issue 4. tsentrohotkontrol, moscow, russia. (in russian). _____, _____, _____, _____, _____, and _____. 2002. moose of the european part of the russian federation. population division, selection, seasonal changes, morphological, and ethological differentiation (by results of the analysis of the license information). pages 153-158 in questions for modern game biologists. tsentrohotkontrol, moscow, russia. (in russian). sablina, t. b. 1955. ungulate belovezhskoi pushi. page 15 in a. n. severtsova, editor. works of the institute of morphology and ecology of animals. moscow, russia. (in russian). safronov, v. m. 2009. regional populations and migration of moose in northern yakutia, russia. alces 45: 17-20. semenov-tjan-shanskij, o. i. 1948. moose on the kola peninsula. works of the lappish reserve 2: 91-162. moscow, russia. (in russian). serzhanin, i. n. 1961. mammals of belarus. minsk publishing house, izdatelstvo akademii nauk, belorusskoi ssr, minsk, russia. (in russian). severtsev, n. a. 1854. moose or sohatii. bulletin of natural sciences 19: 290-299. sipko t. p., v. i. fil, and a. r. gruzdev. 2004. resettlement of moose. the siberian zoological conference, novosibirsk, russia, 15-22 september, 2004. abstract only. _____, _____, and _____. 2008. moving moose on kamchatka. 6th international moose symposium, yakutsk, russia, 1420 august, 2008. abstract only. skalon, v. n., i. scherbakov, and m. bazikin. 1941. nature and socialist facilities. volume 8. moscow, russia. (in russian). sludsky, a. a. 1953. eviction of “taiga” animals in the forest-steppe and steppes of western siberia and kazakhstan. the nature branch of biology 53 (2): 14-32. moscow, russia. (in russian). sobansky, g. g. 1975. territorial accommodation of ungulates in the altay mountains. pages 122-123 in ungulate fauna of the ussr. moscow, russia. (in russian). eurasian moose population – sipko and kholodova alces vol. 45, 2009 34 _____. 1992. ungulates in the altai mountains. novosibirsk, russia. (in russian). sokolov, i. i. 1959. ungulates (perissodactyla and artiodactyla). in fauna of the ussr. mammals. nauka publishing house, moscow, russia. (in russian). tavrovsky, v. a., o. v. egorov, and v. g. krivosheev. 1971. mammals of yakutia. nauka, moscow, russia. (in russian). tikhonov, a. n. 1990. history and systematization moose. 3rd international moose symposium. syktyvkar, russia, 27 august-5 september. abstract only. timofeeva, e. k. 1974. moose: ecology, distribution, and economic importance. leningrad university publishing house, leningrad, russia. (in russian). vereshchagin, n. k. 1967. the geological history of moose and its relationship to man. pages 3-37 in biology and economy of the moose. rosselhozizdat, moscow, russia. (in russian). _____, and a. nikolaev и. 1979. trade animals of neolith tribes of kamchatka. the nature branch of biology 84 (5): 40-44. moscow, russia. (in russian). _____, and o. s. rusakov. 1979. ungulates from the north-western part of the ussr. nauka, leningrad, russia. (in russian). woillard, g. 1979. abrupt end of the last interglacial in north east france. nature 281 no. 5732: 558-562. yanushevich, a. i. 1952. vertebrate fauna of the tuva area. zap-sib publishing house, novosibirsk, russia. (in russian). yazan, j. p. 1966. moose in the northern caucasus. the nature branch of biology 71 (4): 123-125. moscow, russia. (in russian). yurlov, k. t. 1965. change in moose range in the southern west-siberian lowlands. pages 17-27 in biology and economy of the moose. rosselhozizdat, moscow, russia. (in russian). zheleznov, n. k. 1990. wild ungulates in northeastern ussr. vladivostok, russia. (in russian). zhen zuoxin. 1956. animals and birds of china. the nature journal 10: 3-18. moscow, russia. (in russian). zhitkov, b. m. 1914. moose in ussurijskogo kraj. zoological branch of natural sciences: a new series 2 (3): 5-14. st. petersburg, russia. (in russian). alces vol. 48, 2012 broders et al. moose response to temperature 53 ecothermic responses of moose (alces alces) to thermoregulatory stress on mainland nova scotia hugh g. broders1, andrea b. coombs1,2, and j. r. mccarron1 1department of biology, saint mary’s university, halifax, nova scotia, canada b3h3c3; 2current address: department of natural resources, wildlife division, 136 exhibition street, kentville, nova scotia, canada b4n 4e5. abstract: the size of the mainland nova scotia moose (alces alces) population has declined precipitously over the last several decades and their current distribution is discontinuous. in recognition of the state of its moose population, nova scotia declared moose as ‘endangered’ under nova scotia’s endangered species act in 2003. a variety of factors have been attributed to the decline, and the goal of this project was to determine whether thermoregulatory stress may be impacting the viability of the moose population. location and temperature information were collected from gps-collared moose to test predictions related to whether moose behaviour changes in response to high temperatures. overall, our results suggest that moose exhibit behaviours (i.e., ectothermy) that are consistent with thermoregulatory stress, but the actual impacts of this, if any, on population productivity requires further study. the greatest response occurred in the summer during both day and night, when moose moved to areas of lower ambient temperature. further, overall movements were significantly reduced during periods of high temperatures. alces vol. 48: 53-61 (2012) key words: alces alces, ectothermy, endangered, moose, nova scotia, temperature. the incidence of population declines are increasing due to a variety of factors, most notably overexploitation, habitat destruction, and food chain disruption (campbell and reece 2002). large mammals are highly vulnerable to human exploitation and it has been estimated that less than 21% of the earth’s terrestrial surface contains all the large mammals it once did (morrison et al. 2007). in the northern hemisphere, populations along the southern extent of their species range are particularly vulnerable to climate change (renecker and schwartz 1997, lenarz et al. 2009) and may eventually shift north in response to warming temperatures. these and a suite of other threats have been identified as negatively impacting north american moose (alces alces) populations. although some north american moose populations are stable or increasing, population declines have occurred in alaska (timmerman 2003), minnesota (murray et al. 2006), manitoba (v. crichton, manitoba natural resources, pers. comm.), and nova scotia (pulsifer and nette 1995) which has had closures of hunting seasons as a result (parker 2003). the southern range limit of moose may be determined by thermoregulatory stress (renecker and hudson 1986) and links between declining populations and increased ambient temperature associated with climate change have been suggested (murray et al. 2006, lenarz et al. 2009, 2010; but see lankester 2010). marai and haebb (2010) define heat stress as “the state at which mechanisms activate to maintain an animal’s body thermal balance, when exposed to untolerable (uncomfortable) elevated temperatures.” although the initial response may be physiological, behavioural modifications can reduce these physiological stressors (e.g., movement to cooler areas in response to heat stress). of all the extant boreal ungulate species, moose are the most likely candidate to suffer moose response to temperature broders et al. alces vol. 48, 2012 54 from heat stress due to their relatively low, upper critical temperature limit (-5° c in winter and 14° c in summer; karns 1997, renecker and hudson 1986). in addition to panting to ameliorate thermal stress (renecker and hudson 1986), moose often use aquatic areas (renecker and schwartz 1997) or forest stands that buffer from extremes in temperature, a form of ectothermy. such areas of thermal cover (dussault et al. 2004, mysterud and ostbye 2008) could provide conditions that may be as much as 7° c cooler than forest edges (chen et al. 1995). demarchi and bunnell (1995) and dussault et al. (2004) found that nocturnal activity of moose increased in summer and fall as ambient temperature increased, and use of thermal cover was lower at night suggesting that activity of moose may be inversely related to temperature and/or exposure to solar radiation. in late winter heat stress strongly influenced cover selection as moose tend to avoid areas where the temperature exceeds 8° c (the temperature when panting begins to dissipate heat from the body in the late winter months; schwab and pitt 1991). leblond et al. (2010) suggested that when temperatures are cooler, moose chose areas that had less thermal cover and higher amounts of solar energy. contrary to these studies however, lowe et al. (2010) did not find a behavioral response by moose to temperatures in ontario. they suggested that, within their study area, there were no obvious thermal refugia and that moose either were not impacted by the temperature range to which they were exposed, or the resolution of their measurements were not fine enough to detect a response. prior to european settlement of nova scotia, it was believed that the local moose population was large (≈15,000; parker 2003). approximately 100 years ago, the population on cape breton island was extirpated and the current population was founded by the introduction of moose from alberta (pulsifer and nette 1995). today, the remnant native population on the mainland has a discontinuous distribution with a crude population estimate of 1000 (parker 2003). on the mainland, the most significant populations occur in cumberland-colchester counties, pictou-antigonish counties, and in the tobeatic wilderness area. the last hunting season for mainland moose was held in 1981 (parker 2003), and in 2003 the population was classified as endangered under the nova scotia endangered species act. currently, there are no reliable demographic estimates or other data that could be used to justify management decisions. some of the factors believed to be affecting population growth include: parasites such as parelaphostrongylus tenuis, deterioration in the quantity and quality of moose habitat, poaching, predation, and thermal stress (brannen 2004, beazley et al. 2006). the goal of this study was to determine if there is any evidence that moose on mainland nova scotia exhibit signs of heat stress. specifically, our hypothesis was that moose would alter their behavior to reduce physiological stress in response to high temperatures. to assess this hypothesis, we tested 2 emergent predictions: 1) during periods of high temperature moose would select cooler areas, and 2) during periods of high temperature movement would be reduced relative to times when it was cooler. to test these predictions we used data from gps-collars deployed on moose on the mainland of nova scotia, 2002-2006. if our hypothesis is supported, our results may lend justification for further study to quantitatively characterize the population level impacts of increasing temperatures on moose, including population recovery in nova scotia. materials and methods location and temperature data from 12 gps-collared adult moose from cumberlandcolchester counties (n = 5), antigonish county (n = 1), and the chebucto peninsula of halifax county (n = 6) were provided by the nova scotia department of natural resources (nsdnr). the gps collars (lotek alces vol. 48, 2012 broders et al. moose response to temperature 55 gps 2200l; lotek wireless inc., newmarket, ont., canada) were programmed to acquire and store location and temperature data every 2-4 h. although gps collars have the advantage of greater location accuracy and resolution in movement dynamics data relative to conventional telemetry (girard et al. 2002), the ability to record a location at pre-determined times is reduced if the animal is under dense forest cover (rempel et al. 1995, moen et al. 1996, rodgers et al. 1997, dussault et al. 1999) or on steep slopes (gamo and rumble 2000); hence, there is potential for location bias. regardless, location accuracy was expected to be within 10 m under most conditions and times. location data were imported into arcgis geographic information system (gis), version 9.1 (esri, redlands, california) and each location was assigned to 1 of 4 cover types using land-use and forest resource inventory data for the region (interpreted from 1:10,000 aerial photos; nsdnr). the 4 cover types included softwood (75% softwood species by basal area), mixedwood (26-74% softwood species by basal area), hardwood (<25% softwood species by basal area), and other (water and all other land use types that were not forest cover types). for each prediction we controlled for time of day and season effects by conducting separate analysis during the day and night for each of 3 seasons (summer: 15 june-15 september; early winter: 16 november-14 january; late winter: 15 january-15 april). further, because individuals were likely to respond to temperature variations in different ways (e.g., due to differences in age, body condition, location, gender, and reproductive status), data collected from each individual were analyzed separately, but global inferences were based on the results from all individuals. to test our first prediction regarding whether moose selected cooler stands when temperatures were high, we determined the magnitude of the difference between the temperature recorded by the collar and the temperature recorded by the nearest environment canada weather station (collar temperature minus weather station temperature; called ∆t) for each record. positive or negative ∆t values indicated that the animal was in a location warmer or cooler than the temperature recorded at the weather station, respectively. however, complicating this measure was the fact that it was possible that the temperature as recorded on the collar was affected by radiant heat and variation in the degree of shading from the animal. we predicted that the impact of shading might vary between day and night, regardless of season but be relatively consistent all year. further, we expected the impact of radiant heat would be consistent for these endothermic homeotherms across the range of temperatures experienced by the animal in any one season. however, the impact of radiant heat should be greater in the summer due to the lower insulative potential offered by the summer coat. therefore, independent analyses were conducted for day and night as well as during the summer, early winter, and late winter to minimize bias. because of radiant heat, we expected that the collar temperature would be a positively biased measure of local temperature and therefore the power to detect selection of cooler areas based on temperature is reduced. therefore, we expected our results to be conservative. we did not analyze spring and fall data because of our inability to control for variation in the growth or shedding of the winter coat (samuel et al. 1986). to be further conservative and minimize the impacts of equipment malfunction, within each dataset (e.g., summer day data) we first sorted data by ∆t and deleted 5% of the data on each extreme of the continuum so that we only worked with 90% of the data. for prediction 1, we regressed ∆t on the temperature data from the nearest environment canada weather station recorded at the same time, or within an hour, for each individual moose. we predicted that if moose were selecting cooler areas during periods of warmth that moose response to temperature broders et al. alces vol. 48, 2012 56 there would be a significantly negative slope. assumptions of regression were confirmed for all analysis via an examination of residuals for normality and homoscedasticity (sokal and rohlf 1995). to test prediction 2 that moose would move less during periods of high temperatures, we divided the range of environmental temperature values for each of the 3 seasons and for day and night into 2 groups with a 7º c range (arbitrarily chosen based on the range of values in the dataset). we did not use the extremes of the temperature continuum because of low sample sizes and we omitted the records with temperature values in the middle of the distribution to ensure there was opportunity to detect variation between the 2 groups, if indeed there was meaningful variation. during the summer, the temperature ranges used were 10 to 16º c for low temperature and 20 to 26º c for high temperature. during early and late winter the temperature ranges used were -11 to -5º c for low temperature and 0 to 6º c for high temperature. when temperatures were within either the low or high range, for each animal, we calculated the average movement distance (i.e., straight-line distance between successive locations) over all 2 h periods for which we had location data. a one tailed ttest was used to test if the distance travelled during periods of high temperature was less than during periods of low temperature (for that season); we used α = 0.05 for decisionmaking criteria. where we found evidence of ectothermic response to temperature, we characterized the types of sites where moose were located during these times to better understand site type selection. results we recorded 29,964 locations for 12 moose from 2002-2006. summer we examined the temperature response of 7 moose (3 males from halifax county and 2 females from each of halifax and cumberland county) but we only had 2 h movement data from 5 of these animals (all from halifax county), as 2 individuals had collars programmed to record data at 4 h intervals. during the daytime, the slope of the regression lines varied among animals but all were negative and different from zero (all ps < 0.001). there was also a trend in the response to environmental temperature such that the male response (combined results: ∆ t = 8.0 – 0.28 ec temp; p <0.001, df = 1732, β0 se = 0.39, β1 se = 0.02) was greater than that of females (combined results: ∆ t = 4.8 – 0.20 ec temp; p <0.001, df = 1972, β0 se = 0.36, β1 se = 0.02). during the day all 5 moose moved less (all ps <0.001) during periods of high temperature than low temperature. the average movement distance during low temperatures was 2.2 x further than during high temperatures (range = 1.8-2.6 x). during periods of high temperature there was a greater proportion of locations in softwood, and a smaller proportion in mixed wood and open areas than during periods of cooler temperatures (table 1). at night the slopes of the regression lines for all 7 moose were negative and different from zero (all ps < 0.001). as in daytime, there was also a trend in the response to environmental temperature by gender, with the response by males (combined results: ∆ t = 6.6 – 0.39 ec temp; p <0.001, df = 12552, β0 se = 0.47, β1 se = 0.03) greater than that of females (combined results: ∆ t = 2.9 – 0.24 ec temp; p <0.001, df = 1629, β0 se = 0.38 β1 se = 0.02). there were far fewer instances (only 5-16 % as many per individual) of high temperature records than low temperature records. there was no difference (p < 0.05) in the movement distance between periods of high and low temperature for 3 of the 5 moose; 2 others (a male and a female from halifax county) moved more during high temperatures. alces vol. 48, 2012 broders et al. moose response to temperature 57 early winter we examined the temperature response of 5 moose (3 males from halifax county and 1 female from each of halifax and cumberland county). during this time we only had 2 h interval location data from 4 moose (all from halifax county) because the collar for the moose in cumberland county was programmed to record at 4 h intervals. during the day the slopes of the regression lines for all 5 moose were negative, but only 4 of these slopes (range of -0.11 to -0.23) were different from zero (p <0.001; the exception was a halifax county male). one of the individuals for which we have movement data had only 6 records for ‘low’ temperature, therefore analysis was only conducted for 3 individuals (2 males and 1 female), each with ≥20 records for each of ‘high’ and ‘low’ temperature; there was no difference in movement distance between periods of high and low temperature (all ps >0.05). at night there was no consistent trend in temperature response among the 5 moose tracked. regression lines for 2 of the 5 moose (both males from halifax county) were not different from zero (p >0.05), whereas another male and the female from the same area had positive relationships (p <0.05); a female from cumberland county had a negative relationship (p <0.05). one of the 4 moose (a halifax county female) moved more (p <0.05) during high temperature periods; movement distance was not different for the other 3 moose. late winter we had data to examine the temperature response of 11 moose (3 males and 3 females from halifax county, 4 females from cumberland county, and 1 female from antigonish county). however, only 6 animals wore collars programmed to record locations at 2 h intervals; therefore, movement analysis was conducted only with these 6 (2 males and 3 females from halifax county and 1 female from antigonish county). during the day there appeared to be minimal effects of gender in temperature response, but there was a trend with geography. of the 6 moose (3 males and 3 females) from halifax county, only 1 (male) had a regression line different (p <0.05) from zero (it was negative). however, each of the other 5 moose (all female; 4 from cumberland county and 1 from antigonish) had negative regression lines that were all similar to one another (combined results of 5 non-halifax county moose: ∆ t = 0.51 – 0.17 ec temp; p <0.001, df = 892, β0 se = 0.12, β1 se = 0.02). during the day we only had 2 h interval location data from moose in halifax county and one animal in antigonish. although 5 of the 6 moose had average movement distances that were less in high temperature periods than low temperature periods, only 3 were different (p <0.05), the 2 females from halifax and 1 from antigonish. there was a geographic trend in temperature response at night in that each of the 5 non-halifax county moose (all female; 4 from cumberland county and 1 from antigonish) had negative regression lines (combined results of 5 non-halifax county moose: ∆ t = -0.50 – 0.145 ec temp; p <0.001, df = 914, β0 se = 0.132, β1 se = 0.016). of the 6 halifax county moose, 3 had regression temperature site type 20-26 °c 10-16 °c difference softwood 47.6 39.5 8 mixedwood 26.1 30.3 -4.2 open 14.7 19.5 -4.8 water 7 5.7 1.3 hardwood 4.5 4.6 -0.1 other 0.2 0.4 -0.2 # locations 1331 1359 table 1. proportion (%) of the total locations of 11 moose on mainland nova scotia in each of 6 site-types when the temperature recorded at the nearest environment canada weather station was between 20-26 °c and 10-16 °c, 15 june-15 september, 2002-2006. moose response to temperature broders et al. alces vol. 48, 2012 58 lines not different from 0 (p >0.05) and 3 had positive regression lines (p <0.05). during the night there was no difference in the movement distance by 5 of the 6 moose during periods of high and low temperature. one moose (halifax county female) moved less (p <0.05) during high temperature periods relative to low temperature periods. discussion annual movement patterns, home ranges, daily, seasonal, and annual temperature regimes and a number of anthropogenic factors affect moose behavior (andersen 1991, schwab and pitt 1991, courtois et al. 2002, dussault et al. 2004). in this paper we present evidence from 2 predictions that support the hypothesis that moose on mainland nova scotia alter their behavior to reduce physiological stress in response to high temperatures. the extent to which this behavior is able to ameliorate the impacts of heat stress is unknown and would require further investigation. this finding is consistent with dussault et al. (2004) who suggested that moose spend more time under cover during hot days but come out to feed during cool nights. contrary to our results, research along the southern extent of moose range in ontario did not support our hypothesis as moose in their study did not seem to exhibit any behavioral response to increased temperatures (lowe et al. 2010). they found that there was little variation in temperature trends among site types such that there was minimal, if any, thermal advantage to selecting one site type over another. instead, they found that animal movement was negatively related to snow depth. in nova scotia, moose did not always display ectothermy by moving to thermal cover at the same temperature threshold or at the metabolic heat stress temperature threshold identified by renecker and hudson (1986). for example, during summer nights, based on individual regressions, moose moved to cooler areas when temperatures reached 14o c, which is consistent with their findings. however, during summer days there was more inter-individual variation and many moose did not move to cooler areas until around 24o c. during early winter days moose tended to seek cooler areas as day temperatures still generally exceeded their upper critical temperature. this pattern is not reflected during early winter nights or in late winter, possibly due to the lack of microclimate variation, indicating that temperature may be similar in all cover types. dussault et al. (2004) noted that some moose used open, deciduous, or mixed areas even when air temperatures are warm because thermal cover often offers low food availability. given the complexity of an animal’s thermal environment, factors such as wind and solar radiation in combination with ambient temperature presumably influence habitat use and movement. the temperature patterns were not as distinct in winter as they were in the summer, but we found that moose in winter had a greater than expected use of softwood cover based on availability of this cover type in their home range during times of expected heat stress. this finding agrees with other studies which suggest that during some seasons moose will use certain types of forest cover in disproportion to availability (cook et al. 2004). there is an assumption that closed canopy forests such as the softwood stands chosen by moose might provide areas with low snowfall amounts as well as thermal cover which makes them an even more attractive choice (jung et al. 2009). the moose population on the mainland of nova scotia is near the southern periphery of the species range in north america and is listed as ‘endangered’ due to its low population size. the behavioral patterns we document herein are suggestive of moose responding to increased temperature, and although may indicate heat stress, such thermoregulatory behaviour is not unexpected. given the concern for certain moose populations at the alces vol. 48, 2012 broders et al. moose response to temperature 59 southern fringe of their range, it would be prudent for further investigation into the extent to which behavioural thermoregulation (i.e., ectothermy) impacts population productivity. to be cautious, forest managers should be cognizant of, and explicitly address the need to maintain appropriate thermal cover on the landscape that allows moose to use ectothermy to ameliorate the effects of high temperatures at critical times of the year. further study may also be required to quantitatively characterize the relative ability of different cover types to buffer extremes of temperature. acknowledgements arcgis technical assistance was provided by greg baker, research tools technician for the mp_sparc lab. funding for this project was provided by an nserc undergraduate student research award and the nova scotia department of natural resources. this manuscript was improved based on comments provided by b. patterson and 2 reviewers. references andersen, r. 1991. habitat changes in moose ranges: effects on migratory behavior, site fidelity and size of summer home-range. alces 27: 85-92. beazley, k., m. ball, l. isaacman, s. mcburney, p. wilson, and t. nette. 2006. complexity and information gaps in recovery planning for moose (alces alces americana) in nova scotia, canada. alces 42: 89-109. brannen, d. c. 2004. population parameters and multivariate modeling of winter habitat for moose (alces alces) on mainland nova scotia. m. sc. thesis, department of biology, acadia university, wolfville, nova scotia, canada. campbell, n. a., and j. reece. 2002. biology. sixth edition. benjamin cummings, san franscisco, california, usa. chen, j. q., j. f. franklin, and t. a. spies. 1995. growing-season microclimatic gradients from clear-cut edges into oldgrowth douglas-fir forests. ecological applications 5: 74-86. cook, j. g., l. l. irwin, l. d. bryant, r. a. riggs, and j. w. thomas. 2004. thermal cover needs of large ungulates: a review of hypothesis tests. transactions of the 69th north american wildlife and natural resources conference: 708-726. courtois, r., c. dussault, f. potvin, and g. daigle. 2002. habitat selection by moose (alces alces) in clear-cut landscapes. alces 38: 177-192. demarchi, m. w., and f. l. bunnell. 1995. forest cover selection and activity of cow moose in summer. acta theriologica 40: 23-36. dussault, c., r. courtois,j.-p. ouellet, and j. huot. 1999. evaluation of gps telemetry collar performance for habitat studies in the boreal forest. wildlife society bulletin 27: 965-972. _____, j.-p. ouellet, r. courtois, j. huot, l. breton, and j. larochelle. 2004. behavioural responses of moose to thermal conditions in the boreal forest. ecoscience 11: 321-328. gamo, r. s., and m. a. rumble. 2000. gps radio collar 3d performance as influenced by forest structure and topography. biotelemetry 15: proceedings of the 15th international symposium on biotelemetry 15: 464-473. www.fs.fed.us/rm/pubs_other/ rmrs_2000_gamo_r001.pdf <accessed february 2011>. girard, i., j. p. ouellet, r. courtois, c. dussault, and l. breton. 2002. effects of sampling effort based on gps telemetry on home-range size estimations. journal of wildlife management 66: 1290-1300. jung, t. s., t. e. chubbs, c. g. jones, f. r. phillips, and r. d. otto. 2009. winter habitat associations of a low-density moose (alces americanus) population in central labrador. northeastern naturalist 16: 471-480. moose response to temperature broders et al. alces vol. 48, 2012 60 karns, p. d. 1997. population distribution, density and trends. pages 125-140 in a. w. franzmann and c. c. schwartz, editors. ecology and management of the north american moose. smithsonian institution press, washington, d.c., usa. lankester, m. w. 2010. understanding the impact of meningeal worm, parelaphostrongylus tenuis, on moose populations. alces 46: 53-70. leblond m, c. dussault, and j. p. ouellet. 2010. what drives fine-scale movements of large herbivores? a case study using moose. ecography 33: 1102-1112. lenarz, m. s., j. fieberg, m. w. schrage, and a. j. edwards. 2010. living on the edge: viability of moose in northeastern minnesota. journal of wildlife management 74: 1013-1023. _____, m. e. nelson, m. w. schrage, and a. j. edwards. 2009. temperature mediated moose survival in northeastern minnesota. journal of wildlife management 73: 503-510. lowe, s. j., b. r. patterson, and j. a. schaefer. 2010. lack of behavioral responses of moose (alces alces) to high ambient termperatures near the southern periphery of their range. canadian journal of zoology 88: 1032-1041. marai, i. f. m., and a. a. m. haeeb. 2010. buffalo’s biological functions as affected by heat stress a review. livestock science 127: 89-109 moen, r., j. pastor, y. cohen, and c. c. schwartz. 1996. effects of moose movement and habitat use on gps collar performance. journal of wildlife management 60: 659-668. morrison, j. c., w. sechrest, e. dinerstein, d. s. wilcove, and j. f. lamoreux. 2007. persistence of large mammal faunas as indicators of global human impacts. journal of mammalogy 88: 1363-1380. murray, d. l., e. w. cox, w. b. ballard, h. a.whitlaw, m. s. lenarz, t. w. custer, t. barnett, and t. k. fuller. 2006. pathogens, nutritional deficiency, and climate influences on a declining moose population. wildlife monographs 166: 1-30. mysterud, a., and e. ostbye. 2008. cover as a habitat element for temperate ungulates: effects on habitat selection and demography. wildlife society bulletin 27: 385-394. parker, g. 2003. status report on the eastern moose (alces alces americana clinton) in mainland nova scotia. available from www.gov.ns.ca/natr/wildlife/largemammals/pdf/statusreportmoosens.pdf <accessed february 2011>. pulsifer, m. d., and t. nette. 1995. history, status and present distribution of moose in nova scotia. alces 31: 209-219. rempel, r. s., a. r. rodgers, and k. f. abraham. 1995. performance of a gps animal location system under boreal forest canopy. journal of wildlife management 59: 543-551. renecker, l. a., and r. j. hudson. 1986. seasonal energy expenditures and thermoregulatory responses of moose. canadian journal of zoology 64: 322-327. _____, and c. c. schwartz. 1997. food habits and feeding behaviour. pages 403-440 in a.w. franzmann and c. c. schwartz, editors. ecology and management of the north american moose. smithsonian institution press, washington, d.c., usa. rodgers, a. r., r. s. rempel, r. moen, j. paczkowski, c. c. schwartz, e. j. lawson, and m. j. gluck. 1997. gps collars for moose telemetry studies: a workshop. alces 33: 203-209. samuel, w. m., d. a. welch, and m. l. drew. 1986. shedding of the juvenile and winter hair coats of moose (alces alces) with emphasis on the influence of the winter tick, dermacentor albipictus. alces 22: 345-359. schwab, f. e., and m. d. pitt. 1991. moose alces vol. 48, 2012 broders et al. moose response to temperature 61 selection of canopy cover types related to operative temperature, forage, and snow depth. canadian journal of zoology 69: 3071-3077. sokal, r. r., and f. rohlf. 1995. biometry. third edition. w. h. freeman, new york, new york, usa. timmerman, h. r. 2003. the status and management of moose in north america circa 2000-01. alces 39: 131-151. alces28_79.pdf alces26_172distinguishedmoosebio.pdf alces28_215.pdf alces29_225.pdf alces27_12.pdf alces20_3.pdf alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 improving moose population estimates in russia: accounting for distance between residential areas and track sightings vladimir m. glushkov russian institute of game management and fur farming, russian academy of agricultural sciences, kirov, russia. abstract: moose (alces alces) population density in the kirov region of russia is often overestimated when using the relationship between the distance between a residential area and the initial sighting of moose tracks. this paper presents a modified approach to provide better estimates when using this techinique. statistically valid density estimation techniques, standardization of estimation points and routes, landscape characteristics, and time have been addressed in the new approach. moose density is estimated once annually based on the distance to the first track, and annual surveys should maintain alike protocol. this improved method will provide more accurate population density estimates critical to prevent regional overharvest of moose. alces vol. 49: 149–154 (2013) key words: alces alces, density, distribution, moose, population estimate, russia, track surveys. introduction & background spatial distribution of individuals within a population is generally described as 3 types – equal, occasional, and grouped (odum 1986) – that can be affected by regional and temporal influences (naumov 1963). in estonia, moose (alces alces) distribution changes seasonally; moose in summerautumn are evenly distributed but in winter their distribution is sporadic or a “focal type of distribution” (ling 1977). likewise, moose distribution differs between summer and winter in the northeast portion of european russia (i.e., kirov region; glushkov 1982). this seasonal difference is caused by november migration related to forage deficiency on summer range (yazan 1972), as well as increased moose hunting that occurs after snow cover (glushkov 1997, 2001). the relationship between snow cover and increased harvest has not been considered previously relative to population density estimates (i.e., ecological density; bubenik 1965) that are based upon the distance between a residential area and the initial sighting of moose tracks. the unique spatial distribution caused by this relationship is neglected in typical winter route censuses (wrc), creating error in abundance estimates (glushkov 2004) and potential overharvest of moose that threatens population stability (glushkov et al. 2012). this paper provides the rationale for a modified approach to account for this relationship when calculating a population estimate. previous studies provide baseline information about seasonal moose distribution in the kirov region of russia (glushkov 1977). group size is larger in winter (2.8 ± 0.9) than in summer (2.0 ± 0.6), and dispersion:density ratios of 2.4 in november versus 5.1 in march (measured from aerial surveys within 1 min flight range of 60 ha corresponding author: vladimir m. glushkov, russian institute of game management and fur farming, russian academy of agricultural sciences, kirov, russia email: v.m.glushkov@yandex.ru 149 plots [n = 970]) confirms the more uneven winter distribution of moose. these early (november) and late winter (march) data (1976–1985) were used to construct a graph of sighting frequency in plots with varied moose density that described the character and variance of seasonal moose distribution in the kirov region (fig. 1). there were fewer unoccupied plots (0 moose) and plots occupied by ≥4 moose in november than in march when there were fewer plots with 2–3 moose. these seasonal differences are statistically different, and specific to both particular areas (χ2 = 42.7–171.4) and the kirov region as a whole (χ2 = 118.1). these data made it possible to classify summerautumn distribution of moose as “occasional” and winter distribution as “grouped with cluster formation” (glushkov 2001). the distance between human settlements and the initial observation of a moose track was measured during helicopter surveys; the area between was assumed absent of moose. this distance was compared to the sighting frequency of animals and tracks in occupied habitat. in the southern area of the region the correlation was not as strong (r = − 0.50, tr = 2.15) as in the north (r = – 0.65, tr = 3.42). similar tests were conducted with data from 288 terrestrial straight-line survey routes in 27 regional districts (2595 km total length with 288–200 ha sample plots; november 1996); the distance to the initial moose track and the population estimate was inversely related (corr. coeff. = −0.35; p = 0.002). in 10 of 15 districts surveyed, the average distance to the sighting of the first track was >7 km, and in 2 districts it was ∼9 km; tracks were first observed at a distance of 14, 16, and 25 km on the other 3 routes. because physical ability limits the intensity and extent of a terrestrial survey, an equation was developed (glushkov 1999) to calculate the probable distance to the initial track encountered (l) from the length of the route travelled where no tracks were encountered (r0): l ¼ 0.816r0þ2.98 ð1þ the histogram depicting the distribution of theoretical frequencies of plots with various moose densities indicated that the proportion of plots with 2 animals was underestimated 4 times and that of unoccupied plots was overestimated 2 times. in general, the equation to estimate population density (p) from distance (x) had little practical value (p = 9.17 − 0.54 x). the error in density estimates was presumably due to insufficient area in sample plots. a subsequent survey (1999) was carried out on 14 routes with larger sample plots (700–2200 ha) at the end of each route. the following describes this new survey approach that provides more reliable population density estimates. results and discussion figure 2 depicts the relationship between moose population density (d) and the distance (x) from a residential area to the initial (recent) track. the predictive equation was formulated with a logarithmic density fig. 1. the relationship between the sighting frequency and the number of observed moose per plot in early (november) and late (march) winter in the kirov region, russia. 150 moose population estimates in russia – glushkov alces vol. 49, 2013 function that is considered reasonable and acceptable for sample estimates based on the inequality criterion of dispersion and the mean-square deviation of p (draper and smith 1986). d ¼ � 7.7023 lnðxþ þ 20.635; r2¼ 0.8981, p < 0.001 ð2þ a verification of this equation was attempted in 2003–2004 within an experimental hunting farm (63,000 ha) by comparing the “known” moose population estimate with an estimate derived from survey routes and sample plots; the estimate was comparable and deemed satisfactory. however, this comparison is general at best because there was no differentiation between seasonal estimates (early and late winter), and no method to evaluate extrapolation across a larger area. in an attempt to verify the method in practice, and to achieve necessary reduction of the dispersion value, the number of paired observations would need to increase to 40 based on equation (2) and the value of coefficient of determination. in general, the experimental estimates were not contradictory of the hypothesis that moose density is directly related to the distance from a residential area due to their anthropophobic behavior as a result of intensive hunting. this new method is more elementary and easier to implement at the beginning of winter to estimate moose density at both the district and regional scale. its use is intended for determining abundance trends and setting seasonal harvest quotas (glushkov and buldakov 1997). specification of the starting route point, radial direction, and reference to a sample plot at the end of the route removed some associated drawbacks of the traditional wrc method. the independence of the “distance” parameter from weather conditions increases not only accuracy but also comparability of estimates. a relatively even population distribution in early winter predetermines reduction of the estimate error, and defines the “native population which inhabits a given area during summer, autumn, and early winter and is subject to hunting”, a definition critical to determine harvest level. the estimate makes it possible to determine, apart from ecological density, an area that is actually used by moose during early winter (extrapolation area), and animal numbers at the district and regional levels. comparability theory (yurghenson 1970) can be used as the basis to extrapolate population density estimates provided that data are available in a particular region to estimate density in subsequent years. it is possible to use equation (2) initially while simultaneously measuring and calculating plot estimates to improve the population estimate. if necessary, a locally specific equation can be developed from a single estimate from the plots and sample routes; calculations of the average r value and extrapolation areas are provided in glushkov (2001). application of this new method utilizes gis technology and requires preparatory work to organize permanent estimation fig. 2. the relationship between moose population density and the distance to the first track sighting in the kirov region, russia. alces vol. 49, 2013 glushkov – moose population estimates in russia 151 points and placement of routes and plots. the area unused by animals and the extrapolation area are determined with gis technology. an estimation point can be any “standard” residential area – a village, a workers’ settlement, or a farm enterprise with ≥30 people; all are recorded in reference books, marked on maps, and have a post office and permanent approach roads. the principle criteria for selecting these areas are that they are dead-end locations on a year-round motorway and representative of the surveyed lands within the district. the number of estimation points in a district depends on the total area, % forest cover, land cover diversity, and the number of settlements and their distribution. ideally, 4 radial routes with plots at the end must cover the study area completely (see fig. 3); the inner circle corresponds to the anthropogenic zone with zero moose density. moose population density within the ring with sample plots is equal to the density over the whole habitation area (outer ring, fig. 3). in districts with large forest area and lack of human settlements, the width of the ring and the area of land used by animals (extrapolation area) can be correspondingly large. in districts with small fragmented forests and densely populated settlements, the habitation area around proximate human settlements decreases by the value of the overlapping anthropogenic area; i.e., the extrapolation declines. the area standards for one estimation point are 100,000 ha for districts with forest cover >65%, and 60,000 ha otherwise. however, these standards may require a design compromise due to conflict with statistical requirements and the predetermined error value of the estimation data. for example, in densely populated districts with little forest cover, fragmented forests are often isolated by farming lands, settlements, and other man-made features. in this case, an estimation point can be a settlement which is located near a relatively big forest. the width of the forest along the line to the nearest settlement should be ≥2x the average distance to the first moose track; smaller forests and forests located in anthropogenic zones are not subject to estimation. the routes from such estimation points should be oriented into the forest not the cardinal directions (fig. 4). reducing the number of routes and plots to 1–2 per estimation point requires an adequate increase in the number of estimation points. standard route lengths are necessary to carry out the first estimate that is used for further corrections (table 1). the route length is subsequently corrected from the distances to the first track in the experimental estimates. a route is travelled one-fold, once a year, preferably by vehicle. choosing the size and the shape of sample plots is particular to the size of compartments, configuration of forests, and availability of access routes. it is best to use rectangular plots of fig. 3. the principle scheme for establishing 4 radial survey routes with sample plots to estimate moose density in the kirov region, russia. the center typically represents human settlement with zero moose density; density is extrapolated for the area of concentric rings. 152 moose population estimates in russia – glushkov alces vol. 49, 2013 2x4 dimension or plots of other shapes in areas >800 ha (agafonov et al. 1988). the new method of field data collection and subsequent calculation of moose population estimations described here will provide more reliable population estimates than with previous approaches. this is critical in the kirov region of russia that has experienced population overestimates and subsequent overharvest of moose. a coordinated strategy of using better population estimates, and measuring calf survival and non-harvest mortality, including poaching, will benefit regional moose management in russia (glushkov 2009). references agafonov, v. a., s. a. korytin, and i. n. solomin. 1988. winter estimate of game animals on circular routes: rational methods of game animals study. pages 17–25 in russian state research institute of game management and fur farming, kirov, russia. (in russian.) bubenik, a. b. 1965. population density of game animals, feed capacity of game habitat, and forest damage caused by animals. moose biology and hunting 2: 265–280. moscow, russia. (in russian.) draper, n., and h. smith. 1986. applied regression analysis. mir, moscow, russia. (translated from 1981 english version.) glushkov, v. m. 1977. on the method of arial moose count. hunting and game management 12: 14–15. (in russian.) ———. 1982. the moose of vyatka forests. hunting and game management 1: 16– 18. (in russian.) ———. 1997. extrapolation area estimates for moose and wild boar. pages 81-83 in applied ecology issues. proceedings of scientific conference devoted to 75th anniversary of russian state research institute of game management and fur farming. kirov, russia. (in russian.) ———. 1999. moose. pages 117–163 in management of game animals. proceedings of russian state research, institute of game management and fur farming, russian academy of agricultural sciences. kirov, russia. (in russian.) ———. 2001. moose: ecology and management of populations. russian academy fig. 4. a depiction of how survey routes are distributed in irregular fragmented forests; forest width to the nearest settlement should be >2x the average distance to the first moose track. multiple estimation points may be required to achieve a sufficient number of survey routes in an area. table 1. the stand and length of radial routes required in varying proportions of forest cover to estimate moose population density in the kirov region, russia. % forest cover route length (km) 80–100 12 70–79 10 55–69 8 40–54 7 25–39 5 <25 4 alces vol. 49, 2013 glushkov – moose population estimates in russia 153 of agricultural sciences, kirov, russia. (in russian.) ———. 2004. on route standardization for moose estimates. game management bulletin: 195–200. (in russian.) ———. 2009. improving population management and harvest quotas of moose in russia. alces 45: 43–47. ———, and a. l. buldakov. 1997. principles of ungulate estimate with the combined method. pages 83–85 in issues of applied ecology, institute of game management and fur farming. proceedings of scientific conference devoted to 75th anniversary of russian state research institute of game management and fur farming, kirov, russia. (in russian.) ———, m. g. dvornikov, v. v. kolesnikov, v. g. safonov, a. a. sergeyev, m. s. shevnina, and v. v. shiryaev. 2012. factors obstructing wild ungulate management in russia. pages 76–83 in theoretical and applied ecology. (in russian.) ling, h. 1977. experience of system analysis of population structure and dynamics. pages 7–105 in proceedings of tartu university, issue 408. university of tartu, tartu, estonia. naumov, n. p. 1963. animal ecology. moscow, russia. (in russian.). odum, e. 1986. ecology. volume 2. moscow, russia. (translated from english.) smirnov, v. s. 1964. mammal estimation methods. proceedings of the biology institute of ural academy of sciences. (in russian.) yazan, p. 1972. game animals of pechora taiga. research institute of game management and fur farming. kirov, russia. (in russian). yurghenson, p. b. 1970. distribution theory, spatial analysis, and applied ecology of animals. bulletin of moscow society of nature explorers, biology department, moscow, russia. (in russian.) 154 moose population estimates in russia – glushkov alces vol. 49, 2013 improving moose population estimates in russia: accounting for distance between residential areas and track sightings introduction & background results and discussion references in memoriam warren b. ballard jr. warren b. ballard, jr.—beloved husband of heather a. whitlaw and widely published author, editor, and nationally recognized professor at texas tech university's department of natural resources management—passed away peacefully at his lubbock, texas home on january 12, 2012. during warren's long career he produced more than 200 peer-reviewed journal articles and raised over $5.9 million in grant, contract, and research funding, which supported more than 60 graduate students. “his legacy lives on in the students, faculty, and research projects he touched,” said michael galyean, dean of texas tech's college of agricultural sciences and natural resources. warren was born on april 28, 1947 in boston, massachusetts. the family moved to albuquerque, new mexico in the early 1950s, where warren attended st. pius x high school. he earned a bsc. in fish and wildlife management from new mexico state university, his msc in environmental biology from kansas state university, and phd in wildlife science from the university of arizona. after completing his msc in 1971, warren worked for the u.s. fish and wildlife service and then from 1973–1990 as a wildlife biologist and research scientist with the alaska department of fish and game. his groundbreaking research on predator-prey relationships, wolf ecology, and ungulate populations is still widely recognized and considered foundational research. he then worked as a consultant with the national park service and lgl alaska (an ecological research company) while getting his phd. he soon became the director of the new brunswick cooperative fish and wildlife unit (1993–1996) and on june 7, 1995 married the love of his life, fellow wildlife biologist heather whitlaw. warren served as research supervisor with the arizona game and fish department (1996– 1998) before joining the texas tech faculty in 1998 where he achieved the rank of full professor in 2003, was named the bricker chair in wildlife management in 2006, and horn professor in 2008. “horn professors … represent the very best of our faculty,” said texas tech president guy bailey. i a member of more than seven professional societies, warren ballard served as editor of four international journals including the wildlife society bulletin and alces from 1998–2001. he was an associate editor of alces for many years as well as wildlife monographs and the wildlife society bulletin. the quality of ballard's research has been recognized with 22 awards. in 1989, he was honored by his peers with the distinguished moose biologist award. in 2002, he earned the chancellor's council distinguished research award at texas tech as well as a special service recognition award from the wildlife society. ballard was recognized as the outstanding researcher in the college of agricultural sciences and natural resources at texas tech—four times. he became a tws fellow in 2005, and has won various society awards for best professional article, monograph, and publications. warren's legacy in the moose world is certain. his two chapters on population dynamics and predator-prey relationships (coauthored with victor van ballenberghe) in “ecology and management of north american moose” stand as seminal summary works. the moose world is indebted to warren ballard and the species is better off for his being—he will be missed. ii alces24_14.pdf alces22_253.pdf alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces(25)_98.pdf alces22_83.pdf alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces21_215.pdf alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces27_140.pdf alces20_209.pdf alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces(25)_52.pdf alces(23)_227.pdf alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces(25)_175.pdf alces24_188.pdf alces29_281conferenceworkshop.pdf alces29_187.pdf impact of moose browsing on forest regeneration in northeast vermont haley a. andreozzi1, peter j. pekins1, and matt l. langlais2 1department of natural resources and the environment, university of new hampshire, durham, new hampshire 03824, usa; 2department of forests, parks & recreation, st. johnsbury, vt 05819, usa abstract: moose (alces alces) play an important role in the ecological and economic resources of northern new england, a landscape dominated by commercial forests. this study measured the impact of moose browsing on forest regeneration in wildlife management unit e1 in northeastern vermont where moose density was considered high in the 1990–2000s. we surveyed 37 clearcuts categorized into 4 age classes (3–5, 6–10, 11–15, and 16–20 years old). the stocking rate (stems/plot) of commercial species ranged from 74–76% in the 3–5, 6–10, and 11–15 year age classes, increasing to 86% in the 16–20 year age class. the proportion of plots containing a commercial tree without severe damage was above the accepted threshold stocking level of 40–60% in all age classes. the proportion of plots containing a commercial hardwood stem declined with increasing age class; the opposite occurred with softwood stems indicating a possible shift from hardwoodto softwood-dominated stands from selective browsing pressure. height of 11–20 year old stems was less than in new hampshire, indicating that growth was possibly suppressed in vermont due to higher moose density. overall, browsing was not considered a major problem based upon stocking rates. further study is warranted to evaluate whether compensatory growth occurs in response to reduced browsing as forests age and/or moose population density declines. alces vol. 50: 67–79 (2014) key words: alces alces, browsing, clearcut, damage, moose, new england, regeneration, stocking. moose (alces alces) populations have experienced a regional increase in northern new england over the last several decades, making them an increasingly valuable wildlife resource. they play an important role ecologically and economically in vermont, with 78% of the state open to regulated moose hunting and 406 hunting permits issued statewide in 2011 (vtfw 2008, 2011). with forests also covering 78% of vermont's landscape, the state generates over $1.5 billion annually from forest-based manufacturing and forest-related recreation and tourism (nefa 2007). the majority of forestland, 4 million acres, is owned privately or by timber investment management organizations; local, state, and fed‐ eral government owns ∼19% (919,440 acres) (nefa 2007). forest and wildlife management aimed at sustainable forest production is critical for the long-term stability of both vermont's economy and moose population. with adult moose weighing 300–600 kg, substantial browse is required to maintain such large body size (bubenik 1997), estimated at daily dry matter intake of 2.8 kg/ moose/day in january (pruss and pekins 1992). moose have the ability to substantially alter plant communities and are capable of damaging woody plants (renecker and schwartz 1997); repeated browsing can suppress height growth and recruitment of saplings into the canopy (risenhoover and maass 1987). moose browsing has the capability to affect the structure and dynamics of forest ecosystems over the long-term (mcinnes et al. 1992), which has important 67 implications for the management of forests where moose populations are regulated. moose show preference for forage in clearcut and early-successional habitat that is typical of the commercially managed forests of the northeast (thompson et al. 1995, scarpitti et al. 2005). for example, productive moose habitat in new hampshire was linked directly to the early successional forage created by commercial forest harvesting and early-successional browse is a dietary component year-round (scarpitti et al. 2005, scarpitti 2006). clearcuts 5–20 years old provide suitable early winter habitat, as regenerating hardwood and softwood species provide both browse and cover for moose (thompson and stewart 1997). while the impact of moose browsing on forest regeneration has received substantial attention elsewhere, little attention has been paid to the potential and actual effects in northern new england (pruss and pekins 1992, scarpitti 2006, bergeron et al. 2011). in order to manage moose and forest resources with respect to moose density and damage to regeneration, it is important to have extensive ecological knowledge of the relationships among moose, the ecosystems they inhabit, the plants they use as forage (edenius et al. 2002), and the associated impacts on forest production such as timber quality impairment. as moose populations have increased in northern new england, land managers have implied that a relationship exists between high population density and reduced forest regeneration in clearcuts. on isle royale, mcinnes et al. (1992) found that moose browsing affected the structure and dynamics of forest ecosystems on a long-term scale; however, in larger landscapes such impacts are usually more localized and often relate to high seasonal density. in northern new hampshire, bergeron et al. (2011) evaluated the impact of browsing on the regeneration of commercial tree species in 3 regions with different moose population density (0.26–0.83 moose/km2). while regeneration of commercial trees was not considered a regional problem at any density, specific clearcut sites with low regeneration were found adjacent to traditional moose wintering areas. it was predicted that such sites could change from hardwood to softwood dominance over time (bergeron et al. 2011). by the early 2000s, there was anecdotal evidence that the moose population in northeastern vermont, specifically wildlife management unit (wmu) e, was causing measurable damage to forest regeneration; moose densities in wmu e were thought to be well over 1.5 moose/km2 (4 moose/ mile2) (c. alexander, vtfw wildlife biologist, pers. comm.). to achieve the desired population level, hunting permit numbers were dramatically increased by the vermont department of fish and wildlife (vtfw) from 440 to 833 permits in 2004, when it was believed moose had approached their biological carrying capacity (vtfw 2008). the number of hunting permits rose to 1046 in 2005 and continued to increase until 2009, when 1223 permits were issued statewide in an effort to accelerate population reductions to protect forest habitat. by 2008, the population density was approaching the goal set by the 10-year moose management plan (0.7 moose/km2 [1.75 moose/mile2]) and the number of permits was reduced to 765 in 2010 and 405 in 2011. in response, this study was designed to evaluate the impact of moose browsing on the regeneration of commercial tree species in wmu e1 in northeast vermont by conducting qualitative assessments of damage in clearcuts between 3–20 years of age. methods study area the study area was located in north‐ east vermont and encompassed all of 68 browsing and forest regeneration – andreozzi et al. alces vol. 50 vtfw wmu e1, covering an area of 682 km2 bordered by new hampshire and quebec (fig. 1). elevation ranges from ∼250–1,130 m, and it is dominated by maple (acer saccharum, a. pensylvanicum, a. rubrum) and birch (betula alleghaniensis, b. papyrifera) hardwoods, and coniferous stands of balsam fir (abies balsamea) and red spruce (picea rubens.) while heavily forested, timber harvesting is common throughout as the majority of the land is privately owned and commercially harvested (nefa 2007). the 2011 moose density was estimated at 0.77 moose/km2 (1.96 moose/ mi2) based on a rolling 3-year average of moose sightings by early winter (november) deer hunters, and was previously estimated in 2010 as 0.93 moose/km2 (2.41 moose/ mile2) based on aerial surveys (millette et al. 2011). field measurements regeneration surveys were performed in june 2012 to measure the impact of moose browsing on forest regeneration in clearcuts 3–20 years old (leak 2007, bergeron et al. 2011). clear-cuts were separated into 4 age classes (3–5, 6–10, 11–15 and 16–20 years old) in order to assess temporal changes during both the period of typical browsing (0–10 years) and at least 10 years post-browsing (11–20 years). in each age class, 8–11 clear-cuts were located using aerial photography; each was a minimum of 4.1 ha (10 acres) and a maximum of 16.2 ha (40 acres) in size to reflect the typical range in size of clear-cuts in the region (m. langlais, vermont department of forests, parks & recreation county forester, pers. comm.). in certain cases, clear-cuts >16.2 ha were used to achieve appropriate sample sizes within an age class; a section ≤16.2 ha was surveyed. small plot surveys (milacre, ∼2.3 m diameter circle) were evenly spaced on equidistant transects throughout each clear-cut (fig. 2). in each milacre plot, the dominant stem (tallest tree) was recorded as a commercial or non-commercial tree species. if the dominant stem was non-commercial, the plot was searched for the presence of commercial species; commercial species included yellow and white birch, sugar and red maple, american beech (fagus grandifolia), aspen (populus spp.), black cherry (prunus serotina), balsam fir, red and black spruce (picea mariana), and tamarack (larix laricina). stem damage was assessed on a qualitative basis as fork, broom, or crook (fig. 3). the height of the damage above or below breast height (approximately 1.4 m) was recorded, as well as the number of forks and crooks, and the severity of crooks based on angle. light crooks were those ≤30°, moderate crooks were those 30–60°, and severe crooks were those ≥60° from the dominant stem. the relative height of the fig. 1. the location of the study area in vermont used to assess the impact of moose browsing on forest regeneration, 2012. the area included all of wmu e1 in northeast vermont. alces vol. 50 andreozzi et al. – browsing and forest regeneration 69 dominant stem was estimated to the nearest foot when <3.05 m (10 ft), or as ≥3.05 m. data analysis broomed stems and multiple forks above breast height were considered browse defects indicative of a severely damaged tree; otherwise, damage was considered light or moderate. trees with lesser damage are expected to recover during future growth (switzenberg et al. 1955, carvell 1967, trimble 1968, jacobs 1969). stems with single forks above breast height, or multiple forks either above or below breast height were considered to have moderate damage. stems with a single fork below breast height or crooks were considered to have light damage. a fully stocked stand (average density for undisturbed stand) at 80 years was assumed if a minimum of 40–60% of plots (threshold) contained a dominant commercial stem without severe damage (leak et al. 1987). to evaluate relative height between age classes and further assess browse impact, comparisons were made of the proportion of plots containing a dominant commercial stem ≥3.05 m height without severe damage, as vegetation ≥3.05 m was presumed to be above the typical height of moose browsing (bergström and danell 1986). temporal comparisons were made to assess if younger age classes with high initial browse pressure recover to fully stocked stands after 10–15 years. analysis of variance (anova) and pairwise tukey's test were used to look for differences in browse damage between clear-cuts and age classes. analyses were performed with systat v. 13. significance for all tests was assigned a fig. 2. example of the sampling design used to measure browse damage in clearcuts in northeast vermont, summer 2012. equidistant transects were established upon which 100– 400, 2.3 m diameter plots were established to measure the presence of dominant commercial stems, stem quality, and relative height; modeled after bergeron et al. 2011. fig. 3. the 3 qualitative browse categories used to describe browsing damage of dominant stems in milacre sample plots (bergeron et al. 2011). 70 browsing and forest regeneration – andreozzi et al. alces vol. 50 priori at α = 0.05. results are presented throughout as ± se. results a total of 37 clearcuts were surveyed: 11, 8, 8, and 10 in the 3–5, 6–10, 11–15, and 16–20 year age classes, respectively. there were 1709, 1291, 1442, and 1585 milacre plots surveyed in the 4 age classes, respectively. stocking rate of commercial trees (stems/plot) was high in all age classes, and increased with age class (table 1); it ranged from 74–76% in the 3–5, 6–10, and 11– 15 year age classes, increasing to 86% in the 16–20 year age class. the proportion of commercial trees with severe damage was low overall, with <10% damaged severely in all age classes except in the 16–20 age class (11%, table 1). the proportion of plots containing a commercial tree without severe damage was above the defined threshold stocking level of 40–60% in all age classes (table 1, fig. 4), ranging from 67–68% in the 3–5, 6–10, and 11–15 year age classes, and increasing to 75% in the 16–20 year class. the proportion of dominant commercial trees ≥3.05 m without severe damage increased with age class with 1, 25, and 39% in the 6–10, 11–15 and 16–20 year age classes, respectively. the proportion of plots containing a commercial hardwood stem declined with age class, averaging 62, 51, 43, and 40% in the 4 age classes, respectively. conversely, the proportion of plots containing a commercial dominant softwood stem increased with age class, averaging 12, 24, 33 and 46% in the 4 age classes, respectively (fig. 5). the highest stocking rates (>80%) were restricted to softwood-dominated stands. the majority of plots with a dominant non-commercial stem also contained commercial stems (70–81% across age classes). the stocking rate of dominant commercial trees was lower (p = 0.02) in the 3–5 year age class than in the 16–20 year age class, although stocking rate was above the threshold stocking level in all age classes. the proportion of dominant commercial hardwoods was higher (p = 0.014) and the proportion of dominant commercial softwoods lower (p = 0.015) in the 3–5 year age class than in the 16–20 year age class. the proportion of plots beyond browse height (≥3.05 m) and without severe damage in the 6–10 year age class was lower than the 11–15 year (p = 0.022) and the 16–20 year age classes (p < 0.001). at least 3 commercial species accounted for ≥50% of the species composition within each age class (table 2). the majority of these species were classified with light to no damage, and the proportion of noncommercial species declined as age class increased (tables 1 and 2). the proportion of dominant commercial stems classified as hardwood declined with age class, averaging 83 ± 7.6, 69 ± 8.5, 58 ± 8.5 and 49 ± 7.2% in the 4 age classes, respectively; the opposite occurred with the proportion of dominant commercial stems classified as softwood that averaged 17 ± 7.6, 31 ± 8.5, 42 ± 8.5, and 51 ± 7.2%. red maple and yellow birch accounted for 24 and 20% of total species composition in the 3–5 year age class; no other commercial species accounted for more than 6%. in the 6–10 year class, red maple, balsam fir, and yellow birch accounted for the highest proportion of species composition (14–16% each) and in the 11–15 year age class, these 3 species accounted for 11–17% each, and red spruce 11%. red maple, balsam fir, and red spruce accounted for the greatest proportion of dominant commercial stems (21–23% each) in the 16–20 year age class; yellow birch fell to 6% (table 2). discussion overall, the impact of moose browsing on the regeneration of commercial tree alces vol. 50 andreozzi et al. – browsing and forest regeneration 71 table 1. summary values indicating the stocking of commercial tree species, stocking of commercial trees with and without severe damage, the proportion of commercial trees ≥3.05 m in height without severe damage, and the proportion of dominant commercial hardwood and softwood stems in clearcuts in northeastern vermont, 2012. rows with the same letter within columns are not statistically different (p > 0.05). age class stocking rate of dominant commercial trees (%)1 stocking rate of dominant commercial trees w/o severe damage (%)2 stocking rate of dominant commercial trees w/ severe damage (%)3 proportion of dominant commercial trees w/o severe damage and ≥3.05 m tall (%) proportion of dominant commercial hardwoods (%) proportion of dominant commercial softwoods (%) 3–5 74a 67 6 n/a 83a 17a 6–10 75ab 68 7 1a 69ab 31ab 11–15 76ab 67 9 25b 58ab 42ab 16–20 86b 75 11 39b 49b 51b 1proportion of plots containing dominant stems considered to be commercial species 2proportion of plots containing dominant commercial stems not considered to be severely damaged 3proportion of plots containing dominant commercial stems considered to be severely damaged fig. 4. stocking guide for main crown canopy of even-aged hardwood and mixed-wood stands relative to basal area, number of trees per acre, and mean stand diameter. the a-line is fully stocked, the b-line is suggested residual stocking (˜60%), and the c-line is minimum stocking (˜40%) (leak et al. 1987). the proportion (%) of commercial trees without severe damage are plotted by age class; stocking is projected to a 4" mean stand diameter. 72 browsing and forest regeneration – andreozzi et al. alces vol. 50 species in northeast vermont was considered minor. the stocking rate of commercial trees without severe damage was acceptable in all age classes based upon the minimum threshold stocking level of 40–60%, and severe damage from browsing was low in all age classes with regard to acceptable levels, ranging from 6–11% (table 1). while damage was low in all age classes, site-specific severe browsing can shift species composition (edenius et al. 2002). for example, moose drastically altered localized species composition on isle royale, michigan where browsed sites had lower overall tree density than unbrowsed sites due to decline in balsam fir and mountain ash (sorbus americana) and concurrent increase in white spruce (picea glauca) densities (snyder and janke 1976). the increasing proportion of dominant softwood stems with age indicates a possible shift to softwood-dominated stands due to selective browsing of hardwood species (fig. 5). the highest stocking rates (>80%) were restricted to softwood-dominated stands, and stands experiencing the highest levels of damage were stocked predominantly with hardwood species that had much higher damage relative to the softwood species (table 2); softwood species will likely dominate these stands as they mature. the most commercially valuable hardwood species in the study region are yellow birch and sugar maple, and they were dominant species in the youngest 2 age classes, but accounted for only 6 and 5% of dominant stems in the 16–20 year class. conversely, the commercial softwood species, balsam fir and red spruce, were minimal in the youngest age classes, but accounted for a large proportion of the dominant stems (21% each) in the 16–20 year class. red maple, a less valuable commercial species, was the most common deciduous tree species in all age classes ranging from 13–24% of dominant stems (table 2). a fig. 5. proportion (%) of plots containing either dominant commercial hardwood or softwood stems by age class in nothereastern vermont, 2012. alces vol. 50 andreozzi et al. – browsing and forest regeneration 73 table 2. species composition (%) and browse damage category of dominant stems by age class in clearcuts in northeastern vermont, 2012. age class species severe damage moderate damage light damage no damage total 3–5 american beech 0 0 2 1 3 aspen spp. 0 0 3 0 3 balsam fir 1 0 2 3 6 black cherry 1 0 0 0 1 red maple 0 0 15 8 24 red spruce 0 0 1 3 5 sugar maple 0 0 4 2 6 tamarack 0 0 0 1 2 white ash 0 0 0 0 1 white birch 1 0 1 1 3 yellow birch 1 0 13 5 20 non commercial na na na na 26 6–10 american beech 0 0 2 1 3 aspen spp. 0 0 1 0 2 balsam fir 1 1 4 9 15 red maple 1 1 12 3 16 red spruce 0 0 0 9 9 sugar maple 0 0 8 2 10 white ash 1 0 0 0 1 white birch 2 0 2 0 4 yellow birch 1 0 11 3 14 non commercial na na na na 25 11–15 american beech 1 0 1 0 3 aspen spp. 0 0 3 2 6 balsam fir 2 1 7 7 17 black spruce 0 0 0 1 1 red maple 3 2 8 0 13 red spruce 0 0 1 9 11 sugar maple 1 0 3 0 4 white birch 4 1 4 0 9 yellow birch 2 0 8 1 11 non commercial na na na na 24 16–20 american beech 1 1 3 2 6 aspen spp. 0 0 0 1 1 balsam fir 2 0 5 14 21 black spruce 0 0 0 1 1 red maple 7 1 14 1 23 table 2 continued . . . . 74 browsing and forest regeneration – andreozzi et al. alces vol. 50 similar trend occurred in new hampshire (bergeron et al. 2011) where the proportion of dominant commercial hardwood stems also declined with age class. heavy browsing pressure could potentially accelerate successional development by arresting or retarding the height development of preferred browse species in the region (mcinnes et al. 1992, davidson 1993). while previous site compositions are unknown, it is possible a shift from hardwood to softwood dominated stands may be the natural successional trend for these sites. although harvest records were unavailable for most sites, it appeared that many were originally mixed wood stands. the proportion of dominant commercial trees ≥3.05 m (beyond browse height) without severe damage increased with age class, peaking at 39% at 16–20 years (table 1). these stems are expected to recover from any moderate or light damage during future growth without browsing. in contrast, average values in adjacent northern new hampshire were 36, 60, and 71% in the 3 older age classes, suggesting that growth was more suppressed in vermont. intense browsing in areas of high moose density can arrest or retard growth of preferred browse species (bergerud and manuel 1968, angelstam et al. 2000). a study with exclosures on isle royale, michigan indicated that repeated browsing by moose retarded vertical growth of palatable species such as aspen and paper birch, and prevented stems from growing beyond browsing height resulting in a more open canopy (risenhoover and maass 1987). although heavy browsing of the same species in successive years can result in hedgy growth and lower height potential (peek et al. 1976, peek 1997), such stems can compensate if browsing declines or if removed in successive years; for example, after release of a dominant stem in forked stems (jacobs 1969) and the straightening of crooked stems with secondary growth over time (switzenberg et al. 1955, trimble 1968). a clipping study on isle royale indicated that the site-dependent survival and growth of balsam fir were related to suppression brought about by severe browsing in previous years (mclaren 1996). accurate prediction of damage is complicated by this dynamic process that is likely influenced by local site conditions, and seasonal moose density and site fidelity. in studies assessing browse damage in both southern and northern new england, time since harvest was negatively correlated with foraging intensity (faison et al. 2010, bergeron et al. 2011) which may allow compensatory growth by desirable hardwood species beyond the 16–20 year age class. however, an increasing dominance of softwood species coupled with suppressed growth of hardwood species indicates a possible shift in species composition in wmu e1. several studies have indicated change table 2 continued age class species severe damage moderate damage light damage no damage total red spruce 0 0 1 20 21 sugar maple 0 0 4 1 5 white birch 1 0 1 0 1 yellow birch 3 0 3 0 6 non commercial na na na na 14 alces vol. 50 andreozzi et al. – browsing and forest regeneration 75 in forest composition due to heavy moose browsing. in finland, heikkila et al. (2003) measured reduced height of preferred browse species resulting in the release of conifers from competition. on isle royale, moose prevented aspen, birch, and balsam fir from growing into the canopy, with little impact to spruce, resulting in a forest with fewer trees in the canopy, a well-developed understory of shrubs and herbs, and an increase in spruce biomass (mcinnes et al. 1992). similarly, selective pressure resulted in rapid occupation of spruce (picea spp.) as the dominant species in study stands in russia (abaturov and smirnov 2002). a similar trend is possibly occuring in northeast vermont where coniferous species account for >50% of total species composition in the 16–20 year age class (table 2). a reduction in moose density, as implemented in the study area, may also reduce future browsing pressure and provide for the release of preferred hardwood species. a population reduction in newfoundland in the early 1960s resulted in dramatic decline in the proportion of white birch and balsam fir stems browsed in 6–11 and 12–17 year old stands (bergerud et al. 1968). high-density moose populations have the potential to damage preferred forage species (peek 1997), but the negative impacts of over-browsing can be minimized if moose density is kept at low-moderate levels (brandner et al. 1990). in russia, a density of 0.3–0.5 moose/km2 retarded growth of preferred forage species such as aspen, whereas normal stand development occurred at 0.2–0.3 moose/km2 (abaturov and smirnov 1992). in sweden, simulated densities of 0.8–1.5 moose/km2 did not impact winter browse availability; impact was predicted at >2.0 moose/km2 (persson et al. 2005). both northern vermont and new hampshire are classified as a combination of spruce-fir, northern hardwood, and mixed forest types (degraaf and yamasaki 2001), and presumably measurable differences in forest regeneration reflect different moose density. bergeron et al. (2011) found a direct correlation between browse damage and moose density in northern new hampshire; the region with highest density had most damage. densities in northeast vermont were estimated at 1.2–1.8 moose/km2 in 1999–2009 and were probably higher than those in the highest density region of new hampshire estimated at 0.8–1.5 moose/km2 for the same time period (c. alexander, pers. comm., k. rines, nhfg wildlife biologist, pers. comm.). in both states significant differences were found in the stocking rate of dominant commercial trees, and the proportion of both dominant hardwood and softwood commercial tree species between the youngest and oldest age classes. however, the temporal comparisons among age classes indicate that sites with high initial browse pressure are often released from that pressure and recover to commercially valuable stands. in both vermont and new hampshire, stocking rate increased and damage declined over time with relative differences seemingly influenced by local moose density. compensatory growth in the region was measurable in the 16–20 year age class, but likely begins earlier when stems grow beyond browsing height. however, heavy browsing pressure on preferred tree species may result in lower stand height as measured in vermont and a possible shift in forest composition to coniferous species. further assessment is warranted to best evaluate the extent of compensatory tree growth in response to reduction in browsing due to forest aging and/or moose population density. acknowledgements we are grateful to the numerous commercial and private landowners for their cooperation and providing access to their property including, but not limited to, silvio 76 browsing and forest regeneration – andreozzi et al. alces vol. 50 o. conte national fish and wildlife refuge, plum creek, heartwood forestland fund, and devost leasing. we thank d. bergeron, c. alexander (vtfw moose biologist), and k. rines (nhfg moose biologist) for providing useful data. we are also thankful to t. millette, h. stabins, w. leak, and m. yamasaki for providing knowledge and guidance throughout the project, and to j. trudeau, n. fortin, and j. comeau for assistance in the field. references abaturov, b. d., and k. a. smirnov. 1992. formation of stands on clearings in forests with different moose population density. bulletin moskovskava obshestva isputatelij prirodi otdelenie biologii 97: 3–12. ———, and ———. 2002. effects of moose population density on development of forest stands in central european russia. alces supplement 2: 1–5. angelstam, p., p. e. wikberg, p. danilov, w. e. faber, and k. nygren. 2000. effects of moose density on timber quality and biodiversity restoration in sweden, finland, and russian karelia. alces 36: 133–145. bergeron, d. h., p. j. pekins, h. f. jones, and w. b. leak. 2011. moose browsing and forest regeneration: a case study in northern new hampshire. alces 47: 39–51. bergerud, a. t., and f. manuel. 1968. moose damage to balsam fir-white birch forests in central newfoundland. the journal of wildlife management 32: 729–746. ———, ———, and h.whalen. 1968. the harvest reduction of a moose population in newfoundland. the journal of wildlife management 32: 722–728. bergström, r., and k. danell. 1986. moose winter feeding in relation to morphology and chemistry of six tree species. alces 22: 91–112. brandner, t. a., r. o. peterson, and k. l. risenhoover. 1990. balsam fir on isle royale: effects of moose herbivory and population density. ecology 71: 155–164. bubenik, a. b. 1997. behavior. pages 173222 in a. w. franzmann and c. c. schwartz, editors. ecology and management of the north american moose. smithsonian institute press, washington, d.c., usa. carvell, k. l. 1967. the response of understory oak seedlings to release after partial cutting. west virginia university agricultural experiment station, bulletin 553. morgantown, west virginia, usa. davidson, d. w. 1993. the effects of herbivory and granivory on terrestrial plant succession. oikos: 23–35. degraaf, r. m., and m. yamasaki. 2001. new england wildlife: habitat, natural history, and distribution. university press of new england, hanover, new hampshire, usa. edenius, l., m. bergman, g. ericsson, and k. danell. 2002. the role of moose as a disturbance factor in managed boreal forests. silva fennica 36: 57–67. faison, e. k., g. motzkin, d. r. foster, and j. e. mcdonald. 2010. moose foraging in the temperate forests of southern new england. northeastern naturalist 17: 1–18. heikkila, r., p. hokkanen, m. kooiman, n. ayguney, and c. bassoulet. 2003. the impact of moose browsing on tree species composition in finland. alces 39: 203–213. jacobs, r. d. 1969. growth and development of deer-browsed sugar maple seedlings. journal of forestry 67: 870–874. leak, w. b. 2007. accuracy of regeneration surveys in new england northern hardwoods. northern journal of applied forestry 24: 227–229. ———, d. s. solomon, and p. s. debald. 1987. silvicultural guide for northern hardwood types in the northeast (revised). research paper ne-603. u.s. alces vol. 50 andreozzi et al. – browsing and forest regeneration 77 department of agriculture, forest servie, northeastern forest experiment station, broomall, pennsylvania, usa. mcinnes, p. f., r. j. naiman, j. pastor, and y. cohen. 1992. effects of moose browsing on vegetation and litter of the boreal forest, isle royale, michigan, usa. ecology 73: 2059–2075. mclaren, b. e. 1996. plant-specific response to herbivory: simulated browsing of suppressed balsam fir on isle royale. ecology 77: 228–235. millette, t. l., d. slaymaker, e. marcano, c. alexander, and l. richardson. 2011. aims-thermal a thermal and high resolution color camera system integrated with gis for aerial moose and deer census in northeastern vermont. alces 47: 27–37. north east state foresters association (nefa). 2007. the economic importance and wood flows from vermont's forests. north east state foresters association, concord, new hampshire, usa. peek, j. m. 1997. habitat relationships. pages 351–375 in a. w. franzmann and c. c. schwartz, editors. ecology and management of the north american moose. smithsonian institute press, washington, d.c., usa. ———, d. l. urich, and r. j. mackie. 1976. moose habitat selection and relationships to forest management in northeastern minnesota. wildlife monographs 48: 3–65. persson, i. l., k. danell, and r. bergström. 2005. different moose densities and accompanied changes in tree morphology and browse production. ecological applications 15: 1296–1305. pruss, m. t., and p. j. pekins. 1992. effects of moose foraging on browse availability in new hampshire deer yards. alces 28: 123–136. renecker, l. a., and c. c. schwartz. 1997. food habits and feeding behavior. pages 403–439 in a. w. franzmann and c. c. schwartz, editors. ecology and management of the north american moose. smithsonian institute press, washington, d.c., usa. risenhoover, k. l., and s. a. maass. 1987. the influence of moose on the composition and structure of isle royale forests. canadian journal of forest research 17: 357–364. scarpitti, d. 2006. seasonal home range, habitat use, and neonatal habitat characteristics of cow moose in northern new hampshire. university of new hampshire, durham, new hampshire, usa. ———, c. habeck, a. r. musante, and p. j. pekins. 2005. integrating habitat use and population dynamics of moose in northern new hampshire. alces 41: 25–35. snyder, j. d., and r. a. janke. 1976. impact of moose browsing on boreal-type forests of isle royale national park. american midland naturalist 95: 79–92. switzenberg, d. f., t. c. nelson, and b. c. jenkins. 1955. effect of deer browsing on quality of hardwood timber in northern michigan. forest science 1: 61–67. thompson, i. d., and r. w. stewart. 1997. management of moose habitat. pages 377-401 in a. w. franzmann and c. c. schwartz, editors. ecology and management of the north american moose. smithsonian institute press, washington, d.c., usa. thompson, m. e., j. r. gilbert, g. j. matula jr., and k. i. morris. 1995. seasonal habitat use by moose on managed forest lands in maine. alces 31: 233–245. trimble, g. r. 1968. form recovery by understory sugar maple under unevenaged management. u.s. department of agriculture, forest service research note ne-89. northeastern forest experiment station, broomall, pennsylvania, usa. vermont fish and wildlife department (vtfw). 2008. big game management 78 browsing and forest regeneration – andreozzi et al. alces vol. 50 plan 2010-2020: creating a road map for the future. pages 40-52. vermont fish and wildlife department, waterbury, vermont, usa. ———. 2011. 2011 vermont willdife harvest report moose. vermont fish and wildlife department, waterbury, vermont, usa. alces vol. 50 andreozzi et al. – browsing and forest regeneration 79 impact of moose browsing on forest regeneration in northeast vermont methods study area field measurements data analysis results discussion acknowledgements references space use and movements of moose in massachusetts: implications for conservation of large mammals in a fragmented environment david w. wattles1 and stephen destefano2 1massachusetts cooperative fish and wildlife research unit, department of environmental conservation, university of massachusetts, amherst, massachusetts 01003; 2u. s. geological survey, massachusetts cooperative fish and wildlife research unit, university of massachusetts, amherst, massachusetts 01003, usa. abstract: moose (alces alces) have recently re-occupied a portion of their range in the temperate deciduous forest of the northeastern united states after a >200 year absence. in southern new england, moose encounter different forest types, more human development, and higher temperatures than in other parts of their geographic range in north america. we analyzed seasonal minimum convex polygon home ranges, utilization distributions, movement rates, and home range composition of gps-collared moose in massachusetts. seasonal home range sizes were not different for males and females and were within the range reported for low latitudes elsewhere in north america. seasonal movement patterns reflected the seasonal changes in metabolic rate and the influence of the species’ reproductive cycle and weather. home ranges consisted almost entirely of forested habitat, included large amounts of conservation land, and had lower road densities as compared to the landscape as a whole, indicating that human development may be a limiting factor for moose in the region. the size and configuration of home ranges, seasonal movement patterns, and use relative to human development have implications for conservation of moose and other wide-ranging species in more highly developed portions of their ranges. alces vol. 49: 65–81 (2013) key words: alces alces, moose, home range, movements, roads, massachusetts. an animal's home range is the area where it finds the resources it needs for survival and reproduction (burt 1943); essentially it is a measure of spatial use for a given period of time. different home range estimators provide different information regarding how the animal uses space, including total area, areas of intensive use, and areas that are avoided (powell 2000, fieberg and börger 2012). animals have a cognitive map of their home range which allows them to exploit areas of concentrated resources and avoid areas that impart risks or disadvantages (powell 2000, powell and mitchell 2012, spencer 2012). thus home range size, configuration, and use can be influenced by the type, concentration, and distribution of resources, topography and other physical features, human development, and the distribution of other animals such as mates, competitors, and predators (powell and mitchell 2012). further, space use and movement patterns show seasonal changes which can be influenced by temperature and other climatic factors such as snow condition, reproductive status (börger et al. 2006, birkett et al. 2012), and for species that are affected by seasonal changes in forage quantity and quality like moose (alces alces) and other ungulates, foraging times, ruminating times, and metabolic rates (risenhoover 1986, cederlund 1989). knowledge of the size and position of an animal's home range and an individual's 65 movements and use of that area can provide insights into the distribution of resources and limiting factors in the environment (powell 2000, rettie and messier 2000, powell and mitchell 2012, spencer 2012). in areas of high human density, development of the landscape can be a major determinant of landscape use by many wildlife species (forman and deblinger 2000, lykkja et al. 2009, kertson et al. 2011). the concentration and distribution of industries and businesses, residences, roads and other infrastructure, and even the abundance of pets can affect the availability, quality, distribution, and connectivity of wildlife habitats. this is likely true for many or most taxa, but it is especially obvious for large mammals such as moose that require extensive areas to fulfill their life history needs. despite beliefs that temperature (kelsal and telfer 1974, renecker and hudson 1986, peek and morris 1998) and human development (vecellio et al. 1993, peek and morris 1998) might prevent it, moose have recently recolonized and become established in a portion of their historic range in the temperate deciduous forest of southern new england (vecellio et al. 1993, wattles and destefano 2011). this environment provides a number of potential challenges for moose, including forest types that differ from that found in most of its range (westveldt et al. 1956, degraaf and yamasaki 2001, franzmann and schwartz 2007), a thermal environment that could reduce fitness and survival (renecker and hudson 1986; boose 2001; murray et al. 2006; lenarz et al. 2009, 2010), and some of the highest densities of people in the united states (destefano et al. 2005, u. s. census bureau 2010a). habitat use, home range, and movement of moose have been studied throughout much of its range (franzmann and schwartz 2007), including elsewhere in the northeastern u. s. (leptich and gilbert 1989, garner and porter 1990, miller and litvaitis 1992, thompson et al. 1995, scarpitti et al. 2005). however, similar information has been lacking in southern new england where urban and suburban development and high road densities result in fragmentation of much of the landscape and relatively small and scattered natural areas. our objective was to determine how moose use the landscape in the humandominated and developed environment of central and western massachusetts. specifically, we wanted to quantify the seasonal home range size, space use patterns, and movement rates of moose in this recently re-established population. it is well documented that the reproductive cycle (e.g., the rut) and seasonal changes in forage affect movement patterns (belovsky 1981, risenhoover 1986, cederlund 1989, van ballenberghe and miquelle 1990), and we further predicted that movements would be influenced by weather patterns not experienced by moose elsewhere. also, due to the relatively limited number of human-moose conflicts reported in the state (wattles and destefano 2011), we predicted that moose would avoid areas with high levels of human development, locate their home ranges away from people, and that home range size and configuration would be influenced by development level. methods study area our study was conducted in central and western massachusetts, usa and adjacent portions of vermont and new hampshire, between 42° 9’ and 42° 53’ n latitude and 71° 53’ and 73° 22’ w longitude. topography is dominated by glaciated hills underlain by shallow bedrock. glacial activity created abundant small stream valleys, lakes, ponds, and wetlands whose size and nature varies with changes in beaver (castor canadensis) activity. the central and western sections of the study area are separated by the 66 home range and movements – wattles and destefano alces vol. 49, 2013 connecticut river valley which runs n-s through west-central massachusetts. elevation ranges from 100 m above sea level in the connecticut river valley, to 425 m in the hills of central massachusetts and 850 m in the berkshire hills of western massachusetts. the western two-thirds of massachusetts was >80% mixed deciduous, second, or multiple-growth forest, much of it resulting from regeneration of farm fields abandoned in the mid-late 1800s (hall et al. 2002). forest types included spruce-fir-northern hardwoods, northern hardwoods-hemlock (tsuga canadensis)-white pine (pinus strobus), transition hardwoods-white pinehemlock, and central hardwoods-hemlockwhite pine. transitions between forest types can be gradual or distinct depending on localized physiography, climate, bedrock, topography, and soil conditions, resulting in a patchwork of forest types and species groups (westveldt et al. 1956, degraaf and yamasaki 2001). dominant species included spruce (picea spp.), balsam fir (abies balsamea), american beech (fagus grandifolia), birch (betula spp.), trembling aspen (populus tremuloides), eastern hemlock, oaks (quercus spp.), hickories (carya spp.), and maples (acer spp.) depending on area and forest type. early successional habitat was created primarily through logging, and occasionally through wind and other weather events. about 1.5% of the forest was logged annually in 1984–2000, consisting of small (mean = 16.5 ha) cuts of moderate intensity (removal of 27% of timber volume) widely distributed on the landscape (kittredge et al. 2003, mcdonald et al. 2006). the pattern of forest harvest, glaciation, and transitional forest types provided a patchy mosaic of well interspersed forest types, age classes, and wetlands. july is the warmest month when mean daily temperature is 21 °c, and january the coldest when mean daily temperature is −6 °c. mean annual precipitation is 107 cm in central areas and 124 cm in western areas, with all months receiving 7–11 cm and 8–12 cm, respectively (degraaf and yamasaki 2001). the average date of last frost in the region is 15 may; the average day of first frost is 1 october and 15 september in central and western areas, respectively (degraaf and yamasaki 2001). snow depth is typically greater in western than central areas, and depths that restrict moose movement (50–70 cm) can occur in both areas (coady 1974). massachusetts is one of the most densely populated states in the u. s. (destefano et al. 2005; u. s. census bureau 2010a). development intensity is variable throughout the state, but tends to be substantially less in the uplands compared to the valley floors (<15–35 people/km2 in uplands and 35– >360/km2 in valley floors outside of major urban centers; u. s. census bureau 2010b). development in the uplands consists primarily of isolated homes and homes lining roadways within a matrix of forest; agricultural land and medium-to-large towns dominate the valleys. there is a dense road network throughout the area, consisting of state highways, paved, and unpaved municipal roads: 0.78 and 2.22 km of paved roads/km2 and 0.76 and 1.12 km of unpaved roads/km2 for uplands and valleys, respectively. study animals and gps telemetry we captured adult (>1 yr old) moose by opportunistically stalking and darting them from the ground between march 2006 and november 2009. moose were immobilized using either 5 ml of 300 mg/ml or 3 ml of 450 mg/ml xylazine hydrochloride (congaree veterinary pharmacy, cayce, sc, usa; mention of trade names does not imply endorsement by the u. s. government) administered from a 3 or 5 cc type c pneudart dart (pneudart, inc., williamport, alces vol. 49, 2013 wattles and destefano – home range and movements 67 pa, usa). we used tolazolene (100 mg/ml) at a dosage of 1.0 mg/kg as an antagonist. moose were fitted with gps collars, either ats g2000 series (advanced telemetry systems, inc., isanti, mn, usa) or telonics twg-3790 gps collars (telonics, inc., mesa, az, usa). we programmed the collars to attempt a gps fix as frequently as possible while allowing the battery life to extend for at least 1 year; depending on the collar, a gps fix was attempted every 135, 75, or 45 min. collars were also equipped with vhf transmitters, mortality sensors, and mechanisms that released the collars either at a low battery state or a pre‐ programmed date. capture and handling procedures were approved by the university of massachusetts institutional animal care and use committee, protocol numbers 25-02-15, 28-02-16, and 211-02-01. seasons we a priori defined the length and timing of seasons based on several ecological factors including vegetation phenology, weather (including temperature and snow conditions), and the moose reproductive cycle (table 1). the transition between seasons could vary by several days to several weeks depending on weather conditions and other factors. if movements were seen in the data that obviously demonstrated a change in season (e.g., a large increase in movement at the end of the winter when snow had melted or the end of summer indicating the beginning of rutting behavior), the seasons were truncated at that point and the data were included in the following season (fig. 1). home ranges and space use we used 2 methods to calculate space use by moose: minimum convex polygon (mcp) and utilization distributions (ud) by fixed kernel density estimator. we calculated100% mcp home ranges with the create minimum convex polygons tool in hawth's analysis tools (beyers 2006) and uds using the kernel density estimation tool in hrt: home range tools for arcgis (rodgers et al. 2007). all geographic information system (gis) work was performed in arcgis 9.3 (esri 2008). table 1. seasons used for calculating home-range, movements, and core-area habitat analyses for moose in massachusetts, 2006–2011. season breaks were based on phenology of vegetation, temperature, normal snow conditions, and moose reproductive activity. season dates vegetation/browse temperaturea movement moderators season length (d) spring 16 april–31 may growing season; bud-break-leaf out cool-hot potentially temperature 46 calving (females) 8–13 may–15 june growing season; bud-break-leaf out cool-hot newborn calf mobility 30 summer 1 june – 30 aug growing season; full leaf out hot temperature 92 fall 1 sept – 31 oct leaf out to leaf off hot-cool temperature and rut 61 early winter 1 nov – 31 dec dormant season; woody/evergreen warm-cold potentially metabolism 61 late winter 1 jan – 15 april dormant season; woody/evergreen cold-warm potentially snow and metabolism 107 atemperature ranges describing typical temperatures experienced during a season; cold ≤0 °c, cool >0 °c and <14 °c,warm ≥14 °c and <20 °c, hot ≥20 °c. 68 home range and movements – wattles and destefano alces vol. 49, 2013 fig. 1. the y-axis represents mean daily movement rates (m/day, thin line) for female (top; n = 5) and mature male (bottom; n = 10) moose in massachusetts, 2006–2011. the heavy line represents a 10day moving average to remove noise; the vertical dashed lines mark a priori delineated season boundaries. alces vol. 49, 2013 wattles and destefano – home range and movements 69 choice of the kernel bandwidth or smoothing factor (h) is known to have the greatest effect on the resultant utilization distribution when using kernel density estimators (worton 1989). a large h over-smooths the data resulting in a positively biased ud that encompasses unused habitats, whereas a small h under-smooths the data resulting in a fragmented ud (fieberg 2007, fieberg and börger 2012). quantitative methods of determining h can be influenced by sample size, sampling intensity, and the distribution of locations (kie et al. 2010, fieberg and börger 2012), and there is lack of agreement on the best method for calculating h (powell 2000, hemson et al. 2005, gitzen et al. 2006, fieberg 2007, kie et al. 2010, fieberg and borger 2012). we chose a 200-m bandwidth because it strikes a balance between creating a continuous polygon and over-buffering the edges of the utilization distribution. the 200-m bandwidth value merged closely separated locations into a single polygon, but did not merge widely spaced clusters. mitchell and powell (2008) noted that fragmentation of uds may be desired to identify used and unused areas in patchy and fragmented landscapes. increasing the bandwidth beyond 200 m resulted in uds with a larger buffer around all points, but failed to further merge disjointed polygons into a single polygon unless very large values of h were used. smaller values of h resulted in more fragmented uds that did not accurately represent space use. road densities in mcp home ranges and uds were calculated using the masseot (massachusetts executive office of transportation) roads layer (massachusetts office of geographic information 2005). we used a 2005 land use layer (massachusetts office of geographic information 2005) to calculate amount of forest and wetlands, and the protected and recreational open space layer (massachusetts office of geographic information 2005) to calculate amount of protected area. movements we calculated mean seasonal daily movement rates by calculating the distance between successive fixes and summing those distances for each 24-h period beginning at 0:00. mills et al. (2006) showed that decreased gps sampling intensity resulted in reduced observed movement rates in wolves (canis lupus) due to a reduction in tortuosity of the path. we corrected for the variable sampling rate in our collars (135, 75, and 45 min) by subsampling the more intensively sampled datasets (45 min), and taking every other and then every third location to simulate 90 and 135 min intervals, respectively. we saw a consistent reduction in movement rates with increasing sampling interval. therefore, we used this information to weight the movements observed in our 135(n = 23) and 45-min (n = 2) collars to the intermediate 75-min (n = 5) sampling level, making comparisons among individuals possible. statistics we used the r statistical package, version 2.12.2 (r development core team 2005) for all statistical analyses. we used mixed effect models in the r-package lme4 (bates et al. 2012) to analyze the differences in seasonal home range size and movement rates within and between sexes and seasons. we incorporated random intercept in the models to account for unequal sample sizes among sexes and seasons and to account for repeated measures on individual moose and performed post-hoc pairwise comparisons using the r-package lmerconviencefunctions (tremblay and ransijn 2012). we employed one-sample z-tests to compare road densities in the valley bottoms and uplands to home ranges. transformations failed to meet the assumption of normality; 70 home range and movements – wattles and destefano alces vol. 49, 2013 therefore, we used a nonparametric paired wilcoxon's rank-sum test to make comparisons in road density between mcp home ranges and uds. significance level for all analyses was set at 0.05. results capture and deployment of gps collars we deployed gps collars on 21 moose: 5 adult (>3 yr) females, 7 adult males, and 1 immature (<3 yr) male in central massachusetts, and 4 adult and 4 immature males in western massachusetts; 9 were recaptured to replace gps collars. we obtained 127,408 locations of the 21 moose with an overall fix rate of 85%. seasonal data for any animal were only included in the analyses if data were obtained across the entire season. the median number of locations/animal/season ranged from 402 in spring to 1,015 in late winter. the minimum number of locations was 281 for one animal in spring. home ranges and space use mean annual (mcp) home range sizes were not different for mature males (88.8 km2) and females (62.2 km2) (p = 0.28; table 2). ranges of immature males were larger in all seasons and annually (177.5 km2) than either mature males or females, except for females during summer. there were no differences in mean seasonal range sizes for mature males and females (p ≥0.22), with the exception of fall (23.0 and 59.4 km2 for females and males, respectively; p = 0.002) (table 2). seasonal home ranges for females ranged from 23.0 km2 during fall and early winter to 34.8 km2 in summer, with no difference (p ≥0.32) in seasonal home range size. seasonal home range size for mature males ranged from 17.5 km2 in late winter to 59.4 km2 during fall, with fall home ranges larger (p ≤0.01) than all other seasons. mean annual 95% ud sizes were not different between females (26.7 km2) and mature males (28.8 km2) (p ≥0.54; table 3). seasonal ud size for females did not differ among seasons (p ≥0.07; table 3). seasonal ud size for mature males ranged from 8.5 km2 in late winter to 19.6 km2 during fall, with fall larger (p ≤0.01) than summer and early and late winter; additionally, spring and summer uds were larger (p ≤ 0.01) than late winter. mature males had larger uds than central females in fall (p ≤ 0.01). seasonal uds were between 40–51% and 33–63% of seasonal mcp home ranges for females and mature males, respectively. location and composition of home ranges and utilization distributions mcp home ranges consisted of 84% (se = 0.02) forested cover types and 12% wetlands (se = 0.02), and uds were 88% (se = 0.01) forested with 9% (se = 0.01) wetlands. conservation land (state forests, table 2. seasonal and annual mean 100% minimum convex polygon home ranges (km2) for females, mature males (estimated >3 yr old), and immature male moose in massachusetts, 2006–2011. central females mature males immature males season n mean se range n mean se range n mean se range spring 5 26.9 4.2 14.1–39.0 9 28.0 3.2 14–39.0 5 61.4 25.5 15.8–158.1 summer 5 34.8 7.4 18.2–61.4 8 21.9 4.5 6.2–39.5 4 32.5 6.8 16.2–48.5 fall 5 23.0 3.7 12.8–28.8 8 59.4 15.1 31.8–161.3 5 222.6 110.0 6.6–546.8 early winter 4 23.0 2.9 14.9–29.1 10 29.6 5.4 14.3–72.9 5 50.8 11.9 14.6–83.1 late winter 5 25.8 3.8 14.3–38.0 11 17.5 2.7 5.1–31.8 5 33.2 12.8 9.3–80.4 annual 5 62.2 7.7 41.6–78.4 9 88.8 16.8 49.3–199.4 4 177.5 96.0 33.5–458.9 alces vol. 49, 2013 wattles and destefano – home range and movements 71 wildlife management areas, other protected land, and conservation easements) made up more of mcp home ranges (60%, se = 0.05, p ≤0.001) and uds (66%, se = 0.07, p ≤0.001) than was available as a whole in the central uplands (43%), and more in the mcp home ranges (59%, se = 0.1, p ≤0.004) and uds (76%, se = 0.1, p ≤0.001) than was available in the berkshire hills of western massachusetts (32%). additionally, conservation land made up a greater percentage of uds than either the overall mcp home ranges, or the area outside the ud but within the mcps (the unused portion of the mcp home range) (p ≤0.01); however, there was no difference in the amount of conservation land in mcp home ranges compared to the unused portion of the mcp (p = 0.16). all paved road types were at lower density within home ranges and uds compared to both the valley bottoms and uplands overall (p ≤0.001; table 4). additionally, all classes of paved roads (state highways, major local arteries, and local paved roads) were at lower densities within uds than either the overall mcp home ranges, or the unused portion of the mcp home range (p ≤0.04; fig. 2). state highways and local paved roads were also at greater densities in the unused portion of the mcp than in the overall mcp (p ≤0.008). seasonal movement patterns daily movement rates for female moose in central massachusetts were consistently ∼1,000–1,500 m/day in late winter (fig. 2). table 3. seasonal and annual mean 95% fixed kernel utilization distribution (km2) for females, mature males (estimated >3 yr old), and immature male moose in massachusetts, 2006–2011 (smoothing factor (h) = 200 m). females mature males immature males season n mean se range n mean se range n mean se range spring 5 10.8 0.8 8.2–12.8 9 15.5 1.4 10.4–22.0 4 19.6 3.7 13.5–28.8 summer 5 15.9 2.8 8.7–24.4 8 13.9 2.3 5.2–22.5 4 15.4 0.6 13.9–16.4 fall 4 11.4 0.8 10.0–13.6 7 19.6 2.4 10.4–30.6 4 22.2 8.1 6.5–44.8 early winter 5 11.4 1.5 8.4–15.8 9 11.5 0.8 6.3–14.5 4 19.5 1.2 16.5–20.2 late winter 5 13.1 0.7 11.6–15.7 10 8.5 1.3 4.1–15.1 4 13.5 4.4 7.4–26.2 annual 5 26.7 2.0 19.9–32.1 7 28.8 2.4 22.6–41.2 4 37.7 6.8 20.2–51.0 table 4. mean densities (km/km2) (se) of paved and unpaved roads in the valley bottoms, uplands, within maximum convex polygon (mcp) but outside utilization distributions (ud), mcp home ranges, and ud for moose in massachusetts, 2006–2011. valley bottoms uplands mcp outside ud mcp ud interstate highways 0.08 0.01 0.00 (0.00) 0.00 (0.00) 0.00 (0.00) major state highways 0.03 0.00 0.00 (0.00) 0.00 (0.00) 0.00 (0.00) state highways 0.33 0.18 0.13 (0.03) 0.11 (0.03) 0.06 (0.01) major local arteries 0.31 0.09 0.05 (0.02) 0.03 (0.01) 0.01 (0.01) local paved roads 1.48 0.50 0.40 (0.05) 0.30 (0.08) 0.14 (0.09) local unpaved/improved forest roads 0.39 0.48 0.54 (0.06) 0.49 (0.13) 0.44 (0.03) forest roads 0.73 0.28 0.33 (0.04) 0.35 (0.09) 0.38 (0.07) 72 home range and movements – wattles and destefano alces vol. 49, 2013 in spring, daily movement nearly doubled to ∼3,000 m/day prior to calving. there was a sharp decline to 500 m/day the second week of may that corresponded with the observed 8–13 may calving period. mean daily movement rates remained low for may and most of june, before peaking at ∼3,000 m/day in early july and remaining high for the remainder of the summer. movement rate declined in september to about 1,500 m/day and remained fairly consistent for the rest of the year. spring and summer seasonal movement rates for females were greater than all other seasons and calving season movement rates were lower than all other seasons (p ≤0.05; table 5). daily movement rates were lowest (1,000 m/day) for mature males from the fig. 2. road density in annual fixed kernel utilization distribution (dark gray) and minimum convex polygon home range (light gray) for a representative moose in massachusetts. heavy lines are major local roads and state highways, thin solid lines are local paved roads, and dashed lines are forest roads with limited access. alces vol. 49, 2013 wattles and destefano – home range and movements 73 beginning of february until the end of march (table 5). movements increased in early april and peaked at ∼2,500 m/day in late may and early june, before declining as summer progressed. daily movements increased to 3,000 m/day during the second week of september, indicating the start of the rut. movements increased further to a peak of nearly 8,000 m/day the last week of september and remained high through the first week of october, then declined sharply. movement rates remained relatively high at 2,000–2,500 m/day until the beginning of december when they declined to winter levels of 1,000–1,500 m/day. fall seasonal movement rates were greater than in all other seasons for mature males (p ≤0.05; table 5); additionally, spring and summer rates were greater than in late winter, and spring was greater than early winter. male daily movement rates were greater (p ≤0.05) than females during fall and lower during summer. discussion home range as a measure of resource use spatial requirements as measured by home range (second order use; johnson 1980) and uds (i.e., measuring use patterns within the home range; third order use) can provide important information about productivity of available habitat, distribution of resources and limiting factors, and how a species uses resources. this information is critical for conservation planning and habitat protection and connectivity at local and regional scales, and is particularly relevant for large mobile mammals in highly developed landscapes with fragmented patches of protected lands. harris et al. (1990) recommended using at least 2 home range estimators for all animal location data sets, including minimum convex polygon (mohr 1947) because of its prevalent use and comparability among studies. a mcp home range measures the area used by an individual to fulfill its annual or seasonal needs, but it does not describe how the area is used. alternatively, uds created by fixed kernels (worton 1989) describe the pattern and intensity of use within the mcp home range. by examining both, we can quantify areas of actual and relative intensity of use, identify important seasonal habitat patches, and delineate the area of landscape required to provide those patches comparison of uds to mcps shows that moose in southern new england used the table 5. seasonal daily movement rates (m/day) for female and mature male moose in massachusetts. mean seasonal daily movement rates and (se) in light gray, p-values for seasonal comparison between males and females in dark gray, p-values for comparisons among seasons for females above the diagonal and for males below the diagonal. female spring summer fall early winter late winter calving mean 2391 (141.0) 2464 (216.6) 1837 (81.5) 1505 (158.0) 1492 (107.9) 874 (70.6) sp 0.719 0.012 <0.001 <0.001 <0.001 spring 2019 (161.3) sp 0.22 sm 0.006 <0.001 <0.001 <0.001 summer 1731 (120.5) 0.168 sm 0.017 fl 0.112 0.097 <0.001 m a tu re m a le fall 3542 (385.2) <0.001 <0.001 fl <0.001 ew 0.951 0.008 early winter 1514 (107.0) 0.017 0.291 <0.001 ew 0.967 lw 0.009 late winter 1103 (79.8) <0.001 0.004 <0.001 0.051 lw 0.157 74 home range and movements – wattles and destefano alces vol. 49, 2013 landscape in a patchy manner; uds were typically only half the size of mcps, meaning that at any time there was a 95% probability of locating a moose within <50% of the mcp home range. additionally, uds fragmented into multiple polygons, indicating that resources were patchily distributed. maintaining connectivity of used patches within the larger landscape (mcp and larger) is essential for moose and other wide ranging species. rettie and messier (2000) argued that selection at the scale of the home range reflects attempts to reduce the effects of limiting factors. the uds measured here were located almost exclusively on the uplands of the central and western parts of the state, with limited use of valley bottoms. when valley bottoms were included in an mcp home range, they were mostly unused portions that were traversed in movements between ridge tops. overall, uds had greater amounts of forested habitat and conservation land and lower road densities than the landscape as a whole, or than the mcp home ranges. by definition moose spent 95% of their time in these less developed areas and appeared to select for more heavily forested areas away from human development. moose often crossed roads of all types in massachusetts, but seemed to show less avoidance of local residential roads with lower traffic volumes and speed limits than major highways, state highways, and major local arteries. in many instances major roads formed boundaries at the edge of an individual's home range; in other cases home ranges were bisected by highways and main roads. use of higher elevations could also be an attempt to limit thermal stress by taking advantage of reduced ambient temperatures and increased exposure to convective cooling from wind. human development and associated vehicle traffic and high temperatures that result in thermal stress may be limiting factors for moose in massachusetts. seasonal home ranges in central massachusetts, female mcp home ranges were largest during summer when energy demands were greatest because of lactation and seasonal restoration of body condition. mature male home ranges were largest during fall when they search for and attend mates during the breeding season, and smallest during late winter and summer when movements were presumably restricted by the combined effects of lower metabolism, snow conditions, and thermoregulatory constraints. despite the large number of studies on home range size (hundertmark 1997), comparisons to our results must be made with caution. most studies have used traditional vhf telemetry and home ranges were calculated with a small number of locations (e.g., <30), particularly in winter (e.g., <10), which can underestimate home range size (kernohan et al. 2001, börger et al. 2006); further, few vhf locations are collected at night when moose are often active. kernohan et al. (2001) suggested a minimum number of 30 locations, but at least 50 to calculate an accurate home range. additionally, differences in methods and the length, timing, and number of seasons used can make comparisons difficult (kernohan et al. 2001, börger et al. 2006). even with these limitations, our results fall within the range presented by hundertmark (1997) for home range sizes across north america (fig. 3). overall, home range size decreased with decreasing latitude and summer and winter home ranges in massachusetts would be expected at the low end of the scale. in the northeastern united states our results are similar to those of leptich and gilbert's (1989) in maine with >50 locations for 11 of 13 collared moose and an estimated alces vol. 49, 2013 wattles and destefano – home range and movements 75 summer mcp home range of 25 km2 for females. thompson et al. (1995) reported median summer home ranges of 32 km2 for females and 28 km2 for males in maine; their sample sizes in other seasons were too low for comparison. winter ranges were typified by concentrated use of small areas with short movements to other areas of intensive use in minnesota (van ballenberghe and peek 1971) and maine (thompson et al. 1995), a pattern similar to our observations. in northern new hampshire, scarpitti et al. (2005) observed smaller seasonal home ranges for females than our study (≤17 km2 for all seasons), with an earlier study in northern new hampshire (miller and litvaitis 1992) reporting much larger annual home ranges for females (153 km2) with the largest seasonal home ranges during fall (82 km2). garner and porter (1990) reported 36 km2 for summer and 8 km2 for winter home ranges of males in the adirondack mountains of new york. our seasonal results are the opposite of lenarz et al. (2011) who reported smaller home ranges during summer (16 km2) than in winter (33 km2) in minnesota. movements seasonal activity and movement patterns reflect changes in metabolic rate, ruminating time, and activity associated with the annual cycle of vegetation growth in temperate forests (risenhoover 1986, cederlund 1989). increased movement rates in spring corresponded with the start of the growing season and increased abundance and quality of browse. high movement rates in summer have been shown to reflect increased activity associated with more foraging bouts, lower ruminating times, and an attempt by moose to maximize foraging during the growing season (belovsky 1981, cederlund 1989, fig. 3. mean size of winter and summer home ranges in square kilometers for moose in north america relative to latitude (as reported by hundertmark 1997). data for female and male moose added as open symbols. 76 home range and movements – wattles and destefano alces vol. 49, 2013 van ballenberghe and miquelle 1990). we speculate that the periodically reduced rates in movements we observed during spring, summer, and fall were the result of thermoregulatory behavior during periods of high temperatures. the reduced movements during winter were typical of moose throughout their range (phillips et al. 1973, dussault et al. 2005, schwartz and renecker 2007). schwartz and renecker (2007) suggest that the lower winter metabolic rate of moose is an adaptation to counteract reduced forage abundance and quality and the related increased time required to digest a highly fibrous diet, resulting in fewer feeding bouts and lower activity level. movements were further reduced during periods of deep snow; however, snow depth and condition vary annually and across the state with the highest likelihood of deep snow at higher elevations in western massachusetts. when confined by deep snow, moose concentrated their habitat use into as little as 0.5 km2 for up to 3.5 months. the variability in the timing, depth, and condition of snowfall strongly influenced the variability of home range size and movements in early and late winter, as moose moved widely between suitable winter habitats until confined by snow. in addition to the influence of seasonal patterns on movements, changes in daily movement rates were greatest at times of the year corresponding to the annual reproductive cycle, i.e., calving for females and the rut for males. a final important consideration for understanding movements of moose in southern new england is the lack of their major predator, wolves (canis lupus), and the absence of moose hunting. predators and hunters can play important roles in the distribution and movements of their ungulate prey. black bears (ursus americanus) and coyotes (canis latrans) may prey on some moose calves, but in general the influence of predators or hunters on moose movements and distribution is absent in massachusetts. management implications existing distribution of vegetative communities, landscape configurations, and levels of development have allowed moose to re-colonize and establish a low density population throughout central and western massachusetts and into connecticut after 200–300 years of absence. however, southern new england is comprised of some of the most densely populated and highly developed states in the nation, and despite very active and successful conservation agencies and organizations, the trend will continue to move in the direction of more development and increased fragmentation. we have documented key elements of habitat use and movement distances and patterns by this newly re-established moose population. this information can be used to further enhance existing high priority conservation areas and identify new areas for protection and landscape connectivity. massachusetts has many well established biodiversity conservation initiatives (e.g., wildlife biomap and living waters) and planning strategies should recognize and incorporate a suitable scale to accommodate moose. if this large-scale challenge can be met, biodiversity conservation will benefit because moose use a diversity of terrestrial and wetland vegetative types (composition, size, and structure) that provide habitat for a wide array of species. acknowledgments the massachusetts division of fisheries and wildlife through the federal aid in wildlife restoration program (w-35-r) provided funding and support for this research. we appreciate the long-term involvement and support of many people, particularly r. deblinger and t. o'shea. the department of conservation and recreation, u.s. alces vol. 49, 2013 wattles and destefano – home range and movements 77 geological survey, the university of massachusetts, and safari club international provided additional funding and logistical support. capture of moose would not have been possible without the assistance of k. berger and other field assistants and volunteers. references bates, d., m. maechler, and b. bolker. 2012. lme4: linear mixed effects models using s4 classes. <http://lme4.r-forge.rproject.org> (accessed december 2012). belovsky, g. e. 1981. optimal activity times and habitat choice of moose. oecologia 48: 22–30. beyers, h. 2006. hawth's analysis tools for arcgis. version 3.27. <http://www.spatialecology.com/htools> (accessed december 2012). birkett, p. j., a. t. vanak, v. m. r. muggeo, s. m. ferreira, and r. slotow. 2012. animal perception of seasonal thresholds: changes in elephant movement in relation to rainfall patterns. plos one 7: 1–8. boose, e. 2001. fisher meteorological station (since 2001). harvard forest data archive: hf001. petersham, massachusetts, usa. börger, l., n. franconi, f. ferretti, f. meschi, g. de michele, a. gantz, and t. coulson. 2006. an integrated approach to indentify spatiotemporal and individual-level determinants of animal home range size. the american naturalist 168: 471–485. burt, w. h. 1943. territoriality and home range concepts as applied to mammals. journal of mammalogy 24: 346–352. cederlund, g. 1989. activity patterns in moose and roe deer in a north boreal forest. holarctic ecology 12: 39–54. coady, j. w. 1974. influence of snow on behavior of moose. naturaliste canadien 101: 417–436. degraaf, r. m., and m. yamasaki. 2001. new england wildlife: habitat, natural history, and distribution. university press of new england, hanover, new hampshire, usa. destefano, s., r. d. deblinger, and c. miller. 2005. suburban wildlife: lessons, challenges, and opportunities. urban ecosystems 8: 131–137. dussault, c., j. p. ouellet, r. coutois, j. huot, l. breton, and h. jolicoeur. 2005. linking moose habitat selection to limiting factors. ecography 28: 1–10. environmental systems research institute inc. (esri). 2008. arcgis 9.3. redlands, california, usa. fieberg, j. 2007. kernel density estimators of home range: smoothing and the autocorrelation red herring. ecology 88: 1059–1066. ———, and l. börger. 2012. could you please phrase “home range” as a question? journal of mammalogy 93: 890–902. forman, r. t. t., and r. d. deblinger. 2000. the ecological road effect zone of a massachusetts (u.s.a.) suburban highway. conservation biology 14: 36–46. franzmann, a. w., and c. c. schwartz. 2007. ecology and management of the north american moose. second edition. university press of colorado, boulder, colorado, usa. garner, d. l., and w. f. porter. 1990. movement and seasonal home ranges of bull moose in a pioneering adirondack population. alces 26: 80–85. gitzen, r. a., j. t. millspaugh, and b. j. kernohan. 2006. bandwidth selection for fixed-kernel analysis of animal utilization distributions. journal of wildlife management 70: 1334–1344. hall, b., g. motzkin, d. r. foster, m. syfert, and j. burk. 2002. three hundred years of forest and land-use in massachusetts, usa. journal of biogeography 29: 1319–1335. harris, s., w. j. cresswell, p. g. forde, w. j. trewhella, t. woollard, and s. wray. 1990. home-range analysis using 78 home range and movements – wattles and destefano alces vol. 49, 2013 http://lme4.r-forge.r-project.org http://lme4.r-forge.r-project.org http://www.spatialecology.com/htools http://www.spatialecology.com/htools radio-tracking data: a review of problems and techniques particularly as applied to the study of mammals. mammal review 20: 97–123. hemson, g., p. johnson, a. south, r. kenward, r. ripley, and d. macdonald. 2005. are kernels the mustard? data from global positioning system (gps) collars suggests problems for kernel home-range analyses with least-squares cross-validation. journal of animal ecology 74: 455–463. hundertmark, k. j. 1997. home range, dispersal, and migration. pages 303–335 in a.w. franzmann and c.c. schwartz, editors. ecology and management of the north american moose. university press of colorado, boulder, colorado, usa. johnson, d. h. 1980. the comparison of usage and availability measurements for evaluating resource preference. ecology 61: 65–71. kelsal, j. s., and e. s. telfer. 1974. biogeography of moose with particular reference to western north america. naturaliste canadien 101: 117–130. kernohan, b. j., r. a. gitzen, and j. j. millspaugh. 2001. analysis of animal space use and movements. pages 125–166 in j. j. millspaugh and j. m. marzluff, editors. radio tracking animal populations. academic press, san diego, california, usa. kertson, b. n., r. d. spencer, j. m. marzluff, j. hepinstall-cymerman, and c. e. grue. 2011. cougar space use and movements in the wildland-urban landscape of western washington. ecological applications 21: 2866–2881. kie, j. g., j. matthiopoulos, j. fieberg, r. a. powell, f. cagnacci, m. s. mitchell, jm. gaillard, and p. r. moorscroft. 2010. the home-range concept: are traditional estimators still relevant with modern telemetry technology. philosophical transactions of the royal society b 365: 2221–2231. kittredge, d. b., jr., a. o. finley, and d. r. foster. 2003. timber harvesting as ongoing disturbance in a landscape of diverse ownership. forest ecology and management 180: 425–442. lenarz, m. s., m. e. nelson, m. w. schrage, and a. j. edwards. 2009. temperature mediated moose survival in northeastern minnesota. journal of wildlife management 73: 503–510. ———, j. fieberg, m. w. schrage, and a. j. edwards. 2010. living on the edge: viability of moose in northeastern minnesota. journal of wildlife management 74: 1013–1023. ———, r. g. wright, m. w. schrage, and a. j. edwards. 2011. compositional analysis of moose habitat in northeastern minnesota. alces 47: 135–149. leptich, d. j., and j. r. gilbert. 1989. summer home range and habitat use by moose in northern maine. journal of wildlife management 53: 880–885. lykkja, o. n., e. j. solberg, i. herfindal, j. wright, c. m. rolandsen, and m. g. hanssen. 2009. the effects of human activity on summer habitat use by moose. alces 45: 109–124. massachusetts office of geographic information. 2005. <http://www.mass.gov/ anf/research-and-tech/it-serv-and-support/ application-serv/office-of-geographic-in formation-massgis/datalayers/eotroads. html> (accessed december 2012). mcdonald, r.i., g. motzkin, m. s. bank, d. b. kitteridge, j. burke, and d. l. foster. 2006. forest harvesting and landuse conversion over two decades in massachusetts. forest ecology and management 227: 31–41. miller, b. k., and j. a. litvatis. 1992. habitat segregation by moose in a boreal forest ecotone. acta theriologica 37: 41–50. mills, k. j., b. r. patterson, and d. l. murray. 2006. effect of variable sampling frequencies on gps transmitter efficiency and estimated wolf home range alces vol. 49, 2013 wattles and destefano – home range and movements 79 http://www.mass.gov/anf/research-and-tech/it-serv-and-support/application-serv/office-of-geographic-information-massgis/datalayers/eotroads.html http://www.mass.gov/anf/research-and-tech/it-serv-and-support/application-serv/office-of-geographic-information-massgis/datalayers/eotroads.html http://www.mass.gov/anf/research-and-tech/it-serv-and-support/application-serv/office-of-geographic-information-massgis/datalayers/eotroads.html http://www.mass.gov/anf/research-and-tech/it-serv-and-support/application-serv/office-of-geographic-information-massgis/datalayers/eotroads.html http://www.mass.gov/anf/research-and-tech/it-serv-and-support/application-serv/office-of-geographic-information-massgis/datalayers/eotroads.html size and movement distance. wildlife society bulletin 34: 1463–1469. mitchell, m. s., and r. a. powell. 2008. estimated home ranges can misrepresent habitat relationships on patchy landscapes. ecological modeling 216: 409–414. mohr, c. o. 1947. table of equivalent populations of north american small mammals. american midland naturalist 37: 223–249. murray, d. l., e. w. cox, w. b. ballard, h. a. whitlaw, m. s. lenarz, t. w. custer, t. barnett, and t. k. fuller. 2006. pathogens, nutritional deficiency, and climate influences on a declining moose population. wildlife monograph 166: 1–30. peek, j. m., and k. i. morris. 1998. status of moose in the contiguous united states. alces 34: 423–434. phillips, r. l., w. e. berg, and d. b. siniff. 1973. moose movement patterns and range use in northwestern minnesota. journal of wildlife management 37: 266–278. powell, r. a. 2000. animal home ranges and territories. pages 65-110 in l. boitani and t. k. fuller, editors. research techniques in animal ecology. columbia university press, new york, new york, usa. ———, and m. s. mitchell. 2012. what is a home range? journal of mammalogy 93: 948–059. r development core team. 2005. r: a language and environment for statistical computing. r foundation for statistical computing, vienna, austria. <http:// www.r-project.org> (accessed december 2012). renecker, l. a., and r. j. hudson. 1986. seasonal energy expenditures and thermoregulatory responses of moose. canadian journal of zoology 64: 322–327. rettie, w. j., and f. messier. 2000. hierarchical habitat selection by woodland caribou: its relationship to limiting factors. ecography 23: 466–478. risenhoover, k. l. 1986. winter activity patterns of moose in interior alaska. journal of wildlife management 50: 727–734. rodgers, a. r., a. p. carr, h. l. beyer, l. smith, and j. g. kie. 2007. hrt: home range tools for arcgis. ontario ministry of natural resources, centre for northern forest ecosystem research, thunder bay, ontario, canada. scarpitti, d., c. habeck, a. r. musante, and p. j. pekins. 2005. integrating habitat use and population dynamics of moose in northern new hampshire. alces 41: 25–35. schwartz, c. c., and l. a. renecker. 2007. nutrition and energetics. pages 441–478 in a.w. franzmann and c.c. schwartz, editors. ecology and management of the north american moose. university press of colorado, boulder, colorado, usa. spencer, w. d. 2012. home ranges and the value of spatial information. journal of mammalogy 93: 929–947. thompson, m. e., j. r. gilbert, g. j. matula, and k. i. morris. 1995. seasonal habitat use by moose on managed forest lands in northern maine. alces 31: 233–245. tremblay, a., and j. ransijn. 2012. lmerconveniencefunctions: a suite of functions to back-fit fixed and forward-fit random effects, as well as other miscellaneous functions. <http://cran.r-project. org/web/packages/lmerconveniencefunctions/index.html> (accessed decem‐ ber 2012). u. s. census bureau. 2010a. census 2010. resident population data: population density <http://www.census.gov/2010 census/data/apportionment-dens-text.php> (accessed february 2013). ———. 2010b. 2010 census: massachusetts profile map. <http://www2.census.gov/ geo/maps/dc10_thematic/2010_profile/ 2010_profile_map_massachusetts.pdf> (accessed february 2013). 80 home range and movements – wattles and destefano alces vol. 49, 2013 http://www.r-project.org http://www.r-project.org http://cran.r-project.org/web/packages/lmerconveniencefunctions/index.html http://cran.r-project.org/web/packages/lmerconveniencefunctions/index.html http://cran.r-project.org/web/packages/lmerconveniencefunctions/index.html http://www.census.gov/2010census/data/apportionment-dens-text.php http://www.census.gov/2010census/data/apportionment-dens-text.php http://www2.census.gov/geo/maps/dc10_thematic/2010_profile/2010_profile_map_massachusetts.pdf http://www2.census.gov/geo/maps/dc10_thematic/2010_profile/2010_profile_map_massachusetts.pdf http://www2.census.gov/geo/maps/dc10_thematic/2010_profile/2010_profile_map_massachusetts.pdf van ballenberghe, v., and d. g. miquelle. 1990. activity of moose during spring and summer in interior alaska. journal of wildlife management 54: 391–396. ———, and j. m. peek. 1971. radio telemetry studies of moose in northeastern minnesota. journal of wildlife management 35: 63–71. vecellio, g. m., r. d. deblinger, and j. e. cardoza. 1993. status and management of moose in massachusetts. alces 29: 1–7. wattles, d. w., and s. destefano. 2011. status and management of moose in the northeastern united states. alces 47: 53–68. westveldt, m. r., r. i. ashman, h. i. baldwin, r. p. holdsworth, r. s. johnson, j. h. lambert, h. j. lutz, l. swain, and m. standish. 1956. natural forest vegetation zones of new england. journal of forestry 54: 332–338. worton, b. j. 1989. kernel methods for estimating the utilization distribution in home-range studies. ecology 70: 164–168. alces vol. 49, 2013 wattles and destefano – home range and movements 81 space use and movements of moose in massachusetts: implications for conservation of large mammals in a fragmented environment methods study area study animals and gps telemetry seasons home ranges and space use movements statistics results capture and deployment of gps collars home ranges and space use location and composition of home ranges and utilization distributions seasonal movement patterns discussion home range as a measure of resource use seasonal home ranges movements management implications acknowledgments references a novel method of performing moose browse surveys rachel l.w. portinga1,2,3 and ron a. moen1,2 1biology department, university of minnesota duluth, 1049 university drive, duluth, mn 55812; 2university of minnesota duluth natural resources research institute, 5013 miller trunk hwy, duluth, mn, 55811 abstract: we measured browse availability and use along foraging paths of gps radio-collared moose (alces alces) in northeastern minnesota to estimate diet composition and browse species preference. on foraging paths during summer and winter we counted twigs via traditional methods for comparison with a novel method that attempted to better simulate moose foraging behavior. twigs were collected and used to develop diameter at point of browsing – biomass regressions for each browse species. these regressions, different under open and closed canopy, were used to estimate biomass consumption on foraging paths and to compare 4 approaches. the average diets were similar to previously measured regional diets, and importantly, our data identified variance among individual seasonal diets. our field method allowed us to better quantify and compare diet composition and browse selection of individual free-ranging moose directly on foraging paths. alces vol. 51: 107–122 (2015) key words: bite size, browse availability method, browse selection, diameter-biomass regressions, diet composition, minnesota large herbivores like moose (alces alces) view their food resources at the landscape, patch, and feeding station levels (senft et al. 1987). at the landscape level moose choose which patches to visit based on the spatial distribution of browse density and forage availability within each patch. at the patch level, moose must choose where to forage based on the available browse species, and tree and shrub heights at different feeding stations. younger patches can provide large quantities of high quality browse while older patches that have grown out of reach provide less browse (schwartz 1992). within a feeding station, bite size is based on the tradeoff between cropping and processing (spalinger and hobbs 1992). moose need to consume about 130 g dry mass/kg body weight0.75 daily in summer and about 40 g dry mass/kg body weight0.75 daily in winter (renecker and hudson 1985). using an average bite size of 1.02 g/bite (renecker and hudson 1986), this equates to at least 13,000 bites in summer and 4,000 bites in winter for a 454 kg (1000 lb) moose. winter consumption may be up to 50% higher depending on browse availability and species composition (hjeljord et al. 1994). this large demand for forage forces moose to move between patches and feeding stations in order to consume enough biomass. browse availability and bite size have been measured by following moose or moose tracks in snow and counting the number of available twigs per species, the number of bites per species, and measuring diameter at point of browsing, dry mass, and twig length (risenhoover 1987, shipley et al. 1998). locations of moose were found via radio telemetry (risenhoover 1987, hjeljord et al. 1990) or finding a track crossing a road (shipley et al. 1998). these methods were largely opportunistic and data collection was either clumped (location 3present address: hibbing community college, 1515 25th st e, hibbing, mn 55746 107 every hour for 2 days) or spread temporally (1–2 tracks weekly). another typical method is to measure browse availability in plots along randomly placed straight transects instead of following moose foraging paths. this provides an estimate of absolute browse density in a patch, rather than an estimate of browse availability encountered by moose. we measured intensively used feeding patches with 3 different protocols and a randomly placed straight transect in northeastern minnesota. our new protocol (the large feeding station method) attempted to simulate how a moose browses, which we contrasted with measurements along the foraging path and with absolute browse density. study area this study was conducted in northeastern minnesota where moose were previously collared for a vhf telemetry study (fig. 1) (lenarz et al. 2011). these forests transition between the canadian boreal and northern hardwood forests and experience a continental climate with short warm summers and severe winters (heinselman 1996). most was part of the superior national forest with the remaining either state, county, tribal, or industrial forest land fig. 1. the study area was within the superior national forest in northeastern minnesota. each black dot represents one measured foraging path in winter and a dark gray dot represents a summer foraging path. 108 novel browse surveys – portinga and moen alces vol. 51, 2015 (lenarz et al. 2010, moen et al. 2011). more specific details are provided in the minnesota moose research and management plan (mndnr 2011). methods regressions and estimating bite mass summer leaves were collected between july and september 2012, and winter twigs between january and april 2012 and 2013; twigs from both years were combined in the regression analyses. we clipped (standard garden clippers) browsed (∼3 cm below the browse point) and unbrowsed twigs of all browse species (table 1). samples were bagged and labeled with the location, date, and species. all browsed and unbrowsed twigs and leaves were stored at 2–3 °c prior to measurements. these twigs were used to develop diameter-biomass regressions for each season (telfer 1969). in summer we collected stripped twigs of each species which we clipped directly above the first unbrowsed petiole. a winter bite was equal to the twig biomass and a summer bite the leaf biomass from one twig, both with current annual growth >5 cm. on browsed twigs we measured (nearest 0.01 mm) the diameter at point of browsing and on unbrowsed twigs the simulated diameter at point of browsing. in summer, the simulated point of browsing was the diameter underneath the last stripped petiole. the wet weight of winter twigs and stripped summer leaves was weighed to the nearest 0.01 g. after weighing the wet mass of leaves, they were placed in the same bag with the corresponding twig. all unbrowsed table 1. the common and scientific names for each potential browse species in northeastern minnesota and seasons in which each species is consumed. “rare” species make up <1% of the diet at large feeding station paths. “not browsed” species were not consumed along the foraging paths. common name scientific name winter summer balsam fir abies balsamea common not browsed red maple acer rubrum common common mountain maple acer spicatum common common alder alnus rugosa rare rare juneberry amelanchier spp. common common paper birch betula papyrifera common common bog birch betula pumila not browsed rare red-osier dogwood cornus stolonifera common rare hazel corylus cornuta common rare black ash fraxinus niger not browsed rare white pine pinus strobus rare rare balsam poplar populus balsamifera rare rare quaking aspen populus tremuloides common common pin cherry prunus pennsylvanicus common common choke cherry prunus virginianus common common oak quercus spp. not browsed rare willow salix spp. common common elderberry sambucus pubens not browsed rare mountain ash sorbus decora rare common alces vol. 51, 2015 portinga and moen – novel browse surveys 109 twigs in both seasons were stored in labeled bags. all unbrowsed summer and winter twigs were dried at 60 °c for 48 h in a drying oven. dried twigs in winter and dried leaves in summer were stored at room temperature before being weighed to the nearest 0.01 g. most winter twigs (74%) and summer leaves (90%) were measured within 5 days of removal from the drying oven; the remainder was measured 6–9 days later. gps collars we captured adult moose in february and early march 2011 by darting them from helicopters. gps radio-collars (sirtrack ltd. and lotek wireless) fitted to each moose were programmed to transmit a location every 20 min. animal capture and handling protocols met the guidelines recommended by the american society of mammalogists (sikes et al. 2011) and were approved by university of minnesota and national park service animal care and use committees (#0912a75532). measuring browse availability summer browse availability was measured between 25 july and 14 september 2012, and winter browse availability between 3 january and 22 march 2013. browse availability was measured at the patch scale which we identified from the gps locations – patches had a concentrated number of moose locations. we used a handheld garmin gps to reach our pre-identified patches and then searched for a feeding station to identify a foraging path. a feeding station was defined as a plant or clump of plants with browsed twigs that were accessible when the forefeet of a moose are stationary (goddard 1968, novellie 1978, senft et al. 1987). a foraging path was defined as a trail of feeding stations within a patch. summer foraging paths were measured 1 to 15 days after the moose departed, and winter foraging paths were measured 3 to 17 days after departure. patches were considered accessible if they were on public land and we could access them by walking <2 km on a trail and/or <550 m from a trail. we measured winter patches containing 29 foraging paths from 8 moose (6f, 2m), and summer patches containing 31 foraging paths from 7 moose (5f, 2m). we defined a large feeding station as a location that appeared to have ≥10 bites. at all sites we measured browse under 4 different protocols to produce 4 foraging path types: 1) large feeding stations along the foraging path, 2) random plots along the foraging path, 3) random feeding stations along the foraging path, and 4) plots along a straight transect through the area containing the foraging path. each path type consisted of 10 measurement plots. large feeding station plots — the first large feeding station encountered was the first plot of the site and marked as a waypoint on the handheld gps. the plot or feeding station to be measured was a half circle with radius of 99.1 cm (39 in), with the center of the back side (straight line diameter) held at the approximate place where the moose stood. tracks in winter, other sign in either season, or placement of bites relative to open space were also used to determine where the moose stood and the direction it faced. at each large feeding station we counted the unbrowsed and browsed twigs of each browse species between 0.5 and 3 m above the ground (table 1; shipley et al. 1998). each cut-off twig was considered a bite. although an occasional large feeding station had <10 bites, we included it as a large feeding station because the observer estimated it had at least 10 bites. this only occurred at 10 of 290 (3%) large feeding stations in winter and 36 of 297 (12%) in summer. 110 novel browse surveys – portinga and moen alces vol. 51, 2015 we established the foraging path type from the first large feeding station by following tracks and browsing sign to the next large feeding station, marked it as the second waypoint on the gps, and repeated the measurements (fig. 2). plots could not overlap and this process continued until 10 large feeding stations had been measured on the foraging path. random plots on the foraging path — we created the random plot path type by stopping along the foraging path and repeating our browse measurements in random plots. a list of random distances between 5 and 14 m was generated using microsoft excel, and in the field we established the random plots using these distances in the gps “find” feature (fig. 2). random feeding stations — if a random plot had been browsed (evident bites), then that random plot was also defined as a random feeding station. if no browsed bites were in the random plot, we followed the foraging path to the next browsed twig (even if only one bite) and this became the location of the next random feeding station (fig. 2), eventually creating the random feeding path type. straight transect plots — after completing the large feeding station, random plot, and random feeding station measurements, we established a straight line transect that returned to the first plot. along this transect we stopped at random distances between 5 and 14 m until 10 plots were measured. if we reached the first large feeding station plot before completing 10 plots, we lengthened the transect. if, however, the cover type changed past the first plot and <10 plots were measured, we established a new transect in a random direction within the same cover type; 10 of 29 straight transects were angled in winter and 15 of 31 in summer. 3 1 2 8 9 54 7 6 = large feeding station (≥ 10 bites) = random plot = random feeding station (≥ 1 bite) fig. 2. a diagram of how we measured a foraging path. plot 1 is a large feeding station plot with ≥10 bites. plot 2 is a random plot. because plot 2 did not have any bites taken we stop at the next bite which becomes plot 3, a random feeding station plot. plot 4 is the second large feeding station plot. plot 5 is the second random plot with 1–9 bites, so it is also the second random feeding station plot. plot 6 is the third large feeding station plot. plot 7 is the third random plot that had ≥10 bites, so it is also the third random feeding station plot and the fourth large feeding station plot. plot 8 is the fourth random plot. plot 9 is the fourth random feeding station that had ≥10 bites, so it is also the fifth large feeding station plot. we continued until there were 10 plots of each type. alces vol. 51, 2015 portinga and moen – novel browse surveys 111 some cover types had little available browse making the foraging path difficult to follow in summer when 10 of 30 foraging paths had <10 plots in all path types. if no bites were found within 20 m of the previous feeding station when moving forward, we assumed the moose stopped foraging. effectively this meant that there were <10 large feeding stations, random feeding stations, and/or random plots in that foraging path. snow tracking in winter allowed us to more easily identify the foraging path; thus, 10 plots in all path types were measured in 28 of 30 foraging paths. canopy cover was measured 3 times with a densiometer (every 8th plot) to produce an average value in each patch. twigs collected from sites with 0–50% canopy closure were considered grown in open canopy, and twigs from sites with 70–100% canopy closure were considered grown in closed canopy. twigs from sites with 51–69% canopy cover were not used in the regressions or bite size summary statistics. statistical analyses biomass-diameter at point of browsing regressions, anovas on browse density, fig. 3. the percent of random feeding stations measured in each size category (line) and the percent of bites consumed at all feeding stations of a given size category (bar) in winter and summer. the dashed line separates the small feeding stations (≤9 bites) from the large feeding stations. in winter, 57% of the random feeding stations were considered large but they accounted for 86% of the consumed bites. in summer, 49% of the random feeding stations were considered large but they accounted for 82% of the consumed bites. 112 novel browse surveys – portinga and moen alces vol. 51, 2015 kruskal-wallis comparisons of diet, pearson χ2 goodness of fit tests, and bonferroni z-tests were all performed in jmp 10.0. significance level was set at p = 0.05. regressions — simulated diameters at point of browsing and dry masses of twigs from the unbrowsed winter twigs were log10 transformed and used to make 2 separate diameter-biomass regressions for each of the main browse species. the first regression used twigs grown in open canopy (0–50% shaded) and the second twigs from closed canopy (70–100% shaded). similarly, 2 summer regressions were made using leaf dry mass of each browse species. the raw data are found in ward (2014) and only results are presented here. statistics on bite size diameter and bite mass were calculated for each species. a t-test was used to test for statistical differences between the average diameter at point of browsing in open and closed canopy in both seasons for each species. available browse density — browse density was estimated as twig counts and as biomass. to obtain the total number of available twigs per path, we added the number of available twigs and the number of browsed bites. we estimated the total biomass originally available (browsed or unbrowsed) along a foraging path by multiplying the number of twigs of a given species by the average biomass of one bite of that species. for foraging paths in 0–50% shade, we used the average biomass values from open canopy regressions. likewise, we used the average biomass values from closed canopy regressions for foraging paths in 51–100% shade. although the closed canopy regressions were developed with twigs grown in 70–100% shaded areas, we felt the foraging paths in 51–69% shade were better classified as closed canopy than open canopy. balsam fir was not included in summer browse density estimates because it is not typically part of the summer diet. available and consumed browse density along each of the 4 path types were estimated using twig counts and biomass in both seasons. the length of each path was calculated by measuring the length of a line passing through all of the plots of each path type. the area of the foraging path was considered twice this distance to represent the ability of moose to browse either side of the foraging path. to calculate browse density we divided the twig count (available or consumed) by the area of the foraging path. these same calculations were made using biomass and twig counts. the browse density on large feeding station paths was compared with those on the random feeding, random, and straight transect paths using an anova of the log transformed data. diet composition — diet composition was calculated for each moose on the 4 path types in both seasons. we made a weighted average of those diet compositions to estimate diet composition for all moose on each path type in winter and summer. species were considered rare when they made up <1% of the average diet (shipley et al. 1998) at large feeding station paths. the percentage of the diet consisting of rare species is reported in the tables (but not text) to illustrate how a few individual moose consumed many bites of rare species. each individual diet had at least one browse species not identified on the foraging paths. because these data were not normally distributed and no transformation could correct this skewedness, we used a kruskalwallis test to test for significant differences between diet composition on the 4 path types. a kruskal-wallis test was also used to test for differences between each individual diet. browse species selection — we also determined the selection for each browse species from a combined average of all moose and for each individual using the alces vol. 51, 2015 portinga and moen – novel browse surveys 113 data from large feeding station paths. a pearson χ2 goodness of fit test and a bonferroni z-test were performed on the availability and use of all browse species for all moose combined and each individual moose (neu et al. 1974). a species was considered “positively selected”, “negatively selected”, or neither if there was a significantly larger, smaller, or equal proportion of browsed versus available twigs. results regressions all of the twig diameter – biomass regressions had slopes significantly different from zero. the slopes ranged from 0.58– 2.80 in winter and 0.45–2.07 in summer. in winter, 75% of the regressions had an r2 >0.60, and in summer 43% had an r2 >0.60. there was no consistent pattern between the open canopy or closed canopy regression slopes being larger or smaller (ward 2014). bite size across all species in winter, the mean diameter at point of browsing was 3.0 ± 0.02 mm in open canopy (range = 0.5–9.0 mm) and 3.1 ± 0.1 mm in closed canopy (range = 0.2–8.4 mm) (table 2). in summer, the mean across species was 2.3 ± 0.02 mm in open canopy (range = 0.02–11.1 mm) and 2.4 ± 0.04 mm in closed canopy (range = 0.2–6.1 mm) (table 3). using the regressions found in ward (2014), we calculated the average biomass consumed per bite for each browse species (tables 2 and 3). in winter, pin cherry had the largest bite size (2.3 ± 1.4 g) under closed canopy and the smallest bite size under open canopy (0.4 ± 0.1 g). mountain maple had the smallest bite size under closed canopy (0.4 ± 0.2 g). mountain ash had the largest (1.7 ± 1.4 g) and quaking aspen the smallest bite size (0.3 ± 0.2 g) under closed canopy in summer. bite density at feeding stations one purpose of establishing the random feeding station plots was to estimate the frequency of feeding stations of different sizes occurring along foraging paths. in winter 57% of random feeding station plots (n = 281) had ≥10 or more bites, and in summer 49% (n = 267). in both seasons at least 80% of twig consumption on the foraging path was from feeding stations with ≥10 bites, although moose occasionally consumed <10 bites at a station. browse density total available browse density was measured at 29 patches in winter and 30 patches in summer. it was significantly different among the 4 path types in both seasons using either method (winter twigs: f3, 112= 62.7, summer twigs: f3, 118 = 32.5, winter biomass: f3, 112 = 84.3, summer biomass: f3, 120 = 16.8, pall < 0.0001). likewise, density of consumed browse was also significantly different in winter and summer among the 4 path types (winter twigs: f3, 112 = 63.4, summer twigs: f3, 120 = 31.2, winter biomass: f3, 112 = 70.9, summer biomass: f3, 119 = 5.0, pall < 0.0025). as expected, both available and consumed browse densities were highest at large feeding station paths, followed by random feeding station, random plot, and straight transect paths (table 4). the average available browse density estimated by biomass at large feeding stations was 53% higher in summer (15.2 ± 1.7 g/m2) than winter (9.9 ± 1.0 g/m2). conversely, density estimated by twig counts was ∼2.5x larger in winter (15.2 ± 1.6 twigs/m2) than in summer (5.9 ± 0.6 twigs/m2). large feeding station paths had ∼60% more available twigs (727 ± 3) 114 novel browse surveys – portinga and moen alces vol. 51, 2015 in winter than in summer (460 ± 37), whereas the available biomass was ∼2.5x larger in summer (1166 ± 88 g) than winter (471 ± 26 g). the same seasonal differences existed for consumed twigs and biomass. the distance walked in winter to complete the large feeding station paths (27.6 ± 2.0 m, n = 29) was about half that in summer (50.5 ± 4.9 m, n = 31). the available and consumed browse density for each browse species was largest at large feeding station paths followed by random feeding station, random plot, and straight transect paths. the one exception (based on twig count) was that the highest browse density of hazel was found on the straight transect path in summer (when hazel is rarely consumed). table 2. summary statistics on browsed twigs in winter for all browse species. open canopy indicates twigs grown in locations shaded 0–50% and closed canopy indicates twigs grown in locations shaded 70–100%. p-values indicate t-test results between the diameter at point of browsing (dpb) of each species in open and closed canopy. we did not find enough individual twigs of juneberry, paper birch, pin cherry, or willow in closed canopy to calculate reliable averages for those categories. diameter at point of browsing (mm) species canopy average ± se minimum maximum average bite ± se (g) n p balsam fir** open 2.7 ± 0.1 0.9 6.5 1.6 ± 0.3 82 0.002 closed 2.2 ± 0.1 1.0 4.0 1.2 ± 0.2 50 red maple** open 3.5 ± 0.1 1.3 7.4 0.7 ± 0.3 125 0.009 closed 4.1 ± 0.1 2.7 6.9 1.4 ± 0.5 27 mountain maple* open 2.8 ± 0.3 1.5 4.6 0.6 ± 0.3 47 0.019 closed 2.4 ± 0.3 0.4 4.9 0.4 ± 0.2 56 juneberry open 2.4 ± 0.1 0.9 4.5 0.5 ± 0.1 161 0.583 closed na na na na 8 paper birch open 2.7 ± 0.1 0.6 4.8 0.8 ± 0.1 188 na closed na na na na 7 hazel open 2.7 ± 0.1 1.1 5.3 0.6 ± 0.1 301 0.104 closed 2.8 ± 0.1 1.1 4.5 0.6 ± 0.1 132 red-osier dogwood*** open 3.5 ± 0.1 1.5 6.1 1.1 ± 0.1 332 <0.0001 closed 4.3 ± 0.2 2.0 6.6 1.4 ± 0.4 40 quaking aspen open 3.5 ± 0.1 0.9 6.8 0.9 ± 0.1 209 0.155 closed 3.2 ± 0.1 1.0 5.7 0.7 ± 0.4 32 pin cherry open 2.4 ± 0.1 0.6 4.9 0.4 ± 0.1 216 na closed na na na na 6 choke cherry open 3.0 ± 0.3 1.5 4.8 0.7 ± 0.1 53 0.120 closed 2.6 ± 0.4 0.2 4.1 0.4 ± 0.1 20 willow open 3.1 ± 0.1 0.5 6.4 0.9 ± 0.1 501 na closed1 na na na na 0 mountain ash* open 4.3 ± 0.1 1.6 6.8 1.3 ± 0.3 43 0.045 closed 3.7 ± 0.1 1.2 8.4 0.7 ± 0.5 53 combined open 3.0 ± 0.02 0.5 9.0 na 2388 closed 3.1 ± 0.1 0.2 8.4 na 454 alces vol. 51, 2015 portinga and moen – novel browse surveys 115 consumption rate the pattern of consumption rate was similar to that of consumed browse density. the proportion of consumed twigs was highest on the large feeding station paths and declined progressively to the random feeding station, random plot, and straight transect paths. consumption was 45% in summer and 35% in winter on the large feeding station paths. overall, it was 23–45% on all paths except the straight transects where rates were 13% in winter and 9% in summer. diet composition season — at least 70% of all bites (all moose) consumed in winter along the 4 path types consisted of hazel, paper birch, willow, and quaking aspen. the remaining 30% consisted of balsam fir, juneberry, mountain maple, red maple, red-osier dogwood, pin table 3. summary statistics on browsed twigs of all species in summer. open canopy indicates twigs grown in locations shaded 0–50% and closed canopy indicates twigs grown in locations shaded 70–100%. pvalues indicate t-test results between the diameter at point of browsing (dpb) of each species in open and closed canopy. we did not find enough individual twigs of red maple in open canopy or pin cherry, willow, or mountain ash in closed canopy to calculate reliable averages for those categories. diameter at point of browsing (mm) species canopy mean ± se minimum maximum mean bite ± se (g) n p red maple open na na na na 14 0.349 closed 2.8 ± 0.2 1.3 6.0 1.4 ± 0.3 27 mountain maple*** open 2.3 ± 0.03 0.5 4.7 0.7 ± 0.1 675 <0.0001 closed 3.0 ± 0.1 0.5 4.9 1.0 ± 0.1 264 juneberry open 1.6 ± 0.04 0.1 3.2 0.5 ± 0.04 149 0.145 closed 2.1 ± 0.3 0.2 4.2 1.0 ± 0.4 20 paper birch** open 2.3 ± 0.1 0.02 5.1 0.8 ± 0.1 316 0.003 closed 2.0 ± 0.1 0.6 3.8 0.5 ± 0.1 84 hazel open 1.6 ± 0.1 0.5 3.5 0.7 ± 0.04 105 0.739 closed 1.6 ± 0.1 0.6 2.5 0.6 ± 0.1 48 red-osier dogwood*** open 2.9 ± 0.1 1.5 5.7 1.3 ± 0.1 41 0.001 closed 2.1 ± 0.2 0.5 4.4 0.7 ± 0.1 26 quaking aspen*** open 3.1 ± 0.2 0.5 11.1 1.4 ± 0.2 169 <0.0001 closed 1.6 ± 0.1 0.3 4.3 0.3 ± 0.2 53 pin cherry open 2.2 ± 0.1 0.6 4.2 0.8 ± 0.1 53 na closed na na na na 0 choke cherry open 2.2 ± 0.1 1.0 4.1 0.8 ± 0.1 44 0.085 closed 2.0 ± 0.1 0.8 3.9 0.8 ± 0.1 80 willow*** open 2.3 ± 0.1 0.5 5.5 0.9 ± 0.1 242 <0.0001 closed na na na na 14 mountain ash open 4.0 ± 0.1 2.0 7.0 1.1 ± 0.1 72 0.802 closed na na na na 7 all species open 2.3 ± 0.02 0.02 11.1 na 2071 na closed 2.4 ± 0.04 0.2 6.1 na 627 116 novel browse surveys – portinga and moen alces vol. 51, 2015 cherry, and choke cherry. rare species were alder, mountain ash, balsam poplar, and white pine (table 5). in summer 70% of bites consisted of mountain maple, willow, and paper birch on large feeding station, random feeding station, and random plot paths. the remaining 30% was juneberry, red maple, pin cherry, choke cherry, quaking aspen, and mountain ash. rare species were hazel, balsam poplar, red-osier dogwood, balsam fir, alder, bog birch, black ash, oak, elderberry, and white pine. on straight transects at least 70% of consumed twigs were mountain maple, willow, quaking aspen, and species considered rare (table 5). path type — despite the general similarities in diet diversity, all browse species comprised different portions of the winter diet on the 4 path types (kruskal-wallis, h3 > 12.3, p < 0.007) except paper birch and hazel (kruskal-wallis, h3 < 1.2, p > 0.60; table 6). in summer juneberry, quaking aspen, and mountain ash comprised different portions of the diet on all 4 path types in summer (kruskal-wallis, h3 > 8.1, p < 0.045; table 5). no difference existed among the 4 path types for red maple, mountain maple, paper birch, cherry, and willow (kruskal-wallis, h3 < 5.7, p > 0.13). individuals — diets based on twigs consumed on large feeding station paths varied individually and from the pooled average (tables 6 and 7). one winter example of this individual difference was female moose 31180 that consumed 26% red maple and 50% hazel (4 paths combined) compared to the group average of 5% red maple and 26% hazel (table 6); red maple was more available in her foraging patches. an example in summer was male moose 31190 that consumed 10% mountain maple and 61% willow (4 paths combined) compared to the group average of 41% mountain maple and 21% willow (table 7). browse species selection the average diet in winter (all moose combined) was different from that available (v29 ¼ 3122, p < 0.0001). a bonferroni z-test on the combined data indicated that juneberry, red maple, mountain maple, paper birch, red-osier dogwood, and quaking aspen were eaten more than available in summer. hazel was eaten less than available, and cherry and willow were used in proportion to availability (table 8). individual diets were also different from browse availability on their respective foraging paths (all moose: v2�9 � 74:6, p < 0.0001 for all moose). the average summer diet (all moose combined) was also different from available (v28 ¼ 840, p < 0.0001), as were individual diets (all moose: v2�8 � 43:9, p < 0.0001). table 4. available browse density and consumed browse density along four path types in summer and winter measured by twigs/m2 ± se and biomass (g)/m2 ± se. w = winter, s = summer. method season large feeding station random feeding station random plot straight transect available # twigs w 15.4 ± 1.6 2.3 ± 0.2 2.0 ± 0.2 1.4 ± 0.2 s 5.9 ± 0.6 2.0 ± 0.2 1.8 ± 0.3 1.1 ± 0.1 biomass w 9.9 ± 1.0 1.7 ± 0.1 1.5 ± 0.1 1.0 ± 0.1 s 15.2 ± 1.7 6.8 ± 1.9 4.5 ± 0.8 2.9 ± 0.4 consumed # twigs w 5.3 ± 0.6 2.1 ± 0.1 0.5 ± 0.1 0.2 ± 0.03 s 2.7 ± 0.3 1.0 ± 0.3 0.4 ± 0.1 0.1 ± 0.03 biomass w 4.0 ± 0.4 0.5 ± 0.04 0.4 ± 0.04 0.2 ± 0.02 s 6.7 ± 0.7 2.4 ± 0.4 1.0 ± 0.2 0.3 ± 0.04 alces vol. 51, 2015 portinga and moen – novel browse surveys 117 a bonferroni z-test on the combined summer data indicated that red maple, mountain maple, cherry, and mountain ash were eaten more than available in summer, willow less than available, and juneberry, paper birch, and quaking aspen proportional to availability (table 8). discussion we initially chose to measure large feeding stations (≥10 bites) because field observations indicated that these sites were common and theory (senft et al. 1987) supports the strategy of such foraging behavior. by contrasting browse density along a foraging path at large feeding stations with alternate routes, we demonstrated how moose increased effective browse density by selecting a specific foraging path. for example, moose took at least 80% of their bites at large feeding stations with ≥10 bites. the identification of large feeding stations provided a fast and efficient manner to measure browse availability and consumption along presumed foraging paths, and this method can also be used to evaluate the relative quality of browsed and unbrowsed patches (ward 2014, ward and moen 2014) this method avoids 2 potential complications associated with the straight transect table 5. diet composition (average percent of diet ± se) measured on four path types. averages and se were weighted by moose. rare includes species that made up <1% of the diet at large feeding station paths. 29 foraging paths were measured in winter 2013 and 31 were measured in summer 2012. winter species large feeding station random feeding station random plot straight transect hazel 27 ± 7 26 ± 8 27 ± 9 28 ± 8 paper birch 26 ± 7 26 ± 6 25 ± 6 18 ± 6 willow 11 ± 5 14 ± 6 13 ± 6 11 ± 5 quaking aspen 7 ± 3 8 ± 4 10 ± 5 13 ± 6 juneberry 6 ± 2 5 ± 2 4 ± 1 4 ± 2 red maple 5 ± 3 4 ± 2 5 ± 3 4 ± 4 red-osier dogwood 5 ± 4 3 ± 3 3 ± 3 10 ± 11 balsam fir 4 ± 2 6 ± 2 6 ± 3 2 ± 2 mountain maple 4 ± 3 3 ± 1 2 ± 1 2 ± 1 cherry 3 ± 1 2 ± 1 2 ± 1 2 ± 1 rare 2 ± 2 2 ± 1 2 ± 1 5 ± 6 summer mountain maple 42 ± 11 45 ± 10 43 ± 11 25 ± 11 willow 21 ± 8 21 ± 9 28 ± 11 23 ± 11 paper birch 11 ± 3 9 ± 4 6 ± 4 6 ± 5 cherry 9 ± 4 7 ± 4 6 ± 4 3 ± 5 quaking aspen 8 ± 4 10 ± 3 8 ± 3 14 ± 7 mountain ash 4 ± 2 3 ± 2 4 ± 4 0 juneberry 2 ± 1 3 ± 2 2 ± 1 8 ± 5 red maple 1 ± 1 0 0 7 ± 4 rare 1 ± 0.3 1 ± 0.4 0.2 ± 0.1 10 ± 7 118 novel browse surveys – portinga and moen alces vol. 51, 2015 method: 1) measuring random locations, and 2) empty plots. the foraging path approach eliminates these concerns by ensuring plentiful data at actual foraging locations. arguably, it also reflects the browse availability a moose would actually perceive. randomly placed plots in straight transects are often empty, which would mean that many more plots would be required to accurately estimate the availability of patchy browse. our method avoids empty plots, incorporates distance moved between feeding stations, and table 6. diet composition of individual moose in winter 2013 measured by twigs consumed at large feeding station paths. there are diets for eight collared moose. 31189 and 31190 are male, the rest are females. n is the number of foraging paths measured. rare species made up <1% of the combined moose diet at large feeding stations. moose number species all moose 31166 31174 31175 31178 31180 31182 31189 31190 hazel 27 21 38 29 13 50 33 9 68 paper birch 26 14 41 15 57 9 3 56 3 willow 11 5 9 6 5 3 quaking aspen 7 28 16 12 <1 8 2 juneberry 6 18 1 8 9 1 red maple 5 26 9 red-osier dogwood 5 15 <1 38 4 balsam fir 4 2 4 15 1 14 mountain maple 4 16 1 1 25 cherry 3 11 1 5 6 <1 3 6 rare 1 2 5 1 n 29 2 2 3 3 4 2 5 3 table 7. diet composition of individual moose in summer 2012 measured by twigs consumed at large feeding stations only. there are diets for seven collared moose. 31189 and 31190 are male, the rest are females. n is the number of sites measured. rare species made up <1% of the combined moose diet at large feeding stations. moose number species all moose 31166 31168 31175 31178 31180 31189 31190 mountain maple 41 3 57 84 90 36 57 10 willow 21 53 3 9 17 61 paper birch 11 12 1 17 13 7 cherry 9 8 5 3 2 24 1 3 quaking aspen 8 4 36 1 23 2 mountain ash 4 17 5 5 3 juneberry 2 1 12 red maple 1 5 rare 1 1 1 3 n 31 3 2 3 3 3 6 4 alces vol. 51, 2015 portinga and moen – novel browse surveys 119 provides an estimate of effective browse density. a challenge to simulating foraging decision rules when following a foraging path is that humans find large feeding stations by sight, but moose likely use other senses as well. diet composition was statistically different among seasons and path types. the average combined diet in both winter and summer was best categorized as generalist because one genus did not account for >60% of the diet (shipley 2010). the two primary browsed species were hazel and paper birch in winter and mountain maple and willow in summer, hence, moose may forage in different areas in winter and summer. for example, available browse density estimated by twig counts was higher in winter than in summer, with hazel consumed commonly in winter but rarely in summer. use of gps locations may help distinguish seasonal differences in foraging locations and browse species availability. the diet composition was similar to that measured >3 decades previously in the region (peek et al. 1976). the top 5 summer species (percent of diet) were the same in both studies: mountain maple, willow, paper birch, cherry, and quaking aspen. mountain maple was ranked first in our study and fifth by peek et al. (1976), and quaking aspen had the opposite rankings. hazel, willow, and quaking aspen were 3 of the top 5 winter species in both studies. one difference was that paper birch and juneberry were included in our top 5, whereas peek et al. (1976) had balsam fir and red-osier dogwood. during both seasons the primary species consumed table 8. browse species selection in both seasons when data from all moose was combined. if the moose were simply browsing at random, we would expect the 95% confidence interval of the percent browsed to contain the percent available at large feeding stations. season species percent available at large feeding stations 95% confidence interval of percent browsed at large feeding stations selection winter juneberry 4.7 5.1 ≤ – ≥ 6.8 + red maple 3.3 3.8 ≤ – ≥ 5.3 + mountain maple 2.7 4.0 ≤ – ≥ 5.5 + paper birch 19.3 24.7 ≤ – ≥ 27.9 + red-osier dogwood 2.1 3.3 ≤ – ≥ 4.8 + quaking aspen 5.6 5.8 ≤ – ≥ 7.6 + cherry 3.0 2.7 ≤ – ≥ 4.0 0 willow 11.9 11.2 ≤ – ≥ 13.5 0 balsam fir 9.0 2.8 ≤ – ≥ 4.1 − hazel 36.8 26.3 ≤ – ≥ 29.5 − summer red maple 0.5 0.6 ≤ – ≥ 1.3 + mountain maple 27.6 34.6 ≤ – ≥ 38.2 + cherry 7.2 8.3 ≤ – ≥ 10.5 + mountain ash 4.2 8.6 ≤ – ≥ 10.8 + juneberry 3.3 2.2 ≤ – ≥ 3.4 0 paper birch 10.4 9.8 ≤ – ≥ 12.1 0 quaking aspen 8.1 6.1 ≤ – ≥ 8.1 0 willow 28.6 18.9 ≤ – ≥ 21.9 − 120 novel browse surveys – portinga and moen alces vol. 51, 2015 were consistent regardless of path type. because more twigs were counted on the large feeding station paths, they probably provided the better estimate of diet and species consumption rates. this study was unique because we collected data from individual free-ranging moose by using their gps locations to identify their foraging paths shortly after use. presumably each moose selected browse based on availability within the patch they occupied. individual consumption differences occurred in both winter and summer, and though previous studies have not provided for analysis and comparison of individual diet selection, individual differences in habitat selection by moose were documented in british columbia (gillingham and parker 2010). pooling the data from many foraging paths identified the generalized seasonal diets and the most important browse species in this region, and concurred with previous research. it also identified individual diet variation which suggests that moose adapt their diet based on the local composition and availability of browse species. we were able to simulate how a moose browsed in a patch using the large feeding station method. there was some subjectivity in choosing which large feeding station was closest (consecutive) when establishing the foraging path; however, a moose would face the same choice. contrasting browse measurements between simulated and actual foraging paths in the same patch would provide a good evaluation of our approach and potential differences. we offer that incorporating large feeding stations and the distance between adjacent large feeding stations is an efficient method to estimate browse availability at the patch level. acknowledgements we would like to thank the epa great lakes restoration initiative and the minnesota environment and natural resources trust fund (enrtf) for funding and n. bogyo of the 1854 treaty authority for winter field work assistance. references gillingham, m. p., and k. l. parker. 2010. differential habitat selection by moose and elk in the besa-prophet area of northern british columbia. alces 44: 41–63. goddard, j. 1968. food preferences of two black rhinoceros populations. african journal of ecology 6: 1–18. heinselman, m. 1996. the boundary waters wilderness ecosystem. university of minnesota press, minneapolis, minnesota, usa. hjeljord, ø., n. hovik, and h. b. pedersen. 1990. choice of feeding sites by moose during summer, the influence of forest structure and plant phenology. holarctic ecology 13: 281–292. ———, b. e. saether, and r. andersen. 1994. estimating energy intake of freeranging moose cows and calves through collection of feces. canadian journal of zoology 72: 1409–1415. lenarz, m. s. 2011. 2011 aerial moose survey. minnesota department of natural resources, st. paul, minnesota, usa. lenarz, m. s., j. fieberg, m. w. schrage, and a. j. edwards. 2010. living on the edge: viability of moose in northeastern minnesota. journal of wildlife management 74: 1013–1023. (mndnr) minnesota department of natural resources. 2011. minnesota moose research and management plan. minnesota department of natural resources, st. paul, minnesota, usa. moen, r. a., r. peterson, s. windels, l. frelich, d. becker, and m. johnson. 2011. minnesota moose status: progress on moose advisory committee recommendations. nrri technical report no. nrri/tr-2011/41. natural resources research institute, duluth, minnesota, usa. alces vol. 51, 2015 portinga and moen – novel browse surveys 121 novellie, p. a. 1978. comparison of the foraging strategies of blesbok and springbok on the transvaal highveld. south african journal of wildlife research 8: 137–144. neu, c. w., c. r. byers, and j. m. peek. 1974. a technique for analysis of utilization-availability data. journal of wildlife management 38: 541–545. peek, j. m., d. l. urich, and r. j. mackie. 1976. moose habitat selection and relationships to forest management in northeastern minnesota. wildlife monographs 48: 3–65. renecker, l. a., and r. j. hudson. 1985. estimation of dry matter intake of freeranging moose. journal of wildlife management 49: 785–792. ———, and ———. 1986. seasonal foraging rates of free-ranging moose. journal of wildlife management 50: 143–147. risenhoover, k. l. 1987. wintering foraging strategies of moose in subarctic and boreal forest habitats. ph. d. thesis, michigan technical university, houghton, michigan, usa. schwartz, c. c. 1992. physiological and nutritional adaptations of moose to northern environments. alces supplement 1: 139–155. senft, r. l., m. b. coughenour, d. w. bailey, l. r. rittenhouse, o. e. sala, and d. m. swift. 1987. large herbivore foraging and ecological hierarchies. bioscience 37: 789–799. shipley, l. a. 2010. fifty years of food and foraging in moose: lessons in ecology from a model herbivore. alces 46: 1–13. ———, s. blomquist, and k. danell. 1998. diet choices made by free-ranging moose in northern sweden in relation to plant distribution, chemistry, and morphology. canadian journal of zoology 76: 1722–1733. sikes, r. s., and w. l. gannon. 2011. the animal care and use committee of the american society of mammalogists. (2011). journal of mammalogy 92: 235–253. spalinger, d. e., and n. t. hobbs. 1992. mechanisms of foraging in mammalian herbivores: new models of functional response. the american naturalist 140: 325–348. telfer, e. s. 1969. twig weight-diameter relationships for browse species. journal of wildlife management 33: 917–921. ward, r. l. 2014. browse availability, bite size, and effects of stand age on species composition and browse density for moose in northeastern minnesota. m.s. thesis, university of minnesota, minneapolis, minnesota, usa. ———, and r. a. moen. 2014. effects of stand age on species composition and browse density in northeastern minnesota. nrri technical report no. 2014– 36. natural resources research institute, duluth, minnesota, usa. 122 novel browse surveys – portinga and moen alces vol. 51, 2015 a novel method of performing moose browse surveys study area methods regressions and estimating bite mass gps collars measuring browse availability statistical analyses results regressions bite size bite density at feeding stations browse density consumption rate diet composition browse species selection discussion acknowledgements references alces28_223.pdf alces28_89.pdf alces29_235.pdf alces(25)_1.pdf alces27_24.pdf alces20_27.pdf alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces21_403.pdf alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces(23)_61.pdf alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces(25)_104.pdf alces24_22.pdf rodgersar text box alces21_231.pdf alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces27_150.pdf alces22_277.pdf alces vol. 22, 1986 rodgersar text box alces 22 (1986) alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 48, 2012 oehlers et al. visibility of moose 89 visibility of moose in a temperate rainforest susan a. oehlers1,2, r. terry bowyer3, falk huettmann2, david k. person4, and winifred b. kessler5 1yakutat ranger district, tongass national forest, 712 ocean cape drive, yakutat, alaska 99689 usa; 2ewhale lab, institute of arctic biology, and department of biology and wildlife, university of alaska fairbanks, alaska 99775 usa; 3department of biological sciences, 921 south 8th avenue, stop 8007, idaho state university, pocatello, idaho 83209 usa; 4division of wildlife conservation, alaska department of fish and game, 2030 sea level drive, ketchikan, alaska 99901 usa; 526700 west fork road, prince george, british columbia, v2k 5l6 canada. abstract: aerial surveys are the principal methods used to estimate populations of moose (alces alces gigas) in alaska. accounting for missed animals during aerial surveys is problematical, especially in forested habitats; incorporation of a visibility correction factor to account for the proportion of animals missed is known to improve accuracy of population estimates. our purpose was to study factors affecting visibility of radio-collared moose during aerial surveys in a temperate rainforest on the yakutat foreland, alaska, usa. wildlife managers in the area typically assume they observe only 50% of moose during surveys regardless of widely varying conditions. we used logistic regression to examine factors that influenced visibility including vegetation, light conditions, snow cover, and sex, age, and group size of moose. we then used logistic regression to develop a simpler model that only contained variables easily measured during aerial surveys: forest cover, snow cover, light, open versus vegetated habitat, and group size. we used that model to estimate a visibility correction factor. the mean correction factor was 1.304, ranging from1.005-2.138, yielding a population estimate of 699 (90% ci = 671-724) moose from a survey count of 595 animals. our correction factor was within the range reported for other populations of moose, and lower than the correction factor (2.0) currently used in this area. we conclude that application of site and time-specific visibility models is critical when estimating populations of large ungulates, especially in forested habitats. alces vol. 48: 89-104 (2012) key words: aerial surveys, alaska, alces alces gigas, gis, moose, population estimate, radio-telemetry, visibility. population estimates of ungulates based on aerial surveys are subject to error associated with the inability to detect animals that are present (visibility bias; timmerman 1993, anderson and lindzey 1996). environmental factors such as rugged terrain or dense cover may obscure visibility of animals, and differences in habitat selection and morphology by sex and age groups may make some animals more difficult to observe, thereby biasing their visibility (peek et al. 1974, thompson and veukelich 1981, bowyer et al. 2002, bowyer 2004). grouping behavior, activity of individuals (i.e., lying or standing), weather, and ground conditions (e.g., snow cover) can measurably affect visibility of animals. many of these problems are manifest in aerial surveys of moose (alces alces gigas) in temperate rainforests on the yakutat foreland of southeast alaska, usa where snow conditions that facilitate detecting moose can be intermittent, weather conditions for flying are frequently poor, and forest cover is dense and widespread. ideally, population surveys should be conducted during the mating season when moose are more active and sexes aggregated (miquelle et al. 1992, oehlers et al. 2011). because yakutat does not generally receive sufficient snowfall to enhance visibility before sexes spatially segregate after mating visibility of moose oehlers et al. alces vol. 48, 2012 90 and males cast antlers, identification of sex is difficult. sightability (also referred to as detectability or visibility) is the probability that an animal within the field of search for an observer will be seen by that observer (caughley 1974). that probability can be expressed as a scalar, or correction factor for visibility bias, which is then multiplied by the number of moose observed to obtain a more accurate population estimate than an uncorrected count (steinhorst and samuel 1989). correction factors for visibility bias (commonly referred to as sightability correction factors or scfs, and hereafter referred to as correction factors) that account for the proportion of animals undetected during aerial surveys are known to improve the accuracy of population estimates (timmerman 1993), particularly for areas with extensive forest cover and variable weather conditions that occur on the yakutat foreland. survey precision incorporates both the variance of total moose sighted and the variance of the correction factor (timmerman 1993). logistic regression is commonly used to develop correction factors for ungulates (mccorquodale 2001, quayle et al. 2001, mcintosh et al. 2009); this method is designed for use with binomial dependent variables (observed or not), and can accommodate continuous and categorical independent variables (hosmer and lemeshow 2000). we studied factors affecting visibility of moose on the yakutat foreland to improve population estimates from aerial surveys. we derived a series of models predicting correction factors using data from visibility trials from aerial surveys involving radio-collared moose. we examined the influence of temporal and weather-related variables such as month, time of day, cloud cover, light intensity, precipitation, and wind speed on visibility. we considered effects of environmental variables such as snow, forest, and vegetation cover on visibility of moose. in addition, we investigated the influence of sex, age, group size, sex and age composition of groups, activity, and intensity of site use on visibility. logically, we expected that forest cover and lack of snow cover would reduce the probability of moose being observed, and that visibility would increase with increasing snow cover. we also hypothesized that visibility would decline with smaller group size or if moose were bedded. further, we postulated that age or sex would affect visibility, because of morphological differences or if age groups and sexes used different habitats. we derived a model containing all of the covariates that we determined were important predictors of visibility, and a second model that included only those variables for which information could be obtained from routine aerial surveys. the full model was needed to consider all variables, including life-history characteristics such as sex and age and their potential influence on visibility, and would be useful in areas where sex and age composition is known or could be determined during surveys. we considered the second model to be more appropriate for management purposes in our study area, because it did not require data that could only be obtained reliably from radio-collared animals, and is more appropriate for late-winter surveys when sex cannot be accurately determined. finally, we applied the management model to a sample data set to estimate the density of moose within our study area. ours is one of few studies to examine factors influencing visibility of ungulates in a northern temperate rainforest, and our results should be useful to biologists managing ungulates throughout the northern coastal forests of the pacific northwest. study area we conducted research on the yakutat foreland of the tongass national forest, located along the southeast coastline of alaska (fig. 1). our study area of approximately 1,280 km2 encompassed most of the foreland, and included ~80 km of coastline extending from alces vol. 48, 2012 oehlers et al. visibility of moose 91 yakutat bay to dry bay. distance between the coast and mountain ranges varies from 8-24 km. there are several large rivers as well as numerous smaller streams distributed throughout the study area (fig 1). the yakutat foreland (lat. 59°20’ n, long. 139°0’ w) falls within the humid temperate domain, characterized by year-round cloudy, cool, and wet conditions (shephard 1995). the mean annual temperature was 4.1° c and the mean total precipitation was 381 cm (combined snow and rain) from 19712000 (noaa 2005). the mean temperature during this same time period was -3.4° c during january (the coldest month) and 12° c during july (the warmest month). total snowfall during the study was 345 cm; mean daily snowfall was 3.0 cm and the mean snow depth was 20 cm. other than a few rolling bedrock hills, most of the foreland is of low relief (average elevation 20 m; shephard 1995), and is a mosaic of forests, wetlands, and shrublands (shephard 1995). forested areas are dominated by sitka spruce (picea sitchensis), and a small percentage of the upper canopy is composed of black cottonwood (populus trichocarpa), western hemlock (tsuga heterophylla), and mountain hemlock (t. mertensiana). shephard (1995) documented 20 different forest communities on the foreland, with canopy cover ranging from 1-80% and averaging 60% for the common forest communities, with stand heights ranging from 15-47 m. nonforested areas include wetlands and shrublands composed primarily of graminoids, forbs, and shrubs including several species of tall and low willow (salix spp.) ranging from 1-6 m in height, and sitka alder (alnus sinuata) up to 4 m. nonforested areas dominate the coastal areas on the western half of the study area, with patches of spruce dispersed on the heaths and adjacent fig. 1. study area for developing a visibility model for moose on the yakutat foreland, alaska, usa, 2003-2004. visibility of moose oehlers et al. alces vol. 48, 2012 92 to some riparian zones; contiguous forested stands predominate the remainder. the most recent aerial surveys (2002) conducted in the forelands by alaska department of fish and game (adfg) estimated a density of 0.5 moose/km2 with a composition ratio of 19 males:100 females:14 young (n. l. barten, adfg, pers. comm.). total count surveys by parallel transects set approximately 0.4-0.5 km apart are conducted in the nonforested portions of the foreland by adfg in late autumn as soon as snow covers most of the ground, but often those conditions do not occur until well into winter. the adfg assumes that 50% of moose along transects are detected; consequently, the observed number of moose is doubled to estimate population size. that correction factor, however, has never been empirically evaluated or assessed. in addition, adfg does not survey forested portions of the forelands because of low (unknown) visibility, which constitutes about one-half of the study area. other large mammals that occur on the forelands include brown bear (ursus arctos), black bear (u. americanus), and gray wolves (canis lupus). sitka black-tailed deer (odocoileus hemionus sitkensis) occupy some of the islands offshore but are uncommon on the mainland. in addition, moose are an important part of the subsistence economy (ballew et al. 2006, schmidt et al. 2007). methods capture and handling twenty-two female and 16 male moose were darted from a helicopter by adfg personnel with palmer cap-chur equipment with the immobilizing drugs carfentanil and xylazine (roffe et al. 2001) during march and november 2002, and march and december 2003. dosages ranged from 3.0-5.0 mg of carfentanil and 100-130 mg of xylazine depending on time of year, sex, and animal condition. all capture and handling methods followed guidelines established by the american society of mammalogists animal care and use committee (1998) for research on wild mammals. our protocols were approved by independent institutional animal care and use committees at the university of alaska fairbanks (protocol # 04-26) and the adfg (protocol # 03-0001). we fitted moose with gps radio-collars (model 4000, lotek wireless, ontario, canada) that recorded locations 4 times daily, or standard vhf radio-collars (model mp2-mpp4, avm, colfax, california and model 600nh, telonics, mesa, arizona). we programmed both types of collars to release remotely relative to time of deployment (typically 1.5 yr). a lower incisor was removed from each moose to determine age from cementum annuli (gasaway et al. 1978). naltrexone (350-1300 mg) and tolazoline (400-800 mg) were subsequently administered and moose were monitored until they recovered from the immediate effects of immobilization. we also monitored each moose by aerial survey for 1 month post-capture to assess capture-related mortality. three females died or their collars malfunctioned within 1 month of capture, and were not included in the visibility trials. visibility trials we flew surveys to locate collared moose between 24 november 2003 and 18 march 2004 using a cessna® 185 fixed-wing aircraft. timing of sampling and type of aircraft were the same used by adfg when conducting moose surveys. we defined a visibility trial as the effort by the survey crew to count all moose within a 5 km2 sampling quadrat (square survey block) that included a radio-collared moose on a particular day. the aerial-survey crew was composed of the pilot, the primary observer in the front seat, and a secondary observer in the back seat behind the pilot. we attempted to control as many factors as possible, such as using the same aircraft, pilot, and primary observers for all trials. one pilot and 2 primary observers with >150 h of moose alces vol. 48, 2012 oehlers et al. visibility of moose 93 survey experience were used in the trials, with 8 secondary observers ranging in initial experience level of 6-8 (40-150 h of moose survey experience) on a lickart scale of 1-10. our trial procedure was similar to that of quayle et al. (2001). trials to locate individual radio-collared moose were separated by ≥3 days to reduce autocorrelation among locations. the extremely large home ranges of moose on the foreland (mean seasonal home ranges varied from 24.3-86.3 km2; oehlers et al. 2011) made this interval a reasonable choice for attempting to achieve independence among locations. frequencies for the subset of moose to be sampled during a flight were programmed into a receiver (model r4000, ats, isanti, minnesota) and scanned while flying at an altitude of 245-300 m above ground level. once a signal was received, the primary observer obtained the general position of the moose without identifying an exact location. we used a laptop computer equipped with baker geolink sketchmapping software (michael baker corporation, moon township, pa) to record our location and flight path so that the telemetry operator (primary observer) could identify the approximate location of the collared moose on the map without viewing the ground, thereby minimizing observer bias. the pilot and secondary observer also avoided scanning the ground in the immediate survey area to prevent detection of the target animal before beginning the survey. the survey crew noted if the collared moose was accidentally spotted by either observer while obtaining the general location; those observations were eliminated from analyses. the primary observer then delineated a 5 km2 (2.23 km x 2.23 km) quadrat (quayle et al. 2001) centered around the general location of the identified moose on the laptop computer using a 0.4 km grid overlay on the screen. because the location of the moose was inexact, the actual location of the moose was not centered within the quadrat. consequently, observer bias was minimized because none of the observers knew where in the quadrat to expect to find the radio-collared moose. the pilot then flew over the quadrat along transects spaced 0.4 km apart, which were delineated by the grid overlaid on the screen. the laptop screen displayed our flight path, allowing the pilot to navigate and follow the specified transect lines. the pilot flew the aircraft at an altitude of approximately 185 m and speed of about 130 km/h, resulting in a search intensity of approximately 1.0 min/ km2. we circled the location of each moose sighted in the quadrat to identify and record information on all of the variables included in the appendix, and recorded the location of the moose using the sketchmapper software. if the targeted moose was not sighted during the survey, we located that animal via telemetry and recorded the same information. forest cover was measured at 2 scales and recorded as “0” if the predominant vegetation within both a 10 m and 250 m radius of the radio-collared moose was nonforested, and “1” if this same area was predominantly forested (including a range of canopy covers). vegetation cover was defined as “0” if the predominant vegetation was open habitat such as muskeg, meadow, sand, or gravel bar, or “1” if there was vegetation such as tall shrubs or forest that could obscure visibility of the moose. percent vegetation was recorded as a categorical variable (1-3) representing percentage of vegetative cover (shrubs or trees) within a 10 m and 250 m radius of the observed moose that could obscure visibility of that moose. we defined a “group” as 1 or more moose within 50 m of each other (siegfried 1979, molvar and bowyer 1994, bowyer et al. 2001) to encompass the complete range of sociality for this species (monteith et al. 2007). we categorized age of non-collared animals as young (<1 year) or adult (≥1 years old) through visual observation. we expected a high pregnancy rate of yearlings (boer 1992), because preliminary data indicated a predator-limited visibility of moose oehlers et al. alces vol. 48, 2012 94 population (bowyer et al. 2005, oehlers et al. 2011). consequently, we considered yearling females as adults (monteith et al. 2007). moreover, distinguishing between yearlings and adults during aerial surveys in winter was difficult, and further distinguishing of ages beyond yearling or adult was not possible during aerial surveys. we used arcview 3.2 geographic information software (esri, redlands, ca, usa) to plot gps locations for moose and determine elevation and distance to the coast for each moose. elevation was extracted from a raster data layer provided by the u.s. forest service (usfs), which was based on usgs digital elevation model with 20-m resolution. distance from shore was calculated with the usfs shoreline polygon layer for the study area. statistical analyses detection of a radio-collared moose during visibility trials was coded 1 if detected and 0 if not observed. we used sas 9.1 (sas institute, cary, nc) for all statistical tests, and adopted an α = 0.05. we used multivariate logistic regression to model visibility. our suite of potential predictors of detection included parameters such as sex, age, group size, forest cover, snow cover, light conditions, aircraft speed, and experience of observers (appendix). group size was squared because the untransformed covariate was not linear in the logit. we included the identification of individual moose as a coded variable to control for making repeated measures of individual moose. we reduced potential multicollinearity among independent variables by testing for strong correlations between pairs of covariates (│r│≥0.7) and preventing their simultaneous inclusion in logistic regression models. during initial model screening, we also examined variance inflation factors (vif) and tolerance (tol) of independent continuous and discrete variables to identify intercorrelated variables. values of vif <10 and tol >0.40 were considered acceptable (neter et al. 1996, allison 2001). we ultimately considered 16 variables from the initial set of 27 candidate predictor variables. we then screened these remaining covariates using forward step-wise logistic regression (proc logistic; agresti 1990) with an alpha to enter of 0.15 (hosmer and lemeshow 2000, p. 118) and alpha to remove of 0.3, and backward logistic regression with alpha to remove of 0.3, to define a broad initial set of candidate models. we restricted the number of covariates within any candidate model to ≤8, because our sample size of visibility trials was 88; our sample size precluded a global model. our sample size also precluded an all possible regressions approach. we used hosmer and lemeshow tests for goodness-of-fit (hosmer and lemeshow 2000) to determine the appropriateness of the logistic models. once we had established a large set of candidate models, we used akaike’s information criterion (aicc) (burnham and anderson 2002) to select model variables. age and sex were included in most of the top candidate models. classifying moose into discrete age classes (i.e., beyond yearling or adult) is not possible from aerial surveys, and correct classification of sex is difficult once males have cast their antlers, so we repeated this same process omitting age and sex to allow development of models that did not rely on data from captured moose. accordingly, we developed overall explanatory models that included life-history characteristics, as well as management models which included variables that could be measured easily during aerial surveys alone. we used model-averaging procedures to derive composite explanatory and management models (burnham and anderson 2002, giudice et al. 2012). we only considered candidate models with aicc δ values ≤4 for inclusion in composite models. we calculated relative effects (risk ratios) for covariates included in our composite models (farmer et al. 2006). relative effects estimate the change in relative probability of detection alces vol. 48, 2012 oehlers et al. visibility of moose 95 for an incremental change in magnitude of a predictor variable (riggs and pollock 1992). we evaluated relative effects to determine the comparative importance of independent variables in affecting the probability of detection. in general, relative effects >2.0 or <0.5 indicated large effects of covariates on detectability (riggs and pollock 1992). for demonstrative purposes, we applied our composite management model to existing surveys of the moose population that were conducted by adfg on the yakutat foreland from 30 november-4 december 2005 using their survey methodology previously described in study area. model variables were assessed for each individual or group of moose observed during these surveys, and then the corresponding correction factor was calculated for each observation and multiplied by the number of animals in that observation. these corrected estimates were then totaled to derive a mean population estimate and the range of population estimates using the upper and lower correction factor based on the 90% ci (becker and reed 1990, anderson and lindzey 1996, white 2005). these data included 262 observations of single moose or groups and 595 total moose observed. results the median age for both females (n = 22, range = 3-13 yr) and males (n = 16, range = 1-10yr) was 6 years. we conducted 88 trials involving 55 radio-collared females and 33 males; each was surveyed 1-4 times (x = 2.3, sd = 0.70). snow conditions were generally adequate for aerial surveys from novemberjanuary and during the last 20 surveys conducted in march, but comparatively poor during february. we observed 254 groups of moose. radio-collared animals were sighted in 71% of the surveys; males were observed in 76% and females in 66% of the trials. radiocollared animals were detected in 82% of trials in nonforested areas, and in 27% of trials in forested cover. animals 1-3, 4-6, 7-10, and 11-13 years old were detected in 89, 55, 75, and 100% of trials, respectively. mean (± se) group size of collared animals was 3.7 ± 0.4. radio-collared animals were observed in open (31%), shrub (52%), and forested (17%) habitat during the trials. the location of females and males in nonforested and forested habitat was similar; 82 and 85% and 18 and 15%, respectively. logistic regressions forest cover, vegetation cover, and percent cover were each correlated (│r│≥0.7) between the 2 scales of measurement (10 m and 250 m). we considered the 10-m scale more easily estimated and likely to be consistent between observers; consequently, we chose to include the 10-m scale for each of these variables for consideration in our models. following tests for collinearity, variance inflation factors, and tolerance, candidate models for overall visibility included the parameters age, group size, forest cover, light, snow cover, experience secondary, and wind speed start (table 1). age, group size2, forest cover, and snow cover were included in each of the top 3 candidate models. visibility increased by 38% for each additional year of the moose aged, and by 75% for each additional (increasing) experience level of the secondary observer (table 2). overcast skies (versus sun) increased visibility by 175%. visibility increased with group size2 and speed of the plane (flight speed ranged from 129-145 km/h), but effects were small. visibility declined under forested cover (94%), snow cover of 0-33% (76%) or 34-66% (82%), and for females (23%). candidate models derived for management purposes (omitting sex and age) included group size2, forest cover, snow cover, light conditions, and vegetation cover (table 3). similar to the overall model, detectability increased with group size, nonforested and open habitat, overcast skies, and higher snow cover in the composite management model (table visibility of moose oehlers et al. alces vol. 48, 2012 96 4). application of the composite management model to our sample data yielded a range of correction factors from 1.005-2.138 for each observation. the mean correction factor was 1.304, and mean upper and lower (90% ci) correction factors were 1.215 and 1.390, yielding a population estimate of 671-724 animals (x = 699 moose) from an uncorrected count of 595 animals. discussion both our overall and management models included group size, forest cover, and snow cover as covariates of visibility. lack of snow cover strongly reduced visibility of moose and confirmed our hypothesis that visibility would be higher as snow cover increased. nonetheless, that relationship was not linear because visibility was similar between snow cover of 0-33% and 34-66% (57% and 54%, respectively). we believe that snow cover of 34-66% did not improve visibility because snow was still sufficiently patchy to obscure many moose against a dark background. we hypothesize that no snow cover actually may be preferable to patchy snow because patchy snow conditions may fatigue observers more quickly than uniform coverage. forest cover has been included in visibility models for both north american elk (cervus elaphus; samuel et al. 1987, bleich et al. 2001) and moose (peterson and page 1993, anderson and lindzey 1996, drummer and aho 1998, quayle et al. 2001). in our study area, coniferous tree species predominate in the forested areas, obstructing visibility of moose year-round, whereas vegetation in nonforested areas included alders and willows that model k parameters aicc∆i aiccwi a 5 age, group2a, forest cover, light, snow 0.000 0.4428 b 7 age, sex, observer2b, speedc, group2, forest cover, snow 0.7520 0.3041 c 6 age, sex, observer2, group2, forest cover, snow 1.8083 0.1793 d 4 group2, forest cover, snow, light 3.5851 0.0738 e 16 saturatedd 18.8200 0.0000 table 1. number of model parameters (k), differences in akaike’s information criterion (aicc) scores (∆) and aicc weights (wi) for candidate visibility models for moose on the yakutat foreland, alaska, 2003-2004. agroup size2. bexperience secondary. cwind speed start. dincludes survey start time, temperature, group, sex, age, experience primary, experience secondary, wind speed start, flight speed, group size2, forest cover, vegetation cover, percent cover, activity, light, snow cover, and elevation. variable β se rr rr 90% ci intercept -14.284 16.844 n/a n/a age 0.325 0.169 1.384 1.047-1.829 group size2 0.074 0.075 1.077 0.951-1.219 forest cover -2.849 0.945 0.058 0.0120.275 light 1.010 1.157 2.746 0.407-18.524 snow cover 1 (0-33%)a -1.419 1.201 0.242 0.033-1.755 snow cover 2 (34-66%)a -1.661 1.059 0.190 0.033-1.090 sexb -0.256 0.352 0.774 0.433-1.384 flight speed 0.063 0.075 1.065 0.941-1.205 experience secondary 0.562 0.693 1.754 0.559-5.504 table 2. regression coefficients and risk ratios (rr) for selected composite overall explanatory model for visibility of moose on the yakutat foreland, alaska, 2003-2004. confidence intervals did not overlap 1 in the individual models. asnow cover is relative to the reference variable of level 3, 67-100%. bsex is relative to the reference variable of male. alces vol. 48, 2012 oehlers et al. visibility of moose 97 do not retain leaves during winter and have less effect on visibility. group size was less influential on visibility than either forest or snow cover. group size affects visibility of elk (samuel et al. 1987, bleich et al. 2001, mccorquodale 2001), feral horses (equus caballus; ransom 2012), and mule deer (odocoileus hemionus; ackerman 1988); logically, larger groups are generally more visible. moose tend to aggregate in open areas in alaska during rut (miquelle et al. 1992, molvar and bowyer 1994); therefore, if snow conditions are adequate, visibility would be highest during the peak of rut. visibility did not differ when moose were standing or bedded. light condition also was an important predictor of visibility as moose were more visible in overcast conditions when glare and shadows were minimized. fox (1977) noted similar issues with glare from snowfields during mountain goat (oreamnos americana) surveys conducted in clear weather in southeast alaska. visibility increased with increasing age of moose, and was higher for males than for females, although the relative effect of sex was small. greater visibility of males could distort male:female ratios and result in the underestimation of the female population, unless a correction for differential visibility is incorporated. although several other studies of ungulate visibility reported that sex or group composition was accounted for in multivariate models because of correlation with other covariates such as group size or vegetation (anderson and lindzey 1996, bleich et al. 2001, mccorquodale 2001), sex in our model was not correlated with any other variable. the effect of sex on visibility probably occurred because of physical differences between the sexes; larger body size, darker color, and presence of antlers in early winter likely explain the higher visibility of males. solberg et al. (2010) also reported that male moose were observed by hunters with a 1.26 higher probability than females during the hunting season, and suggested that this difference was reflective of fundamental differences in antipredator behavior, including risk taking (such as use of open habitat), activity level, and space use. although age was not model k parameters aicc∆i aiccwi a 4 group2a, forest cover, snow, light 0.0000 0.5404 b 5 group2, forest cover, vegetation cover, snow, light 1.0609 0.3180 c 3 group2, forest cover, snow 2.6844 0.1442 d 14 saturatedb 14.4800 0.0004 table 3. number of model parameters (k), differences in akaike’s information criterion (aicc) scores (∆), and aicc weights (wi) for candidate visibility management models for moose on the yakutat foreland, alaska, 2003-2004. agroup size2. bincludes survey start time, temperature, group, experience primary, experience secondary, wind speed start, flight speed, group size2, forest cover, vegetation cover, percent cover, activity, light, snow cover, and elevation. variable β se rr rr 90% ci intercept 0.048 0.905 n/a n/a group size2 0.070 0.038 1.073 1.007-1.142 forest cover -2.551 2.190 0.078 0.002-2.894 light 1.441 0.920 4.225 0.926-19.279 snow cover 1 (0-33%)a -1.028 0.935 0.358 0.076-1.673 snow cover 2 (34-66%)a -1.377 0.897 0.252 0.057-1.109 vegetation cover -0.284 0.383 0.753 0.400-1.416 table 4. regression coefficients and risk ratios (rr) for selected composite management model for visibility of moose on the yakutat foreland, alaska, 2003-2004. confidence intervals did not overlap 1 in the individual models. asnow cover is relative to the reference variable of level 3, 67-100%. visibility of moose oehlers et al. alces vol. 48, 2012 98 significantly correlated with other variables, all observations of the oldest animals were in large groups in non-forested habitat; therefore, other covariates besides age were likely more influential on visibility. we did not detect an influence of age and sex composition of groups on visibility. we attempted to standardize flight speed during surveys, and weather conditions resulted in a minimal range of speeds (130145 km/h). remarkably, visibility of moose increased with speed of the plane. nonetheless, flight speed was in only 1 of the top 4 candidate overall explanatory models, its relative effect was small, and the 90% ci included 0; within the range of speeds we flew, this variable was likely of minimal importance. experience level (1-10) of the second observer increased visibility by 75% in the explanatory model; however, the effect was highly variable, and was not included in the management model. although experienced observers have developed a search image, and therefore may be more likely to observe moose, observer experience is difficult to quantify, and experience level changes over the course of visibility trials. previous studies have documented differences in visibility related to observer experience (leresche and rausch 1974, caughley et al. 1976); however, recent studies have noted little effect on visibility when observers were experienced (ackerman 1998) or when observer experience correlated with other variables in the model (samuel et al. 1987, anderson and lindzey 1996). all second observers in our study were experienced in moose surveys (i.e., 40-150 h of moose survey experience); consequently, our model will be most effectively applied when using experienced observers, a conclusion also reached by quayle et al. (2001). our overall visibility of moose was 70.5% and similar to that in quebec (crête et al.1986; 73%), alberta (rolley and keith 1980; 64%), and isle royale, michigan (peterson and page 1993; 78%), and higher than in minnesota (giudice et al. 2012; 38-56%), michigan (drummer and aho 1998; 39%), wyoming (anderson and lindzey 1996; 59%), and alaska (leresche and rausch 1974; 43-68%). correction factors for moose range from 1.03-3.2 (oosenberg and ferguson 1992, timmerman and buss 1998) and are generally higher in areas of denser cover and higher moose density (gasaway et al. 1986, peterson and page 1993). comparisons of visibility rates may be tenuous, however, because of differences in aircraft type (crête et al. 1986), number of observers, search intensity, and habitat (anderson and lindzey 1996). our results are within the range of correction factors reported for moose, but emphasize the variability in visibility and the need to develop correction factors specific to a particular area and time frame. the use of a dynamic correction factor, such as that developed with a visibility model, is superior to the use of a static correction factor. our modeled correction factor is offered as an alternative to the use of both a calculated scf (scfc) and an observed scf (scfo) as described by gasaway et al. (1986). observed scfs must be calculated for each survey (preferable daily), and are cost prohibitive in areas dominated by dense coniferous forests and areas of low moose density (gasaway et al. 1986), both of which occur in our study area. our results confirm that visibility of moose from aircraft varies with environmental factors and group size. therefore, application of the visibility model, combined with an appropriate sampling strategy, and with sophisticated analytical methods such as machine learning (‘non-linear statistics’; breiman 2001), may improve the accuracy and precision of population estimates over the use of a static correction factor. our method could be extended to other areas of similar environmental conditions such as the remainder of coastal alaska and british columbia (and could be tested for applicability to interior alaska) if protocols associated with alces vol. 48, 2012 oehlers et al. visibility of moose 99 the chosen model are followed (mccorquodale 2001). because visibility may differ among types of aircraft used (crête et al 1986), surveys should be conducted using a cessna® 185 or similar fixed-wing aircraft at approximately 185 m above ground elevation, as used in model development (samuel et al. 1987, anderson and lindzey 1996). additionally, observers should be experienced and their observation skills constantly calibrated in aerial surveys of moose. conducting surveys when moose are likely to be most visible (i.e., with nearly continuous snow cover and overcast light conditions) will provide the most precise population estimates. improved population estimates will allow for more knowledgebased and effective management decisions by state and federal managers. acknowledgements this research was funded primarily by the u.s. forest service, with additional funding from the department of interior bureau of indian affairs, and in-kind support from the adfg. the institute of arctic biology, department of biology and wildlife, and the alaska fish and wildlife cooperative research unit of the university of alaska fairbanks were all instrumental in the educational and funding portion of the project. w. eastland contributed technical support. many thanks to t. o’connor, c. grove, and e. campbell of the u.s. forest service for project support. special thanks to adfg biologists j. crouse, s. jenkins, n. barten, and k. white for their support in capture and handling of moose. pilots d. russel, l. hartley, b. bingham, and j. liston contributed to captures of moose, and our aerial survey efforts. thanks also to helicopter pilots of temsco helicopter and u.s. forest service helicopter managers a. stearns, j. schlee, and d. andreason. we acknowledge u.s. forest service personnel n. catterson, k. schaberg, b. lucey, d. gillikin, s. mehalick, m. moran, and c. wiseman for field and logistical support. references ackerman, b. r. 1988. visibility bias of mule deer aerial census procedures in southeast idaho. ph. d. dissertation, university of idaho, moscow, usa. agresti, a. 1990. categorical data analysis. john wiley & sons, new york, new york, usa. allison, p. d. 2001. logistic regression using the sas system: theory and application. sas institute inc., cary, north carolina, usa. anderson, c. r., and f. g. lindzey. 1996. moose sightability model developed from helicopter surveys. wildlife society bulletin 24: 247-259. animal care and use committee. 1998. guidelines for the capture, handling, and care of mammals as approved by the american society of mammalogists. journal of mammalogy 74: 1416-1431. ballew, c., a. r.tzilkowski, k. hamrick, and e .d. nobmann. 2006. the contribution of subsistence foods to the total diet of alaska natives in 13 rural communities. ecology of food and nutrition 45: 1-26. becker, e. f., and d. j. reed. 1990. a modification of a moose population estimator. alces 26: 73-79. bleich, v. c., c. s. y chun, r. w. anthes, t. e. evans, and j. k. fischer. 2001. visibility bias and development of a sightability model for tule elk. alces 37: 315-327. boer, a. h. 1992. fecundity of north american moose (alces alces): a review. alces supplement 1: 1-10. bowyer, r. t. 2004. sexual segregation in ruminants: definitions, hypotheses, and implications for conservation and management. journal of mammalogy 85: 1039-1052. _____, d. r. mccullough, and g. e. belovsky. 2001. causes and consequences of sociality in mule deer. alces 37: 371402. visibility of moose oehlers et al. alces vol. 48, 2012 100 _____, d. k. person, and b. m. pierce. 2005. detecting top-down versus bottom-up regulation of ungulates by large carnivores: implications for conservation of biodiversity. pages 342-261 in j. c. ray, k. h. redford, r. s. steneck, and j. berger, editors. large carnivores and biodiversity conservation. island press, covelo, california, usa. _____, k. m. stewart, s. a. wolfe, g. m. blundell, k. l. lehmkuhl, p. j. joy, t. j. mcdonough, and j. g. kie. 2002. assessing sexual segregation in deer. journal of wildlife management 66: 536-544. breiman, l. 2001. statistical modeling: the two cultures. statistical science 16: 199-231. burnham, k. p., and d. r. anderson 2002. model selection and multimodel inference: a practical information-theoretic approach. second edition. springer-verlag inc., new york, new york, usa. caughley, g. 1974. bias in aerial survey. journal of wildlife management 38: 921-933. _____, r. sinclair, and d. scott-kemmis. 1976. experiments in aerial survey. journal of wildlife management 40: 290-300. crête, m., l. rivest, h. jolicoeur, j. brassard, and f. messier. 1986. predicting and correcting helicopter counts of moose with observations made from fixed-wing aircraft in southern quebec. journal of applied ecology 23: 751-761. drummer, t. d., and r. w. aho. 1998. a sightability model for moose in upper michigan. alces 34: 15-19. farmer, c. f., d. k. person, and r. t. bowyer. 2006. risk factors and survivorship of black-tailed deer in a managed forest landscape. journal of wildlife management 70: 1403-1415. fox, j. l. 1977. summer mountain goat activity and habitat preference in coastal alaska as a basis for the assessment of survey technique. pages 190-199 in w. samuel and w. g. macgregor, editors. proceedings first international mountain goat symposium, 19 february, 1977, kalispell, montana, usa. gasaway, w. c., s. d. dubois, d. j. reed, and s. j. harbo. 1986. estimating moose population parameters from aerial surveys. biological paper of the university of alaska fairbanks 22: 1-108. _____, d. b. harkness, and r. a. rausch. 1978. accuracy of moose age determinants from incisor cementum layers. journal of wildlife management 42: 558-563. guidice, j. h., j. r. fieberg, and m. s. lenarz. 2012. spending degrees of freedom in a poor economy: a case study of building a sightability model for moose in northeastern minnesota. journal of wildlife management 76: 75-87. hosmer, d. w., and s. lemeshow. 2000. applied logistic regression. second edition. john wiley and sons, new york, new york, usa. leresche, r. e., and r. a. rausch. 1974. accuracy and precision of aerial moose censusing. journal of wildlife management 38: 175-182. mccorquodale, s. m. 2001. sex-specific bias in helicopter surveys of elk: sightability and dispersion effects. journal of wildlife management 65: 216-225. mcintosh, t. e., r. c. rosatte, j. hamr, and d. l. murray. 2009. development of a sightability model for low-density elk populations in ontario, canada. journal of wildlife management 73: 580-585. miquelle, d. g., j. m. peek, and v. van ballenberghe. 1992. sexual segregation in alaskan moose. wildlife monographs 122: 1-57. molvar, e. m., and r. t. bowyer. 1994. costs and benefits of group living in a recently social ungulate: the alaskan moose. journal of mammalogy 75: 621-630. alces vol. 48, 2012 oehlers et al. visibility of moose 101 monteith, k. l., c. l. sexton, j. a. jenks, and r. t. bowyer. 2007. evaluation of techniques for categorizing group membership of white-tailed deer. journal of wildlife management 71: 1712-1716. national oceanic and atmospheric administration (noaa). 2005. local climatological data. <http://cdo.ncdc.noaa.gov> (accessed may 2007). neter, j. w., m. h. kutner, c. j. nachtsheim, and w. wasserman. 1996. applied linear statistical models. fourth edition. the mcgraw-hill companies, inc., new york, new york, usa. oehlers, s. a., r. t. bowyer, f. huettmann, d. k. person, and w. b. kessler. 2011. sex and scale: implications for habitat selection by alaska moose. wildlife biology 17: 67-84. oosenberg, s. m., and s. h. ferguson. 1992. moose mark-recapture survey in newfoundland. alces 28: 21-29. peek, j. m., r. e. leresche, and d. r. stevens. 1974. dynamics of moose aggregations in alaska, minnesota, and montana. journal of mammalogy 55: 126-137. peterson, r. o., and r. e. page. 1993. detection of moose in midwinter from fixedwing aircraft over dense forest cover. wildlife society bulletin 21: 80-86. quayle, j. f., a. g. machutchon, and d. n. jury. 2001. modeling moose sightability in south-central british columbia. alces 37: 43-54. ransom, j. i. 2012. detection probability in aerial surveys of feral horses. journal of wildlife management 76: 299-307. riggs, m. r., and k. h. pollock. 1992. a risk ratio approach to multivariable analysis of survival in longitudinal studies of wildlife populations. pages 74-89 in d. r. mccullough and r. h. barrett, editors. wildlife 2001: populations. elsevier applied science, new york, new york, usa. roffe, t. j., k. coffin, and j. berger. 2001. survival and immobilizing moose with carfentanil and xylazine. wildlife society bulletin 29: 1140-1146. rolley, r. e., and l. b. keith. 1980. moose population dynamics and winter habitat use at rochester, alberta, 1965-1979. canadian field-naturalist 94: 9-18. samuel, m. d., e. o. garton, m. w. schlegel, and r. g. carson. 1987. visibility bias during aerial surveys of elk in northcentral idaho. journal of wildlife management 51: 622-630. schmidt, j. i., j. m. ver hoef, and r. t. bowyer. 2007. antler size of alaskan moose: effects of population density, harvest intensity, and use of guides. wildlife biology 13: 53-65. shephard, m. e. 1995. plant community ecology and classification of the yakutat foreland, alaska. us forest service, r10-tp-56. alaska region, juneau, alaska, usa. siegfried, w. r. 1979. vigilance and group size in springbok. madoqua 12: 151154. solberg, e. j., c. m. rolandsen, m. heim, j. d. c. linnell, i. herfindal, and b. seather. 2010. age and sex-specific variation in detectability of moose (alces alces) during the hunting season: implications for population monitoring. european journal of wildlife research 56: 871-881. steinhorst, r. k., and m. d. samuel. 1989. sightability adjustment methods for aerial surveys of wildlife populations. biometrics 45: 414-425. thompson, i. d., and m. f. veukelich. 1981. use of logged habitats in winter by moose with calves in northeastern ontario. canadian journal of zoology 59: 2103-2114. timmerman, h. r. 1993. use of aerial surveys for estimating and monitoring moose populations a review. alces 29: 35-46. _____ , and m. e. buss. 1998. population and harvest management. pages 559-615 visibility of moose oehlers et al. alces vol. 48, 2012 102 in a.w. franzmann and c.c. schwartz, editors. ecology and management of the north american moose. smithsonian institute press, washington, d. c., usa. white, g. c. 2005. correcting wildlife counts using detection probabilities. wildlife research 32: 211-216. alces vol. 48, 2012 oehlers et al. visibility of moose 103 variable type description method/time of collection month discrete month of visibility trial aerial survey age discrete age of collared moose capture sex discrete sex of collared moose capture group indicator 0 = single moose, 1 = ≥ 2 moose aerial survey group size discrete total number of moose seen within 50 m of collared moose aerial survey composition indicator 0 = single-sex group; 1 = both sexes in group aerial survey males discrete number of adult males in group aerial survey females discrete number of adult females in group aerial survey calves discrete number of calves in group aerial survey unknown sex discrete number of unknown sex adults in group aerial survey forest cover 10 m indicator 0 = nonforested, 1 = forested, within 10 m of moose aerial survey forest cover 250 m indicator 0 = nonforested, 1 = forested, within 250 m of moose aerial survey vegetation cover 10 m indicator 0 = open habitat such as muskeg, 1= shrub or forested habitat within 10 m of moose aerial survey vegetation cover 250 m indicator 0 = open habitat such as muskeg, 1= shrub or forested habitat within 250 m of moose aerial survey percent vegetation 10 m indicator 1 = 0-33%, 2 = 34-66%, 3 = 67-100% vegetative cover within 10 m of moose aerial survey percent vegetation 250 m indicator 1 = 0-33%, 2 = 34-66%, 3 = 67-100% vegetative cover within 250 m of moose aerial survey elevation continuous elevation above sea level in meters gis distance from coast continuous straight-line distance from coastline to center of moose group in meters gis activity indicator 0 = bedded, 1 = active (any moose in group) aerial survey site use indicator 0 = no beds, few tracks, 1 = beds and multiple tracks aerial survey cloud cover indicator 0 = clear, 1 = partly cloudy, 2 = overcast aerial survey precipitation indicator 0 = none, 1 = mist, 2 = light rain, 3 = hard rain, 4 = snow aerial survey snow cover indicator 1 = 0-33%, 2 = 34-66% ,3 = 67-100% aerial survey wind speed start continuous wind speed (km/h) at beginning of survey aerial survey wind speed end continuous wind speed (km/h) at end of survey aerial survey appendix candidate predictor variables considered during initial modeling for visibility of moose on the yakutat foreland, alaska, 2003-2004. visibility of moose oehlers et al. alces vol. 48, 2012 104 flight speed continuous average flight speed (km/h) during survey (excludes circling) aerial survey (plane instrumentation) temperature continuous average temperature (celsius) during survey aerial survey start time discrete survey start time; military time rounded to hour aerial survey light indicator 0 = sunny, 2 = flat light/even shadows aerial survey experience primary continuous previous experience level of primary observer, scale of 1-10 collected from each surveyor prior to visibility trials number flights primary discrete number of previous visibility trials by primary observer collected from each surveyor prior to visibility trials experience secondary continuous previous experience level of secondary observer, scale of 1-10 collected from each surveyor prior to visibility trials number flights secondary discrete number of previous visibility trials by secondary observer tabulated throughout visibility trials alces(25)_178.pdf alces(25)_58.pdf alces20_223.pdf alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces24_195.pdf alces28_1.pdf alces29_197.pdf alces24_126.pdf moose habitat use throughout gros morne national park krystal kerckhoff1, brian e. mclaren1, shane p. mahoney2 and tom w. knight3 1lakehead university, faculty of natural resources management, 955 oliver road, thunder bay, on, canada p7b 5e1; 2government of newfoundland and labrador, sustainable development and strategic science division, 2 canada drive, st. john’s, nl, canada a1b 4j6; 3gros morne national park, p.o. box 130, rocky harbour, nl, canada a0k 4n0. abstract: previous research indicated high variability in availability and habitat use by female moose in the lowlands of gros morne national park (gmnp), newfoundland and labrador, an area dominated by bogs and forest. here, we extend the earlier analysis with an additional 7 female moose (alces alces americana) occupying the park highlands, a region dominated by heath and shrub vegetation with forest limited to sheltered valleys, typical of interior and highland parts of the province. resource selection function (rsf) models with differences in habitat use between moose resident in the 2 regions and 2 moose that migrated from the lowlands in winter to the highlands in summer were rejected. in summer, more use of closed-canopy forest types occurred on the lowlands, while more use of non-forest habitat types occurred on the highlands. as before, we found that selection of disturbed forest is a winter phenomenon on the lowlands of gmnp; the same series of habitat types associated with disturbance were avoided in summer. summer migration by about 20% of gmnp moose to the highlands suggests that foraging opportunities are better during that season than in winter, a motivation for migration perhaps augmented by an overabundance of moose on the lowlands and unfavourable temperatures in disturbed areas that might otherwise serve as lowland foraging areas. an observation of more clustered relocations of moose on the highlands than on the lowlands of gmnp is consistent with our conclusion that moose use habitats within the highlands and lowlands of newfoundland and labrador very differently. we recommend 2 approaches to moose management for these different landscapes, both within gmnp and elsewhere in newfoundland and labrador. alces vol. 49: 113–125 (2013) key words: gros morne national park, habitat selection, moose, newfoundland and labrador, resource selection functions. according to habitat selection theory (fretwell and lucas 1970), individuals distribute themselves in a manner proportional to the quantity or quality of limiting resources available in each of several foraging patches, larger habitat units, and still larger landscapes. for ungulates, habitat selection should be driven by an individual’s ability to sense and select higher-quality food items or foraging areas (mcnaughton 1985, fryxell 1991). habitat selection that involves migration between 2 different landscapes can arise in a seasonal climate where different fitness opportunities (or forage availability) are offered by each landscape, but where the difference is less during the growing season (holt and fryxell 2011). moose (alces alces americana) in newfoundland and labrador, canada presumably distribute themselves optimally according to habitat selection theory in each of 2 typical landscapes in this province, the “highlands” and the “lowlands.” we explore this idea with analysis of summer and winter location data from gps-collared moose, using the assumption that more forested, lowland landscapes are, on average (i.e., throughout the year), superior to the 113 highlands where some moose migrate during summer. this paper is motivated by previous study of gros morne national park (gmnp), an area of 1,805 km2 in western newfoundland, where approximately 20% of the female moose population migrates within the park from forested, coastal lowlands (< 400 m above sea level) in winter to relatively open highlands (between 400 and 800 m) during summer (mclaren et al. 2000). our interpretation is that spending summer in a highland landscape offers an advantage to this fraction of the moose population. we compared seasons of activity of the resident moose in the highlands and lowlands of gmnp, and described their finer-scale activity in terms of frequency of smaller movements and the densities in which these smaller movement clusters occur throughout, by comparing the 2 landscapes. during winter, snow limits accessibility to forage more on the highlands than in the coastal lowlands (martin 2004), a motivation for migration that is consistent with empirical evidence from other studies of moose (reviewed by ball et al. 2001). as an additional explanation for the moose migration within gmnp, as suggested in mclaren et al. (2000), summer migration from the lowlands may be a means to avoid black bear (ursus americana) predation on calves, because the highlands may offer easier escape from this predator given the longer sightlines in open habitats. this second idea would be similar to the explanation for why woodland caribou (rangifer tarandus caribou) often migrate up mountain slopes (bergerud et al. 1984), and for why elk (cervus canadensis) migrate between high and low elevations in alberta (hebblewhite and merrill 2007, hebblewhite et al. 2008). however, because moose densities are about tenfold higher in newfoundland than in other parts of their range in north america (mclaren et al. 2004), creating obvious effects on hampering regeneration in the forests of the coastal plain of gmnp (connor et al. 2000, mclaren et al. 2004, gosse et al. 2011, humber and hermanutz 2011), and because only a fraction of the population migrates, we favour limited forage availability in the lowlands as the primary factor for summer moose migration. to explore this hypothesis, which is consistent with the holt and fryxell (2011) model for migration, we compare the frequency of the fine-scale summer movement on the lowlands and the highlands, and compare resource selection functions (rsfs) for highland and lowland moose in gmnp. study area gmnp is located on the gulf of the st. lawrence on the northern peninsula of newfoundland. its lowlands, which encompass parts of the western newfoundland forest and the coastal plain sub-region of the northern peninsula forest (damman 1983), are characterized by weather influences from the gulf, producing moderate levels of annual precipitation (900–1000 mm) and cold and snowy winters (300–350 mm is in the form of snow; hare 1952). its highlands, which are situated in the long range barrens (damman 1983), are similarly influenced by the gulf, but with an orographic effect that creates a harsher climate, having annual precipitation and snowfall on average double that of the lowlands (watson 1974). the mean annual temperature on the highlands is 4.5 °c colder than that of the lowlands (banfield 1983). in 1878, one female and one male moose were introduced to newfoundland from nova scotia, and in 1904, 2 male and 2 female moose were introduced from new brunswick (pimlott 1953). moose first inhabited the northern peninsula of newfoundland by the 1940s (caines and deichmann 1989). while moose are currently found in all ecoregions of newfoundland, 114 habitat use throughout gros morne park – kerckhoff et al. alces vol. 49, 2013 their density varies considerably. in the late 1970s when gmnp was being established, moose increased first on the highlands and by the 1980s moose had increased throughout the park (connor et al. 2000). at the time of the gps collaring, surveys using stratified random blocks estimated the moose population at 7,377 ± 1,249 (4.1 ± 0.7 moose/km2; mclaren et al. 2000; gmnp, unpublished data). in 2007, population size was estimated separately for the two landscapes, at 3,975 ± 1,287 in the lowlands (4.2 ± 1.4 moose/km2) and 788 ± 223 sd in the highlands (0.9 ± 0.3 moose/km2); densities in partial surveys of the park were estimated in 2009 at 5.9 moose/km2 on the lowlands and 1.1 moose/km2 on the highlands (gmnp, unpublished data). methods habitat classification taylor and sharma (2010) classified habitat types on the lowlands and highlands of gmnp from a single-image subset of 2, 10-m multispectral spot-5 satellite images (recorded 20 june 2006) with a k-means unsupervised classification. classes were reorganized and described using information from aerial photographs and forest inventories, and local expert knowledge and field visits. ten habitat types resulted for the lowlands (table 1), and 6 for the highlands (table 2). collectively, the lowlands comprise 938 km2 or 52% of the park, of which 417 km2 or 44% is moose habitat in forest or disturbed forest types; the highlands comprise 867 km2 or 48% of the park, 641 km2 or 74% of which is moose habitat, but only a fraction of which is forest (table 3). the classifications in table 1 and 2 are the reference for our description of habitat use by moose. moose locations in june 1997, 12 adult female moose (11 with at least one calf) were immobilized and fitted with gps collars (lotek engineering, inc.; mclaren et al. 2000; table 4). the collars were set to attempt a fix at 3-h intervals. remote downloading occurred in september 1997, november 1997, and march 1998. the collars were removed in november 1998, and the remaining data records were collected at that time. location accuracy was found to be dependent on collar position in relation to topography and canopy, but 95% of all differentially corrected data from test collars had ± 25 m accuracy (moen et al. 1997, mclaren et al. 2000). all 2-dimensional fixes were removed from the dataset, and only differentially corrected locations were used in the current analysis. depending on collar functioning, locations were recorded over a 4–15.5 month period (table 4). five of the collared moose were year-round residents in the lowlands, 5 were year-round residents in the highlands, and the remaining 2 migrated seasonally between the 2 landscapes. data analysis the dataset was divided into summer and winter seasons following vander wal and rodgers (2009). six moose were used for calculation of seasonal transition dates; 3 in the lowlands and 3 in the highlands with sufficient data records to span most of a calendar year. for these moose, cumulative distance travelled was calculated in arcview version 9 (esri, redlands, california) and plotted against time beginning with 1 january. winter was defined as the period when rate of travel was less than the mean rate, estimated from the points of inflection of the best-fit logistic curves to the plots, where the estimated changes from winter to summer and from summer to winter are symmetric around the inflection points. curvefitting used the logistic regression program in the statistical package for the social sciences (spss), version 18 (also used for all subsequent analysis). the median dates alces vol. 49, 2013 kerckhoff et al. – habitat use throughout gros morne park 115 for the start and end of winter were estimated from the 3 curves for each of the 2 landscapes and used to define the seasons for all subsequent analysis. summer and winter home ranges and core-use areas were calculated using the fixed-kernel method in home range tools (rodgers et al. 2007) with gaussian (bivariate normal) distributions, reporting the 95% and 50% isopleths for ranges and cores, respectively. the bandwidth size was determined by finding the smallest proportion of the reference bandwidth that allowed one continuous outer line to encompass the table 1. habitat descriptions from a lowlands classification of gros morne national park, newfoundland, canada. habitat type description category mature softwood forest softwood dominated, especially balsam fir (abies balsamea); some mixed stands with white birch (betula papyrifera). closed-canopy forest closed spruce forest softwood dominated (balsam fir and black spruce, picea mariana); other species include tamarack (larix laricina), trembling aspen (populus tremuloides) and alder (alnus spp.); site condition can be wet. some stands of scrub forest. closed-canopy forest closed mixed forest balsam fir dominated with some mixed stands (balsam fir, white birch). stem density can be very high. younger mixed stands (∼30 years since disturbance) are included. closed-canopy forest young softwood forest softwood dominated with high content of hardwoods; canopy >50% and 6–9 m in height. closed-canopy forest open softwood forest balsam fir dominated with 25–50% open canopy; white birch can be significant; some tree regeneration (heights of 1–4 m). open-canopy forest open mixed forest softwood dominated with 25–50% open canopy. sometimes wet. trees shorter than in closed mixed forest; some tree regeneration. open-canopy forest open hardwood forest hardwood dominated with 25–50% open canopy. often originally a mixed forest where regeneration of balsam fir does not occur. open-canopy forest sparse softwood forest softwood dominated (balsam fir, black spruce) with <25% canopy; limited regeneration; ferns and grass very prominent (<50% of ground cover); forest canopy is very broken consisting of mostly remnant forest from past disturbance; low density young black spruce <6 m height; pockets of conifer regeneration <4 m height can be present. disturbed forest: sparse canopy with herb/grass ground cover herb-hardwood forest dominant plants include ferns, grass and raspberry (rubus spp.) >50% of ground cover; very sparse forest canopy; some remnant white birch with alder or elderberry (sambucus racemosa). very little balsam fir. scattered spruce <4 m height. includes forested areas that have not regenerated after severe disturbance. disturbed forest: sparse canopy with herb/grass ground cover herb forest dominant plants include ferns and grass (>50% of ground cover); exposed soil is common; large amounts of dead material (standing or fallen) and scattered remnant trees. little regeneration >30 cm height. mostly forested areas that have not regenerated after severe disturbance. disturbed forest: sparse canopy with herb/grass ground cover 116 habitat use throughout gros morne park – kerckhoff et al. alces vol. 49, 2013 polygons (worton 1989). areas of open water, wetlands, and rock barrens were excluded from each of the resulting polygons and the remaining area was divided into the habitat types appropriate to the landscape. fine-scale habitat use examined areas where a minimum of 3 consecutive gps locations < 24 h apart occurred, with distances between them of < 50 m. this definition of an important habitat patch was arbitrary, but based on an inference that foraging and other activities such as bedding take place with shorter travel distances. mean weekly travel distances, as well as distances between the habitat patches, were calculated for each moose, for summer and winter separately, and then compared across seasons using repeatedmeasures analysis of variance (anova). minimum travel distances were calculated in all cases as straight lines between successive location points. rsfs (manly et al. 2002) were modelled 6 times each using logistic regression from pooled locations of all individuals: 1) based on number of locations in each habitat type within the home range, compared to its area on the surrounding landscape, for describing summer habitat use by residents and migrants using the highlands in a marginal model; 2) in a similar marginal model for describing winter habitat use by residents and migrants using the lowlands; 3) in a conditional model based on number of locations for each moose in each habitat type within its table 2. habitat descriptions from a highlands classification of gros morne national park, newfoundland, canada. habitat type description category open softwood forest balsam fir and some black spruce in a closed canopy ranging to <75% open; dense pockets of krummholz (locally known as tuckamore). open heath and fen and bog interspersed. closedto opencanopy forest scrub forest trees <4 m height. open heaths, fens, and bogs throughout (>50% of area). open-canopy forest shrub predominantly low shrubs (<1 m height), interspersed with fens, bogs, and small pockets of scrub forest. associated with transition from fen and tundra to scrub forest. can be wet. non-forest tundra low heath vegetation comprised of sedges (carex spp.), caribou moss (cladonia spp.) and crowberries (empetrum spp.); <20% rock, but few shrubs or trees. fairly dry. non-forest fen sedge meadows with fens throughout. non-forest rock barren boulder fields and exposed rock. very little vegetation. non-forest table 3. habitat availability in gros morne national park, newfoundland, canada. habitat type availability on landscape area (km2) percent lowlands mature softwood forest 69.6 7 closed spruce forest 27.6 3 closed mixed forest 65.5 7 young softwood forest 56.7 6 open softwood forest 62.5 7 open mixed forest 43.6 5 open hardwood forest 39.0 4 sparse softwood forest 19.3 2 herb-hardwood forest 20.0 2 herb forest 13.5 1 highlands open softwood forest 184.7 21 scrub forest 133.5 15 shrub 130.0 15 tundra 130.3 15 fen 62.6 7 alces vol. 49, 2013 kerckhoff et al. – habitat use throughout gros morne park 117 table 4. first and last dates of collaring, record length, and home range area in summer and winter from fixed-kernel estimates using a 95% isopleth for 12 gpscollared moose. this table also shows median seasonal transition dates, and lengths of summer and winter for moose using the two landscapes year-round in gros morne national park, newfoundland, canada. migrating moose are identified by an asterisk. id landscape first day collared last day recording record length (days) home range size (km2) median seasonal transition dates median season length (days) summer winter winter to summer summer to winter summer winter 15 lowlands 25-jun-97 13-oct-98 468 11.9 12.1 18-apr-98 11-oct-98 173 181 16 lowlands 25-jun-97 13-oct-98 468 13.2 13.1 18-apr-98 11-oct-98 173 181 19 lowlands 25-jun-97 05-nov-97 130 2.9 1.8 18-apr-98 11-oct-98 173 181 21* lowlands 25-jun-97 16-jan-98 201 — 5.5 18-apr-98 11-oct-98 173 181 22* lowlands 26-jun-97 18-jun-98 352 — 12.3 18-apr-98 11-oct-98 173 181 25 lowlands 26-jun-97 21-jun-98 355 8.3 12.1 18-apr-98 11-oct-98 173 181 26 lowlands 26-jun-97 15-nov-97 139 4.2 2.4 18-apr-98 11-oct-98 173 181 17 highlands 25-jun-97 13-oct-98 468 11.3 10.8 30-apr-98 24-oct-98 174 180 18 highlands 25-jun-97 27-feb-98 242 6.8 8.7 30-apr-98 24-oct-98 174 180 20 highlands 25-jun-97 17-mar-98 262 6.6 8.2 30-apr-98 24-oct-98 174 180 21* highlands 25-jun-97 16-jan-98 201 5.7 — 30-apr-98 24-oct-98 174 180 22* highlands 26-jun-97 18-jun-98 352 8.0 — 30-apr-98 24-oct-98 174 180 23 highlands 26-jun-97 13-oct-98 467 9.2 7.2 30-apr-98 24-oct-98 174 180 24 highlands 26-jun-97 01-jun-98 335 7.0 4.6 30-apr-98 24-oct-98 174 180 1 1 8 h a b it a t u s e t h r o u g h o u t g r o s m o r n e pa r k – k e r c k h o f f e t a l . a l c e s v o l . 4 9 , 2 0 1 3 home range, compared to its area on the surrounding landscape, for describing summer and winter habitat use by residents of the lowlands; 4) in a similar conditional model for describing summer and winter habitat use by residents of the highlands; 5) in a conditional model based on number of locations for each moose in each habitat type within its core-use area, compared to its area in the home range, for describing finer-scale summer and winter habitat use by residents of the lowlands; and 6) in a similar conditional model for describing summer and winter habitat use within the core-use areas of residents of the highlands. in the first 2 (marginal) models, one moose resident on the highlands was removed because of too few locations (id 18, table 4). habitat use by residents in the 2 landscapes, habitat use by the 2 migrant moose, and differences in habitat use between the winter and summer seasons were statistically compared in a mixed-effects model with random intercepts and coefficients (gillies et al. 2006). to determine the most parsimonious regression models, corrected akaike’s information criteria (aicc) and model deviance were compared to a model with random variables for each individual moose. a compound symmetric structure was assumed, meaning that covariance among all responses of an animal was assumed constant (skrondal and rabe-hesketh 2004) and habitat availability was also assumed constant over time (manly et al. 2002). these assumptions limit the applicability of the rsfs to the time period studied. random intercepts and coefficients for all habitat types experiencing some use were estimated, and coefficients significantly > 1 were defined as selection of a habitat type. calculations were all relative to open softwood forest as a reference habitat type, which was defined similarly for both landscapes. results home-range size varied considerably among individual moose, and there was no consistent size difference by landscape for either winter (f1,11 = 0.57, p = 0.58) or summer (f1,11 = 1.53, p = 0.06; table 4). there was also no difference in winter and summer home-range sizes on either the lowlands (f1,11 = 0.26, p = 0.88) or the highlands (f1,11 = 0.33, p = 0.67). the mean distances travelled during a one-year period were 309 km for residents on the lowlands and 267 km for residents on the highlands. moose travelled less in winter than in summer. weekly travel distances varied according to season (f1,11 = 106.35, p < 0.001; fig. 1). there was no difference in weekly travel distances by landscape (f1,11 = 0.75, p = 0.47). the summer season differed in length between the 2 landscapes, but only by a day (table 4). summer, defined by moose travel rates, started and ended close to 2 weeks earlier on the lowlands. the best-fit marginal rsf models describing habitat use by residents and migrants showed consistent selection of fig. 1. weekly distance travelled (km) for moose in gros morne national park, newfoundland, canada. alces vol. 49, 2013 kerckhoff et al. – habitat use throughout gros morne park 119 habitats in summer when they occupied the highlands together, and variable selection of habitats in winter when they occupied the lowlands together (table 5); however, for both seasons, models with differences in habitat use between residents and migrants were rejected. for the remaining 4 rsfs, conditional models with residents and migrants pooled, the best-fit were those including seasonal differences. in summer, habitat types used less than expected based on availability at both the home-range and core-use scales were herb-hardwood forest and herb forest (table 6). closed spruce forest, closed mixed forest, and young softwood forest were among the top habitat types selected in summer relative to open softwood forest, but were not selected more than expected. the pattern was generally reversed in winter on the lowlands; herbhardwood forest and herb forest, along with open hardwood forest and sparse softwood forest, were all selected by resident moose at both the home-range and core-use scales. at the home-range scale, young softwood forest and both closed forest types were also selected in winter. in the rsfs calculated for moose resident on the highlands, the fen, tundra, and shrub habitat types were selected in summer at the home-range scale, while only the fen and tundra types were selected at the core-use scale. the pattern was similar in winter, but scrub forest was also selected at the home-range scale and shrub, not tundra, was selected at the core-use scale. the overall trend in summer was more use of closed-canopy forest types on the lowlands and more use of non-forest habitat types on the highlands. selection of disturbed forest is a winter phenomenon on the lowlands of gmnp; the same category of habitat types is avoided in summer. defined by repeated occupation of an area with travel distances < 50 m apart, most fine-scale habitat patches on the lowlands were categorized as disturbed forest (50/127) or as open-canopy forest (47/127). there were an additional 22 fine-scale habitat patches identified in young softwood forest, while only 8 of the 127 fine-scale habitat selections on the lowlands were in closedcanopy forest. there were 13.5 habitat patches per 100 km2 on the lowlands, but more on the highlands (18.3/100 km2) where the majority were in open softwood forest (70/159). straight-line distances travelled between fine-scale habitat patches were table 5. ranking of habitat types, from most to least selected, for highland residents (n = 5; n = 3,252) and migrants (n = 2; n = 2,919) during summer on the highlands, and lowland residents (n = 5; n = 2,013) and migrants (n = 2; n = 1,018) during winter on the lowlands, where lower-case n refers to total number of locations in home ranges used to calculate resource selection functions (rsfs); gros morne national park, newfoundland, canada. habitat types significantly selected (p < 0.05) by at least 4 of 5 residents or both of the migrants are shown in boldface. the open softwood forest is a reference habitat (shown in italics). residents migrants summer on the highlands (1) fen fen (2) tundra tundra (3) shrub shrub (4) open softwood forest open softwood forest (5) scrub forest scrub forest winter on the lowlands (1) closed spruce forest herb-hardwood forest (2) herb forest closed spruce forest (3) closed mixed forest herb forest (4) mature softwood forest mature softwood forest (5) young softwood forest closed mixed forest (6) herb-hardwood forest sparse softwood forest (7) sparse softwood forest open hardwood forest (8) open hardwood forest young softwood forest (9) open mixed forest open mixed forest (10) open softwood forest open softwood forest 120 habitat use throughout gros morne park – kerckhoff et al. alces vol. 49, 2013 greater in summer than in winter (f1,120 = 36.28, p = 0.01), a consistent pattern for moose in both landscapes (f1,120 = 0.08, p = 0.93); there was no difference in travel distances between patches by landscape (f1,120 = 0.01, p = 0.99). discussion despite variation among individual moose in habitat selection in the park’s more diverse and forested lowlands, as reported earlier (mclaren et al. 2009), we are able to show with rsfs that more use of closed-canopy forest occurs in summer, likely as a means of heat avoidance. conversely, selection of disturbed and open-canopy forest is a winter phenomenon on the lowlands; the same category of habitat types is avoided during summer. in the cooler highlands, selection of non-forest habitat types may reflect less need to escape heat in the summer relative to the lowlands, and perhaps a means to escape insects. in winter, where moose populations are locally at higher densities according to both aerial surveys (gmnp, unpublished data) and the frequency of our identified winter habitat patches, selecting disturbed and opencanopy forest on the lowlands may be matched to optimal foraging, while selecting fen and shrub on the highlands may be matched to travel through areas where snow is packed along trails that reduces the energy cost of locomotion (telfer and kelsall 1979). table 6. ranking of habitat types, from most to least selected, within home ranges and core-use areas for resident moose in gros morne national park, newfoundland, canada: 5 moose in lowlands and 4 moose in highlands; lower-case n refers to total number of locations in home ranges or in core-use areas used to calculate rsfs. habitat types selected or avoided (p < 0.05) in proportion to their available area are shown in boldface. the open softwood forest is a reference habitat (shown in italics). summer habitat ranking winter habitat ranking lowlands home range (n = 3,765) core-use area (n = 1,485) home range (n = 2,013) core-use area (n = 1,679) (1) closed spruce forest closed mixed forest (1) herb forest herb-hardwood forest (2) young softwood forest closed spruce forest (2) herb-hardwood forest herb forest (3) closed mixed forest young softwood forest (3) young softwood forest sparse softwood forest (4) open softwood forest mature softwood forest (4) open hardwood forest open hardwood forest (5) mature softwood forest open softwood forest (5) closed mixed forest open mixed forest (6) sparse softwood forest open hardwood forest (6) closed spruce forest closed spruce forest (7) open hardwood forest open mixed forest (7) sparse softwood forest young softwood forest (8) open mixed forest sparse softwood forest (8) open mixed forest closed mixed forest (9) herb-hardwood forest herb-hardwood forest (9) mature softwood forest open softwood forest (10) herb forest herb forest (10) open softwood forest mature softwood forest highlands home range (n = 2,954) core-use area (n = 1,609) home range (n = 1,914) core-use area (n = 1,619) (1) fen fen (1) fen fen (2) tundra tundra (2) shrub shrub (3) shrub shrub (3) tundra tundra (4) open softwood forest open softwood forest (4) scrub forest scrub forest (5) scrub forest scrub forest (5) open softwood forest open softwood forest alces vol. 49, 2013 kerckhoff et al. – habitat use throughout gros morne park 121 in areas where snow is deep, reducing the energy cost of travel may be more important than avoiding competition for food. the most straightforward way of describing habitat use is in terms of density (holt and fryxell 2011). to approximate local moose densities, the total area in habitat types selected by moose could be substituted for an average density over the entire landscape areas. if habitat types selected during winter, based on the core-use areas of gps-collared moose occupying the lowlands in this season, are used to represent the best habitat types (herb-hardwood, herb, sparse softwood, and open hardwood forests), winter density would be 20.6 moose/km2, almost 5 x larger than the density estimate across the lowlands in the march 2007 survey (4.2 moose/km2). if moose remaining on the highlands in winter similarly used only those habitat types selected in core-use areas by the gps-collared subset (fen and shrub), their density would be 1.7 moose/km2, about twice the landscape density estimate (0.9 moose/km2) for the highlands. the lowland winter habitat types are essentially abandoned during summer in favour of habitat types providing thermal cover (closedcanopy forest); this change, combined with 20% of the population migrating to the highlands (mclaren et al. 2000), reduces effective summer density on the lowlands. meanwhile, summer migrants, according to the 2007 winter lowland population estimate, should double the corresponding winter estimate for the highlands, where tundra, roughly equal in area to shrub, is simply substituted as a preferred habitat in summer. presumably, seasonal abundance of forage is one benefit to spending the summer on the highlands. thus, 2 landscapes in gmnp provide insight into habitat selection by moose in newfoundland and labrador. we find that moose adapt seasonally to the park’s lowlands and highlands. moose adopting either of 2 strategies, year-round residence in one landscape or migration between landscapes, do not appear to select habitat differently when they occupy the same landscape. this point parallels the consensus for migration in a review and study in sweden (ball et al. 2001) that concluded that snow depth is the likely driver for moose migration. what differs in our study is insight into the advantages in summer for the fraction of moose opting to return to an otherwise less hospitable landscape, that being the snowy highlands. if we accept a conclusion from a québec study that movement rates for moose are better indicators of forage availability than home range size (dussault et al. 2005), and that the habitat patches in open softwood forest on the highlands offer more forage in summer than that provided on average in the disturbed or open-canopy forest on the lowlands, we are describing a situation similar to what has been described for predatorfree svaldbard reindeer (rangifer tarandus platyrhynchus) (bremset hansen et al. 2009). in this case, populations in overgrazed range move to areas of higher forage biomass, not higher forage quality. further, although plant phenology from spring through early summer is generally associated with increasing forage quality (klein 1990), the nitrogen content in forage declines initially after snowmelt (van der wal et al. 2000). thus, migrant moose may travel upland in gmnp to maximize biomass consumption while tracking delayed plant phenology in the cooler highlands climate. it is recommended that moose management in gmnp consider 2 landscapes (the lowlands and the highlands) as separate management units due to differences both in habitat types they offer and densities of moose they support. park management plans should ensure landscape connectivity for moose migrating between the highlands and lowlands. on this note, management across newfoundland and labrador that is both 122 habitat use throughout gros morne park – kerckhoff et al. alces vol. 49, 2013 effective and adaptable need not be dependent on defining discrete populations of moose, but should be in the context of the 2 very different landscapes the province offers to moose. acknowledgements dr. a. rodgers, centre for northern forest ecosystem research, ontario ministry of natural resources, and dr. d. eastman, university of victoria, provided advice and constructive criticism on portions of this paper submitted by its lead author as a msc thesis in forestry at lakehead university. two anonymous reviewers and alces editor dr. e. addison provided helpful comments to improve this manuscript. s. taylor, gmnp, provided invaluable assistance in interpreting habitat types and in analyzing gps data. this project was funded and coordinated by the institute for biodiversity, ecosystem science and sustainability of the government of newfoundland and labrador, with additional funding and in-kind support provided by parks canada and lakehead university. the original fieldwork and preliminary data analysis occurred under the guidance of the inland fish and wildlife division of the government of newfoundland and labrador, with assistance by d. anions and c. mccarthy (formerly of gmnp) and technical advice from dr. a. rodgers. references ball, j. p., c. nordengren, and k. wallin. 2001. partial migration by large ungulates: characteristics of seasonal moose alces alces ranges in northern sweden. wildlife biology 7: 39–47. banfield, c. e. 1983. climate. pages 37– 106 in g. r. south, editor. biogeography and ecology of the island of newfoundland. w. junk publishers, boston, massachusetts, usa. bergerud, a. t., h. e. butler, and d. r. miller. 1984. anti-predator tactics of calving caribou: dispersion in mountains. canadian journal of zoology 62: 1566– 1575. bremset hansen, b., i. herfindal, r. aanes, b.-e. sæther, and s. henriksen. 2009. functional response in habitat selection and the trade-offs between foraging niche components in a large herbivore. oikos 118: 859–872. caines, p., and h. deichmann. 1989. resource description and analysis: gros morne national park. gros morne national park, parks canada, rocky harbour, newfoundland, canada. connor, k. j., w. b. ballard, t. dilworth, s. mahoney, and d. anions. 2000. changes in structure of a boreal forest community following intense herbivory by moose. alces 36: 111–132. damman, a. w. h. 1983. an ecological subdivision of the island of newfoundland. pages 163–205 in g. r. south, editor. biogeography and ecology of the island of newfoundland. w. junk publishers, boston, massachusetts, usa. dussault, c., r. courtois, j.-p. ouellet, and i. girard. 2005. space use of moose in relation to food availability. canadian journal of zoology 83: 1431–1437. fretwell, s. d., and h. l. lucas. 1970. on territorial behaviour and other factors influencing habitat distribution in birds. acta biotheoretica 19: 16–36. fryxell, j. m. 1991. forage quality and aggregation by large herbivores. american naturalist 138: 478–498. gillies, c. s., m. hebblewhite, s. e. nielsen, m. a. krawchuk, c. l. aldridge, j. l. frair, d. j. saher, c. e. stevens, and c. l. jerde. 2006. application of random effects to the study of resource selection by animals. journal of animal ecology 75: 887–898. gosse, j., l hermanutz, b. mclaren, p. deering, and t. knight. 2011. degradation of boreal forests by non-native herbivores in newfoundland’s national parks: recommendations for ecosystem alces vol. 49, 2013 kerckhoff et al. – habitat use throughout gros morne park 123 restoration. natural areas journal 31: 331–339. hare, f. k. 1952. the climate of the island of newfoundland: a geographical analysis. geographic bulletin 2: 36–88. hebblewhite, m., and e. h. merrill. 2007. multiscale wolf predation risk for elk: does migration reduce risk? oecologia 152: 377–387. ———, ———, and g. mcdermid. 2008. a multi-scale test of the forage maturation hypothesis in a partially migratory ungulate population. ecological monographs 78: 141–166. holt, r., and j. fryxell. 2011. theoretical reflections on the evolution of migration. pages 17–31 in e. milner-gulland, j. fryxell, and a. sinclair, editors. animal migration: a synthesis. oxford university press, oxford, united kingdom. humber, j. m., and l. hermanutz. 2011. impacts of non-native plant and animal invaders on gap regeneration in a protected boreal forest. biological invasions 13: 2361–2377. klein, d. r. 1990. variation in quality of caribou and reindeer forage plants associated with season, plant part, and phenology. rangifer special issue 3: 123–130. manly, b. f. j., l. l. mcdonald, d. l. thomas, t. l. mcdonald, and w. p. erickson. 2002. resource selection by animals, second edition. kluwer academic publishers, dordrecht, netherlands. martin, c. 2004. climatology and historical snowcover of the big level plateau, gros morne national park, newfoundland. msc thesis, memorial university of newfoundland, st. john’s, newfoundland, canada. mclaren, b., c. mccarthy, and s. mahoney. 2000. extreme moose migrations in gros morne national park, newfoundland. alces 36: 217–232. ———, b. a. roberts, n. djan-chékar, and k. p. lewis. 2004. effects of overabundant moose on the newfoundland landscape. alces 40: 45–59. ———, s. taylor, and s. h. luke. 2009. how moose select forested habitat in gros morne national park, newfoundland. alces 45: 125–135. mcnaughton, s. j. 1985. ecology of a grazing ecosystem: the serengeti. ecological monographs 55: 259–294. moen, r., j. pastor, and y. cohen. 1997. accuracy of gps telemetry collar locations with differential correction. journal of wildlife management 61: 530–539. pimlott, d. h. 1953. newfoundland moose. transactions of the north american wildlife conference 18: 563–581. rodgers, a. r., a. p. carr, l. b. hawthorne, l. smith, and j. g. kiem. 2007. home range tools for arcgis. version 1.1. ontario ministry of natural resources, centre for northern forest ecosystem research, thunder bay, ontario, canada. skrondal, a., and s. rabe-hesketh. 2004. generalized latent variable modeling: multilevel, longitudinal, and structural equation models. chapman and hall, new york, new york, usa. taylor, s., and r. sharma. 2010. assessing the status of forest recovery in disturbance affected areas using spot 5 imagery in gros morne national park. gros morne national park, parks canada, rocky harbour, newfoundland, canada. telfer, e. s., and j. p. kelsall. 1979. studies of morphological parameters affecting ungulate locomotion in snow. canadian journal of zoology 57: 2153–2159. vander wal, e., and a. r. rodgers. 2009. designating seasonality using rate of movement. journal of wildlife management 73: 1189–1196. van der wal, r., n. madan, s. v. lieshout, c. dormann, r. langvatn, and s. d. albon. 2000. trading forage quality for quantity? plant phenology and patch 124 habitat use throughout gros morne park – kerckhoff et al. alces vol. 49, 2013 choice by svalbard reindeer. oecologia 123: 108–115. watson, w. b. 1974. the climate of gros morne national park, newfoundland. atmospheric environmental service, department of environment, st. john’s, newfoundland, canada. worton, b. j. 1989. kernel methods for estimating the utilization distribution in home range studies. ecology 70: 164–168. alces vol. 49, 2013 kerckhoff et al. – habitat use throughout gros morne park 125 moose habitat use throughout gros morne national park study area methods habitat classification moose locations data analysis results discussion acknowledgements references alces26_30.pdf alces22_449_distinguishedmoosebio.pdf alces vol. 22, 1986 alces vol. 22, 1986 alces28_235.pdf alces28_95.pdf alces29_243.pdf browse removal, plant condition, and twinning rates before and after short-term changes in moose density thomas f. paragi1, c. tom seaton1, kalin a. kellie1, rodney d. boertje1,2, knut kielland3, donald d. young, jr.1, mark a. keech1,4, and stephen d. dubois5,6 1alaska department of fish and game, 1300 college road, fairbanks, alaska 99701, usa; 3institute of arctic biology, university of alaska-fairbanks, fairbanks, alaska 99775, usa; 5alaska department of fish and game, p.o. box 605, delta junction, alaska 99737, usa abstract: we monitored forage-based indices of intraspecific competition at changing moose (alces alces) densities to gauge short-term, density-dependent environmental feedback and to ultimately improve management of moose for elevated sustained yield. in 4 areas of interior alaska where moose density recently changed, we evaluated the magnitude of change among 4 browse indices: proportional offtake of current annual growth biomass (oftk), proportion of current twigs that were browsed (ptb), mean twig diameter at point of browsing (dpb), and proportion of plants with broomed architecture. in 1 area where moose density increased 100% in 6 years following effective predation control, browse removal increased 138% for oftk, 20% for ptb, and 16–42% for dpb of primary browse species, with a 44% increase in brooming. we also studied 3 areas where moose density declined 31–41% following elevated antlerless harvests of 2–4 years duration. in these areas (with intervals of 3–12 years between browse surveys) we found declines of 30–40% in oftk, 26–68% in ptb, and 11–37% in dpb, but changes in plant architecture were inconsistent. the proportion of parturient cows with neonate twins did not change between browse surveys, presumably because of a substantial lag time influenced by life history of the dominant reproductive cohorts and little change in browse nutrient content and digestibility. of the 4 browse indices studied, proportional oftk most consistently reflected the direction and magnitude of short-term changes in moose density. area-specific measures of habitat and animal conditions at high moose density provided an objective means for gauging the capacity of the respective ecosystems to support moose and maintain forage plants. we used these measures of winter forage and moose condition to justify implementing harvest strategies and to ultimately reduce high moose densities below levels of strong negative feedback. alces vol. 51: 1–21 (2015) key words: alaska, density-dependent, forage, intraspecific competition, moose, nutritional condition. moose (alces alces) management becomes increasingly challenging for populations at the extremes of the nutritional gradient. at low densities managers may consider predator control to increase abundance (gasaway et al. 1992). at high densities habitat enhancement may be an option to increase forage, or antlerless harvests could reduce abundance or population growth rates (boertje et al. 2009, young and boertje 2011). wildlife managers in alaska are often required to estimate harvestable surplus and nutritional status of wild moose populations over large areas (≤15,000 km2) of remote forested and subalpine habitats. it is difficult to estimate the 2present address: 220 blue bird, kerrville, texas 78028, usa 4present address: p.o. box 84634, fairbanks, alaska 99708, usa 6present address: p.o. box 702, delta junction, alaska 99737, usa 1 capability of habitats to support moose because of limited studies on the physiological requirements based on captive animals (reviewed in schwartz and renecker 1997) and the inherent variability in habitat and other environmental factors. biologists must either quantify forage production (kg/ha) in the context of daily food requirements for an absolute estimate of carrying capacity (e.g., wolff and zasada 1979, crete 1989, maccracken et al. 1997) or use indices to assess the relative nutritional status of the moose population and/or condition of the range. there are no standardized economical methods for assessing landscape carrying capacity in remote areas of alaska, so biologists use nutritional status of the moose population or indices related to winter forage use (boertje et al. 2007, seaton et al. 2011). these indices presumably reflect the negative feedback in nutrition from increased intraspecific competition for food resources at increasing moose density and indicate if competition is reduced as density declines. negative feedback reduces productivity and thus sustainable harvest among age and sex classes (mccullough 1984), poses a heightened risk of unsustainable forage removal, and can lead to dramatic population declines, often facilitated through winter-related mortality (gasaway et al. 1983). the most established index of nutritional status of a moose population in interior alaska is twinning rate, the proportion of parturient females with 2 neonatal calves (franzmann and schwartz 1985, keech et al. 2000, boertje et al. 2007). dressed weights of harvested calves (e.g., cederlund et al. 1991), age at first reproduction, short-yearling live mass, and browse removal rate (boertje et al. 2007) have also been used to estimate or gauge nutritional status of moose populations. however, few studies have examined how well these indices respond following intended changes in abundance through management actions. measuring animal indices can be constrained by sample size at low density, limiting their usefulness in monitoring change in abundance. it can be difficult or infeasible to observe an adequate sample of random parturient females from aircraft for estimating twinning rate in areas of low moose density (e.g., stout 2010) or in dense cover that hinders viewing of calves. conversely, browse sampling is not constrained by moose observations. seaton et al. (2011) documented an inverse correlation between proportional browse biomass removal and twinning rate across a 10-fold range in density (0.1–1.2 moose/km2) among 8 game management units of interior alaska. that study demonstrated the utility of a habitat metric for indirectly judging nutritional condition of adult female moose, which helped substantiate prior conclusions by boertje et al. (2007) based on smaller sample sizes. in this study we sought to document shortterm, landscape-level changes in browse removal rates and architecture of winter forage species following short-term, managementinduced changes in moose density. seaton (2002) reviewed methods of estimating browse removal by moose and used a modified technique to characterize “apparent” browse production (prod; kg/ha above snow) and browse offtake by moose (oftk; kg/ha), and to estimate proportional oftk (oftk/prod). the technique quantifies woody biomass through measuring twig diameter at the proximal end of current annual growth (cag) and the diameter at point of browsing (dpb) in late winter, just prior to the new growing season. earlier studies reported a correspondence between proportional oftk and the proportion of twigs browsed by moose (ptb = number of dpb > 0 divided by number of cag) (regelin et al. 1987, maccracken and viereck 1990), which suggested that the simpler twig count would be more efficient. however, seaton (2002:32) cited another 2 browse indices to moose density – paragi et al. alces vol. 51, 2015 study (k. kielland and t. osborne, unpublished data) where ptb was insensitive to a large (8-fold) change in moose abundance in western interior alaska. thus, estimating the biomass produced and removed with diameter measurements is important because moose may clip twigs at a range of diameters, and the nutritional value (e.g., digestibility and nutrient concentrations) decreases as cag diameter increases (vivås and sæther 1987, kielland and osborne 1998). the smallest diameter twigs provide the most nutrient gain per unit of mass but extend rumen fill time, whereas the largest diameter twigs provide less nutrient gain per unit mass and extend rumen processing time (gasaway and coady 1974, shipley and spalinger 1992). seaton (2002) also sought to incorporate forage plant architecture to gauge the longer-term effects of moose browsing. this information might additionally help managers and the public understand negative feedback at higher moose densities and characterize relatively less use at lower densities. in this study we followed a recommendation by seaton et al. (2011) to evaluate the utility of browse indices for detecting short-term changes in intraspecific competition following intended management actions. in 4 areas with baseline browse data and subsequent changes in moose density, we examined the magnitude of changes in browse removal and plant architecture as gauges of density-dependent feedback. we compared changes in browse metrics with changes in an established index of nutrition (twinning rates) to better understand how browse removal may or may not reflect changes in moose nutrition. we assumed that reducing moose abundance in a defined area where forage production changed relatively little over time will reduce intraspecific competition for preferred species, with an inverse response following an increase in abundance. we predicted that increased moose density would cause increased proportional oftk, increased ptb, increased mean dpb for at least the dominant or preferred browse species, and an increased proportion of plants with architecture partly or heavily affected by moose foraging. these conditions are presumed to be coincident with a decrease in moose nutrition, resulting in a lower twinning rate at higher density. where moose density decreased, we predicted the inverse responses, with one exception; that short-term reversal of trend in plant architecture as affected by moose browsing at high density (broomed → unbrowsed) would be unlikely on existing plants in the absence of widespread disturbance to regenerate young plants, such as fire or flooding. also, moose nutrition, as indexed by twinning rates, would not likely increase immediately following a decline in moose abundance unless browse quantity and quality increased substantially. intentionally changing reproductive rates is more likely a long-term proposition based on changing calf weights and eventually the life history of the dominant reproductive cohorts (females 4–10 yr old; boertje et al. 2007). study areas and moose abundance the 4 study areas (unit 19d, unit 20a central hills, unit 20a western flats, and unit 20d) were located in the boreal forest of interior alaska, usa (seaton et al. 2011, fig. 1). management actions were implemented in these areas to influence moose abundance. to illustrate the magnitude of density change in each area, we calculated moose abundance and confidence intervals for areas approximating the extent of browse sampling before and after management actions. population increase unit 19d in the remote kuskokwim valley is comprised of large floodplains and alces vol. 51, 2015 paragi et al. – browse indices to moose density 3 fig. 1. continued on next page. 4 browse indices to moose density – paragi et al. alces vol. 51, 2015 fig. 1. post-treatment sampling grids and browse plot locations before and after management actions intended to affect moose density for the 4 study areas in interior alaska. alces vol. 51, 2015 paragi et al. – browse indices to moose density 5 forested uplands within 40 km of mcgrath (62° 57′ n, 155° 36′ w). in this study area we sampled browse on 3 occasions: broadly over 10,600 km2 in 2001, narrowly in a 1368 km2 experimental area (keech et al. 2011) in 2003, and over a 2896 km2 moose survey area that included the experimental area in 2009 (keech 2012). we used the 2003 browse data for pre-treatment in unit 19d because sampling scale (paragi et al. 2008:36) was closer to that from 2009 (fig. 1a), but we provided 2001 data for additional pre-treatment context on variation in browse metrics when moose density was low (paragi et al. 2008:9). in unit 19d, keech et al. (2011) described predation control beginning in 2003 that caused a gradual increase in moose abundance over the next 6 years. in the 2009 browse sampling area, moose density estimated from fall aerial surveys doubled from 0.30/km2 in 2001 to 0.62/km2 by 2009 (keech 2012:14). population decrease units 20a and 20d are near the road system in the tanana valley near the city of fairbanks and the town of delta junction, respectively. the unit 20a central hills study area was comprised of forested uplands and subalpine shrubs 100 km south of fairbanks in the foothills of the alaska range (64° 08′ n, 147° 55′ w). this area was primarily used during fall and winter by moose that migrated to lowland flats to the north during the summer (keech et al. 2000), inclusive of when hunting and abundance surveys occurred. we sampled 600 km2 of the unit 20a central hills as part of a larger study by seaton (2002) in 2000 and sampled 790 km2 in 2012 that conformed to a harvest-reporting boundary and largely overlapped the earlier browsesampling area (fig. 1b). the unit 20a western flats (64° 26′ n, 148° 50′ w) were forested lowlands with interspersed wetlands. we sampled 1100 km2 of the western flats in 2006 and 1625 km2 in 2009; the latter survey largely overlapped the earlier survey area, but nearly half of it was influenced by large fires in 2006 and 2009 (fig. 1c). the unit 20d study area (63° 46′ n, 145° 15′ w) varied from agricultural lands and forest with several upland areas that had burned in the last 20 years near delta junction to subalpine scrub in the foothills of the alaska range 50 km south. we sampled browse in southwestern unit 20d over 3250 km2 in 2007 and 2010 (fig. 1d). in unit 20a, antlerless harvests were implemented or expanded to reduce moose abundance during 2004–07 by use of hunt zones and extended seasons, including the central hills and western flats (young and boertje 2011). we subsampled data from the larger unit 20a surveys before and after antlerless harvests to estimate post hoc moose abundance in browse survey areas in the hills and flats (kellie and delong 2006). in southwest unit 20d, antlerless harvest reduced moose abundance in fall 2007 (dubois 2008:397) and fall 2008 (dubois 2010:390). the unit 20d antlerless hunts were short duration prior to substantive snowfall, with female harvest predominantly occurring in the lowland flats north of the foothills (s. dubois, unpublished data) that were accessible by all-terrain vehicles. similar to the abundance estimates for unit 19d (keech et al. 2011, keech 2012), in the other 3 study areas we multiplied estimated abundance by a sightability correction factor (scf) for moose not observed using radiomarked individuals. we used scf = 1.21 for unit 20a (boertje et al. 2009) and scf = 1.1 for unit 20d (s. dubois, unpublished data), and the scf variance was incorporated with that of the geospatial population estimator (gspe; kellie and delong 2006) into 90% confidence limits (goodman 1960, keech et al. 2011). 6 browse indices to moose density – paragi et al. alces vol. 51, 2015 methods browse removal we used moose distribution from fall abundance surveys during shallow (<40 cm) snow as the sampling extent for browse surveys and attempted to minimize sampling bias at the landscape and vegetative stand scales. our landscape sampling design and procedures for selecting plot locations began with pre-selected random points among vegetation strata (seaton 2002) but evolved with logistical experience in the field (paragi et al. 2008:2–4). since 2006 we have primarily used rectangular gspe cells based on 2 minutes of latitude and 5 minutes of longitude (ca. 3.7 × 4.1 km) from recent moose surveys for stratified random sampling at a 3:2 ratio of high:low moose density (e.g., kellie and delong 2006:21). most plot access in remote areas was by helicopter, but we used vehicles where portions of unit 20d were near a highway or forest road (paragi et al. 2008:4). the landscape sampling protocol was developed in boreal forest, but we accommodated the linear nature of riparian browse distribution when we began sampling subalpine habitats. the helicopter flew on the designated course within gspe cells until the first patch of browse ≥0.5 m tall and above snow was encountered, at which point we selected a randomized distance (30–100 m from the nearest safe landing spot) and direction (3 tries to select a site with browse before sampling cell was skipped) for choosing the plot center in the vegetation stand. one exception to stratified sampling with gspe cells was the addition in 2009 of ad hoc systematic plots to ensure adequate sampling of the riparian zone along the kuskokwim river and takotna river in unit 19d for comparison to earlier sampling stratified by vegetation type (paragi et al. 2008). we chose a random starting point along the kuskokwim river near the eastern boundary of the sampling area and landed at the nearest willow bar every 10 km downriver (straight line by helicopter) and also every 10 km upriver on the takotna from its confluence. in 2012 we began defining study areas in unit 20a by polygon boundaries based on drainages that are used for cataloging moose harvest location from hunter reports so that inference about browse could be more directly related to changes in reported harvest. our objective was to sample at least 30 plots per study area with browse above snow to optimize precision and cost (seaton et al. 2011), so we typically selected at least 40 sample cells because some random plot locations near helicopter landing spots do not contain browse. this sample size was not achieved in unit 20a western flats in 2006, where we omitted several sites due to absence of browse near safe landing zones. however, the 15 plots achieved in this survey (table 1) are expected to accurately reflect the biomass removal level but with potentially high variance (seaton et al. 2011, fig. 3). where clumps of randomly chosen sample cells occur, sampling at least one cell in the clump provided landscape coverage if logistics became limiting (e.g., degraded flying weather or distance from fuel). we analyzed proportional oftk over the winter to describe the interaction between moose and their winter forage. the rationale for plot sampling and browse metric analysis is described elsewhere (seaton 2002, paragi et al. 2008). snow depth >70 cm can restrict access to forage, increase energetic requirements for locomotion, and influence habitat selection; snow depth >90 cm greatly restricts movement, potentially hindering adequate forage intake (coady 1974). consequently, we recorded snow depth at plots during browse surveys for a context of winter severity, particularly as a confounding factor between sampling events. we sampled only plants above snow alces vol. 51, 2015 paragi et al. – browse indices to moose density 7 with measurable cag between 0.5 and 3.0 m above ground level in a 15-m radius plot near the end of browse removal in late winter (late march or early april, before leaf emergence). we randomly selected 3 plants per species present, using plants as the sample unit for inference on browse removal at the scale of study area. plant taxonomy followed collet (2004) for willows and viereck and little (2007) for other species, with winter willow identification aided by an unpublished guide (d. simpson, alaska department of fish and game [adfg] 1986). for each randomly selected plant within a species, we randomly selected 10 twigs. for each twig we recorded to 0.1 mm precision the dpb if applicable, and cag (lyon 1970). we then counted the total number of twigs with cag on each of the 3 plants. we used the regression coefficients relating diameter to dry mass (paragi et al. 2008:40–41) and the number of twigs with cag per plant to estimate prod and oftk (telfer 1969). an exception was salix lasiandra for which 3 plants were measured on each of 2 plots in unit 20a western flats in 2006. we did not have a regression equation for s. lasiandra, so we used s. bebbiana equations for biomass analysis because these twigs have a similar morphology. we estimated oftk based on sampled twigs only (mean twig per sampled plant) with plants as the sample unit. we extrapolated prod and oftk from sampled twigs to the plot level for comparison among study areas, recognizing that this may introduce sampling bias through variation in the proportion of total plants sampled per species and variability in plant counts within plots. we used software written in r language (r development core team 2008, version 2.1.1; code and instructions available under project 5.10 at <http://www.adfg.alaska.gov/ index.cfm?adfg=librarypublications.wildlife research#habitat> to read a microsoft® access® (version 2003) database containing plot counts, twig diameters, diameter– biomass pairs, and dry-weight conversions. we used this software to estimate the diameter–biomass relationships, prod, and oftk on the basis of plant, species, plot, and study area (paragi et al. 2008). we applied binomial 95% confidence limits (cochran 1977:58) with n as the number of plants measured, rather than twigs, to avoid table 1. sampling details and estimates of apparent browse production (sampled twigs above snow extrapolated to plot composition) by study area within game management units in interior alaska. game management unit browse sampling year sampling area (km2) browse samples (n) apparent production (kg / ha) plots plants twigs x ̄ 95% ci 19d 2001 10,600 36 251 2420 201 19 19d 2003 1368 39 298 2377 689 52 19d 2009 2896 42 278 2746 343 26 20a c. hills 2000 600 49 235 2504 745 154 20a c. hills 2012 790 37 177 1799 30 3.8 20a w. flats 2006 1100 15 109 1099 75 9.0 20a w. flats 2012 1625 44 312 2945 14 1.0 20d 2007 3250 75 437 4312 52 4.7 20d 2010 3250 57 431 4108 73 8.0 8 browse indices to moose density – paragi et al. alces vol. 51, 2015 http://www.adfg.alaska.gov/index.cfm?adfg=librarypublications.wildliferesearch#habitat http://www.adfg.alaska.gov/index.cfm?adfg=librarypublications.wildliferesearch#habitat http://www.adfg.alaska.gov/index.cfm?adfg=librarypublications.wildliferesearch#habitat pseudo-replication (unequal proportion of plants sampled per species and per plot) and to portray a more conservative variation. where moose density changed between browse evaluation periods, we evaluated significant probability of increase or decrease in twig metrics with 1-tailed tests, where direction of change associated with removal was predicted to be the same as direction of change in moose abundance. we tested for difference between proportions of oftk and of ptb using a z-test (zar 1984:396) with the smaller number of plants for degrees of freedom in the t distribution. we tested for difference in snow depth and in dpb before and after management actions for each browse species using mann-whitney u (conover 1980) because data distributions were often non-normal (lilliefor’s test, p < 0.05). browse architecture seaton (2002:19) classified forage plants based on their history of browsing by moose and the resulting compensatory growth, termed “architecture.” in contrast to removal of cag in a specific winter, architecture describes multi-year growth history and is unidirectional (unbrowsed → browsed → broomed) unless disturbance resets plants to an unbrowsed canopy. thus, in addition to historic moose density, architecture in study areas is influenced by fire or vegetation management in prior years that stimulates young growth. three categories of plant architecture were defined from evidence of browsing prior to the current year for each plant: “unbrowsed” (no evidence of browsing prior to the current year); “browsed” (browsing in past years but <50% cag twigs between 0.5 and 3.0 m arose from lateral stems that were produced as a result of browsing); and “broomed” (>50% of cag twigs between 0.5 and 3.0 m arose as lateral stems). to reduce measurement error, architecture was classified by the first 3 authors or under their direct supervision. we used a chi-square test for independence in proportions of the plant architecture classes and portrayed variation in the proportions of plants in architecture classes with binomial confidence limits using n as the number of plants. twinning rate boertje et al. (2007) described estimation of moose twinning rates from aerial surveys shortly after peak of calving in late may. we obtained data from area or research biologists that conducted surveys annually in our 4 study areas. to evaluate trend in twinning rate, we used r script to estimate the mean rate and 95% confidence limits using a parametric bootstrap (100,000 repetitions). results moose density in the unit 20a central hills was 2.4 moose/km2 prior to antlerless harvest and was reduced 33% by the 2nd browse survey (table 2). the 2003 abundance estimate likely represented the peak density 3 years after the first browse survey in 2000 as inferred from abundance estimates in the larger unit 20a (young and boertje 2011); thus, moose density during the first browse survey may have been slightly lower. the 2012 abundance estimate was 5 years after the end of liberal antlerless harvest, a period of reduced and comparatively stable moose abundance in all of unit 20a (young 2012, table 2). in unit 20a western flats, the evidence for a 31% decline in density from 1.2 moose/km2 in 2006 was weaker because the estimate had twice the proportional variance of the other study areas (table 2). the 2006 browse survey occurred after 2 years of liberal antlerless harvest, thus likely reflected a reduced moose density (lesser expected difference in browse removal) from the peak in fall 2003 for all of unit 20a (young and boertje 2011). this population may have experienced a relatively smaller change in alces vol. 51, 2015 paragi et al. – browse indices to moose density 9 abundance than the other 3 study areas, so we expected that the magnitude of change in browse metrics and twinning rate might be ambiguous with respect to the other 3 study areas. moose density in unit 20d was 2.1 moose/km2 before antlerless harvest; and was reduced 41% by 2010. willows (salix spp.) dominated or codominated apparent browse production in most instances (fig. 2). s. alaxensis composed the majority of browse biomass in active riparian floodplains regardless of elevation (e.g., including incised drainages in subalpline), whereas s. pulchra often dominated or co-dominated production with betula neoalaskana or populus tremuloides in upland sites, particularly after recent fires or logging. prod ranged greatly among study areas and within study areas between years (14–745 kg/ha; table 1). within a study area, the greatest change was in unit 20a central hills that was possibly influenced by different browse sampling stratifications before and after antlerless harvest. sampling in the 2000 browse survey in the 20a central hills included 1 plot of extremely high production (22,148 kg/ha; paragi et al. 2008:53) that boosted mean apparent production from 329 to 745 kg/ha. lower prod in post-treatment browse surveys for units 19d and 20a western flats may also reflect slow vegetative recovery from recent fires (fig. 2a and 2c). oftk exceeded 45% for dominant willow species at higher moose densities within study areas (fig. 2): s. alaxensis in unit 19d (2009) and unit 20a central hills (2000), and s. pulchra in unit 20a central hills (2000) and unit 20a western flats (2006). mean snow depth (7–33 cm) during browse surveys differed little between table 2. estimates of moose density and browse removal (sampled twigs only) by moose reported by study area within game management units in interior alaska. moose abundance surveys were in early winter prior to the associated browse surveys unless otherwise noted. proportions of offtake and twigs browsed were predicted to change in the direction of trend in moose density. game management unit browse survey year moose density (no./km2) proportional browse offtake proportion of twigs browsed x ̄ 90% cl x ̄ 95% cla prop. 95% ci 19d 2001 0.30 0.25, 0.35 0.159 0.112, 0.195 0.110 0.039 19d 2003 0.30b 0.25, 0.35 0.171 0.144, 0.221 0.287 0.051 19d 2009 0.62 0.48, 0.72 0.405 0.332, 0.471 0.346 0.056 20a central hills 2000 2.36c 1.84, 2.87 0.433 0.394, 0.462 0.401 0.063 20a central hills 2012 1.59 1.32, 1.86 0.303 0.227, 0.357 0.130 0.050 20a western flats 2006 1.19 0.71, 1.68 0.307 0.112, 0.442 0.161 0.069 20a western flats 2012 0.82 0.50, 1.14 0.190 0.147, 0.228 0.119 0.036 20d 2007 2.11 1.63, 2.60 0.253 0.191, 0.323 0.167 0.035 20d 2010 1.24 1.00, 1.48 0.153 0.106, 0.199 0.117 0.030 abootstrapped confidence intervals may be asymmetrical. bdensity in fall 2003 assumed to be similar to that from survey in fall 2001 (keech 2012:13–14). csurvey in fall 2003; this was likely the period of maximum abundance prior to liberal antlerless harvest (young and boertje 2011, fig. 2) and the closest period with enough abundance sample units in the browse study area to permit a post hoc analysis. for comparison, the estimate with visibility correction for all of unit 20a (12,900 km2 of moose habitat <1350 m elevation) was 1.05 moose/km2 in 1999 (prior to 1st browse survey); 1.37 moose/km2 in 2003 (prior to liberal antlerless harvest); 0.98 moose/km2 in 2008 (after liberal antlerless harvest), and 0.98 moose/ km2 in 2011 (prior to 2nd browse survey) (young 2012, table 2). 10 browse indices to moose density – paragi et al. alces vol. 51, 2015 sampling periods within study areas and remained <70 cm by late winter in 3 of 4 study areas. unit 19d was the exception where mean snow depth for the winter post-treatment (107 cm) was 56 cm higher than pre-treatment (mann-whitney u = 18, p < 0.001). oftk was more consistent and precise than ptb in reflecting direction and magnitude of changes in moose density, or lack thereof. change in oftk from before to after management actions (table 2) was significant in all 4 areas and in the expected direction of change in moose density (z ≥ 2.7, p < 0.01, 1-tailed). change in ptb (table 2) was significant for unit 20a central hills (z = 6.0, p < 0.0005) and unit 20d (z = 2.1, 0.01 < p < 0.025), but not for unit 19d (z = 1.5, 0.05 < p < 0.1) or unit 20a western flats (z = 1.1, 0.2 < p < 0.5). whereas the pre-treatment moose density was similar between the 2001 and 2003 browse surveys in unit 19d (0.38 and 0.41 moose/km2 in the smaller 2003 browse study area; keech 2012:13), the lack of difference in oftk (z = 0.35, p > 0.5) was in marked contrast with the unexpected difference in ptb (z = 6.2, p < 0.001, 2-tailed; table 2). 0 100 200 300 400 k g/ ha unit 19d (a) (b) (c) (d) prod01 oftk01 prod03 oftk03 prod09 oftk09 prod00 oftk00 prod12 oftk12 prod06 oftk06 prod12 oftk12 prod07 oftk07 prod10 oftk10 0 100 200 300 400 500 600 700 k g/ ha unit 20a c hills 0 10 20 30 40 50 60 k g/ ha unit 20a w flats 0 10 20 30 k g/ ha unit 20d fig. 2. apparent production (prod) and offtake (oftk; both in kg/ha) as extrapolated from sampled twigs to plot composition in 4 study areas within interior alaska. moose abundance increased (unit 19d) or decreased (all others) coincident with intended outcome of management actions. years of prod and oftk are represented by last 2 digits of year starting with 2000. note differences in y-axis scale among areas; error bars are 95% confidence limits. an additional year of pre-treatment data (2001) was available for comparison in unit 19d. species codes: bene (betula neoalaskana; formerly b. papyrifera), cost (cornus stolonifera), poba (populus balsamifera), potr (p. tremuloides), saal (salix alaxensis), saar (s. arbusculoides), sabe (s. bebbiana), sagl (s. glauca), sain (s. interior), sapu (s. pulchra), and sari (s. richardsonii). alces vol. 51, 2015 paragi et al. – browse indices to moose density 11 further, ptb was highly variable for various degrees of oftk whether a moose population increased or decreased (fig. 3). direction of species-level change in dpb generally corresponded with the direction of change in both oftk and ptb. the strongest correspondence existed for those species composing the dominant (or codominant) biomass in a study area (table 3). among the 4 study areas, proportional magnitude of change in dpb (p < 0.05 by species) was 11–35% (x̄ = 22%, n = 6) in the predicted direction for each study area. the relative importance of dpb as a component of oftk is evident in scaling among browse metrics during the 100% increase in moose in unit 19d; 138% increase in oftk corresponded to increases of 20% in ptb (all species combined) and 16–42% in dpb of the primary browse species. where moose density decreased 31–41% in the other 3 study areas, decreases were documented in oftk (30% in unit 20a central hills to 40% in unit 20d), ptb (26% in unit 20a western flats to 68% in unit 20a central hills), and dpb of the primary browse species (11–37%, both extremes in unit 20d). plant architecture also responded to changes in moose density. there was a 44% increase in broomed plants following a moose population increase in unit 19d (fig. 4). in the absence of vegetative disturbance such as fire, the proportion of unbrowsed plants increased 6-fold (2% to 16%) in the unit 20a central hills after the moose population declined from the highest density in our 4 study areas; this was also the longest period between end of liberal antlerless harvest (presumed greatest point of density reduction; table 2) and the post-treatment browse survey (5 years later). there was an 83% increase in proportion of unbrowsed plants in unit 20a western flats following a reduction in moose density, where fires created new unbrowsed plants during and after antlerless hunts. however, we found no change in plant architecture in unit 20d (fig. 4) despite a 41% decline in moose density over 3 years. when compared with the 4 browse metrics, twinning rate showed no changes or trend in these moose populations during the intervening period of reduced moose density between browse surveys or soon thereafter (fig. 5). (a) (b) 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 devo mer ssa moibfo noitr oporp propor�on of twigs browsed unit 19d 2003 2009 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 devo mer ssa moibfo noitroporp propor�on of twigs browsed unit 20a c hills 2000 2012 fig. 3. examples of greater range in proportion of twigs browsed for a given range of offtake where moose populations had increased (a) and decreased (b) in interior alaska. the dashed lines illustrate a 1:1 relationship for comparison to the yearspecific plot data. 12 browse indices to moose density – paragi et al. alces vol. 51, 2015 discussion proportional oftk was a more comprehensive metric than ptb or dpb for detecting short-term, landscape-level changes in intraspecific competition for winter forage and potential effects on plants following management actions intended to affect moose density. proportional oftk consistently reflected change in moose density despite substantial variation in prod (total and among species) before and after management actions. we infer this relationship of oftk and density change as evidence that proportional oftk is unbiased, likely because moose distribution reflects browse distribution and oftk reflects available prod. our estimates of prod and oftk were complicated by 4 factors: 1) a change in landscape sampling design between preand post-treatment for unit 19d and unit 20a central hills, 2) by differences in size of some preand post-treatment study areas because of changing management issues, 3) by measuring a relatively limited number of plots over large diverse landscapes that met precision objectives for proportional oftk (seaton et al. 2011) but increased chance of sampling error, and 4) by measuring a limited number of forage plants without regard to nutrition or digestibility. we attribute the relatively high variation in the relationship between oftk and ptb (fig. 3) to condensing species with different twig diameters to a simple count of browsed table 3. change in mean diameter (mm) at point of browsing (dpb) by moose on the primary winter forage species in areas where moose populations increased (unit 19d, 2003–09) or decreased (all others) in accordance with intended outcome of management actions, interior alaska. dpb was predicted to change in the direction of trend in moose abundance. bold text represents dominant species (>50% of total estimated apparent production extrapolated from sampled plants to plot). trend in dpb (positive or negative) is inferred from the mann-whitney u statistic (p < 0.05). an additional year of pre-treatment data (2001) in unit 19d is shown for comparison. species codes: bene (betula neoalaskana), saal (salix alaxensis), sabe (s. bebbiana), and sapu (s. pulchra). area year bene saal sabe sapu unit 19d 2001 2.7 3.9 3.0 2.1 unit 19d 2003 2.8 4.3 3.4 3.2 unit 19d 2009 3.0 5.0 3.2 3.1 trend none pos none none p 0.43 <0.001 0.11 0.7 unit 20a central hills 2000 2.9 4.3 3.3 3.2 unit 20a central hills 2012 2.8 4.7 3.2 2.5 trend none none none neg p 0.67 0.28 0.2 <0.001 unit 20a western flats 2006 3.1 5.2 2.8 3.2 unit 20a western flats 2012 2.8 3.4 2.6 2.2 trend none neg none neg p 0.06 0.03 0.25 <0.001 unit 20d 2007 2.5 4.5 2.4 2.8 unit 20d 2010 2.5 3.8 2.6 2.5 trend none neg none neg p 0.58 <0.001 0.32 <0.001 alces vol. 51, 2015 paragi et al. – browse indices to moose density 13 twigs of all species combined. the effect of seemingly small change in dpb is relatively more important than changes in ptb in explaining change in oftk because the species-specific mass-diameter relationship is based on non-linear twig geometry that 296 277 235 177 109 310 431 295 0 0.2 0.4 0.6 0.8 1 19d 2003 19d 2009 20a c hills 2000 20a c hills 2012 20a w flats 2006 20a w flats 2012 20d 2007 20d 2010 unbrowsed browsed broomed χ2 = 1.8, p = 0.40 χ2 = 6.2, p = 0.045 χ2 = 26.1, p <0.001 χ2 = 29.9, p <0.001 fig. 4. the proportional changes in categories of browse plant architecture where the moose populations increased (unit 19d) or decreased (all others) coincident with the intended outcome of management actions in interior alaska. the binomial confidence interval (95%) and sample size are shown above bars. (b) 0.00 0.10 0.20 0.30 0.40 1998 2002 2006 2010 tw in ni ng ra te unit 20a c hills(a) 0.00 0.20 0.40 0.60 0.80 2001 2003 2005 2007 2009 2011 2013 tw in ni ng ra te unit 19d (d)(c) 0.00 0.10 0.20 0.30 0.40 2005 2008 2011 tw in ni ng ra te unit 20a w flats 0.00 0.10 0.20 0.30 0.40 2005 2008 2011 tw in ni ng ra te unit 20d fig. 5. moose twinning rates in late spring for (a) 1 moose population that increased and (b-d) 3 populations that decreased coincident with intended outcome of management actions in interior alaska. open circles indicate years of browse surveys; confidence intervals were 95% bootstrapped. 14 browse indices to moose density – paragi et al. alces vol. 51, 2015 does not scale equally with the linear proportion of twigs browsed. our data may be the first to report how changes in moose foraging behavior, as shown by changes in stem cropping diameter (dpb), is influenced by moose density across a broad geographic range. these observational and experimental data strongly suggest that variation in dpb reflects variation in competition which translates into variation in demographic processes (e.g., twinning rates). however, dpb of several species is not readily condensed to a single numeric score, and using dpb in isolation could complicate comparisons of browse removal before and after management actions if preferences for forage species change over time. oftk incorporates both ptb and dpb, thus reduces potential confounding of either component alone. over relatively short periods of change in moose density (2 years in unit 20d to 6 years in unit 19d), we expected change in architecture reflecting plant life histories to lag behind (or be of lesser magnitude) than change in the 3 metrics based on cag and dpb. however, magnitude of changes in architecture reflecting increases in moose density (more broomed plants) or decreases in moose density (more unbrowsed plants) were often of equal magnitude to change in cag and dpb metrics in as few as 5 years after change in density (unit 20a western flats). this rate of change is comparable to recovery of broomed willows after elk (cervus elaphus) reductions in wyoming (singer and zeigenfuss 2003:80–81). the lack of architectural changes in unit 20d despite a 41% decline in moose density may be explained by mismatch in scale of abundance surveys and distribution of antlerless harvest, where the latter primarily occurred in the flat and relatively accessible northern portion of the study area (s. dubois, unpubl. data). we had stratified unit 20d for browse surveys into flats and hills (sampling design before liberal antlerless harvest described in paragi et al. [2008]) and found no change in browse plant architecture in the flats. however, we noted a significant increase in brooming in the hills (t. paragi, unpublished data), indicating the continued effect of high browsing pressure where moose density had not been reduced. this experience highlighted the importance of scaling the browse sampling appropriately to the extent of abundance estimates and management actions. browse removal rate at the moose population level is not an absolute measure of carrying capacity for a given moose density, nor a meaningful demographic parameter linked in real time to changes in birth and death rates. browse removal rate is best used in concert with nutritional indices such as twinning rate that are also demographic parameters conveying population status relative to carrying capacity. singer and zeigenfuss (2003:67–70) studied consumption of willows as a product of percent leader use and percent twig use and found that moose density alone had little value in describing willow consumption where moose (x̄ = 1.9/km2) shared winter range with elk (x̄ = 16.3/km2) in wyoming. similarly, månsson (2009) found no positive relationship between betula spp. biomass removal and moose density where biomass removal of the dominant browse species, scots pine (pinus sylvestris), was related to moose density in sweden. however, we found that oftk of all browse species combined correlated with the direction and magnitude of moose density change. the highest biomass removal for all species combined approached 45% of cag where moose density was highest (unit 20a central hills). we have focused on how browse removal relates to animal condition, but high levels of offtake warrant consideration of sustainable plant health or productivity as another trigger for habitat enhancement or prudent reduction in herbivore density. alces vol. 51, 2015 paragi et al. – browse indices to moose density 15 low to moderate levels of browse removal can stimulate browse production (suter 1992), but removal beyond a threshold causes decline in production (danell et al. 1985, persson et al. 2005a). persson et al. (2007) found that production response of betula spp. to simulated browsing in northern sweden was highest at 25–40% biomass removal (representing ca. 3 moose/km2 on the winter range) on sites with moderate to high soil nutrients but lower on sites with low soil nutrients. singer and zeigenfuss (2003:70) observed a range (0–47%) in willow removal among study sites, with the highest growth response occurring at moderate (ca. 21%) consumption levels. they surmised that repeated consumption >30% is likely detrimental to plants and >45% removal is exceptionally high. we observed proportional oftk >45% for browse species (fig. 2) when moose in units 20a and 20d were at relatively high density (>2/km2). the poor nutritional condition indicated by low twinning rate of these populations (fig. 5; also boertje et al. 2007) suggests that such levels of browsing intensity have a negative effect on browse production. despite uncertainty in the threshold of sustainable browse removal, the recent antlerless harvests in units 20a and 20d intended to reduce relatively high moose densities and negative effects on the winter forage base seemed prudent regardless of whether proportional oftk is considered at the level of study area or plant species. we acknowledge that factors independent of change in moose density may influence estimates of browse metrics. we caution that our estimates of “apparent” production were for twigs above late winter snow depth in a limited number of plots, rather than a rigorous landscape estimate of total biomass production. the actual biomass available to moose might vary over time independent of moose abundance because of plant density in response to disturbance, plant structural change since disturbance, compensatory growth over a life history of exposure to browsing, growing season limits to cag (e.g., drought or insect defoliation), and competing herbivory by hares. diet quality can multiply effects on intake rate to produce greater removal on better quality forages (white 1983, mcart et al. 2009). study areas should remain large enough so that browse sampling allows inference at the population level and reduces annual variation in estimates of production and removal inherent at smaller scales (ahlén 1975:111, 134–135, 165; mackie 1976, månsson et al. 2007). managers wishing to monitor plant health should consider sampling designs focused on total production during the snow-free period and stratifying by vegetation type in addition to moose density. finally, we urge others to replicate our evaluation of management actions in a more rigorous and balanced experimental design, recognizing the risk that public regulatory bodies can interrupt the duration or type of management “treatment” once baseline data are collected. our only case study of deep snow (107 cm in unit 19d, 2009) was also our only case study of population increase. we do not know if unit 19d would have had removal levels equally as high in 2009 had snow depth been <90 cm. deep snow likely contributed to the increased 2009 browse removal in unit 19d by concentrating moose and exacerbating the effect of increased density (k. kellie, unpublished data). testa (2001) noted that the proportional number of s. alaxensis twigs browsed on 2 important wintering areas in the nelchina basin of southcentral alaska was 60–82% during winters with snow >70 cm compared with 12–35% during winters <70 cm. collins (2002:11) further defined a positive relationship between snow depth and biomass removal for this population. repeating surveys among winters of markedly different 16 browse indices to moose density – paragi et al. alces vol. 51, 2015 snow depth for a relatively stable moose population would be instructive as to the effect of snow-depth on spatial distribution and degree of browse removal. twinning rates did not respond in an expected fashion during the intervening period between the end of management actions causing moose density to change and subsequent browse surveys (2–6 yr); presumably, immediate consequences to reproduction were small. we surmise that increased body weight of the youngest, non-reproductive cohorts may be a better short-term index to improved nutrition following reduced intraspecific competition. lag in nutritional condition of wild ruminants has been documented following density reduction (blood 1974, albon et al. 1987, boertje et al. 2007). reduced twinning rate following prolonged high density might persist until enough more robust female calves born during periods of lower food competition enter the breeding population and affect the birth rate (e.g., solberg et al. 2004). in 3 of our 4 study areas, estimates of apparent production also decreased, possibly reflecting a decline in per capita forage that could dampen reproductive responses (solberg et al. 2012). our relevant case study in interior alaska occurred in unit 20a moose, where twinning rate took 12 years to increase despite a dramatic decline from peak abundance in the 1960s (fig. 5 in boertje et al. 2007). the lower twinning rate in the 1990s in unit 20a, despite lower moose density than in the 1960s, may have been evidence of degraded range capacity from having moose at prolonged high density in the 1960s and less extensive wildfires after the 1960s (less forage per capita). unfortunately, we do not have historic data on forage abundance or quality to evaluate this speculation. time lag in reproductive response may differ among periods, or among populations, in part due to differences in vegetation recovery rate (sand et al. 1996:242) that is potentially influenced by negative effects on soil fertility from prolonged high biomass removal (persson et al. 2005b). monitoring systems that quantify densitydependent responses of ungulates to their habitat as a correlate of population density have existed for decades (e.g., aldous 1945), but few jurisdictions manage ungulate abundance primarily based on monitoring indices of habitat (e.g., keigley and fager 2006) or nutrition. dubois (2008:388) first proposed use of twinning rate thresholds developed by boertje et al. (2007) rather than a population objective to recommend management of moose population trend in unit 20d. our study demonstrated the value of browse biomass offtake for corroborating intended reductions in intraspecific competition and gauging relationship of moose abundance to carrying capacity at higher densities. offtake and twinning rate provide managers with objective means to recommend and monitor effectiveness of forage enhancement, or timely reduction in moose density through harvest across age and sex classes to reduce forage competition and avoid prolonged negative effects of high density (boertje et al. 2007, young and boertje 2011). we urge managers and public regulatory bodies to utilize an empirical monitoring and decision framework for moose population management that incorporates measures of plant and animal condition in addition to population objectives. when reporting metrics on forage or animal condition, managers need to clearly identify that maintaining higher moose densities incurs an increased risk of strong negative feedback after severe winters (gasaway et al. 1983, boertje et al. 2009). when the public desires high moose densities, we encourage managers to discuss the risks associated with various management options and acceptable means for achieving proposed harvest alces vol. 51, 2015 paragi et al. – browse indices to moose density 17 objectives (adfg 2011, young and boertje 2011). acknowledgements our work was funded by federal aid in wildlife restoration and the alaska legislature. several adfg employees (particularly l. parrett), other agency personnel, and volunteers helped with field work. c. maurer, q. slade, a. shapiro, r. swisher, and m. terwilliger safely and efficiently piloted helicopters. b. taras estimated bootstrap confidence limits for twinning rates and provided formulas for incorporating visual detection variance on moose population estimates. we thank j. benson, s. brainerd, t. hanley, and d. james for constructive comments. references adfg (alaska department of fish and game). 2011. intensive management protocol. division of wildlife conservation. juneau, alaska, usa. ahlén, i. 1975. winter habitats of moose and deer in relation to land use in scandinavia. viltrevy 9: 45–192. albon, s. d., t. h. clutton-brock, and f. e. guinness. 1987. early development and population dynamics in red deer. ii. density-independent effects and cohort variation. journal of animal ecology 56: 69–81. aldous, c. f. 1945. a winter study of mule deer in nevada. journal of wildlife management 9: 145–151. doi: 10.2307/ 3795893. blood, d. a. 1974. variation in reproduction and productivity of an enclosed herd of moose (alces alces). international congress of game biologists 11: 59–66. boertje, r. d., m. a. keech, d. d. young, k. a. kellie, and c. t. seaton. 2009. managing for elevated yield of moose in interior alaska. journal of wildlife management 73: 314–327. doi: 10.2193/ 2007-591. ———, k. a. kellie, c. t. seaton, m. a. keech, d. d. young, b. w. dale, l. g. adams, and a. r. aderman. 2007. alaska moose nutrition: signals to begin liberal antlerless harvests. journal of wildlife management 71: 1494–1506. doi: 10.2193/2006-159. cederlund, g. n., h. k. g. sand, and a. pehrson. 1991. body mass dynamics of moose calves in relation to winter severity. journal of wildlife management 55: 675–681. doi: 10.2307/38 09517. coady, j. w. 1974. influence of snow on behavior of moose. naturaliste canadien 101: 417–436. cochran, w. g. 1977. sampling techniques. third edition. john wiley and sons, new york, ny, usa. collet, d. m. 2004. willows of interior alaska. u.s. fish and wildlife service, yukon flats national wildlilfe refuge, fairbanks, alaska, usa. collins, w. b. 2002. interrelationship of forage and moose in game management unit 13. federal aid in wildlfe restoration, research final performance report, grants w-24-4 to w-27-4, study 1.50. alaska department of fish and game, juneau, alaska, usa. conover, w. j. 1980. practical nonparametric statistics. second edition. john wiley and sons, new york, ny, usa. crete, m. 1989. approximation of k carrying capacity for moose in eastern quebec. canadian journal of zoology 67: 373– 380. doi: 10.1139/z89-055. danell, k., k. huss-danell, and r. bergström. 1985. interactions between browsing moose and two species of birch in sweden. ecology 66: 1867–1878. doi: 10.2307/2937382. dubois, s. d. 2008. unit 20d moose. pages 386–423 in p. harper, editor. moose management report of survey and inventory activities, 1 july 2005 through 30 june 2007. project 1.0. alaska 18 browse indices to moose density – paragi et al. alces vol. 51, 2015 department of fish and game, juneau, alaska, usa. ———. 2010. unit 20d moose. pages 380–409 in p. harper, editor. moose management report of survey and inventory activities, 1 july 2007 through 30 june 2009. project 1.0. alaska department of fish and game, juneau, alaska, usa. franzmann, a. w., and c. c. schwartz. 1985. moose twinning rates: a possible population condition assessment. journal of wildlife management 49: 394–396. doi: 10.2307/3801540. gasaway, w. c., r. d. boertje, d. v. grangaard, d. g. kelleyhouse, r. o. stephenson, and d. g. larsen. 1992. the role of predation in limiting moose at low densities in alaska and yukon and implications for conservation. wildlife monographs 120. ———, and j. w. coady 1974. review of energy requirements and rumen fermentation in moose and other ruminants. naturaliste canadien 101: 227–262. ———, r. o. stephenson, j. l. davis, p. e. k. shepherd, and o. e. burris. 1983. interrelationships of wolves, prey, and man in interior alaska. wildlife monographs 84. goodman, l. a. 1960. on the exact variance of products. journal of the american statistical association 55: 708–713. doi: 10.1080/01621459.1960.10483369. keech, m. a. 2012. response of moose and their predators to wolf reduction and short-term bear removal in a portion of unit 19d. federal aid in wildlife restoration, final wildlife research report. project 1.62. adf&g/dwc/ wrr-2012-7, grants w-33-4 to w-3310. alaska department of fish and game, fairbanks, alaska, usa. ———, r. t. bowyer, j. m. ver hoef, r. d. boertje, b. w. dale, and t. r. stephenson. 2000. life-history consequences of maternal condition in alaskan moose. journal of wildlife management 64: 450–462. doi: 10.2307/3803243. ———, m. s. lindberg, r. d. boertje, p. valkenburg, b. d. taras, t. a. boudreau, and k. b. beckmen. 2011. effects of predator treatments, individual traits, and environment on moose survival in alaska. journal of wildlife management 75: 1361–1380. doi: 10.1002/jwmg.188. keigley, r. b., and c. w. fager. 2006. habitat-based adaptive management at mount haggin wildlife management area. alces 42: 49–54. kellie, k. a., and r. a. delong. 2006. geospatial survey operations manual. alaska department of fish and game, fairbanks, alaska, usa. kielland, k., and t. osborne. 1998. moose browsing on feltleaf willow: optimal foraging in relation to plant morphology and chemistry. alces 34: 149–155. lyon, j. l. 1970. lengthand weight-diameter relations of serviceberry twigs. journal of wildlife management 34: 456–460. doi: 10.2307/3799033. maccracken, j. g., v. van ballenberghe, and j. m. peek. 1997. habitat relationships of moose on the copper river delta in coastal south-central alaska. wildlife monographs 136. ———, and l. a. viereck 1990. browse regrowth and use by moose after fire in interior alaska. northwest science 64: 11–18. mackie, r. j. 1976. evaluation of range survey methods, concepts, and criteria (effectiveness of the key browse survey method). job progress report. federal aid project w-120-r-6, study 28.01, job bg-2.01, 1 july 1974-30 june 1975. montana department of fish and game, helena, montana, usa. månsson, j. 2009. environmental variation and moose alces alces density as determinants of spatio-temporal heterogeneity in browsing. ecography 32: 601–612. doi: 10.1139/z07-015. alces vol. 51, 2015 paragi et al. – browse indices to moose density 19 ———, h. andren, a. pehrson, and r. bergström. 2007. moose browsing and forage availability: a scale-dependent relationship? canadian journal of zoology 85: 372–380. mcart, s. h., d. e. spalinger, w. b. collins, e. r. schoen, t. stevenson, and m. bucho. 2009. summer dietary nitrogen availability as a potential bottom-up constraint on moose in south-central alaska. ecology 90: 1400–1411. doi: 19537559doi: 10.1890/08-1435.1. mccullough, d. r. 1984. lessons from the george reserve, michigan. pages 211– 242 in l. k. halls, editor. white-tailed deer ecology and management. wildlife management institute, washington, dc, usa. paragi, t. f., c. t. seaton, and k. a. kellie. 2008. identifying and evaluating techniques for wildlife habitat management in interior alaska: moose range assessment. federal aid in wildlife restoration, project 5.10. final research technical report, grants w-33-4 through w-33-7. alaska department of fish and game, juneau, alaska, usa. persson, i.-l., r. bergström, and k. danell. 2007. browse biomass production and regrowth capacity after biomass loss in deciduous and coniferous trees: response to moose browsing along a productivity gradient. oikos 116: 1639–1650. doi: 10.1111/j.0030-1299.2007.15946.x. ———, k. danell, and r. bergström. 2005a. different moose densities and accompanied changes in tree morphology and browse production. ecological applications 15: 1296–1305. doi: 10.1890/04-0499. ———, ———, j. pastor, j. danell, and r. bergström. 2005b. impact of moose population density on the production and composition of litter in boreal forests. oikos 108: 297–306. doi: 10.1111/ j.0030-1299.2005.13844.x. regelin, w. l., c. c schwartz, and a. w. franzmann. 1987. effects of forest succession on nutritional dynamics of moose forage. swedish wildlife research supplement 1: 247–262. sand, h., r. bergström, g. cederlund, m. östergren, and f. stålfelt. 1996. density-dependent variation in reproduction and body mass in female moose alces alces. wildlife biology 2: 233–245. schwartz, c. c., and l. a. renecker. 1997. nutrition and energetics. pages 441–478 in a. w. franzmann and c. c. schwartz, editors. ecology and management of the north american moose. smithsonian institution press, washington, dc, usa. seaton, c. t. 2002. winter foraging ecology of moose in the tanana flats and alaska range foothills. m.s. thesis, university of alaska-fairbanks, fairbanks, alaska, usa. ———, t. f. paragi, r. d. boertje, k. kielland, s. dubois, and c. l. fleener. 2011. browse biomass removal and nutritional condition of moose, alces alces. wildlife biology 17: 55–66. doi: 10.2981/10-010. shipley, l. a., and d. e. spalinger. 1992. mechanics of browsing in dense food patches: effects of plant and animal morphology on intake rate. canadian journal of zoology 70: 1743–1752. doi: 10.1139/ z92-242. singer, f. j., and l. c. zeigenfuss. 2003. part ii. a survey of willow communities, willow stature and production, and correlations to ungulate consumption and density in the jackson valley and the national elk refuge. pages 58–86 in l. c. zeigenfuss and f. j. singer, editors. ecology of native ungulates in the jackson valley: habitat selection, interactions with domestic livestock, and effects of herbivory on grassland and willow communities. final report. natural resource preservation progress project #00-03, and interagency agreement nos. 1460-0013 and 1460-01-005 20 browse indices to moose density – paragi et al. alces vol. 51, 2015 between grand teton national park and the u.s. geolgical survey. united sates geological survey, fort collins, colorado, usa. solberg, e. j., a. loison, j.-m. gaillard, and m. heim. 2004. lasting effects of conditions at birth on moose body mass. ecography 27: 677–687. doi: 10.1111/ j.0906-7590.2004.03864.x. ———, o. strand, v. veiberg, r. andersen, m. heim, c. m. rolandsen, r. langvatn, f. holmstrøm, m. i. solem, r. eriksen, r. astrup, and m. ueno. 2012. moose, red deer and reindeer – results from the monitoring program for wild cervids, 1991–2011 (abstract in english). nina report 885. nina, trondheim, norway. stout, g. w. 2010. unit 24b moose. pages 572–610 in p. harper, editor. moose management report of survey and inventory activities, project 1.0, 1 july 2007–30 june 2009. alaska department of fish and game, juneau, alaska, usa. suter, s. m. 1992. the morphology and chemistry of two willow species in relation to moose winter browsing. m.s. thesis, university of alaska-fairbanks, fairbanks, alaska, usa. telfer, e. s. 1969. twig weight-diameter relationships for browse species. journal of wildlife management 33: 917–921. doi: 10.2307/3799325. testa, j. w. 2001. population dynamics of moose and predators in game management unit 13. federal aid in wildlife restoration, research final performance report, study 1.49, grants w-24-3 to w-27-3. alaska department of fish and game, juneau, alaska, usa. viereck, l. a., and e. l. little, jr. 2007. alaska trees and shrubs. second edition. university of alaska press, fairbanks, alaska, usa. vivås, h. j., and b.-e. sæther. 1987. interactions between a generalist herbivore, the moose (alces alces) and its food resources: an experimental study of winter foraging behavior in relation to browse availability. journal of animal ecology 56: 509–520. white, r. g. 1983. foraging patterns and their multiplier effects on productivity of northern ungulates. oikos 40: 377– 384. doi: 10.2307/3544310. wolff, j. o., and j. c. zasada. 1979. moose habitat and forest succession on the tanana river floodplain and yukontanana upland. alces 15: 213–244. young, d. d., jr. 2012. unit 20a moose. project 1.0. moose management report of survey and inventory activities, 1 july 2009–30 june 2011. alaska department of fish and game, juneau, alaska, usa. ———, and r. d. boertje. 2011. prudent and imprudent use of antlerless moose harvests in interior alaska. alces 47: 91–100. zar, j. h. 1984. biostatistical analysis. second edition. prentice-hall, upper saddle river, new jersey, usa. alces vol. 51, 2015 paragi et al. – browse indices to moose density 21 browse removal, plant condition, and twinning rates before and after short-erm changes in moose density study areas and moose abundance population increase population decrease methods browse removal browse architecture twinning rate results discussion acknowledgements references 61 a possible source of brain abscesses in bull moose vince crichton1 and rick wowchuk2 11046 mcivor ave., winnipeg, manitoba r2g 2j9, canada; 2box 2217, swan river, manitoba r0l 1z0, canada abstract: the presence of cranial infections and abscessations is well documented in males of multiple cervids in north america. the preponderance of such infections is related directly to antlers and all processes from antler growth, fighting, and through to casting. one proposed infection pathway is through an open wound at the pedicle formed at casting. moose generally do not cast antlers in synchrony, and we propose that males irritated by the imbalance of a remaining antler are more likely to actively remove that antler by striking trees. this behavior is a possible explanation for the occurrence of cast antlers with attached bone and that antlers from bulls of all ages can have substantial amounts of parietal bone attached. the force of this activity may cause breakage of the parietal bone leaving either an opening to the meninges in the cranial vault or a significant depression in the bone. we propose that shed antlers with measurable parietal bone attached, estimated as high as 10% of cast moose antlers, would create abnormally large wounds and possibly an enhanced route of cranial infection and subsequent abscessations. alces vol. 55: 61–65 (2019) key words: abscesses, antlers, brain, cervids, males, moose. antlers are principally a secondary sexual characteristic critical in the reproductive behavior of cervids. they occasionally have structural abnormalities associated with either injury during the growth period or in response to a previous physical injury. another type of abnormality is associated with the annual process of casting antlers. specifically, multiple reports document cast white-tailed deer (odocoileus virginianus), elk (cervus elaphus), and moose (alces alces) antlers, and to a lesser extent caribou (rangifer tarandus) antlers, with measurable parietal bone attached (fig. 1). surprisingly, during filming of deer engaged in antler sparring for an outdoor program, the host displayed a broken-off antler with part of the skull attached. of consequence is that in this or a similar outcome after casting, a possible infection route into the cranial cavity and meninges may occur in open wound sites at the pedicle (w. samuel, university of alberta, retired). intracranial abscessations in male white-tailed deer are well documented in north america, seasonally focused in september-april, and presumed to be associated with breeding behaviors involving antlers (i.e., sparring, rubbing, and casting) (davidson et al. 1990, baumann et al. 2001, cohen et al. 2015). these studies report a frequency of <10%, but karns et al. (2009) documented a 35% rate of abscessations in a high-density deer population in maryland; albeit, their sample size was less robust than the other studies. infections are principally associated with arcanobacterium pyogenes (davidson et al. 1990, baumann et al. 2001, karns et al. 2009), a common bacterium that invades superficial wounds of ungulates (zulty and montali 1988). the infection brain abscesses in moose – crichton and wowchuk alces vol. 55, 2019 62 pathway is presumed to be through the open wound associated with either normal casting or abnormal antler breakage at the pedicle. in an examination of 4953 male deer from georgia, cohen et al. (2015) found 91 abscesses (1.8%) and higher probability of an abscess with increasing age; no cranial abscesses were found in 2562 females. although they looked at site-specific variables, none were strongly associated with observations of infection. van ballenberghe (1982) noted an infection at the pedicle of an alaskan moose that had prematurely cast an antler in early september prior to the peak of the rutting period. although documented reports are rare with moose, maccracken et al. (1994) found a high frequency of cast moose antlers with attached pedicle bone in the copper river delta in alaska; however, the authors attributed this to genetic and/or local geographic causes. alternate hypotheses for this anomaly include trauma associated with antler rubbing, male confrontations during the rut, behavior and activity during the casting process (davidson et al. 1990), and physiological stress associated with relative nutritional condition following harsh winters (landete-castillejos et al. 2010). regardless of origin, these wounds and skull fractures presumably open a pathway for intracranial infection by arcanobacterium pyogenes that could prove fatal. it is suggested that the skeletal fractures, abscessations, and infection generally confined to male cervids are directly related to antlers. the origin of injuries to the pedicle area is not clear other than the fact that parietal bone is observed on some cast antlers. both authors (biologist and licensed antler dealer) have observed multiple cast antlers (>1000; range – 1–10% annually) from moose, elk, and deer with a substantial piece of the skull attached (fig. 1 and 2). we find this more common in moose and deer than elk, and although we have examined fewer, have documented erosion of the parietal bone in a caribou skull (fig. 3). we and others searching fig. 1. cast antler from a bull moose showing a portion of the parietal bone attached to the cast antler. alces vol. 55, 2019 brain abscesses in moose – crichton and wowchuk 63 for cast antlers occasionally locate moose antlers at the base of trees which is suggestive of using trees to physically remove antlers. the process as to how breakage occurs is not entirely clear given the multiple behavioral and nutritional explanations. for example, although breakage could occur from blunt force during rutting activity (sparring and fighting), antlers are not necessarily shed in synchrony when cast, fig. 2 . antler from mature bull moose showing pedicle area totally covered with skull bone. fig. 3. caribou skull showing erosion of the parietal bone at the point of antler attachment. brain abscesses in moose – crichton and wowchuk alces vol. 55, 2019 64 as single-antlered moose are frequently observed during the casting period. we suggest that some moose (1.5 years and older) facilitate casting of the remaining antler, which may be an irritant due to imbalanced weight distribution, by striking or rubbing the attached antler against trees or a rigid object in an attempt to dislodge it. newsom (1937) observed that moose may facilitate casting by knocking their antlers against trees and such behavior may explain why some cast antlers are occasionally found at the base of trees. in moose, the torque force caused by the massive, horizontal antlers with a center of gravity far from the skull likely exceeds that in other cervids (nygren et al. 1992). although we have observed this phenomenon in all-aged bulls, larger antlers from animals > 2.5 years old generally have more attached parietal bone than those from younger animals. we did not examine antlers for mineral content as done by maccracken et al. (1994) and recognize that rutting behavior may cause weakening of or hairline fractures at the pedicle (davidson et al. 1990), and that age (hindelang and peterson 1996) and environmental factors (landetecastillejos et al. 2010) influence antler growth, relative bone strength, and presence of osteolytic lesions on the skull. nevertheless, we propose that a contributing factor to substantial pieces of parietal bone on cast moose antlers may be the physical force used to shed the remaining contralateral antler when striking it against solid objects, most often trees. in turn, associated larger wounds at the pedicle provide a possible route of infection for arcanobacterium pyogenes to the meninges resulting in intracranial abscessations. will an antler grow in subsequent years if the pedicle is damaged? field observations suggest that antlers may not develop normally from a damaged pedicle the following year, and bulls with a single antler are observed occasionally in summer (pers. observation, crichton). however, permanent damage or failure to grow antlers is considered rare in advanced cervid species such as moose in which pedicle wounds have multiple months to heal between casting and regrowth (bubenik 1982, goss 1983). regardless, reduced health and survival of moose, elk, deer, and caribou can occur from a cranial infection, and physically-forced antler casting may enhance that possibility. consideration of such is warranted with regard to management strategies aimed at regulating or banning the use of antler traps that ensnare or knock off antlers forcibly. references baumann, c. d., w. r. davidson, d. e. roscoe, and k. beheler-amass. 2001. intracranial abscessation in whitetailed deer of north america. journal of wildlife diseases 37: 661–670. doi:10.7589/0090-3558-37.4.661 bubenik, g. a. 1982. the endocrine regulation of the antler cycle. pages 73–107 in r. d. brown, editor. antler development in cervidae. caesar kleberg wildlife research institute, kingsville, texas, usa. cohen, s., e. h. belser, c. h. killmaster, j. w. bowers, b. j. irwin, m. j. yabsley, and k. v. miller. 2015. epizootiology of cranial abscess disease in whitetailed deer (odocoileus virginianus) of georgia. journal of wildlife diseases 51: 671–679. doi:10.7589/2014-05-129 davidson, w. r., v. f. nettles, l. e. hayes, e. w. howerth, and c. e. couvillion. 1990. epidemiologic features of an intracranial abscessation/suppurative meningoencephalitis complex in white-tailed deer. journal of wildlife diseases 26: 460–467. doi:10.7589/0090-3558-26.4.460 goss, r. j. 1983. deer antlers: regeneration, function, and evolution. academic press, new york, new york, usa. alces vol. 55, 2019 brain abscesses in moose – crichton and wowchuk 65 hindelang, m., and r. o. peterson. 1996. osteoporotic skull lesions in moose at isle royale national park. journal of wildlife diseases 32: 105–108. doi:10.7589/0090-3558-32.1.105 karns, g. r., r. a. lancia, c. s. deperno, m. c. conner, and m. k. stoskopf. 2009. intracranial abscessation as a natural mortality factor for adult male deer (odocoileus virginianus) in kent county, maryland, usa. journal of wildlife diseases 45: 196–200. doi:10.7589/ 0090-3558-45.1.196 landete-castillejos, t., j. d. currey, j. a. estevez, y. fierro, a. calatayud, f. ceacero, a. j. garcia, and l. gallego. 2010. do drastic weather effects on diet influence changes in chemical composition, mechanical properties and structure in deer antlers? bone 47: 815–825. doi:10.1016/j.bone.2010. 07.021 maccracken, j. g., t. r. stephenson, and v. van ballenberghe. 1994. peculiar antler cast by moose on the copper river delta, alaska. alces 30: 13–19. newsom, w. m. 1937. winter notes on the moose. journal of mammalogy 18: 347–349. doi:10.2307/1374210 nygren, k., r. silvennoinen, and m. karna. 1992. antler stress in the nasal bone region of moose. alces supplement 1: 84–90. van ballenberghe, v. 1982. growth and development of moose antlers in alaska. pages 37–48 in r. d. brown, editor. antler development in cervidae. caesar kleberg wildlife research institute, kingsville, texas, usa. zulty, j. c., and r. j. montali. 1988. actinomyces pyogenes infection in exotic bovidae and cervidae: 17 cases (1976–1986). journal of zoo animal medicine 19: 30–32. doi:10.2307/ 20094849 screenposition alces27_31.pdf alces20_47.pdf alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces(25)_112.pdf alces21_419.pdf alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces(23)_89.pdf alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces24_34.pdf alces27_161.pdf alces21_253.pdf alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces vol. 21, 1985 alces(25)_182distinguishedmoosebio.pdf alces22_303.pdf alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces27_85.pdf epizootiology of elaphostrongylus alces in swedish moose margareta stéen1, ing-marie olsson ressner2, bodil olsson3, and erik petersson4 1department of anatomy, physiology and biochemistry, swedish university of agricultural sciences, p. o. box 7090, se-750 07 uppsala, sweden; 2swedish chemicals agency (kemi), p. o. box 2, se-172 13 sundbyberg, sweden; 3tns sifo, p.o. box 115 00, se-404 30 gothenburg, sweden; 4department of aquatic resources, swedish university of agricultural sciences, se-178 93 drottningholm, sweden abstract: a total of 961 harvested and 241 unharvested moose (alces alces) carcasses and parts from throughout sweden were examined for elaphostrongylus alces from 1985 to 1989. when available, the central nervous system and skeletal muscles were searched for adult nematodes, and lungs and feces were examined for first-stage larvae. the parasite was distributed throughout sweden with highest prevalence (56%) in the central region and lowest in the south (13%). prevalence was highest in calves and old moose (>9 years) and lowest in middle-aged animals (5–9 years), with no statistical difference between sexes, although prevalence trended higher in young males. body condition and abundance of elaphostrongylus alces were negatively correlated, and condition was poorer in unharvested than harvested moose. a short (39–73 days) prepatent period was documented, and calves as young as 1.5 months were infected. these results indicate the importance of continued surveillance of elaphostrongylus alces, particularly because a warming climate will likely increase abundance of intermediate mollusk hosts and possibly cause increased infection of moose. alces vol. 52: 13–28 (2016) key words: alces alces, climate, body condition, elaphostrongylus alces, intermediate host, gastropods, moose, prepatent period, protostrongylidae, sweden the moose (alces alces) population in scandinavia began to rise in the 1970s, peaking in the mid-1980s in sweden. with few large predators at that time, it was not unusual to find dead or sick animals (hörnberg 2001, stéen et al. 2005), and in the 1980–1990s, high mortality was noted in both swedish and norwegian moose, as well as in semidomestic reindeer (rangifer tarandus). a previously unknown disease, elaphostrongylosis (stéen and rehbinder 1986, stuve 1986), was reported in the 1980s and sick animals were characterized by locomotive abnormalities such as ataxia, incoordination, swaying of the hindquarters, broad and stamping gait, and a certain way of hypermetria that suggested paralysis of ascending proprioceptive nerve fibers (stéen and roepstorff 1990). a previously undescribed species of elaphostrongyline nematode with a dorsal-spine larva, elaphostrongylus alces (stéen et al. 1989) was invariably associated with sick and dead moose (stéen and rehbinder 1986). parasites of the genera parelaphostrongylus and elaphostrongylus belong to the subfamily elaphostrongylinae (protostrongylidae, metastrongyloidea, nematoda). species of the genus parelaphostrongylus (p. tenuis, p. odocoilei, p andersoni) affect the central nervous system (cns) and skeletal muscle corresponding author: margareta stéen, department of anatomy, physiology and biochemistry, swedish university of agricultural sciences, po. box 7068, se-750 07 uppsala, sweden, margareta.steen@slu.se 13 mailto:margareta.steen@slu.se fasciae of nearctic cervids in north america including white-tailed deer (odocoileus virginianus), black-tailed deer (o. hemonious hemonious), mule deer (o. h. columbianus), and occasionally wapiti (cervus canadensis) and moose (alces alces spp.). species of elaphostrongylus (e. alces, e. cervi, e. panticola, e. rangiferi) affect the cns, the peripheral nerve system (pns), and the skeletal muscle fasciae in eurasian cervids including moose, red deer (cervus elaphus), maral deer (c. e. sibiricus), roe deer (capreolus capreolus), and reindeer (lankester 2001). in the new world, as in the old world, central nervous disorders and mortality occur in wild cervids infected with elaphostrongyline nematodes (anderson 1964, lankester 2001, 2010). representatives of the genera elaphostrongylus and parelaphostrongylus are also harmful to domestic ruminants (lankester 2001). both elaphostrongylus spp. and parelaphostrongylus spp. develop from the first to third larval stage (l1–l3) in their gastropod intermediate host, and develop from the l3 to the adult (l5) stage in their cervid (final) host (olsson et al. 1998, olsson 2001). specific identification of adult protostrongylids and first-stage larvae (l1) in feces of swedish moose was a result of multiple studies. the morphology of e. alces was initially described by stéen et al. (1989) and (stéen and johansson 1990), and subsequent comparison of specific proteins in protostrongylid l1 indicated that l1 and adult e. alces had the same protein pattern in moose, but differed from the l1s and adult protostrongylid parasites in other wild ruminants (stéen et al. 1993). experimental infection of captive moose indicated that l1 collected from wild moose caused elaphostrongylosis, and l1 excreted from infected and sick moose and transmitted to terrestrial snails (arianta arbustorum) in which larvae develop (lankester et al. 1998), were identified as e. alces using genomic dna (gajadhar et al. 2000) and single-strand conformation polymorphism (sscp) analysis (chilton et al. 2005, huby-chilton et al. 2006). collectively, these studies indicate that protostrongylid larvae in swedish moose are e. alces. given the prevalence and deleterious effect of this disease, our objective was to determine if e. alces is related to age, sex, condition, and geographic distribution of moose in sweden. study area sweden was divided into 6 regions from the far north (69°03′36″n 20°32′55″e) to the far south (55°20′13″n 13°21′34″e) to determine the distribution and prevalence of elaphostrongylosis (fig. 1). sweden is sheltered by the scandinavian mountains and has a continental climate with large differences in temperature and precipitation between summer and winter, and a relatively small amount of precipitation (swedish meteorological and hydrological institute [smhi]). summer temperatures are similar to those in north america and asia at similar latitude, although due to the gulf stream, winter in sweden is typically milder (smhi 2015). methods hunting begins on the first monday of september in northern sweden, on the second monday of october in the south, and seasons end in december or january (swedish association for hunting and wildlife management 2015). prior to the hunting seasons (1986 and 1987), we sent hunters report cards (hunting site, sex, and approximate age) and wrapping materials to pack body parts (i.e., lungs, feces, spinal cords, and mandibles). moose carcasses and parts (n = 1137) were examined in 5 consecutive years (1985– 1989); 896 (79%) were associated with harvested moose (1986, 1987, 1989) and 241 (21%) were from non-harvested animals (i.e., euthanized or found dead; 1985–1989). 14 epizootiology of elaphostrongylus alces – stéen et al. alces vol. 52, 2016 we used 1020 lungs and 1084 fecal samples to identify presence of elaphostrongylid l1, 655 spinal cords (membranes) to identify presence of adult worms, and 636 mandibles to measure fat content (table 1). age was determined by dental wear (gasaway et al. 1978) or from information provided by hunters; 151 animals were not aged due to lack of information. five age classes were established: 1) calves were ≤12 months), 2) yearlings were >12 and ≤24 months, 3) young animals were >24 months and ≤5 years, 4) middle-aged animals were >5 and ≤9 years, and 5) old animals were >9 years. sex was determined from the whole carcass or hunter information. evaluation of body condition was by visual inspection, location, and appearance of body fat (n = 948), and/or by measuring fat content (%) in mandible bone marrow (n = 636; engelsen etterlin et al. 2009). three categories of condition were established: bd ac z y x w s t u c b d eo m h f arc�c circle 66°33'39" n treriksröset 69°03'36''n, 20°32'55''e smygehuk 55 °20'13''n, 13°21'34'' region 1: n= 135 region 2: n=558 region 3: n=37 region 4: n=100 region 5: n=40 region 6: n=26 n k g fig. 1. map of sweden with county codes and 6 regions: region 1 = the laplandic counties (ac, z, bd, and y), region 2 = southern part of norrland (w, s, and x county), region 3 = northern svealand (c, u, t, and b county), region 4 = southern svealand and northern götaland (d, p, r, e, and o county), region 5 = småland (f, h, and g county), and region 6 = southern götaland (n, l, k, and m county). dots represent locations of moose infected with elaphostrongylus alces; n = harvested moose. alces vol. 52, 2016 stéen et al. – epizootiology of elaphostrongylus alces 15 normal, poor (below normal), and emaciated (lack of adipose tissue). the fat content in bone marrow was measured with standard techniques under specified assay conditions and techniques (nmkl no 131, nordic committee on food analysis 1989) and also used to assign condition: normal = 75– 94%, poor = 16–<75%, and emaciated = 0.4–<16% fat content. assigning condition from visual inspection (without measuring fat content) was considered reliable because of the strong correlation between the condition category assigned from fat measurements and visual inspection of the same animals (rs = −0.801, p < 0.001, n = 592). bodies/parts were inspected for adult e. alces worms and l1 with necropsy procedures described previously (stéen and rehbinder 1986, stéen et al. 1997, 1998) and included examination of muscle fasciae, the cranial cavity, brain, and spinal cord membranes and epidural space of the spinal cord (stéen and rehbinder 1986, stéen et al. 1997). lungs from all animals were palpated and inspected for nodules, and 20 g samples of minced lungs and feces were processed to detect l1 (baermann 1917). l1s were identified as protostrongylids quantified in a counting chamber under a stereo microscope, expressed as larvae per gram of wet feces (lpg), and classified into 7 levels of relative abundance ranging from none (0) to heavy (6) (national veterinary institute, sweden). there were 4 categories of infection: 0 = uninfected, 1 = in the epidural space but not in lungs or feces, 2 = in lungs but not feces, and 3 = in feces. animals were categorized as either infected or uninfected (presence or absence of l1 and/or e. alces worms) for certain statistical comparisons (e.g., sex or age groups, prevalence in population or region), data management and statistics data were tested for normal distribution and seasonal variation, and if not normally distributed, normality was achieved with log-transformation. a peak function analysis was used to identify the best fit to the relationship between bone marrow fat and season (tablecurve software, systat 2002). a mean value was calculated for harvested animals and this value was applied together with the individual values for remaining table 1. total number of moose (harvested/unharvested) and sample location/type – epidural space of the spinal cord (epidural), lungs, feces, mandibles – used to study elaphostrongylus alces in sweden, 1985–1989. moose epidural lungs feces mandibles sex males 386/87 173/85 343/84 369/85 260/36 females 456/147 223/144 404/145 427/145 262/57 unknown 54/7 23/7 37/7 49/7 21/– age group calves 457/107 187/103 392/102 434/103 270/48 yearlings 227/36 113/36 52/12 220/36 182/11 young 40/27 23/27 39/27 38/27 36/10 middle aged 31/19 27/19 28/19 26/19 30/5 old 6/37 4/37 6/37 6/37 5/16 unknown 135/14 65/13 110/14 123/14 20/1 total 896/241 419/236 784/236 847/237 543/93 16 epizootiology of elaphostrongylus alces – stéen et al. alces vol. 52, 2016 animals. the residuals for all animals were calculated (harvested moose were not combined as above), and adjusted values were calculated by adding the residual to the common mean. this produced a few values >100% that were not further corrected in subsequent analyses. bone marrow fat (adjusted for seasonal variation) was subsequently analysed using generalized linear models. body condition was also analysed with generalized linear models, modeling the probability of being in normal condition (see above) assuming a binary distribution of the response variable. the total parasite infection or parasites found in either the lungs, feces, or in the epidural space were similarly corrected, and the probability of being infected was tested with respect to 3 predictors (age, sex, region). the age when calves were infected was estimated with birth date information from each county. comprehensive data were available from 5 counties: västerbotten (ac in region 1), västra götaland (o in region 4), kalmar (h in region 5), kronoberg (g in region 5), and södermanland (d in region 4) (fig. 1). in 3 counties (h, g, and d) the mean value + sd (malmsten 2014) was used as the birth date, and in 2 counties (ac and o) the mean value + sd was estimated (broberg 2004). birth dates for the counties without data were estimated using a multiple imputation (proc mi in sas statistical software, sas 2014) with a markov chain monte carlo method in which longitude and latitude of resident cities were used with the number of imputations set to 60. other than 3 counties with a minor inconsistency (3–4 days), the approach produced an acceptable trend of earlier birth dates in southern sweden, and the dates corresponded well with the span of birth dates reported by a national hunting organization (swedish association for hunting and wildlife management 2015) (table 2). the mean category of infection (0–3) in each age group was calculated to illustrate the relationships among age (mean age of group), category of infection, and body condition. these values were used to develop a contour graph using sigmaplot software (systat 2008) where body condition, age group, and infection category were interpolated. results infection, age and sex age of moose was skewed towards young animals (table 1), and age in the two groups (harvested and unharvested) was not distributed evenly. unharvested moose were older than those harvested for combined age classes, calves, and by sex (table 3). the average age of harvested animals (n = 761) was 10.4 months (95% ci = 9.6 – 10.4; range = 0–15 years), and 22.3 months (ci = 17.4–28.4; range = 0–20 years) for unharvested animals (n = 227). females were older in the yearling, middle-aged, and combined age groups. a slight majority (57%) of the harvested sample (n = 896) was infected with l1 and/or adult e. alces worms. the prevalence was similar between sexes in each age class for l1 in lungs, l1 in feces, and adult worms in the epidural space of the spinal cord (fig. 2). there was a tendency (p = 0.074) toward higher prevalence in males than females in the young age class. worms were found in the epidural space of the spinal cord in animals 3 months to 2 years old, but not in animals 3 to 9 years old; worms were found in a single 10-year old moose. the abundance of l1 in lungs (n = 784) was high in calves and yearlings, lower at 3–4 years of age, and minimal in adults. nearly the entire sample (98%) of unharvested moose (n = 241) was infected with e. alces (fig. 3). worms were found in the epidural space of the spinal cord in 3 month to 4 year-old animals. the average age of alces vol. 52, 2016 stéen et al. – epizootiology of elaphostrongylus alces 17 infected calves was 4.8 months (95% ci = 4.7–4.9). no worms were found in the epidural space of the spinal cord in 5–9 yearold moose, but worms reappeared at 10–16 years of age. the prevalence of adult worms in the epidural space of the spinal cord was 36% in the combined data (harvested and unharvested, n = 655); the prevalence of l1 in lungs (n = 1020) and feces (n = 1084) was 64 and 53%, respectively. the prevalence (worms/ l1) was 66% overall; 88% in old moose, 74% in yearlings, 67% in calves, 55% in young, and 48% in middle-aged animals. there were differences (p < 0.001) in frequency of infection among age groups; the oldest animals had the highest frequency of infection (l1) and the middle-aged the lowest. the frequency of worms in the epidural space of the spinal cord was high in calves/ yearlings, leveled out at 4 years, and then was not identified until 10–16 years at low frequency. the abundance of l1 in lungs of old animals was at the highest level (6). body condition body condition of harvested animals (n = 981) was either normal (40% overall, 24% calves) or poor (59%, 75% calves). in unharvested moose (n = 227), body condition was normal in 38% overall, with calves and old animals lower; 25% calves, 45% young, and 29% old animals were in normal condition. for all moose, body condition and category of infection were correlated (rs = 0.215, p < 0.001). in separate age classes, this table 2. prevalence of elaphostrongylus alces (adjusted for julian date and age of the sampled moose) and birth date of moose in swedish counties. birth dates marked with an asterisk are observed values; others are estimated (see data management and statistics). region county mean prevalence (%) n birth date (julian date) birth date 1 ac västerbotten 43.8 82 167* 16 june 1 bd norrbotten 51.4 13 168 17 june 1 z jämtland 53.5 8 171 20 june 1 y västernorrland 58.2 21 164 13 june 2 w dalarna 63.2 32 160 9 june 2 x gävleborg 67.7 457 157 6 june 2 s värmland 49.4 36 159 8 june 3 b stockholm 61.7 2 149 29 may 3 c uppsala 100.0 3 152 1 june 3 t örebro 67.4 17 156 5 june 3 u västmanland 79.3 15 154 3 june 4 d södermanland 100.0 5 148* 28 may 4 e östergötland 27.6 12 150 30 may 4 o västra götaland 53.9 12 153* 2 june 5 f jönköping 39.1 16 151 31 may 5 g kronobergs 28.2 10 144* 24 may 5 h kalmar 32.2 11 143* 23 may 6 k blekinge 37.4 5 140 20 may 6 m skåne 0 8 144 24 may 6 n halland 41.5 12 146 26 may 18 epizootiology of elaphostrongylus alces – stéen et al. alces vol. 52, 2016 correlation was found in yearlings (rs = 0.262, p < 0.001, n = 239), young (rs = 0.463, p < 0.001, n = 67), middle-aged (rs = 0.441, p = 0.002, n = 47), and old animals (rs = 0.456, p = 0.003, n = 40), but not in calves (rs = 0.054, p = 0.239, n = 471). for all moose, body condition was correlated inversely with category of infection (rs = 0.084, p = 0.025, n = 721); separate correlations were found in yearlings (rs = 0.213, p = 0.002, n = 227) and young animals (rs = 0.398, p = 0.011, n = 40). figure 4 illustrates the probability of normal body condition relative to age and category of infection, indicating that calves have poor body condition regardless of category of infection, and that some middle-aged animals have normal body condition despite high abundance of l1 in feces. old individuals were generally in normal body condition if not infected, although few were without infection. bone marrow fat content (n = 615) varied annually (table 4, fig. 5). on average, harvested animals had higher fat content (93%, 95% ci = 91 – 96) than unharvested animals (70%, ci = 66 – 75) with values corrected for time of year, sex, and age class (table 4). in a combined sample, a negative correlation was found between bone marrow fat content and category of infection (rs = −0.212, p < 0.001, n = 635). this negative correlation was found in calves (rs = −0.131, p = 0.020, n = 319), yearlings (rs = −0.223, p = 0.002, n = 193), young (rs = table 3. age in months (mean and 95% ci) of swedish moose examined for elaphostrongylus alces, 1985–1989. the column to the far right gives level of significance between harvested and euthanized + dead moose (for the last three rows the t-tests are performed on log transformed data; the data presented in the table are back-transformed values). if all sexed animals are combined, the sexes differed in age (p < 0.05). age class sex harvested unharvested t-value calves females 4.3 (4.1–4.4) 8.5 (8.1–8.8) 22.3, p < 0.001 males 4.2 (4.0–4.4) 8.1 (7.8–8.4) 21.1, p < 0.001 all calves‡ 4.2 (3.5–5.0) 8.3 (6.6–9.9) 4.28, p < 0.001 yearlings* females 18.4 (17.7–19.0) 19.2 (17.5–20.9) 0.91, p = 0.362 males 17.7 (17.0–18.4) 17.3 (15.6–19.0) 0.36, p = 0.716 combined 18.0 (16.9–19.1) 18.3 (15.4–21.1) 0.17, p = 0.863 young females 46.5 (42.8–50.4) 47.4 (43.0–51.7) 0.28, p = 0.782 males 42.0 (36.9–47.1) 44.6 (37.4–51.8) 0.58, p = 0.562 combined‡ 44.7 (42.0–47.4) 46.2 (42.9–49.5) 0.70, p = 0.481 middle-aged* females 90.3 (85.0–95.5) 93.4 (86.7–100.2) 0.74, p = 0.461 males 84.0 (74.4–93.6) 76.8 (65.5–88.1) 0.98, p = 0.333 combined‡ 89.0 (86.0–92.1) 89.1 (85.1–93.0) 0.01, p = 0.994 old females 144.0 (106–181.7) 154.2 (141.3–167.2) 0.52, p = 0.606 males 120.0 (144.6–195.4) 144.0 (100.5–187.5) 0.56, p = 0.580 combined‡ 146.0 (139.1–152.9) 153.4 (150.6–156.2) 1.94, p = 0.053 all females§ 12.3 (10.7–14.1) 29.6 (21.0–41.6) 4.73, p < 0.001 all males§ 9.0 (8.0–10.0) 14.2 (10.7–18.7) 3.92, p = 0.004 all moose 10.4 (9.6–11.4) 22.3 (17.4–28.4) 5.74, p < 0.001 *sexes differ by age class (all causes of death included). ‡includes individuals not sexed. §sexes differ with all age classes combined. alces vol. 52, 2016 stéen et al. – epizootiology of elaphostrongylus alces 19 −0.618, p < 0.001, n = 46), and old (rs = −0.736, p < 0.001, n = 21), but not middleaged moose (rs = −0.319, p = 0.062, n = 35). time of infection the earliest identification of a calf diagnosed with elaphostrongylosis was at ~1.5 months on 21 july in region 3, county of calves yearlings young mid. age old 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 epidural space fr eq u en cy 0.0 0.2 0.4 0.6 0.8 1.0 females males lungs 0.0 0.2 0.4 0.6 0.8 1.0 feces total e. alces fig. 2. the mean abundance (frequency) of elaphostrongylus alces measured in adult worms in the epidural space of the spinal cord, and larvae in lungs and feces of harvested moose, sweden, 1985– 1989. the values are least-squared means and 95% confidence limits from a generalized linear model. the interaction between age groups and sex was a categorical predictor and julian date was a continuous predictor. 20 epizootiology of elaphostrongylus alces – stéen et al. alces vol. 52, 2016 uppsala (table 5). the abundance of l1 was category 6 in the lungs and 4 in feces, and the calf was in normal body condition. the earliest calf death where worms were found in the epidural in the spinal cord was on 10 october (~4 months old) in region 2, county of värmland; the abundance of l1 was category 6 in the lungs and 3 in feces. worms were first found in harvested calves on 15 september (~3 months old). fig. 4. contour plot of the probability of normal body condition in moose (n = 613) versus age and 3 categories of elaphostrongylus alces infection intensity, sweden, 1985–1989. ca lve s ye arl ing s yo un g mid dle ag ed old p ro po rt io n (% ) 0 20 40 60 80 100 uninfected e. alces worms in epidural space only e. alces l1 in lungs e. alces l1 in feces h hh h hd d d d d fig. 3. the 4 categories of elaphostrongylus alces infection within 5 age groups of swedish moose, 1985–1989. h bars represent harvested moose (n = 761) and d bars represent unharvested moose (euthanized or found dead; n = 227). alces vol. 52, 2016 stéen et al. – epizootiology of elaphostrongylus alces 21 table 4. analysis of bone marrow fat (%) and body condition of harvested and unharvested moose, sweden, 1985–1989. values are mean ± se with sample size in parentheses. pair-wise comparisons (t-values) of harvested and unharvested animals are provided in each age group (raw). means denoted by the same letter in each column (percent fat and body condition separately) were not different (p < 0.05). the values for percent fat are adjusted for time of year causing certain values to be >100% (see data management and statistics). variable age class harvested euthanized or dead statistic bone marrow fat (%) calves 83.6 ± 0.8a (270) 60.5 ± 1.8a (48) t = 11.6 p < 0.001 yearlings 97.1 ± 0.9b (182) 69.9 ± 3.8b (11) t = 6.88 p < 0.001 young 98.7 ± 2.1b (36) 83.6 ± 4.0c (10) t = 3.32 p < 0.001 middle aged 99.6 ± 2.3b (30) 77.4 ± 5.7bc (5) t = 3.62 p < 0.001 old 104.6 ± 5.7b (5) 63.3 ± 3.2ab (16) t = 6.34 p < 0.001 all animals 96.7 ± 1.3 (523) 70.9 ± 1.7 (92) t =11.6 p < 0.001 probability of normal condition calves 0.17 ± 0.03a (368) 0.25 ± 0.06a (102) z = 0.81 p = 0.420 yearlings 0.48 ± 0.04b (204) 0.30 ± 0.12c (35) z = 1.19 p = 0.235 young 0.50 ± 0.11bc (40) 0.47 ± 0.15b (27) z = 0.25 p = 0.800 middle aged 0.74 ± 0.12c (30) 0.64 ± 0.21b (17) z = 0.34 p = 0.734 old 0.84 ± 0.17abc (6) 0.12 ± 0.06ab (34) z = 2.53 p = 0.012 all animals 0.33 ± 0.02 (648) 0.29 ± 0.05 (215) z = 0.56 p = 0.578 julian date 25 50 75 100 125 150 175 200 225 250 275 300 325 350 b od y co nd iti on (b on e m ar ro w fa t % ) 0 20 40 60 80 100 fig. 5. the dependence of moose body condition (expressed as percent fat) on day of the year (julian date). in this analysis harvested moose are represented by a single value (julian date = 292.80; percent fat = 66.05). the line in the figure is the estimated peak function; a lorentzian peak function; y = 45.58 + [− 34.65/(1 + (((x – 115.05)/62.57))^2)]; f3,89 = 6.23, r2 = 0.174, p < 0.001. the peak are estimate to julian date = 115.05, which is 25 april. 22 epizootiology of elaphostrongylus alces – stéen et al. alces vol. 52, 2016 the abundance of l1 was category 0 in the lungs and 6 in feces. first stage larvae (infection intensity = 6) were found in lungs from 14 august (~2 months old) to 4 june the following year (~12 months old). abundance of l1 in calves ranged from categories 1–6 by 2 months old, and the lung infection remained high; 81% had an l1 abundance category of 4–6 in the first year. the excretion of larvae began at a low level (2) on 14 august, and calves continued to excrete larvae throughout the first year at all levels of abundance (1–6). the prevalence of infection in harvested moose (n = 896) differed among regions (fig. 6), ranging from 13% in southernmost region 6 to 56% in region 3 (fig. 1 and table 2). infection was most prevalent in central sweden, least prevalent in southern sweden, and similar (p < 0.05) in southern and northern sweden. discussion although parasites at low abundance are generally less harmful to their host, when the host population increases rapidly, as with swedish moose in the 1970–80s (hörnberg 2001, stéen et al. 2005), an increasing risk to the individual and host population is possible (toft 1991). the proportion of elaphostrongylosis (symptoms of nervous disorder and/or emaciation) varies among age-classes in moose, with young animals more prone to illness (stéen et al. 2005). similarly, we found that e. alces worms located in the epidural space of the spinal cord were more prevalent in calves and yearlings, and only occasionally found in adults. the high abundance measured in young animals may simply reflect that the swedish moose population is skewed towards young animals (sand et al. 2011). conversely, abundance of l1 in lungs and feces was highest in old moose, and lowest in young and middle-aged moose. both stuve (1986) and stéen et al. (2005) suggested that e. alces most frequently infects males and young animals; however, we found no difference in the abundance within the epidural space, lungs, or feces between sexes or age groups of harvested moose, only a tendency toward males in the young age group. similarly, male reindeer calves with dominant mothers had higher abundance of e. rangiferi than female calves, and it was suggested that because these calves had better access to forage, they were at greater risk of ingesting infected gastropods (halvorsen 1986a). calf weight is dependent on summer browse availability in a table 5. age (in days) of moose calves infected by elaphostrongylus alces, sweden, 1985–1989. parasite location n mean age ± sd min age max age epidural 281 140.8 ± 20.0 50 215 lung 348 140.2 ± 20.8 50 215 feces 416 141.0 ± 17.6 101 215 region 1 2 3 4 5 6 p re va le nc e 0.1 0.2 0.3 0.4 0.5 0.6 0.7 a ab a b b b fig. 6. the prevalence of elaphostrongylus alces in regions of sweden. the values are calculated with logistic regression, regions was a categorical predictor and julian date a continuous predictor. the values are mean ± se; means with the same letter are not different (p > 0.05). only harvested moose were used in the analysis. alces vol. 52, 2016 stéen et al. – epizootiology of elaphostrongylus alces 23 cow’s home range, with access to and quality of forage related to its relative status (saether and heim 1993). stuve (1986) attributed the difference in infection rate between sexes in older moose to physiological changes associated with the rut, as suggested with reindeer (halvorsen 1986b). a novel finding of our study was that l1 were found in lungs and feces of calves by 21 july, and adult worms in the epidural space by 15 september, or ~50–100 days after birth (broberg 2004, malmsten 2014). this prepatent period aligns with experimental infections of e. alces in moose in which patent infection was realized 39–73 days post-infection (stéen et al. 1997). because calves sample vegetation in the first days of life to promote development of rumen microbes (syroechkovsky et al. 1989), their potential to exposure to e. alces l3 is almost immediate. not surprisingly, adult e. alces were identified in the epidural space of the caudal vertebral canal in 2 other calves harvested in september (handeland and gibbons 2001). further, calves and yearlings were most frequently infected in the epidural space of the spinal cord which seemingly corroborates that moose shed most e. alces l1 during their early years, after which a sharp drop in larval shedding and low numbers of adult worms in older animals occur (stuve 1986, stéen et al. 2005). in both harvested and unharvested moose, e. alces worms were found in the epidural space of the spinal cord of animals aged 3–4 months to 4 years, not in middleaged animals, and again at 10–16 years. conversely, high levels of larvae were found in lungs and feces irrespective of age. we believe that the low frequency of worms in older animals, despite having l1 in lungs and shed larva, is due to migration from the cns/pns into the muscle fasciae, as with some other elaphostrongylins (lankester 2001). the pattern of e. alces adults migrating out to the muscle fasciae, presumably due to an immune response in the epidural space (stéen et al. 1997, 1998), differs somewhat from that of e. rangiferi, e. cervi, and p. tenuis. the latter are believed to remain in the cns as adult worms during their entire life (in the subdural or subarachnoid space, inside the meninges), although e. rangiferi also migrates to the muscle fasciae (hemmingsen et al. 1993). e. rangiferi, e. cervi, and p. tenuis may realize an immunological harbor within the cns, as might p. andersoni that is associated with blood vessels and connective tissues where females deposit eggs (lankester 2001). we hypothesize that e. alces worms are attacked by the immune system in the epidural space, and they migrate to the muscle fasciae where, with lower immunological defense, they deliver most of their larvae. after ingestion, l3 migrate from the gastrointestinal (gi) tract to the perineal cavity along the mesenchyme nerves, and into the abdominal wall associated with the more posterior lateral nerves. it is likely that e. alces does not need to enter the cns parenchyma to develop to the 5th stage (adult), as other elaphostrongylus spp., but remains epidurally-associated with lateral nerves of the pns and finally migrates to the muscle fasciae (olsson et al. 1998). the lack of worms in the epidural space of the spinal cord in moose during their prime could be explained by this migration; however, it could also reflect an immune response to prevent reinfection as described for p. andersoni that realizes declining larval output as deer age with few adult worms in deer >1 year old. further, repeated infection in whitetailed deer resulted in sharp decline in larval numbers and a strong cellular response to adult worms (lankester 2001). worms in the epidural space of older moose could simply be a reinfection associated with a weaker immune system, or an initial infection. whether some l3s migrate directly to muscle 24 epizootiology of elaphostrongylus alces – stéen et al. alces vol. 52, 2016 fasciae without being associated with neural tissue is unknown. infected animals, on average, had lower body condition than uninfected animals except for middle-aged animals in their prime. calves were in poorer condition regardless of category of infection (as expected for young, growing animals), middle-aged were likely in normal condition despite high shedding rate of l1 in feces, and old individuals were in normal condition if uninfected. thus, infection, not age per se, seemed to reflect relative body condition. however, individual variation of immunological response to the parasite presumably exists because some individuals die young, others remain in normal condition through prime, and old animals are increasingly susceptible. in contrast with e. alces, no protostrongylid l1 of e. cervi were recovered from iberian red deer fawns (cervus elaphus hispanicus) (vicente and gortázar 2001). prevalence of e. cervi l1 increased with age of deer (vicente et al. 2006) which is opposite to our findings with e. alces in moose; both had higher infection rates in young males than females. the e. cervi pattern corresponds with that in reindeer in which e. rangiferi infects the host late in the season, remaining at the same intensity for at least 3 years (halvorsen et al. 1985). it appears that e. cervi and e. rangiferi have more similar and longer evolutionary relationships to each other and their respective hosts than e. alces. moose have a long, independent evolutionary history from the alceini and the plio-pleistocene, suggesting a peculiar adaption and habitat restriction of the species (niedziałkowska et al. 2014), and presumably, a relatively short evolutionary period with e. alces that could be less adapted with its host than e. cervi and e. rangiferi. it is possible that e. alces is more pathogenic to its host because both harvested and unharvested moose of below normal or emaciated body condition were infected with e. alces. in 2-year old moose, stuve (1986) found that infected moose were lighter (carcass weight) than uninfected moose, yet conversely, stéen et al. (1997) found that moose experimentally infected with e. alces retained normal weight when fed ad libitum. it remains unclear, however, if poor body condition is an indirect or direct effect of the parasite, that emaciation is either directly caused by an inflammatory response due to an epidural localization, or that elaphostrongylosis causes locomotor disorders making it difficult to move and feed (stéen and rehbinder 1986, stéen and roepstorff 1990, stéen et al. 2005). in summary, different morphology (stéen et al. 1989, stéen and johansson 1990, gibbons et al. 1991, lankester et al. 1998), genetics (gajadhar et al. 2000, chilton et al. 2005, huby-chilton et al. 2006), location (stéen et al. 1997, 1998) (epidural for e. alces, subdural/subarachnoid for e. rangiferi), and life span and host age relationships with infection (lankester 2001) suggest different, and perhaps, ongoing evolutionary adaption in elaphostrongylus species with their hosts. of further consequence is that rising temperatures, and a warmer and wetter climate are predicted to increase habitat, distribution, and abundance of mollusk hosts (halvorsen and skorping 1982, halvorsen et al. 1985), which in turn could lead to higher infection rates in cervids (handeland and slettback 1994, halvorsen 2012). although moose are not necessarily in poor condition when infected with e. alces, condition and parasite abundance were correlated. we therefore suggest continued surveillance of this disease and its specific consideration in management of moose in sweden. acknowledgements we thank w. e. faber, department of natural resources, central lakes college, brainerd, minnesota, usa for his feedback on the manuscript. we are grateful for all alces vol. 52, 2016 stéen et al. – epizootiology of elaphostrongylus alces 25 technical help and assistance from employees, and earlier employees h. mann, s. persson, and i. forssell at the department of parasitology, national veterinary institute and swedish university of agricultural sciences, uppsala, sweden. last, but not least, we thank swedish hunters for their help and cooperation in data collection. financial support for this study was provided by the swedish environmental protection agency, stockholm, sweden. references anderson, r. c. 1964. neurological disease in moose experimentally infected with pneumostrongylus tenius from white-tailed deer. veterinary pathology 1: 289–322. doi: 10.1177/030098586400100402. baermann, g. 1917. eine einfache metode zur auffindung von ancylostoma(nematoden-) larven aus erdproben. mededeel uithet geneesk lab. te weltevreden, feestbundel, batavia, pp. 41–47 (in german). broberg, m. 2004. reproduction in moose: consequences and conflicts in timing of birth. doctoral thesis, gothenburg university, gothenburg, sweden. chilton, n. b., f. huby-chilton, m. w. lankester, and a. a. gajadhar. 2005. a method for extracting genomic dna from individual elaphostrongyline (nematoda: protostrongylidae) larvae and differentiation of elaphostrongylus spp. from parelaphostrongylus spp. by pcr assay. journal of veterinary diagnostic investigation 17: 585–588. doi: 10.1177/104063870501700612. engelsen etterlin, p., a. neimanis, d. gavier-widen, and c. hård af segerstad. 2009. postmortal hullbedömning av hull hos tamdjur och vilda djur (postmoral examination of body condition in pet animals and wildlife). national veterinary institute, uppsala, sweden (in swedish). gajadhar, a., t. steeves-gurnsey, j. kendall, m. lankester, and m. stéen. 2000. differentation of protostrongylid dorasal-spined larvae by pcr amplication of its-2 rdna. journal of wildlife diseases 36: 713–722. doi: 10.7589/0090-3558-36.4.713. gasaway, w. c., d. b. harkness, and r. a. rausch. 1978. accuracy of moose age determinations from incisor cementum layers. journal of wildlife management 42: 558–563. doi: 10.2307/3800818. gibbons, l. m.,o. halvorsen, and g. stuve. 1991. revision of the genus elaphostrongylus cameron (nematoda, metastrongyloidea) with particular reference to species of the genus occurring in norwegian cervids. zoologica scripta 20(1): 15–26. doi: 10.1111/j.1463-6409.1991.tb00272.x. halvorsen, o. 1986a. on the relationship between social status of host and risk parasitic infection. oikos 47: 71–74. doi: 10.2307/3565921. –––. 1986b. epidemiology of reindeer parasites. parasitology today 12: 334–339. doi: 10.1016/0169-4758(86)90053-0. –––. 2012. reindeer parasites, weather and warming of the arctic. polar biology 35: 1209. doi: 10.1007/s00300-012-1209-0. –––, and a. skorping. 1982. the influence of temperature on growth and development of the nematode elaphostrongylus rangiferi in the gastropods arianta abustrum and euconulus fulvus. oikos 38: 285–290. doi: 10.2307/3544666. –––, a. skorping, and k. hansen. 1985. seasonal cycles in the output of first stage larvae of the nematode elaphostrongylus rangiferi from reindeer, rangifer tarandus tarandus. polar biology 5: 49–54. doi: 10.1007/bf00446045. handeland, k., and l. m. gibbons. 2001. aspects of the life cycle and pathogenesis of elaphostrongylus alces in moose (alces alces). journal of parasitology 87: 1054–1057. doi: 10.1645/0022-3395 (2001)087[1054:aotlca]2.0.co;2. –––, and t. slettbakk. 1994. outbreaks of clinical cerebrospinal elaphostrongylosis in reindeer (rangifer tarandus tarandus) 26 epizootiology of elaphostrongylus alces – stéen et al. alces vol. 52, 2016 in finnmark, norway, and their relation to climatic conditions. journal of veterinary medicine b 41(1–10): 407–410. doi: 10.1111/j.1439-0450.1994.tb00244.x. hemmingsen, w., o. halvorsen, and a. skorping. 1993. migration of adult elaphostrongylus rangiferi (nematoda: protostrongylidae) from the spinal subdural space to the muscles of reindeer (rangifer tarandus). journal of parasitology 79: 728–732. doi: 10.2307/3283612. huby-chilton, f., n. b. chilton, m. w. lankester, and a. a. gajadhar. 2006. single-strand conformation polymorphism (sscp) analysis as a new diagnostic tool to distinguish dorsal-spined larvae of the elaphostrongylinae (nematoda: protostrongylidae) from cervids. veterinary parasitology 135: 153–162. doi: 10.1016/j.vetpar.2005.08.001. hörnberg, s. 2001. changes in population density of moose (alces alces) and damage to forests in sweden. forest ecology and management 149: 141–151. doi: 10.1016/s0378-1127(00)00551-x. lankester, m. w. 2001. extrapulmonary lungworms of cervids. pages 228–278 in w. m. samuel, a. a. kocan, and m. pybus, editors. parasitic diseases of wild mammals, 2nd edition. iowa state university press, ames, ia. –––. 2010. understanding the impact of meningeal worm, parelaphostrongylus tenuis, on moose population. alces 46: 53–70. –––, i.-m. c. olsson, m. stéen, and a.a. gajadhar. 1998. extra-mammalian larval stages of elaphostrongylus alces (nematoda: protostrongylidae), a parasite of moose (alces alces) in fennoscandia. canadian journal of zoology 76: 33–38. doi: 10.1139/z97-168. malmsten, j. 2014. reproduction and health of moose in southern sweden. doctoral thesis, swedish university of agricultural sciences, uppsala, sweden. niedziałkowska, m., k. j. hundertmark, b.jezdrzejewska, k.niedziałkowski, v. e. sidorovich, m. gorny, r. veeroja, e. j. solberg, s. laaksonen, h.sand, v. a.solovyev, m. shkvyria, j.tiainen,i.m.okhlopkov,r.juskaitis, g. done, v. a. borodulin, e. a. tulandin, and w. jezdrzejewski. 2014. spatial structure in european moose (alces alces): genetic data reveal a complex population history. journal of biogeography 41: 2173–2184. doi: 10.1111/jbi.12362. nordisk metodik-kommitté för livsmedel (nordic committee on food analysis). 1989. sbr (schmid-bondzynski-ratslaff), no. 131. esbo, finland. http://www.evira.fi (accessed july 2015). olsson, i.-m. 2001. elaphostrongylus alces – transmission, larval morphology and tissue migration. veterinary medicine licentiate thesis, swedish university of agricultural sciences, uppsala, sweden. –––, m. w. lankester, a. a. gajadhar, and m. steén. 1998. tissue migration of elaphostrongylus spp. in guinea pigs (cavia porcellus). journal of parasitology 84: 968–975. doi: 10.2307/ 3284629. saether, b. e., and m. heim. 1993. ecological correlates of individual variation in age at maturity in female moose (alces alces); the effects of environmental variability. journal of animal ecology 62: 482–489. doi: 10.2307/5197. sand, h., n. jonzén, h. andrén, and j. månsson. 2011. strategier för beskattning av älg. (strategies for moose management). forest facts, report from the swedish university of agricultural sciences, no. 2 4, p 4 (in swedish). sas 2014. version 9.4 ts level 1m0, x64_7pro platform, sas institute inc., cary, north carolina, usa. stéen, m., c. g. m. blackmore, and a. skorping. 1997. cross-infection of moose (alces alces) and reindeer (rangifer tarandus) with elaphostrongylus alces and elaphostrongylus rangiferi (nematoda, protostrongylidae): effects on parasite morphology and prepatent period. alces vol. 52, 2016 stéen et al. – epizootiology of elaphostrongylus alces 27 http://www.evira.fi veterinary parasitology 71: 27–38. doi: 10.1016/s0304-4017(97)00013-7. –––, a. g. chabaud, and c. rehbinder. 1989. species of the genus elaphostrongylus, parasite of swedish cervidae. a description of e. alces n. sp. annales de parasitologie, humanie et comparee 64: 134–142. –––, and c. johansson. 1990. elaphostrongylus spp. from scandinavian cervidae–a scanning electron study (sem). rangifer 10: 39–46. doi: 10.7557/2.10.1.789. –––, i.-m. olsson, and e. broman. 2005. diseases in a moose population subjected to low predation. alces 41: 37–48. –––, s. persson, and l. hajdu. 1993. protostrongylidae in cervidae and ovibus moscatus; a clustering based on isoelectric focusing on nematode proteins. rangifer 13: 221–223. doi: 10.7557/ 2.13.4.1121. –––, and c. rehbinder. 1986. nervous tissue lesions caused by elaphostrongylosis in wild swedish moose. acta veterinaria scandinavia 27: 326–342. –––, and l. roepstorff. 1990. neurological disorder in two moose calves (alces alces l.) naturally infected with elaphostrongylus alces. rangifer, special issue 3: 399–406. doi: 10.7557/2.10.3.887. –––, i. y. warsame, and a. skorping. 1998. experimental infection of reindeer, sheep and goats with elaphostrongylus spp. (nematoda, protostrongylidae) from moose and reindeer. rangifer 18: 73–80. doi: 10.7557/2.18.2.1448. stuve, g. 1986. the prevalence of elaphostrongylus cervi infection in moose (alces alces) in southern norway. acta veterinaria scandia 27: 397–409. syroechkovsky, e. e., e. v. rogacheva, and l. a. renecker. 1989. moose husbandry. pages 369–386 in r. j. hudson, k. r. drew, and l. m. baskin, editors. wildlife production systems. economic utilization of wild ungulates. cambridge university press, cambridge, england. swedish association for hunting and wildlife management. 2015. http:// jagareforbundet.se (accessed july 2015). swedish meteorological and hydrological institute (smhi). 2015. http:// www.smhi.se (accessed september 2015). systat. 2002. table curve 2d, version 5.01. systat software inc., san jose, ca. –––. 2008. sigmaplot for windows version 11.0, build 11.0.0.75. systat software inc., san jose, california, usa. toft, c. a. 1991. an ecological perspective: the population and community consequences of parasitism. pages 319–343 in c. a. toft, a. aeschlimann, and l. bolis, editors. parasite-host: association coexistence or conflict? oxford university press, oxford, england. vicente, j., i. g. fernández de mera, and c. gortazar. 2006. epidemiology and risk factors analysis of elaphostrongylosis in red deer (cervus elaphus) from spain. parasitology research 98: 77–85. doi: 10.1007/s00436-005-0001-2. –––, and c. gortázar. 2001. high prevalence of large spiny-tailed protostrongylid larvae in iberian red deer. veterinary parasitology 96: 165–170. doi: 10.1016/ s0304-4017(00)00425-8. 28 epizootiology of elaphostrongylus alces – stéen et al. alces vol. 52, 2016 http://jagareforbundet.se http://jagareforbundet.se http://www.smhi.se http://www.smhi.se epizootiology of elaphostrongylus alces in swedish moose study area methods data management and statistics results infection, age and sex body condition time of infection discussion acknowledgements references alces28_137.pdf alces(25)_63.pdf alces28_15.pdf alces20_245.pdf alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces26_104.pdf alces24_133.pdf alces20_259.pdf alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 diversity and abundance of terrestrial gastropods in voyageurs national park, mn: implications for the risk of moose becoming infected with parelaphostrongylus tenuis tim cyr1, steve k. windels2, ron moen3, and jerry w. warmbold2,4 1integrated biosciences graduate program and natural resources research institute, university of minnesota duluth, duluth, minnesota 55811; 2voyageurs national park, 360 highway 11 east, international falls mn, 56649; 3natural resources research institute, 5013 miller trunk highway, duluth, minnesota 55811; 4present address: university of south dakota, 414 east clark st, vermillion, south dakota 57069. abstract: voyageurs national park (vnp) has a stable population of about 40–50 moose (alces alces). recent declines in moose abundance in adjacent areas in northern minnesota raise concerns about the long-term viability of moose in vnp. the parasitic nematode parelaphostrongylus tenuis has been documented in moose in vnp and has been implicated in moose declines in other populations. terrestrial gastropods are the intermediate hosts for p. tenuis, and describing spatial and temporal differences in their abundance should increase understanding about the risk of p. tenuis infection for vnp moose at the individual and population levels. we used cardboard sheets to estimate species composition and abundance of terrestrial gastropods in representative vegetation communities in vnp. we collected a total of 6,595 gastropods representing 25 species, 22 terrestrial snails and 3 slugs; 8 are known vectors of p. tenuis, including the slug deroceras laeve, the most common species found. gastropods were more abundant in september than july, and in upland forests (maximum = 555 gastropods/m2) more than in wetter lowlands (20 gastropods/m2). we used location data from gpscollared moose in vnp to estimate the relative exposure of moose to gastropods that could be infected with p. tenuis larvae. the boreal hardwood forest and northern spruce-fir forest ecotypes had the highest use by moose and high abundance of p. tenuis vectors in summer, and may pose the greatest risk for infection. habitat use and the related risk of ingesting gastropod vectors varied by individual moose. our method can be extended in moose range to estimate the relative risk of p. tenuis infection. alces vol. 50: 121–132 (2014) key words: alces, meningeal worm, minnesota, moose, p. tenuis, parasite introduction the parasitic nematode parelaphostrongylus tenuis can be fatal to moose (alces alces) (anderson 1964), and was the probable cause of 5% of mortality of radio-collared moose in northwestern minnesota and >20% of incidentally-recovered moose in northern minnesota (murray et al. 2006, wünschmann et al. 2014). the infection causes weakness in the hindquarters, circling, tilting of the head, and increased fearlessness of humans (anderson and prestwood 1981). infections can be lethal and cause mortality indirectly through increased risk of predation or accidents (lankester et al. 2007, butler et al. 2009, wünschmann et al. 2014). voyageurs national park (vnp) in northern minnesota maintains a stable, low-density population of about 40–50 moose, and p. tenuis infection and associated mortality has been documented in and surrounding vnp (windels 2014). though the effect on moose at the population level in vnp is unknown, previous studies suggest it is unlikely to be a major 121 mortality source at the currently low whitetailed deer (odocoileus virginianus) density (3–6 deer/km2) (whitlaw and lankester 1994b). the normal lifecycle of p. tenuis includes white-tailed deer as the definitive host and terrestrial gastropods as intermediate hosts (lankester and anderson 1968). white-tailed deer ingest infected gastropods while foraging and gastropods become infected with p. tenuis by crawling over or near infected deer feces (lankester 2001). however, only 0.1–4.2% of gastropods collected in minnesota and ontario were infected with p. tenuis larvae (lankester and anderson 1968, lankester and peterson 1996). at those infection rates, a white-tailed deer would need to consume up to 1000 gastropods to encounter a single larva (lenarz 2009). however, lankester and peterson (1996) reasoned that even at such low rates of infection in gastropods, the high rates of infection measured in white-tailed deer in the region (≤91%, slomke et al. 1995) is explained by the large volume of vegetation eaten on and near the ground over a few months in the autumn. infection rates in white-tailed deer derived from winter fecal samples have ranged from 67–90% from the 1970s to the present in vnp (gogan et al. 1997, vanderwaal et al. 2014). white-tailed deer are the definitive host of p. tenuis but moose, an aberrant host, also ingest infected gastropods during foraging and become infected. initial signs of p. tenuis infection can appear in moose as early as 20 days after experimental infection (lankester 2002). gastropods are necessary for p. tenuis to complete its life cycle. therefore, knowledge of gastropod populations in vnp may help managers better understand the role of p. tenuis in local moose population dynamics. the distribution and habitat preferences of terrestrial gastropods in vnp have not been studied previously. extrapolation from studies of gastropod communities in different regions of minnesota and the surrounding areas is possible (e.g., from northwestern minnesota [nekola et al. 1999] or rock outcrops in northeastern minnesota [nekola 2002]). gastropods exhibit habitat preferences that result in variation in presence or density across vegetation communities or other habitat features, and few studies have examined their abundance and diversity at fine spatial scales (moss and hermanutz 2010). the risk of p. tenuis infection is presumably influenced by vector density and could vary within a population because individual moose demonstrate differential habitat use (gillingham and parker 2008). fine-scale habitat use derived from gps collars can help clarify the risk of p. tenuis infection to individuals and populations of moose. combined, individual differences in habitat use and variability among habitat types in gastropod diversity and abundance may result in differential risk of moose and other cervids to p. tenuis infection (vanderwaal et al. 2014). in this study we surveyed terrestrial gastropod species on the kabetogama peninsula in vnp. our objectives were to 1) estimate the abundance and diversity of terrestrial gastropods in different ecotypes, with particular focus on known vectors of p. tenuis, 2) document changes in gastropod abundance over the growing season, and 3) compare the use of cover types by gps-collared moose to density of p. tenuis vectors to estimate the encounter risk of individual moose. study area voyageurs national park (48.50° n, 92.88° w) is an 882 km2 protected area comprised of a mixture of forested land (61%) and large lakes (39%) along the u.s.-canada border. moose are primarily restricted to the kabetogama peninsula (windels 2014), a 300 km2 roadless area in the center of vnp, and have remained relatively stable 122 gastropod vectors of p. tenuis in vnp – cyr et al. alces vol. 50, 2014 since the 1990s with density ranging from 0.14–0.19 moose/km2 (windels 2014). white-tailed deer density in winter during the study ranged between 3–6/km2 (gogan et al. 1997, unpublished data of vnp). vegetation is a mix of southern boreal and laurentian mixed conifer-hardwood forests comprised primarily of a mosaic of quaking aspen (populus tremuloides), paper birch (betula papyrifera), balsam fir (abies balsamea), white spruce (picea alba), white pine (pinus strobus), red pine (p. resinosa), jack pine (p. banksiana), and black spruce (picea mariana) (faber-landgendoen et al. 2007). soils range from thin, sandy loams over bedrock to poorly draining clays at lower elevations (kurmis et al. 1986). beaver-created wetlands and associated seral stages are abundant (johnston and naiman 1990). temperatures vary from −40 to 36 °c, with an average annual temperature of 1.4 °c. mean annual precipitation is 62 cm, with most precipitation falling between may and september (kallemeyn et al. 2003). methods we used the “ecotype”-level vegetation classification derived from the usgs-nps vegetation map (hop et al. 2001) to select the 10 most common terrestrial ecotypes on the kabetogama peninsula to sample for gastropods. we excluded 4 of these because they were too wet to sample with our methodology: poor conifer swamps, rich hardwood swamps, wet meadows, and shrub bogs. the remaining 6 ecotypes comprised 80% of the non-aquatic vegetation communities (table 1); 4 were dry uplands (rock barrens with trees, northern spruce-fir forests, boreal hardwood forests, and northern pine forests) and 2 wet lowland ecotypes (northern shrub swamp and rich conifer swamp). we randomly selected 5 patches (polygons) within each of the 6 ecotypes within a restricted area to facilitate access to sampling sites (fig. 1) assuming that these sites were representative of those across the entire peninsula. at each site we sampled during a single over-night period at approximately 1-month intervals in each of 4 periods: 6– 20 june, 29 july–3 august, 18–25 august, and 9–14 september. we used 0.25 m2 cardboard sampling squares (50 � 50 cm) placed on ground vegetation to collect gastropods (lankester and peterson 1996, hawkins et al. 1998, nankervis et al. 2000, maskey 2008). we randomly selected a starting sample point and direction within each polygon such that a 100-m sampling transect would fit entirely within the polygon. we placed 10 corrugated cardboard squares on the 100-m transect and verified that all were in the same ecotype. the cardboard was placed directly on the soil or duff layer after rocks and branches were cleared from the sampling site. the cardboard was saturated with water and covered with a 0.36 m2 sheet of 3-mm thick clear plastic. sheets were set in the morning and retrieved ∼24 h later. the wetness of each sheet was estimated as the percentage of the bottom that was visibly damp. all slugs table 1. area (km2) and % total area covered by each of 6 terrestrial vegetation ecotypes sampled on the kabetogama peninsula, voyageurs national park (vnp), minnesota, usa, juneseptember 2011. area calculations exclude lakes and ponds. ecotype classifications are according to the us-national vegetation classification system applied to vnp (hop et al. 2001). ecotype area (km2) % northern spruce-fir forest 66 23 boreal hardwood forest 62 21 northern pine forest 52 18 treed rock barrens 39 13 northern shrub swamp 8 3 rich conifer swamp 5 2 total 232 80 alces vol. 50, 2014 cyr et al. – gastropod vectors of p. tenuis in vnp 123 and snails on the underside of the cardboard were collected and stored in plastic jars with damp paper towels. subsequent identification was to the lowest taxonomic level possible using available keys (burch 1962, nekola 2007, j. nekola, minnesota department of natural resources, pers. comm.). in 3 cases, we lumped 2 closely related species together that could not be reliably differentiated by morphological characteristics: zonitoides nitidus and z. arboreus, nesovitrea electrina and n. binneyana, and euconulus alderi and e. fulvous. we identified potential gastropod vectors of p. tenuis based on a literature review (lankester and anderson 1968, gleich et al. 1977, upshall et al. 1986, rowley et al. 1987, platt 1989, lankester and peterson 1996, whitlaw et al. 1996, nankervis et al. 2000, lankester 2001). we considered the 100-m sample transect the sample unit and tested for the effects of ecotype and sample period on abundance of gastropod groups (total gastropods, snails only, slugs only) using factorial anova. we also tested for an interaction between ecotype and sampling period. we used bonferroni corrections when making post-hoc comparisons between main effects (ecotype and sample period) and set statistical significance at p = 0.05. we obtained gps locations at 15-min intervals from 11 adult moose (9f:2m) wearing gps collars to measure habitat use during june-september 2010. spatial data were displayed using arcgis 10.1 with arcgis spatial analyst (esri, redlands, ca, usa 2012), and home ranges were calculated in the geospatial modeling environment fig. 1. primary moose range (dashed line) and terrestrial gastropod sampling area (crosshatched area) in voyageurs national park, minnesota, usa, 6 july-14 september 2010. 124 gastropod vectors of p. tenuis in vnp – cyr et al. alces vol. 50, 2014 (2012 spatial ecology llc) running via arcgis 10.1 and r 3.0.1. we calculated the proportion of locations in each ecotype for individual moose. we calculated a relative measure of p. tenuis transmission risk to moose in different ecotypes by comparing the abundance of gastropods in each ecotype to habitat use in each ecotype. mean monthly habitat use (i.e., proportion of all locations within an ecotype) varied little from june to september; all differences were <5% between months for any ecotype. we therefore used the mean proportion of use for the entire june-september period to estimate an overall risk of p. tenuis infection by ecotype during summer. we also evaluated variation in relative risk of p. tenuis infection to individual moose. risk value was calculated by multiplying the proportion of each ecotype used by a moose by the mean density of potential p. tenuis gastropod vectors measured in each ecotype. we scaled the risk value for each moose to the highest individual risk value to compare relative risk of infection among individual moose. our indices of risk assume that 1) gastropod infection rates (i.e., proportion of gastropods infected with p. tenuis larvae) did not vary among gastropod species, among habitat types, or over the sampling time, and 2) the likelihood of a moose ingesting a potentially infected vector gastropod in a given ecotype is proportional to the density of known vectors of p. tenuis in that ecotype. our index of risk does not consider morbidity or mortality for infected moose, because the severity and duration of the infection can be highly variable (lankester 2002, 2010). results we collected 6,595 gastropods representing 9 families and 25 species (3 slug species and 22 snail species; table 2), and successfully classified 62% of slugs and 50% of snails. we could not identify 3,116 (47%) of the gastropods because they were damaged beyond recognition during collection and storage, or were juveniles that can be difficult to identify accurately even to the family level (j. nekola, pers. comm.). the total number of snails/m2 (including unidentified) increased from july to september in all ecotypes combined (anova, f3,29 = 8.7, p < 0.001). the treed rock barren cover type had the lowest snail density (7.1/m2) for all sampling periods combined. the northern pine forest and northern spruce-fir forest ecotypes had the most snails for all periods combined, increasing from 7.3 and 10.2 snails/m2 in july to 23.7 and 22.8 snails/m2 in september, respectively (fig. 2). overall, slug density was relatively constant over time within each ecotype, and at lower density than snails. slug density (including unidentified) was more variable over time than snail density (fig. 3). slug density in all 4 sampling periods combined was lowest (1.3/m2) in the rich conifer swamp ecotype and highest in the northern pine (6.2/m2) and northern spruce-fir forests (6.9/m2). northern shrub swamp (3.3/m2) and rich conifer swamp (1.3/m2) had lower slug densities than the other 4 ecotypes (anova, f5,29 = 20.88, p < 0.001). cardboard wetness increased as the survey progressed (anova, f3,29 = 165.8, p < 0.001); for example, mean wetness was 47% in survey 1 and 90% in survey 4. within ecotypes, cardboard wetness in the treed rock barren ecotype was lower (51%) than in the other 5 ecotypes (range = 75– 82%; anova, f5,29 = 44.3, p < 0.001). eight of the collected species are known vectors of p. tenuis and comprised 32% of the sample. the slug deroceras laeve was the most common vector collected (26% of total captures), was present in every ecotype, and most common in the northern spruce-fir forest ecotype. two other slug vectors were pallifera hemphili and a deroceras specimen that we could not identify to species, but alces vol. 50, 2014 cyr et al. – gastropod vectors of p. tenuis in vnp 125 assumed was a p. tenuis vector like its congener d. leave. the snails discus cronkhitei, zonitoides nitidus+arboreas, strobilops spp., and cochlicopa sp., known vectors of p. tenuis, were ∼11% of the sample and found across all surveys and sample sites (table 2). risk of p. tenuis infection was highest in northern spruce-fir forests (fig. 4). the northern spruce-fir ecotype had the highest use by moose (35% of total locations) and also had the second highest estimated density of p. tenuis vectors. treed rock barrens had the fourth highest use by moose (8%) table 2. composition of terrestrial gastropods collected in voyageurs national park, minnesota, usa, june-september 2011. gastropod species were identified to the lowest taxonomic level possible; 62% of slugs and 50% of snails were classified. group family species count % total captures p. tenuis vectors slug limacidae deroceras laeve 906 26.0 slug limacidae deroceras sp. (but not d. leave) 13 0.4 slug philomycidae pallifera hemphili 6 0.2 snail endodontidae discus cronkhitei 55 2.0 snail strobilopsidae strobilops spp. 145 4.0 snail valloniidae cochlicopa sp. 6 0.2 snail zonitidae zonitoides (nitidus+arboreas) 159 4.6 total 1290 37.4 non-vectors snail endodontidae helicodiscus parallelus 7 0.2 snail endodontidae punctum californicum 7 0.2 snail endodontidae punctum minutissimum 2 <0.1 snail endodontidae punctum spp. 4 0.1 snail oxychilidae nesovitrea (electrina+binneyana) 105 3.0 snail pupillidae columella simplex 6 0.2 snail pupillidae gastrocopta pentodon 6 0.2 snail pupillidae gastrocopta sp. 11 0.3 snail pupillidae vertigo spp. 319 9.0 snail pupillidae unknown 143 4.0 snail succineidae oxyloma retusa 19 0.5 snail valloniidae cochlicopa lubricella 11 0.3 snail valloniidae zoogenetes harpa 62 2.0 snail zonitidae euconulus (alderi + fulvous) 638 18.0 snail zonitidae guppya sterkii 6 0.2 snail zonitidae striatura milium 29 0.8 snail zonitidae striatura exigua 7 0.2 snail zonitidae striatura ferrea 6 0.2 snail zonitidae vitrina limpida 326 9.0 snail zonitidae unknown 461 13.0 total 2175 61.4 126 gastropod vectors of p. tenuis in vnp – cyr et al. alces vol. 50, 2014 and the third highest p. tenuis vector density, suggesting moderate risk. boreal hardwood forests were also a moderate risk ecotype based on their relatively high use and low vector density. rich conifer swamps and northern shrub swamps were low risk fig. 2. mean (+se) number of snails/m2 (including unidentified) measured in each of 6 ecotypes for a single over-night period in each of 4 sampling periods in juneseptember, 2011 in voyageurs national park, minnesota, usa. sample periods were: survey 1 = 6–20 june, survey 2 = 29 july – 3 august, survey 3 = 18–25 august, survey 4 = 9–14 september. fig. 3. mean (+se) number of slugs/m2 (including unidentified) measured in each of 6 ecotypes for a single over-night period in each of 4 sampling periods in juneseptember, 2011 in voyageurs national park, minnesota, usa. sample periods were: survey 1 = 6–20 june, survey 2 = 29 july – 3 august, survey 3 = 18–25 august, survey 4 = 9–14 september. alces vol. 50, 2014 cyr et al. – gastropod vectors of p. tenuis in vnp 127 ecotypes because of their relatively low use (5% and 7%) and low density of p. tenuis vectors (fig. 4). moose displayed variability in individual risk of infection as a result of differential habitat use. ten of 11 moose had relative risk scores of 0.68–1.0, and relative risk differed by ≤32% for the majority of moose. moose v09 was an exception as it spent little time in gastropod rich habitats and had a much lower risk of infection (0.21) relative to the other moose (table 3). discussion gastropod density, and more specifically density of known vectors of p. tenuis, differed among the ecotypes and sample periods. similar to previous studies, ecotypes of mixed conifer-deciduous forest types had the highest gastropod densities (gleich et al. 1977, kearney and gilbert 1978, nankervis et al. 2000). the increasing density of gastropods and potential p. tenuis vectors from summer to fall is also consistent with previous studies in northern minnesota (lankester and peterson 1996). d. laeve was the most abundant gastropod found in our study area, and is likely the most important vector of p. tenuis. most larvae in infected gastropods are presumably in the infective stage (i.e., third stage) by early july (lankester and peterson 1996) corresponding to our sampling period between mid-june and september. the cardboard sampler method is meant to provide a relative measure of gastropod diversity and abundance, and it is critical that they be as uniform as possible in shape, thickness, and wetness. all were saturated with water at the time of deployment but dried at different rates depending on habitat features (e.g., soil moisture, rockiness, exposure) and weather conditions (e.g., dry and windy vs. calm and humid). cardboard wetness varied from 0–100% at collection and this wide variation could skew the estimates of gastropod abundance because they are less likely to be found on dry cardboard (unpublished data, vnp). variation in cardboard wetness could be minimized by distributing the cardboard after the warmest part of the day and checking them before the warmest part of the next day, which would be especially important in the longer and warmer days in july and early august. past studies indicate lower boreal hardwood forest northern pine forest northern shrub swamp northern sprucefir forest rich conifer swamp treed rock barren 5 7 9 11 13 15 17 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 g as tr o p o d v ec to rs /m 2 o f tr ap propor�on of habitat use fig. 4. relative risk of moose encountering p. tenuis gastropod vectors in 6 ecotypes in voyageurs national park, minnesota, usa, july-september 2010. 128 gastropod vectors of p. tenuis in vnp – cyr et al. alces vol. 50, 2014 gastropod abundances in early summer (lankester and anderson 1968, kearney and gilbert 1978, lankester and peterson 1996), and although these studies did not report the relative wetness of cardboard sheets, they may be biased low if sheets were drier in early summer. cardboard samplers may underestimate the total density of gastropods in an area, as the number of gastropods in the soil underneath cardboard samplers has been reported higher than those attached to the cardboard samplers (hawkins et al. 1998). by combining information about gastropod density and relative habitat use, we assessed the relative risk of p. tenuis infection for moose in different habitat types (fig. 4). we likewise calculated risk values for individual moose (table 3). these methods can also be used to compare risk of infection between different geographic areas or populations. however, we caution that the assumptions associated with our methods need to be considered carefully because seasonal variation of infection rates in gastropod hosts is not well understood (lankester and anderson 1968, kearney and gilbert 1978, lankester and peterson 1996). high whitetailed deer density has been correlated with increased infection rates of gastropods (lankester and peterson 1968) and moose (whitlaw and lankester 1994a) at larger spatial scales. a recent study found no correlation between white-tailed deer abundance and p. tenuis infection at smaller spatial scales within vnp (vanderwaal et al. 2014), although the range of deer abundance was limited across sites. risk of p. tenuis infection varies among individual moose because of differences in habitat use within respective home ranges. it will also be influenced by landscape composition and the availability of different habitats within an area. for example, the western half of the kabetogama peninsula has more area covered by the higher risk table 3. proportional habitat use and individual risk of moose encountering p. tenuis infected gastropods in the kabetogama peninsula, voyageurs national park, minnesota, usa, june-september 2010. risk value is calculated by multiplying the proportion of each ecotype used by a moose by the mean density of p. tenuis gastropod vectors measured in each ecotype. the relative index of risk is the risk value scaled to the highest risk value found for an individual moose in 2010 (i.e., moose v05). proportion habitat use moose # northern pine forest northern sprucefir forest treed rock barren boreal hardwood forest northern shrub swamp rich conifer swamp risk value relative index of risk v05 0.43 0.18 0.02 0.24 0.00 0.03 9.50 1.00 v06 0.42 0.10 0.02 0.17 0.03 0.05 8.96 0.94 v07 0.09 0.34 0.11 0.18 0.07 0.04 8.91 0.94 v14 0.05 0.35 0.14 0.19 0.10 0.04 8.76 0.92 v07 0.04 0.37 0.10 0.18 0.06 0.03 7.80 0.82 v10 0.05 0.33 0.10 0.19 0.06 0.03 7.66 0.81 v18 0.07 0.19 0.26 0.19 0.02 0.00 7.64 0.80 v17 0.13 0.25 0.08 0.18 0.03 0.02 7.44 0.78 v12 0.10 0.32 0.03 0.21 0.06 0.05 7.23 0.76 v08 0.01 0.37 0.00 0.27 0.02 0.01 6.49 0.68 v09 0.00 0.17 0.00 0.13 0.09 0.11 2.04 0.21 alces vol. 50, 2014 cyr et al. – gastropod vectors of p. tenuis in vnp 129 boreal hardwood and northern spruce-fir ecotypes, and conversely, the eastern half of the park contains more of the drier, low risk treed rock barrens and northern pine ecotypes. vanderwaal et al. (2014) found that p. tenuis infection rates in white-tailed deer increased as the proportion of vector-rich habitats such as mixed conifer-hardwood forest increased within a local area. while our methods only considered coarse habitat use in our risk index, moose behavior within individual ecotypes is presumably also important. moose may prefer to bed in certain ecotypes (e.g., in lowland habitats in hot weather) and feed in others (peek 1997), and even if gastropods are abundant in certain ecotypes, the risk of p. tenuis infection should be less in areas less preferred for foraging. risk of infection may also be affected by individual preferences for forage choice, previous exposure to p. tenuis, health status, genetics, body mass/longevity (ezenwa et al. 2006), and other factors not considered here. acknowledgments we thank m. lankester and two anonymous reviewers for comments that improved our manuscript. we thank n. walker, b. olson, and w. chen for project assistance. funding for this work was provided by voyageurs national park, usgsnps national resource preservation program, the environment and natural resources trust fund, the integrated biosciences graduate program at the university of minnesota duluth, and the natural resources research institute. the north american moose conference and workshop newcomers travel grant provided a travel grant to the senior author. references anderson, r. c. 1964. neurologic disease in moose infected experimentally with pneumostrongylus tenuis from whitetailed deer. pathologia veterinaria 1: 289–322. ———, and a. k. prestwood. 1981. lungworms. pages 266–317 in w. r. davidson, f. a. hayes, v. f. nettles, and f. e. kellogg, editors. diseases and parasites of the white-tailed deer. miscellaneous publications number 7, tall timbers research station, tallahassee, florida, usa. burch, j. b. 1962. how to know the eastern land snails. w. c. brown company, dubuque, iowa, usa. butler, e., m. carstensen, m. schrage, d. pauly, m. lenarz, and l. cornicelli. 2009. preliminary results from the 2007–2008 moose herd health assessment project. 2009 summaries of wildlife research findings. minnesota department of natural resources, st. paul, minnesota, usa. ezenwa, v., s. a. price, s. altizer, n. d. vitone, and c. cook. 2006. host traits and parasite species richness in even and odd toed hoofed mammals, artiodactyla and perissodactyla. oikos 115: 526–537. faber-landgendoen, d., n. aaseng, k. hop, and m. lew-smith. 2007. field guide to the plant community types of voyageurs national park. u.s. geological survey techniques and methods 2–a4. gillingham, m. p., and k. l parker. 2008. the importance of individual variation in defining habitat selection by moose in northern british columbia. alces 44: 7–20. gleich, j. g., f. f. gilbert, and n. p. kutscha. 1977. nematodes in terrestrial gastropods from central maine. journal of wildlife diseases 13: 43–46. gogan, p. j. p., k. d. kozie, and e. m. olexa. 1997. ecological status of moose and white-tailed deer at voyageurs national park, minnesota. alces 33: 187–201. hawkins, j. w., m. w. lankester, and r. r. a. nelson. 1998. sampling terrestrial 130 gastropod vectors of p. tenuis in vnp – cyr et al. alces vol. 50, 2014 gastropods using cardboard sheets. malacologia 39: 1–9. hop, k., d. faber-landgendoen, m. lewsmith, n. aaseng, and s. lubinski. 2001. voyageurs national park, minnesota: usgs-nps vegetation mapping program. u.s. department of interior, u.s. geological survey, upper midwest environmental sciences center, la crosse, wisconsin, usa. johnston, c. a., and r. j. naiman. 1990. aquatic patch creation in relation to beaver population trends. ecology 71: 1617–1621. kallemeyn, l. w., k. l. holberg, j. a. perry, and b. y. odde. 2003. aquatic synthesis for voyageurs national park. usgs information and technology report 2003-0001. kearney, s. r., and f. f. gilbert. 1978. terrestrial gastropods from the himsworth game preserve, ontario, and their significance in parelaphostrongylus tenuis transmission. canadian journal of zoology 56: 688–694. kurmis, v., s. l. webb, and l. c. merriam. 1986. plant communities of voyageurs national park, minnesota, u.s.a. canadian journal of botany 64: 531–540. lankester, m. w. 2001. extrapulmonary lungworms of cervids. pages 228–278 in w. m. samuel, m. j. pybus, and a. a. kocan, editors. parasitic diseases of wild mammals, 2nd edition. iowa state university press, ames, iowa, usa. ———, 2002. low-dose meningeal worm (parelaphostrongylus tenuis) infections in moose (alces alces). journal of wildlife diseases 38: 789–795. ———, 2010. understanding the impact of meningeal worm, parelaphostrongylus tenuis, on moose populations. alces 46: 53–70. ———, and r. c. anderson. 1968. gastropods as intermediate host of meningeal worm, pneumostrongylus tenuis, dougherty. canadian journal of zoology 46: 373–383. ———, and w. j. peterson. 1996. the possible importance of wintering yards in the transmission of parelaphostongylus tenuis to white tailed deer and moose. journal of wildlife diseases 32: 31–38. ———, ———, and o. ogunremi. 2007. diagnosing parelaphostrongylosis in moose (alces alces). alces 43: 49–59. lenarz, m. s. 2009. a review of the ecology of parelaphostrongylus tenuis in relation to deer and moose in north america. pages 70–75 in m. w. doncarlos, r. o. kimmel, j. s. lawrence, and m. s. lenarz, editors. 2009 summaries of wildlife research findings. minnesota department of natural resources, st. paul, minnesota, usa. maskey, j. 2008. movements, resource selection, and risk analyses for parasitic disease in an expanding moose population in the northern great plains. ph. d. thesis, university of north dakota, grand forks, north dakota, usa. moss, m., and l. hermanutz. 2010. monitoring the small and slimy – protected areas should be monitoring native and nonnative slugs (mollusca: gastropoda). natural areas journal 30: 322–327. murray, d. l., e. w. cox, w. b. ballard, h. a. whitlaw, m. s. lenarz, t. w. custer, t. barnett, and t. k. fuller. 2006. pathogens, nutritional deficiency, and climate influences on a declining moose population. wildlife monographs 166: 1–30. nankervis, p. j., w. m. samuel, s. m. schmitt, and j. g. sikarskie. 2000. ecology of meningeal worm, p. tenuis (nematode), in white-tailed deer and terrestrial gastropods of michigan's upper peninsula with implications for moose. alces 36: 163–181. nekola, j. c. 2002. distribution and ecology of terrestrial gastropods in northwestern minnesota. final report to 2001– 2001 nongame and natural heritage research program, minnesota alces vol. 50, 2014 cyr et al. – gastropod vectors of p. tenuis in vnp 131 department of natural resources, st. paul, minnesota, usa. ———. 2007. key to the terrestrial gastropod genera of wisconsin and nearby states. <http://www.uwlax.edu/biology/ faculty/perez/perez/perezlab/research/ snailkey%20-%20april%2027%20 updates.pdf> (accessed may 2012). ———, m. barthel, p. massart, and e. north. 1999. terrestrial gastropod survey of igneous outcrops in northeastern minnesota. final report to 1998 nongame and natural heritage research program, minnesota department of natural resources st. paul, minnesota, usa. peek, j. m. 1997. habitat relationships. pages 351–376 in a. w. franzmann and c. c. schwartz, editors. ecology and management of the north american moose. wildlife management institute, washington d. c., usa. platt, t. r. 1989. gastropod intermediate hosts of parelaphostrongylus tenuis (nematoda: metastrongyloidea) from northwestern indiana. journal of parasitology 75: 519–523. rowley, m. a., e. s. loker, and j. f. pagels. 1987. terrestrial gastropod hosts of parelaphostrongylus tenuis at the national zoological park's conservation and research center, virginia. journal of parasitology 73: 1084–1089. slomke, a. m., m. w. lankester, and w. j. peterson. 1995. infrapopulation dynamics of parelaphostrongylus tenuis in white-tailed deer. journal of wildlife diseases 31: 125–135. upshall, s. m., m. d. b. burt, and t. g. dilworth. 1986. parelaphostrongylus tenuis in new brunswick: the parasite in terrestrial gastropods. journal of wildlife diseases 22: 582–585. vanderwaal, k., s. k. windels, b. t. olson, t. vannatta, and r. moen. 2014. spatial epidemiology of liver fluke and meningeal worm in white-tailed deer in northern minnesota, usa. international journal of parasitology. in press. wünshmann, a., a. g. armien, e. butler, m. schrage, b. stromberg, j. b. bender, a. f. firshman, and m. carstensen. 2014. necropsy findings in 62 opportunistically collecting freeranging moose (alces alces) from minnesota, usa (2003–2013). journal of wildlife diseases. doi: 10.7589/201402-037. windels, s. k. 2014. 2014 voyageurs national park moose population survey report. natural resource data series nps/voya/nrds-2014/645. national park service, fort collins, colorado, usa. whitlaw, h. a., and m. w. lankester. 1994a. a retrospective evaluation of the effects of parelaphostongylosis on moose populations. canadian journal of zoology 72: 1–7. ———, and ———. 1994b. the co-occurrence of moose, white-tailed deer, and parelaphostrongylus tenuis in ontario. canadian journal of zoology 72: 819–825. ———, ———, and w. b. ballard. 1996. parelaphostrongylus tenuis in terrestrial gastropods from white-tailed deer winter and summer range in northern new brunswick. alces 32: 75–83. 132 gastropod vectors of p. tenuis in vnp – cyr et al. alces vol. 50, 2014 http://www.uwlax.edu/biology/faculty/perez/perez/perezlab/research/snailkey%20-%20april%2027%20updates.pdf http://www.uwlax.edu/biology/faculty/perez/perez/perezlab/research/snailkey%20-%20april%2027%20updates.pdf http://www.uwlax.edu/biology/faculty/perez/perez/perezlab/research/snailkey%20-%20april%2027%20updates.pdf http://www.uwlax.edu/biology/faculty/perez/perez/perezlab/research/snailkey%20-%20april%2027%20updates.pdf diversity and abundance of terrestrial gastropods in voyageurs national park, mn: implications for the risk of moose becoming infected with parelaphostrongylus tenuis introduction study area methods results discussion acknowledgments references alces22_preface.pdf alces vol. 22, 1986 alces28_243.pdf alces21_501distmoosebio.pdf alces vol. 21, 1985 alces26_64.pdf the changing role of hunting in sweden—from subsistence to ecosystem stewardship? sara lindqvist1,3, camilla sandström1,2, therese bjärstig2, and emma kvastegård1 1swedish university of agricultural sciences, department of wildlife, fish, and environmental sciences, skogsmarksgränd, 901 83 umeå, sweden; 2umeå university, department of political sciences, umeå universitet, 901 87 umeå, sweden. abstract: although hunting served traditionally to supply game meat, and that is still important in sweden, recreation is the most common reason for hunting moose (alces alces) today. hunting also serves an important management purpose in regulating moose populations to control crop and forest damage. this study used semi-structured interviews with key stakeholders and officials involved in the recently implemented ecosystem-based, adaptive local moose management system where hunters and landowners become environmental stewards responsible for managing moose in context with forest damage, vehicular collisions, large carnivores, and biodiversity. our study found that participation and collaboration in reaching management objectives was perceived as positive by stakeholders, although their stewardship is jeopardized if specific management responsibilities are not clarified regarding monitoring. further, it is important to find long-term funding solutions for monitoring activities that are critical for adequate data collection and to support the stakeholder role as steward. the importance of monitoring must be communicated to individual hunters and landowners to achieve an ecosystem-based moose management system that effectively incorporates both social and ecological values. alces vol. 50: 53–66 (2014) key words: adaptive, alces alces, biodiversity, knowledge-based, management areas, monitoring, moose, sweden. historically, hunting was primarily associated with human subsistence and livelihood, whereas today it is mostly associated with recreation (hendee 1974, barnard 2004), and has gradually developed into a tool to meet sustainable wildlife and ecosystem management objectives (holsman 2000, fischer et al. 2012). with this shift, hunters and other stakeholders should be directly or indirectly involved in environmental stewardship; i.e., the responsible use and protection of the natural environment through conservation and sustainable practices (leopold 1950, holsman 2000, chapin et al. 2009). moose (alces alces) hunting in sweden embodies this development where utility and leisure lately have become closely intertwined with management objective such as game population control (holsman 2000, council of europe 2007). a new moose management system was implemented in sweden in january 2012 that emphasizes the stewardship role of hunters and landowners. it provides a unique opportunity to analyze the extent that stakeholders support this institutional change and whether the new system offers the resources necessary for hunters and landowners to exercise ecological stewardship. the swedish moose management system has evolved several times in past decades to balance the interest of hunters (i.e., high moose populations) with other societal 3present adress: sjövägen 96, 834 34 brunflo, sweden 53 interests, most notably the commercial forestry sector concerned with browsing on economically valuable tree species. an increasing number of moose-vehicle collisions (mvc), negative effects on agriculture, and biodiversity are also of high concern (e.g., angelstam et al. 2000, edenius et al. 2002, lavsund et al. 2003, seiler 2003). because these incremental changes did not resolve these conflicts, the swedish parliament decided, in one step, to move from a top-down administrative system consisting of a patchwork of organizational management units, into an ecosystem-based, adaptive local management system incorporating moose biology components like home range size and seasonal migration in combination with stakeholder engagement (swedish government 2010; see also broman 2003, wennberg-digasper 2008). to avoid repeating previous management failures, the main objective of this reform was to establish a knowledge-based and adaptive moose management system with the capacity to balance different interests (sandström et al. 2013). in particular, monitoring that was performed irregularly and incoherently across sweden, is now considered a key focus in the new system to integrate knowledge for establishing and evaluating management objectives. accordingly, the extent to which the new management system will succeed is dependent on 3 central elements of ecological stewardship: the will among hunters and landowners to 1) support wildlife management program goals designed to balance social and ecological values, 2) support and participate in the development of institutions for defining and implementing stewardship goals, and 3) participate in monitoring activities related to meeting social or ecological objectives (holsman 2000, chapin et al. 2009). our objective was to study the implementation phase to assess stakeholder acceptance and capacity to handle the objectives of ecosystem-based institutions and monitoring as part of knowledge-based management. background compared to previous management systems, the current system has a pronounced national objective for long-term balance of the moose population with forest resources and societal interests (swedish official investigation 2009). to reach this objective, an official investigation identified the need to overcome institutional deficiencies, primarily the lack of collaboration between key stakeholders and an ecosystembased approach, but also the lack of systematic monitoring of moose and forest resources (swedish official investigation 2009). the institutional change included some redistribution of tasks between the management levels, but also adding a new management body at the ecosystem level (i.e., moose management areas) that covers the equivalent of a moose population (at least 50,000 ha in the south and 100,000 ha in the north). this approach should bridge the regional level with moose management units and license areas at the local level (table 1). the moose management areas are governed by a moose management group consisting of 3 landowners and 3 hunters. this group is responsible for 1) making an ecosystembased and adaptive moose management plan for their respective area (stretching over maximum 3 years), 2) advising hunters and landowners in creating local management plans within the moose management units, and 3) coordinating and evaluating monitoring activities (swedish official investigation 2009). the institutional amendment of the moose management system resulted in redistribution of funding as well, from primary administration use by the county administrative boards to include all monitoring of moose and forest resources (swedish government 2010). 54 hunting role in sweden – lindqvist et al. alces vol. 50, 2014 the lack of systematic monitoring is addressed by science-based recommendation to focus on 4 accepted monitoring methods to provide annual information about the moose population including harvest statistics, hunter observation rates, pellet-group counts, and calf weights (bergström et al. 2011, danell et al. 2011, ericsson and kindberg 2011, kindberg et al. 2011). jointly, these methods require long-term, standardized implementation to function as reliable indices either singly or in combination (bergström et al. 2011, månsson et al. 2011). to further meet the broader goals of the ecosystem-based system, information about large carnivore populations and mvcs will be evaluated, and assessment of forage and browsing damage are also necessary to address the primary management objective (swedish government 2010). moose damage survey methods and forage forecasts are suggested as the basis of monitoring and are preferably conducted every 3rd and 5th year, respectively (kalén and bergquist 2011, rolander et al. 2011). the damage survey methods estimate the proportion of old and fresh stem damage within a given height interval and area (rolander et al. 2011), and the forage forecasts estimate the availability and quality of food resources through combination of satellite mapping of clearcuts and field sampling (kalén and bergquist 2011). study area we conducted our study in 5 counties (västerbotten, dalarna, södermanland, västra götaland, and kronoberg) distributed across sweden: 62°00' n, 15°00' e (fig. 1). these counties were selected because of their differences in ecology, landownership, and use patterns that present varied challenges to fully implement the new ecosystem-based, adaptive local management system. the counties cover all swedish vegetation types including alpine, boreal, boreonemoral, and nemoral zones. forests are dominated by commercially valuable scots pine (pinus sylvestris) and norway spruce (picea abies), and by deciduous tree table 1. the institutional framework of the old and new swedish moose management systems. the swedish environmental protection agency has the ultimate responsibility at the national level with the swedish forest agency functioning primarily as an advisory and supporting authority. at the regional level the county administrative boards have authoritative responsibility for moose management. at the regional level in the new system, wildlife management delegations with members from all interest groups are included. at the ecosystem level in the new system are moose management areas consisting of moose management groups with stakeholder representatives (3 hunters and 3 landowners). the different categories of license areas in the old system are removed, and license areas or moose management units exist only at the local level in the new system. moose management units also include stakeholders at the local level. old system new system national level swedish environmental protection agency /swedish forest agency regional level county administrative boards county administrative boards including wildlife management delegations ecosystem level (50,000–100,000 ha) moose management areas led by a moose management group with 3 landowners and 3 hunters local level (10,000–15,000 ha) license areas (a, b, c, and e) wildlife management units moose management units (hunters and landowners) moose management units (hunters and landowners) license areas (in exceptional cases) alces vol. 50, 2014 lindqvist et al. – hunting role in sweden 55 species such as birch (betula spp.), aspen (populus tremula), and rowan (sorbus aucuparia), as well as broadleaved trees like oak (quercus spp.) in the south. the 5 counties also differ in several other ways that could affect local and regional moose management, including moose population density (hörnberg 2001) and health; for example, moose tend to be larger in the north (sand et al. 1995). predation by brown bears (ursus arctos) is mostly in västerbotten and dalarna, whereas wolves (canis lupus) occur in dalarna and västra götaland. the effect of predation on a local moose population varies among and within these counties depending on the composition and number of predators (sand et al. 2011). while roe (capreolus capreolus) and red deer (cervus elaphus) exist in all 5 counties, some of sweden’s densest fallow deer (dama dama) populations exist in västra götaland and södermanland; södermanland also has a small population of mouflon sheep (ovis aries orientalis). only these 2 counties were considered with potential for interspecific competition between moose and other ungulates. land ownership among the counties also differ with more commercial forest companies in the north, and more private land in the south; forest ownership in sweden is ∼51% private and 42% commercial (bergman and åkerberg 2006). methods representative officials and stakeholders within moose management areas were interviewed in each county. in total, 29 semi-structured, qualitative interviews were conducted in october–december 2012. phone interviews, except 2 face-to-face, were used to ensure a high response rate. an interview manual with a vast spectrum of qualitative questions regarding the moose management system guided the interviews, and all respondents were asked identical questions. the recorded interviews lasted 45–120 min and were transcribed in full. respondents were given the opportunity to read their transcribed interview, and to clarify and/or alter any content to ensure the information was as valid as possible. the first interview in each county was with the wildlife manager of the county administrative board, which is the regional authority responsible for wildlife management issues, including moose management. given their local knowledge, each was asked to suggest a typical management area that was neither more collaborative, nor more conflicted and turbulent, than any other management area in the county. the stakeholders (3 landowner and 3 hunters) in this management area were contacted for further fig. 1. the 5 counties that served as the study area; from the north, västerbotten, dalarna, södermanland, västra götaland and kronoberg counties. the figure in each county represents the number of moose management areas in that county. 56 hunting role in sweden – lindqvist et al. alces vol. 50, 2014 interviews. in addition, the swedish forest agency (sfa) forest manager who was responsible for wildlife issues in each county was interviewed, since the sfa is an important advisory authority for the stakeholders. there were 40 potential interviews in the study: 5 wildlife managers, 5 forest managers, and 6 stakeholders per county (n = 30) of which 19 were interviewed (9 landowners and 10 hunters) with 1–2 nonrespondents (unreachable or unwilling) in each county. interviews were conducted with at least one hunter and one landowner in each management group; therefore, the data were regarded as reasonably balanced and useful to analyze the new adaptive moose management system. the interviews were thoroughly read and all information regarding or related to monitoring was extracted from the material. specific quotes that strengthen, clarify, or illustrate general (or divergent) responses are provided in the results. to ensure integrity, respondents are anonymous and we only describe their stakeholder group or agency and county. the interviews were conducted in swedish and we present interview quotes based upon our translation to english. results willingness and capacity to support nationwide objectives we asked respondents about their attitudes towards the overarching objective of the new moose management system, the need for collaboration, and the ecosystembased approach. the management of moose in larger geographical areas, as opposed to smaller management units, and the comprehensive ecosystem approach were considered positive in all counties. all but one hunter and landowner felt they would be more able to actively influence the management process. the single hunter from västerbotten claimed that “local decisions have been moved even further away”. hunters and landowners also emphasized the increased collaboration between them as important in fostering management legitimacy and trust among all participants. there was a strong conviction among officials and stakeholders that the new system would enable fact-based rather than assumption-based management, and that information derived from moose monitoring, forage forecasts, and browsing pressure estimates would allow more local, detailed, and nuanced management decisions. two example quotes were: “an advantage is that we will be able to get a clear view of the moose population, and that we, with determined effort, will achieve a high-quality moose population” (hunter from dalarna), and “we will have a moose population that is adjusted to the forage production—that’s what’s important, that we do not have too many moose, but we use the resources we have in our forests” (landowner in kronoberg). the stakeholders were developing moose management area plans in all counties. the plan was considered an important tool towards realizing management objectives and was regarded as a living document which could be altered if conditions changed. the perceptions of system resilience and how quickly management might respond to change differed among respondents; some believed that a yearly revision was reasonable, while others thought it possible to make immediate amendments during a hunting season. rapid changes in the moose population would primarily be based on hunter observations, and to some degree on harvest statistics, but this assumes that reporting is relatively fast and that management groups meet frequently to assess the situation. how such reports would be used differed among respondents; some claimed that if hunter observations differed from specified sexratios (e.g., equal male:female ratio), then restrictions might be implemented. others alces vol. 50, 2014 lindqvist et al. – hunting role in sweden 57 observed that reporting is generally quite slow, and that the hunting season would likely be over before information could reach the moose management groups and subsequently to hunting teams. willingness and capacity to implement stewardship goals we asked respondents to describe the ongoing implementation process to understand the degree of support and willingness to participate in the development of institutions for defining and implementing stewardship goals. despite pervasive belief in the new moose management system and a willingness among stakeholders and officials to participate in the implementation process, several obstacles were identified in fully implementing the system. the first obstacle was the short time period between the political decision in april 2011 and when the moose management groups would initiate their work (january 2012); both officials and stakeholders claimed they had inadequate time. specifically, more time was needed to acquire monitoring information, to fully understand the function and implementation of the new moose management system, and to adjust work arrangements before the first 3-year plan was due. a landowner in kronoberg summarized this with: “the process has been too fast, we do not have enough knowledge, we have too little knowledge regarding our game populations, we have no monitoring methods we all agree on applying, we have poor knowledge of browsing damage, we have no good overview of forage forecasts either, we rule and believe that we can manage the moose population, and we almost become overconfident and imagine we can calculate, down to nearest decimal, how many moose we can harvest.” the second obstacle was the lack of available and sufficient knowledge from monitoring to define management plans and objectives. the need for better data and the desire to develop a better overview of resource status was apparent in all counties. a new database (algdata.se) providing easy access to monitoring information and statistics important for developing their management plans was an identified need. certain regional county data requirements differed. västra götaland respondents stressed the lack of resolution in forage forecasts, moose damage surveys, predator densities, and mvcs. they also preferred that numbers and data be available for each management unit or even each hunting team, rather than an average number at the county or management area level. in södermanland, the importance of including other ungulate species was highlighted with regard to species-specific browsing damage surveys. the third obstacle was the vagueness of management responsibilities regarding data collection, statistics, and monitoring. despite the high degree of awareness of the need for monitoring, the responsibilities for such activities were especially perceived as confusing among the respondents. the main emphasis of responsibility for implementation, interpretation, and evaluation of monitoring lies with the moose management groups in their respective areas; the moose management units are obliged by regulations to participate in monitoring (environmental protection agency 2011). yet, there was confusion concerning role and authority among both officials and stakeholders. a forest manager in västerbotten stated: “we should have a locally based management. and what does it mean? does it mean that the management unit or maybe even the level below [hunting team] holds the steering capacity? or does it mean that if the moose management group does not approve the unit’s management plan then you have to do it all over again?”. the county administrative board, the agency with actual decision-making power to demand participation 58 hunting role in sweden – lindqvist et al. alces vol. 50, 2014 http://www.algdata.se in monitoring (swedish official investigation 2009), referred to the management groups and claimed the task was theirs to solve. they in turn claimed a lack of instruments and authority of decision-making and cannot demand monitoring participation from the management units. it is unclear if the management unit participation refers to all specified monitoring activities or if they can choose among suggested actions. stakeholders also claimed that hunters had the actual capability for control. license areas that cover parts of many moose management areas in all counties were another issue because they are not obligated by law to follow any management plan or participate in monitoring activities; consequently, they risk counteracting or interfering with plans in management units or management areas. respondents were apprehensive that the vagueness about management responsibility might undermine the purpose and role of management areas and the crucial monitoring required for ecosystem-based management. there were also uncertainties in all counties regarding high costs associated with decisions, prioritization, and the interval of monitoring activities to provide management with sufficient information. the lack of financial resources to support the moose management system was the fourth obstacle. originally, the moose management system was intended to be economically self-sufficient through harvest fees, including the cost of monitoring programs (environmental protection agency 2011). both officials and stakeholders found it unrealistic that, in accordance with the decision made by the swedish parliament, the entire moose management system (administration, management group members, and monitoring) should be funded by harvest fees alone; cost of monitoring was of greatest concern. most funding during the implementation phase was used for administrative work with little left for monitoring. most monitoring methods listed here are relatively inexpensive and require minimal effort of hunters, with pellet-group counts the exception as field work should occur twice annually (bergström et al. 2011). to save money, an official and a stakeholder suggested that monitoring be conducted in 4–5 year intervals rather than annually. however, this approach would reduce the ability to detect trends in data and recent information would not be unavailable during a 3-year plan undermining the adaptive advantage of the management plan. neither wildlife or forest managers could financially support monitoring activities and higher harvest fees were not considered a viable solution; in södermanland and västra götaland, this might have a contrary effect where hunters would refocus efforts on other game. one perception was that monitoring expenditures should be shared among stakeholders, versus exclusively funded by hunters, by having landowners responsible for forest resources and hunters responsible for moose. stakeholders also indicated that the sfa should be responsible for funding and providing forage forecasts and moose damage surveys. in all counties but västerbotten, another problem was how to fairly subdivide monitoring costs among many small landowners in the fragmented ownership common in sweden. one solution was that all monitoring be conducted by volunteers; however, certain deficiencies were identified relative to voluntary monitoring including lack of trust among some stakeholders regarding the credibility of monitoring data, and the time that stakeholders were willing to spend on voluntary activities. circumventing such concerns requires hiring professionals for monitoring which would increase costs substantially. the money available from harvest fees in each specific county differs because of variable harvest fees, but more alces vol. 50, 2014 lindqvist et al. – hunting role in sweden 59 importantly, on the annual moose harvest in each county; for example, harvest is about ∼1,000 moose in södermanland versus 6,000 in dalarna. raising harvest fees to increase management finances was not considered as an option, especially in södermanland and västra götaland where increasing costs for hunters might have an unintended effect. rather than investing money and efforts in moose hunting, officials and hunters suggested that other game species might instead become increasingly important to the hunters. willingness and capacity to participate in monitoring activities we asked respondents to describe monitoring methods used previously, new methods to be implemented in the adaptive system (fig. 2), and their general knowl‐ edge about different monitoring methods. information about local moose populations during implementation of the new system was mostly obtained from harvest statistics and hunter observations, but data from all base-monitoring methods were used to some extent when moose management areas defined their initial plans. throughout sweden in 2012, ∼50% of management areas used hunter observation rates of moose during the first week of moose hunting. rates within our management areas were higher than the national average (unpublished data, swedish hunting association), although none of the 5 counties considered this sufficient information to manage their moose population. pellet-group counts were conducted in all counties; however, they did not provide complete data for any county or moose management area due to their fragmented application. only södermanland collected data on calf weight but it was regarded as a monitoring method for future use. hunters still recorded calf weights and future use of these data might be facilitated by a fully developed database (algdata.se). this fig. 2. the monitoring methods planned to be used in each county (cab) and selected moose management areas (mma). additional monitoring will be conducted in mmas than at the county level in västerbotten, dalarna and södermanland. 60 hunting role in sweden – lindqvist et al. alces vol. 50, 2014 database will gather data from all monitoring effort and other areas of interest, and will presumably serve as an important management tool in facilitating stewardship. forage forecasts were provided by the sfa in all counties but a comprehensive forage forecast should include both satellite mapping and field visits; only satellite images were available in kronoberg and södermanland. the quality of forecasts in other counties could not be ascertained from responses. a countywide moose damage survey was conducted only in västerbotten, the only county with a long-term tradition of monitoring browsing damage. browsing was monitored on a smaller scale in dalarna. the need and desire to collect more detailed and comprehensive information was evident in the planned monitoring activities identified by officials and stakeholders in all counties. the necessity for more information and data about forest resources was specifically highlighted by a landowner in västerbotten: “in order to achieve a functioning adaptive management based on more facts and local knowledge that consider both the quality of the moose population, but also take moose damages into account”. there was confusion and opinions were divided about the applicability of monitoring methods and related validity of derived estimates. respondents in västra götaland desired estimates at the management unit level or lower, believing estimates at the management area level were too coarse. officials and stakeholders indicated their intention to address mvcs in management plans. they generally believed that mvcs were not only of societal importance, but reflected the relative size and trend of the moose population. stakeholders in västra götaland claimed that moose impacts on crops must be assessed, since moose damage, especially to oat (avena sativa) fields, is a recurring issue within their area. all respondents identified the need to include prominent county-specific conditions such as other ungulate species, moose migration patterns and predation, and habitat changes in the moose management system. in the 3 counties with established populations of large carnivores (i.e., brown bears in västerbotten and dalarna, wolves in dalarna and västra götaland), information about predator populations and predation rates was considered an important com‐ ponent of moose management plans. res‐ pondents in södermanland stressed the importance of co-managing other deer and ungulate species, rather than focus on moose singly, arguing that moose and deer interact and share several resources (e.g., habitat and food). they also questioned why deer and wild boar (sus scrofa) were not already considered in the ecosystem-based management system. in västerbotten, seasonal migration patterns of moose were reflected in the management plan. in kronoberg, habitat changes resulting from the 2005 hurricane gudrun received special attention, as it felled 75 million cubic meters of forest in southern sweden (swedish hydrological and meteorological institute 2013), creating beneficial habitat for moose and roe deer. discussion the respondents generally perceived the increased participation and collaboration between hunters and landowners as positive. despite some concern that the new mana‐ gement approach might remove decisionmaking from the local to ecosystem level and lose legitimacy with individual hun‐ ters and landowners, the majority of officials and stakeholders supported the program goals and believed the opposite. most envisioned a decentralized influence from the county to ecosystem level or down to the local level, and believed that enhanced alces vol. 50, 2014 lindqvist et al. – hunting role in sweden 61 stakeholder stewardship would consolidate management legitimacy. despite the positive spirit and increased collaboration, most respondents found the specific management responsibilities to be unclear. the term “local” seemed to create confusion about division of power, at what scale, and where power resided to influence decision-making. lack of funding and the actual time provided for implementation of the new system were other obstacles identified by respondents. while time constraints would gradually be remedied, the funding issue risks undermining the entire management system since it reduces the possibilities for stakeholders to contribute to ecosystem stewardship. stakeholders were aware of the need to maximize information to successfully balance moose and forest resources. all acknowledged the need for reliable monitoring and appropriate sampling as part of adaptive management, and that the purpose of monitoring might be lost without using appropriate techniques. however, stakeholder expectations and scientific recommendations regarding data acquisition and monitoring differed. the suggested monitoring methods were most useful for moose management areas, and less so for smaller local units where the goal is to use accurate estimates (ericsson 2011). pellet-group counts are an exception, but using this method in a smaller management unit would not be as useful when managing a large population, a primary goal for achieving ecosystem-based management (swedish official investigation 2009). understanding the limitations of monitoring techniques is crucial to achieve transition from an assumption-based to a knowledge-based management system. the forest monitoring methods are applicable at both the management area and local scales (kalén and bergquist 2011, rolander et al. 2011). however, local damage surveys are often only roughly correlated with population density (rolander et al. 2011), and applying such data with moose population estimates at larger scale risks error in interpretation within the larger management area. further, data collected at different resolutions probably poses risk to the overall assessment that includes various monitoring techniques, potentially complicating stakeholder stewardship further. the ability to modify the management plan was perceived as a core and essential feature of the moose management system; however, perceptions varied as to whether it was possible to alter management decisions during an ongoing hunting season. indeed, invoking change given new information is essential in adaptive management (allen et al. 2011). the question remains as to the extent of adaptability in the new moose management system. the regulations (environmental protection agency 2011) provide for changes during a hunting season, yet this would require continual monitoring and rapid response by stakeholders. unless reporting efforts improve, it will be difficult to adjust during a hunting season. even if these components function flawlessly, most moose in sweden are harvested in the first weeks of the hunting season (ball et al. 1999); therefore, from a practical standpoint, non-emergency changes and deviations would likely occur the following year. accounting for mvcs and predation was relatively well developed in all counties. predation, for example, could be estimated precisely from wolf monitoring (sand et al. 2011). conversely, with respect to other ungulates—such as in södermanland where red deer, roe deer, fallow deer, and wild boar (sus scrofa) are abundant—current monitoring of populations, damage, and forage forecasts were considered highly insufficient; neither wild boar nor deer are systematically monitored in sweden (apollonio et al. 2010). deer populations could 62 hunting role in sweden – lindqvist et al. alces vol. 50, 2014 be assessed with pellet-group counts, but density estimates are dependent upon species-specific identification of pellet groups and defecation rates (neff 1968). it might also be possible to expand hunter observations to include other deer with the methods of kindberg et al. (2009), where effort-corrected observations of brown bears was suggested as a useful monitoring technique. adequate evaluation would be required before implementing this approach into moose management plans. despite progress with genetic monitoring of browsing damage (spong 2011), information about different ungulate forage selection and food overlap is generally lacking. overall, our results suggest that critical knowledge-gaps exist with both hunters and landowners that preclude their effective participation in, and use of many monitoring techniques. conclusions we found strong stakeholder support for the moose management goals to balance social and ecological values. although the willingness to embrace ecosystem stewardship was pervasive among stakeholders, the moose management program faces several challenges in fully implementing an ecosystem-based, adaptive moose management system. this includes the need to clarify primary concepts like “ecosystem-based” and “local”, and to have all stakeholders clarify and agree about definitions within the plan and the role and responsibility at each level; such an approach should help mitigate conflicts and avoid further confusion. several obstacles were identified concerning monitoring, the key tool enabling environmental stewardship. specifically, unclear responsibilities and inadequate funding threaten local and regional data collection, both of which could jeopardize stakeholder stewardship. the importance and benefits of monitoring and reporting, as emphasized by the respondents, must be communicated to the grass-root level to enhance future participation; i.e., the responsibility of individual stewardship. demonstrating that monitoring is a worthwhile effort of individual hunters or landowners is perhaps the most difficult challenge. it is important to create a sustainable moose management system in balance with forage resources, traffic safety, and large carnivore populations, as well as other land uses and biodiversity objectives (moller et al. 2004). unless these challenges are resolved, the primary objective of the new moose management system risks the same failure as with previous plans. to address the monitoring issues, an addition or revision of the regulations (environmental protection agency 2011) regarding the purpose of monitoring is suggested; this approach would be positive for all stakeholders and should facilitate their commitment to the system. the issue of monitoring costs, specifically regarding forest resources, needs to be addressed directly with full support of stakeholders to achieve adequate participation. lastly, information regarding coexistence and co-management of ungulates in management areas needs to be incorporated into the system, an approach similar to that for large carnivores in sweden (andrén et al. 2011). acknowledgements this research was funded through future forests, a multi-disciplinary research program supported by mistra (the foundation for strategic environmental research), the swedish forestry industry, the swedish university of agricultural sciences, umeå university, and the forestry research institute of sweden. we thank the two anonymous reviewers for valuable comments on an earlier draft of the manuscript. alces vol. 50, 2014 lindqvist et al. – hunting role in sweden 63 references allen, c. r., j. j. fontaine, k. l. pope, and a. s. garmestani. 2011. adaptive management for a turbulent future. journal of environmental management 92: 1339–1345. andren, h., a. jarnemo, h. sand, j. mansson, l. edenius, and p. kjellander. 2011. ekosystemaspekter på älgförvaltning med stora rovdjur. ecosystem aspects on moose management with large carnivores. fakta skog 12/2011. swedish university of agriculture, umeå, sweden. (in swedish). angelstam, p., p. e. wikberg, p. danilova, w. e. faber, k. nygren, w. b. ballard, and a. r. rodgers. 2000. effects of moose density on timber quality and biodiversity restoration in sweden, finland, and russian karelia. alces 36: 133–145. apollonio, m., r. andersen, and r. putman. 2010. european ungulates and their management in the 21st century. cambridge university press, new york, new york, usa. ball, j. p., g. ericsson, and k. wallin. 1999. climate changes, moose and their human predators. ecological bulletins 47: 178–187. barnard, a. 2004. hunter-gatherers in history, archaeology and anthropology. berg publishers, oxford, united kingdom. bergman, m., and s. åkerberg. 2006. moose hunting, forestry, and wolves in sweden. alces 42: 13–23. bergstrom, r., j. mansson, j. kindberg, å. pehrson, g. ericssonr, and k. danell. 2011. spillningsinventering för älg. pellet-group count for moose. fakta skog 12/2011. swedish university of agriculture, umeå, sweden. (in swedish). broman, e. 2003. environment and moose population dynamics. ph.d. thesis. department of environmental sciences and conservation, göteborg university, göteborg, sweden. chapin, s., g. kofinas, and c. folke. 2009. principles of ecosystem stewardship: resilience-based natural resource management in a changing world. springer science and business media, new york, new york, usa. council of europe. 2007. european charter on hunting and biodiversity. adopted by the standing committee of the bern convention at its 27th meeting in strasbourg, 26–29 november, 2007. <http:// www.cicwildlife.org/uploads/media/ hunting_charter_en.pdf> (accessed april 2013). danell, k., j. p. ball, r. bergstrom, g. ericsson, j. kindberg, and h. sand. 2011. älgkalvvikter – ett konditionsmått. moose calves’ weights – an index of fitness. fakta skog 13/2011. swedish university of agriculture, umeå, sweden. (in swedish). edenius, l., m. bergman, g. ericsson, and k. danell. 2002. the role of moose as a disturbance factor in managed boreal forests. silva fennica 36: 57–67. environmental protection agency. 2011. nfs 2011:7. naturvårdsverkets föreskrifter och allmänna råd om jakt efter älg och kronhjort. environmental protection agency’s regulations and general guidelines on hunting for moose and red deer. (in swedish). ericsson, g. 2011. inventering för adaptiv älgförvaltning i älgförvaltningsområden (äfo). monitoring for an adaptive moose management within moose management areas (mma). inventeringsmanualens förord. manual preface. swedish university of agriculture, umeå, sweden. (in swedish). ———, and j. kindberg. 2011. älgobservationer (älg-obs). hunter observations. fakta skog 11/2011. swedish university of agriculture, umeå, sweden. (in swedish). fischer, a., c. sandstrom, m. delibesmateos, b. arroto, d. tadie, d. randall, f. hailu, a. lowassa, m. msuha, 64 hunting role in sweden – lindqvist et al. alces vol. 50, 2014 http://www.cicwildlife.org/uploads/media/hunting_charter_en.pdf http://www.cicwildlife.org/uploads/media/hunting_charter_en.pdf http://www.cicwildlife.org/uploads/media/hunting_charter_en.pdf v. kerezi, s. reljic, j. linnel, and a. majic. 2012. on the multifunctionality of hunting – an institutional analysis of eight cases from europe and africa. journal of environmental planning and management 56: 531–552. hendee, j. c. 1974. a multiple-satisfaction approach to game management. wildlife society bulletin 2: 104–113. holsman, r. h. 2000. goodwill hunting. exploring the role of hunters as ecosystem stewards. wildlife society bulletin 28: 808–816. hornberg, s. 2001. the relationship between moose (alces alces) browsing utilization and the occurrence of different forage species in sweden. forest ecology and management 149: 91–102. kalen, c., and j. bergquist. 2011. fodpro, foderprognoser. skogliga inventeringsmetoder i en kunskapsbaserad älgförvaltning. fodpro, forage forecasts, moose damage survey methods in a knowledge-based moose management. version 1.0. skogsstyrelsen (swedish forest agency). (in swedish). kindberg, j., g. ericsson, r. bergstrom, and k. danell. 2011. avskjutningsstatistik för älg. harvest statistics for moose. fakta skog 10/2011. swedish university of agriculture, umeå, sweden. (in swedish). ———, ———, and j. e. swenson. 2009. monitoring rare or elusive large mammals using effort-corrected voluntary observers. biological conservation 142: 159–165. lavsund, s., t. nygren, and e. j. solberg. 2003. status of moose populations and challenges to moose management in fennoscandia. alces 39: 109–130. leopold, a. 1950. a sand county almanac and sketches here and there, illustrated by charles w. swartz. oxford university press, new york, new york, usa. mansson, j., c. e. hauser, h. andren, and h. p. possingham. 2011. survey method choice for wildlife management: the case of moose alces alces in sweden. wildlife biology 17: 176–190. moller, h., f. berkes, p. o. b. lyver, and m. kislalioglu. 2004. combining science and traditional ecological knowledge: monitoring populations for comanagement. ecology and society 9(3): 2. http://www.ecologyandsociety.org/vol 9/iss3/art2/ (accessed april 2103). neff, d. j. 1968. the pellet-group count technique for big game trend, census, and distribution: a review. the journal of wildlife management 32: 597–614. rolander, m., c. kalen, and j. bergquist. 2011. äbin, skogliga inventeringsmetoder i en kunskapsbaserad älgförvaltning (moose damage survey methods in a knowledge-based moose management). älgbetesinventering (äbin) version 1.0. skogsstyrelsen (swedish forest agency). (in swedish). sand, h., h. andren, j.e. swenson, and j. kindberg. 2011. flera jägare på älgpopulationen – predationsmönster hos varg och björn. several hunters of the moose population – predation patterns of wolf and brown bear. fakta skog 25/2011. swedish university of agriculture, umeå, sweden. (in swedish). ———, g. cederlund, and k. danell. 1995. geographical and latitudinal variation in growth patterns and adult body size of swedish moose (alces alces). oecologica 102: 433–442. sandstrom, c., s. wennberg-digasper, and k. öhman. 2013. conflict resolution through ecosystem-based management: the case of swedish moose management. international journal of the commons 7: 549–570. seiler, a. 2003. the toll of the automobile. ph. d. thesis. swedish university of agricultural sciences, umeå, sweden. spong, g. 2011. dna-analyser och viltövervakning. dna analysis and wildlife surveillance. fakta skog 17/2011. swedish university of agriculture, umeå, sweden. (in swedish). alces vol. 50, 2014 lindqvist et al. – hunting role in sweden 65 http://www.ecologyandsociety.org/vol9/iss3/art2/ http://www.ecologyandsociety.org/vol9/iss3/art2/ swedish government. 2010. proposition 2009/10:239. älgförvaltningen. 2010. proposition 2009/10:239 moose management. (in swedish). swedish hydrological and meteorological institute. 2013. <http://www. smhi.se/kunskapsbanken/meteorologi/gudrun-januaristormen-2005-1.5300> (accessed march 2013). swedish official investigation. 2009. 2009:54. uthållig älgförvaltning i samverkan. 17th june 2009. sustainable moose management in collaboration. fritzes. stockholm, sweden. (in swedish). wennberg-digasper, s. 2008. natural resource management in an institutional disorder: the development of adaptive comanagement systems of moose in sweden. ph. d. thesis. division of political science, department of business administration and social sciences, luleå university of technology, sweden. 66 hunting role in sweden – lindqvist et al. alces vol. 50, 2014 http://www.smhi.se/kunskapsbanken/meteorologi/gudrun-januaristormen-2005-1.5300 http://www.smhi.se/kunskapsbanken/meteorologi/gudrun-januaristormen-2005-1.5300 http://www.smhi.se/kunskapsbanken/meteorologi/gudrun-januaristormen-2005-1.5300 the changing role of hunting in sweden&#x2014;from subsistence to ecosystem stewardship? background study area methods results willingness and capacity to support nationwide objectives willingness and capacity to implement stewardship goals willingness and capacity to participate in monitoring activities discussion conclusions acknowledgements references alces19_14.pdf alces vol. 19, 1983 alces vol. 19, 1983 alces vol. 19, 1983 alces vol. 19, 1983 alces vol. 19, 1983 alces vol. 19, 1983 alces vol. 19, 1983 alces vol. 19, 1983 alces vol. 19, 1983 alces vol. 19, 1983 alces vol. 19, 1983 alces22_139.pdf alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces vol. 22, 1986 alces29_249.pdf alces19_148.pdf alces vol. 19, 1983 alces vol. 19, 1983 alces vol. 19, 1983 alces vol. 19, 1983 alces vol. 19, 1983 alces vol. 19, 1983 alces vol. 19, 1983 alces(25)_118.pdf alces(23)_107.pdf alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 alces vol. 23, 1987 moose habitat in massachusetts: assessing use at the southern edge of the range david w. wattles1 and stephen destefano2 1massachusetts cooperative fish and wildlife research unit, department of environmental conservation, university of massachusetts, amherst, massachusetts, 01003 usa; 2u. s. geological survey, massachusetts cooperative fish and wildlife research unit, university of massachusetts, box 34220, amherst, massachusetts, 01003 usa. abstract: moose (alces alces) have recently re-occupied a portion of their range in the temperate deciduous forest of the northeastern united states after a more than 200 year absence. in southern new england, moose are exposed to a variety of forest types, increasing development, and higher ambient temperatures as compared to other parts of their geographic range. additionally, large-scale disturbances that shape forest structure and expansive naturally occurring shrub-willow communities used commonly elsewhere are lacking. we used utilization distributions to determine third order habitat selection (selection within the home range) of gps-collared moose. in central massachusetts, forests regenerating from logging were the most heavily used cover type in all seasons (48 63% of core area use). habitat use of moose in western massachusetts varied more seasonally, with regenerating forests used most heavily in summer and fall (57 and 46%, respectively), conifer and mixed forests in winter (47 65%), and deciduous forests in spring (41%). this difference in habitat selection reflected the transition from northern forest types to more southern forest types across the state. the intensive use of patches of regenerating forest emphasizes the importance of sustainable forest harvesting to moose. this study provides the first assessment of habitat requirements in this southern portion of moose range and provides insights into re-establishment of moose in unoccupied portions of its historic range in new york and pennsylvania. alces vol. 49: 133–147 (2013) key words: alces alces, forest regeneration, habitat use, massachusetts, moose. moose (alces alces) have recently recolonized a portion of their historic range in the temperate deciduous forest of southern new england after more than 200 years absence (vecellio et al. 1993, wattles and destefano 2011). this environment is unique in moose range and provides a number of potential challenges, including forest types that differ from most of the geographic range (westveldt et al. 1956, degraaf and yamasaki 2001, franzmann and schwartz 2007), a thermal environment that could reduce fitness and survival (renecker and hudson 1986, boose 2001, murray et al. 2006, lenarz et al. 2009, 2010), and high levels of human development (u. s. census bureau 2000, destefano et al. 2005). habitat use and diet have been studied throughout much of moose range (franzmann and schwartz 2007), including elsewhere in the northeastern united states (crossley and gilbert 1983, leptich and gilbert 1989, garner and porter 1990, miller and litvaitis 1992, thompson et al. 1995, corresponding author: david wattles, massachusetts cooperative fish & wildlife research unit, 225 holdsworth natural resources center, university of massachusetts, amherst, ma 01003. 617-2567055, (fax) 413-545-4358, email: dwattles@eco.umass.edu 133 scarpitti et al. 2005, scarpitti 2006). however, similar information has been lacking in the transitional forests of southern new england. the recolonization of southern new england saw moose push the southern extent of their range from spruce-fir and northern hardwood forests into transitional and more southerly forest types, which lack many of the plant species preferred by moose further north. massachusetts provides a unique environment to examine the effects of this transition in use of forest types and habitat selection over a relatively small area. the objectives of this study were to 1) determine how moose use the temperate deciduous forest of southern new england, 2) compare habitat use among seasons, and 3) assess whether suitable habitat exists to support long-term occupation of southern new england. study area the study area was located in central and western massachusetts, usa (fig. 1). topography is dominated by glaciated hills underlain by shallow bedrock. glacial activity created abundant small stream valleys, lakes, ponds, and wetlands whose size and nature vary with beaver (castor canadensis) activity. the central and western sections of the study area are separated by the connecticut river valley which runs n-s through westcentral massachusetts. elevation ranges from 100 m above sea level in the connecticut river valley to 425 m in the hills of central massachusetts, and 850 m in the berkshire hills of western massachusetts. the western two-thirds of massachusetts was >80% mixed deciduous, secondor multiple-growth forest, much of it resulting from regeneration of farm fields abandoned fig. 1. area used to study moose-habitat relationships in northeastern usa, specifically westcentral massachusetts depicted in blow-up with dashed line, and bordered by southern vermont and new hampshire. figure also depicts the forest types of massachusetts (after westveldt et al. 1956 and degraaf and yamasaki 2000). 134 habitat in massachusetts – wattles and destefano alces vol. 49, 2013 in the mid-to-late 1800s (hall et al. 2002). with the exception of wetlands and smallscale logging, the undeveloped portion of the massachusetts landscape was nearly 100% closed canopy mixed-coniferousdeciduous forest. massachusetts represents a forest transition zone, where forests shift from those common in northern new england to more southern forest types. moose transition across 4 forest types in massachusetts, including spruce-fir-northern hardwoods, northern hardwoods-hemlock (tsuga canadensis)-white pine (pinus strobus), transition hardwoods-white pine-hemlock, and central hardwoods-hemlock-white pine (fig. 1). the spruce-fir-northern hardwoods type is dominated by spruce (picea spp.), balsam fir (abies balsamea), american beech (fagus grandifolia), birch (betula spp.), trembling aspen (populus tremuloides), eastern hemlock, and maple (acer spp.). in the northern hardwoods forest, white pine and hemlock largely replace spruce and fir. transition hardwoods-white pine-hemlock forests contain most of the species in the northern hardwoods type; in addition, oaks (quercus spp.) and hickories (carya spp.) become increasingly common. in the central hardwoods-hemlock-white pine forest, beech, sugar maple (a. saccharum), and yellow birch (b. alleghaniensis) are rare, replaced by oaks and hickories. transitions between forest types can be gradual or distinct depending on localized physiography, climate, bedrock, topography, land-use history, and soil conditions, resulting in a patchwork of forest types and species groups (westveldt et al. 1956, degraaf and yamasaki 2001). early successional habitat was created primarily through timber harvest practices, and occasionally through wind and other weather events. from 1984 to 2000, about 1.5% of the forest was logged annually, consisting of small (mean = 16.5 ha) cuts of moderate intensity (removal of 27% of timber volume) widely distributed on the landscape (kittredge et al. 2003, mcdonald et al. 2006). the pattern of forest harvest, glaciation, and transitional forest types provided a patchy mosaic of well interspersed forest types, age classes, and wetlands. july was the warmest month when mean daily temperature was 21.1 °c, and january the coldest when mean daily temperature was −6.1 °c. mean annual precipitation was 107 cm in central areas and 124 cm in western areas, with all months receiving 7–11 cm and 8–12 cm, respectively (the weather channel 2011a, 2011b). the average date of last frost in the region was 15 may; the average day of first frost was 1 october and 15 september in central and western areas, respectively (degraaf and yamasaki 2001). maximum snow depth was typically greater in western massachusetts than central areas and reaches depth in both areas (50–70 cm) that can restrict moose movement (coady 1974). methods study animals and gps telemetry adult (>1 yr old) moose were captured by locating, stalking, and darting them from the ground in state forests, wildlife management areas, and other conservation areas between march 2006 and november 2009. moose were immobilized using either 5 ml of 300 mg/ml or 3 ml of 450 mg/ml xylazine hydrochloride (congaree veterinary pharmacy, cayce, south carolina, usa; mention of trade names does not imply endorsement by the u. s. government) administered from a 3 or 5 cc type c pneudart dart (pneudart, inc., williamport, pennsylvania, usa). tolazolene (100 mg/ml) at a dosage of 1.0 mg/kg was used as an antagonist. moose were fitted with gps collars; either ats g2000 series (advanced telemetry systems, inc., isanti, minnesota, usa) or telonics twg-3790 gps collars (telonics, inc., mesa, arizona, usa). alces vol. 49, 2013 wattles and destefano – habitat in massachusetts 135 we programmed the collars to attempt a gps fix as frequently as possible while allowing the battery life to extend for at least 1 year; depending on the collar, a gps fix was attempted every 135, 75, or 45 min. collars were equipped with very high frequency (vhf) transmitters, mortality sensors, and release mechanisms that opened the collars either at a low battery state or a preprogrammed date. capture and handling procedures were approved by the university of massachusetts institutional animal care and use committee, protocol numbers 25-02-15, 28-02-16, and 211-02-01. seasons we defined the length and timing of seasons based on several ecological factors including timing of the growing season of vegetation, weather (including temperature and snow conditions), and the moose reproductive cycle (table 1). the transition between seasons can vary by several days to several weeks depending on weather conditions and other factors. if movements were identified in the location data for an animal that obviously demonstrated a change in season (e.g., a large increase in movements at the end of the winter when snow had melted, or at the end of summer indicating the beginning of rutting behavior), the data were truncated at that point and included in the following season. habitat availability and core area habitat use we compared vegetation and land cover types in the home range cores of moose to that available in larger mcp home ranges (third-order habitat selection; johnson 1980). we used a fixed kernel density estimator (kde) (worton 1989) and the kernel density estimation tool in hrt: home range tools for arcgis (rogers et al. 2007) to calculate utilization distributions (ud). we then used the create minimum convex polygons tool in hawth's tools (beyers 2006) to calculate 100% minimum convex polygon (mcp) home ranges (mohr 1947). all geographic information system (gis) work was performed in arcgis 9.3 (esri 2008). table 1. seasons used for calculating home-range, movements, and core-area habitat analyses. season dates vegetation/ browse temperaturea movement season length (d) spring 16 april – 31 may growing season; bud-break-leaf out cool-hot not snow restricted, potentially temperature restricted 46 calving (females) 8–13 may – 15 june growing season; bud-break-leaf out cool-hot restricted by newborn calf 30 summer 1 june – 30 aug growing season; full leaf out hot restricted by temperature 92 fall 1 sept – 31 oct leaf out to leaf off hot-cool rut and temperature influenced 61 early winter 1 nov – 31 dec dormant season; woody/evergreen warm-cold not snow restricted, potentially metabolism restricted 61 late winter 1 jan – 15 april dormant season; woody/evergreen cold-warm potentially snow and metabolism restricted 107 atemperature ranges describing typical temperatures experienced during a season; cold ≤0°c, cool >0°c and <14°c, warm ≥14°c and <20°c, hot ≥20 °c. 136 habitat in massachusetts – wattles and destefano alces vol. 49, 2013 the kernel bandwidth or smoothing factor (h) is known to have the greatest effect on uds (worton 1989). a large h oversmooths the data, resulting in a more biased ud that encompasses unused habitats, while a small h under-smooths the data, resulting in a fragmented ud (fieberg 2007). there is lack of agreement on the best method for calculating h (powell 2000, hemson et al. 2005, gitzen et al. 2006, fieberg 2007, kie et al. 2010); therefore, we used 2 values (80 m and 30 m) of h to calculate uds. we used the 50-percent isopleth of the 80 m ud to identify home range cores. however, the 80 m bandwidth still resulted in over-smoothed uds with large buffers around gps locations that incorporated unused habitat. as a result we used a second h value of 30 m, based on the median distance between gps locations for our most intensively sampled animals, approximating within-patch movement of the animals. the resulting uds incorporated little unused habitat and were used to assess habitat use within the core areas calculated with the 80 m h. we classified habitats into 8 categories: coniferous forest (mostly coniferous with minimal deciduous component), deciduous forest (mostly deciduous with minimal coniferous component), mixed forest (mixed deciduous and coniferous), regenerating forest (logged areas <20 years old and powerline right-of-ways), wooded wetlands (conifer, mixed, and deciduous wooded wetlands), other wetlands (grassy fens, shrub swamps, bogs, deep wetlands, and open water), open (e.g., fields and meadows), and developed. we set the age restriction of regenerating forest at 20 years because, while logged areas >20 years may still provide browse, these stands more closely resembled mature forest. in addition, older harvests were difficult to distinguish or map accurately. open and developed were absent from almost all core area habitat use and were later dropped from the analysis. we manually digitized cover and land use within the cores in arcgis 9.3 (esri 2008) using a compilation of available gis base-layers from the massachusetts office of geographic information (massgis; massgis 2011) and other sources, including 2005 and 2009 orthophotos, department of environmental protection wetland layers, forest harvest information from the massachusetts department of conservation and recreation (dcr) and harvard forest (mcdonald et al. 2006), 2003 and 2009 national agricultural imagery program (naip) satellite imagery, and mid-1990s black and white orthophotos, as well as state wetland layers for vermont and new hampshire. we assessed habitat availability within the home range by generating sets of 250 random points within 100% annual mcp home ranges using the generate random points tool in hawth's tools (beyers 2006, wattles 2011) and manually classifying cover and land use. we calculated use:availability ratios by comparing cover and land use within 30 m ud core areas (used) to mcp home ranges (available) (aebischer et al. 1993). use:availability ratios >1 indicated a cover type was used more than available; a ratio <1 indicated use was less than available. calving sites were identified based on large decreases in daily movement of cows followed by a concentration of gps locations during the calving season (may– june) (poole et al. 2007). analyses we used analysis of variance (anova) to analyze the differences in habitat availability, core area habitat use, and use:availability ratios within and between sexes, seasons, and portions of the study area. we used type iii anova to account for unequal sample sizes among groups and seasons. we performed pairwise comparisons using alces vol. 49, 2013 wattles and destefano – habitat in massachusetts 137 tukey's contrasts with adjusted p-values using the single-step method. significance level for all analyses was set at 0.05. we used r, version 2.12.2 (r development core team 2005) for all statistical analyses. results capture and deployment of gps collars we deployed gps collars on 26 adult moose (7 females and 19 males); 5 were excluded due to mortality, suspected infection with brainworm (parelaphostrongylus tenuis), or collar failure. data analysis included 5 females and 8 males in central and 8 males in western massachusetts. nine moose were recaptured and recollared when the batteries in their initial gps collars ran low. we obtained 127,408 locations from the 21 moose with an overall fix rate of 85%. seasonal data for any animal were included in the analyses only if data were obtained across the entire season. the median number of locations per animal per season ranged from 402 in spring to 1,015 in late winter. the minimum number of locations was 281 for one animal in spring. home range core area habitat use habitat use within seasons. regenerating forest was used more than all other cover types by both central males and females during all seasons (proportion of use 0.48 to fig. 2. mean proportional seasonal core area habitat use for female (n = 5, 5, 4, 5, and 5 individuals for spring, summer, fall, early winter, and late winter, respectively) and male (n = 7, 7, 7, 6, and 7 individuals for spring, summer, fall, early winter, and late winter, respectively) moose in central massachusetts and male moose in western massachusetts (n = 7, 6, 4, 8, and 7 individuals for spring, summer, fall, early winter, and late winter, respectively). error bars represent standard errors of the means. 138 habitat in massachusetts – wattles and destefano alces vol. 49, 2013 0.63; p ≤ 0.006), with the exception of wooded wetlands, mixed forest, and conifer forests during spring by females (fig. 2). no other differences in seasonal core area habitat use were significant for central males or females. both central males and females showed selection for regenerating forest during all seasons (table 2), except for females during spring. central males also showed selection for wooded wetlands during fall. all other habitat types were either used in proportion to or less than their availability. the lack of selection of regenerating forests by females in spring was likely because calving areas dominated female spring habitat use; calving sites varied among individuals and included wooded wetlands, mature mixed and conifer-dominated mixed stands, and mixed and conifer shelter cuts. the importance of vegetative cover type varied with season for western males (table 2, fig. 2), that selected for deciduous forest and used it (proportional use = 0.41) more than all other habitat types except regenerating forest during spring (p ≤ 0.03). regenerating forest (0.22) was also used more than other wetlands at this time of year (p = 0.03). during summer (0.57) and fall (0.46) regenerating forest use was greater than all other habitat types (p ≤ 0.02); however, it was used more than its availability only during summer. no other habitat types were used more than their availability at any other time of year. forest table 2. p-values for anova of use:availability ratios. dark gray indicates use:availability >1, light gray <1, and white use not significantly different than availability. spring summer fall early winter late winter females coniferous 0.688 0.329 0.066 0.027 0.611 mixed 0.219 <0.001 0.001 0.030 0.070 deciduous 0.008 0.045 0.228 0.368 0.561 regenerating 0.089 0.008 0.010 0.021 0.028 wooded wetland 0.445 0.213 0.523 0.113 0.090 other wetland 0.061 0.958 0.956 0.004 0.006 central males coniferous 0.508 0.458 0.035 0.940 0.898 mixed <0.001 <0.001 <0.001 0.786 0.145 deciduous 0.687 0.059 0.088 0.072 0.066 regenerating 0.004 0.002 <0.001 0.039 0.003 wooded wetland 0.077 0.063 0.002 0.358 0.004 other wetland 0.007 0.190 0.482 0.001 0.045 western males coniferous 0.037 0.024 0.002 0.089 0.188 mixed 0.369 <0.001 0.165 0.393 0.053 deciduous 0.014 0.165 0.809 0.300 0.505 regenerating 0.885 <0.001 0.083 0.249 0.360 wooded wetland 0.677 0.988 0.181 0.027 0.049 other wetland 0.139 0.609 0.475 0.722 0.019 alces vol. 49, 2013 wattles and destefano – habitat in massachusetts 139 types with a conifer component (mixed and coniferous forest) combined to be most used in early (0.47) and late winter (0.65). high use of regenerating forest continued in early winter (0.32). habitat use among seasons. there were no differences in the use of various vegetative cover types by females among seasons. central males used more mixed forest in early winter than fall (p = 0.046). central males used wooded wetlands more during fall than early and late winter (p < 0.001), and more during both spring and summer than in late winter (p ≤ 0.02). western males used less conifer forest in their home range cores during spring, summer, and fall than in early winter (p ≤ 0.03), and less in fall than late winter (p ≤ 0.04). similarly, they used more mixed forest in late winter cores than all other seasons, but only significantly more in fall (p≤ 0.01). western males use regenerating forests more during summer than spring (p = 0.02) or late winter (p = 0.01), while use of deciduous forest was greater during spring than summer and early or late winter (p ≤ 0.04). wooded wetland use was greater during fall than early and later winter (p ≤ 0.03). habitat use based on gender and region. there were no differences in seasonal core area habitat use between central males and females. however, western males used deciduous forest more than central males or females during spring and fall (p ≤ 0.01), and more coniferous forest during early winter (p ≤ 0.02). western males also used more mixed forest than central males during late winter (p = 0.03), but less regenerating forest (p = 0.04) than central males during spring, and less regenerating forest than either central males or females in late winter (p ≤ 0.01). discussion not all areas within a home range hold equal importance to the animal. if food and other resources are unevenly distributed, areas of higher densities of critical resources should be more important than areas with lower levels of that resource (powell 2000). if animals focus their use in some portion of the home range where resources are concentrated, those areas represent centers of activity or cores of the home range (hayne 1949, kaufmann 1962, samuel et al. 1985, powell 2000). due to the concentrated use of these areas, home range cores may be critically important to an individual's survival and reproductive success. identifying home range core areas and core area habitat can provide important insights into the ecology of a species and its survival strategies. this is particularly important for managers in southern new england, where moose have only recently re-established after many decades of absence and where habitat differs from much of the rest of their geographic range. the typical annual pattern of habitat use by moose reflects the seasonal availability of resources (peek 2007). sites that optimize forage quantity and quality vary by forest type and season and are a main driver of the vegetative cover types that moose select (telfer 1988, westworth et al. 1989, mccracken et al. 1997, poole and stuartsmith 2005, peek 2007). as a result, habitat use follows a familiar pattern across their geographic range (peek 2007). for example, moose in our study were extremely reliant on young, regenerating forest for browse (phillips et al. 1973, pierce and peek 1984, bangs et al. 1985, mccracken et al. 1997, poole and stuart-smith 2005, peek 2007, gillingham and parker 2008); used wetlands for thermal cover (renecker and hudson 1986) and some summer forage (ritcey and verbeek 1969, jordan et al. 1973, crossley 140 habitat in massachusetts – wattles and destefano alces vol. 49, 2013 and gilbert 1983, morris 2002); browsed conifers such as balsam fir (where available) and hemlock in winter (crossley and gilbert 1983, thompson et al. 1995); and used conifers as cover during warm periods (schwab and pitt 1991, dussault et al. 2004) and periods of deep snow (peek et al. 1976, monthey 1984, thompson et al. 1995). however, we found large differences in the availability and distribution of some vegetative cover types compared to the rest of the species range, which resulted in differences in habitat selection. probably the most important difference was the amount and distribution of early successional forest habitat and the processes that create these habitats. while this cover type was heavily used by moose, large disturbances that create it – either natural (fire, wind, insects) or human-caused (logging) – are rare and becoming rarer in southern new england. the amount and distribution of timber harvesting activities is minimal as compared to many other regions, and large-scale natural processes such as flooded river deltas, sup-alpine and riparian shrub communities, and avalanche corridors do not exist. in addition, some key woody species such as willows (salix spp.), aspen, mountain-ash (sorbus americana), and other shade-intolerant species, all of which provide high quality browse for moose in more northern regions, are not abundant in southern new england. with the exception of wetlands and small-scale logging, the undeveloped portion of the massachusetts landscape is nearly 100% closed canopy mixedconiferous-deciduous forest. as a result, moose use the various cover types of closed canopy forest, small wetlands, and patches of young forest created by logging. additionally, while wetlands that supported aquatic vegetation were used throughout spring, summer, and fall, and these sites likely provided critical nutrients, their importance as feeding sites was relatively low compared to regenerating forests in our study area and to wetlands elsewhere in moose range (jordan et al. 1973, crossley and gilbert 1983, ritcey and verbeek 1989, morris 2002). similarly, while roadside salt licks are commonly used by moose in northern new hampshire (miller and litvaitis 1992, scarpitti et al. 2005), we saw no indication of the use of roadside wetlands that would indicate their use as salt licks. we also saw clear differences in forest cover use between central and western massachusetts, and by extension between the forest types of southern and northern new england. the most important factor was likely the transition across the state from spruce-fir-northern hardwoods and northern hardwoods-hemlockwhite pine forest to the transition hardwoodwhite pine-hemlock and central hardwoodhemlock-white pine forest types, and the associated changes in plant communities and structure. the forests in the berkshire mountains of western massachusetts are similar to forests in southern vermont and new hampshire (degraaf and yamasaki 2001), and use of these forests reflected many similar habitat patterns that have been reported in northern new england (crossley and gilbert 1983, leptich and gilbert 1989, thompson et al. 1995, scarpitti 2006). conifer and mixed-coniferous-deciduous stands, with balsam fir and hemlock, were important cover types during winter in western massachusetts, as in northern new england (crossley and gilbert 1983, thompson et al. 1995, scarpitti 2006). balsam fir occurred in the spruce-fir-northern hardwood forests at the highest elevations in western massachusetts, but it was absent in central massachusetts and lower elevations in western massachusetts. with the absence of balsam fir, hemlock was the only conifer that was a large portion of the winter diet of moose; white pine was avoided (faison et al. 2010). while the use of stands of hemlock and mixed stands with hemlock alces vol. 49, 2013 wattles and destefano – habitat in massachusetts 141 and deciduous shrubs and saplings increased in central massachusetts during winter, the lack of balsam fir increased reliance on high-density regenerating stands of hardwoods. additionally, typically less restrictive snow conditions in central massachusetts (20–60 cm in late winter) may have played a role in the increased use of regenerating stands in late winter, while deep snow (80–110 cm in late winter) in western massachusetts may have forced moose into the shelter of spruce-fir stands. similarly, western males used deciduous forests more in spring and fall compared to moose in central massachusetts. favored deciduous species, such as hobblebush (viburnum lantanoides), striped maple (a. pensylvanicum), beech (early in the growing season), and aspen were less common in central massachusetts, and the reduced availability of these key species seemed to limit use of deciduous forest in central compared to western areas. the dominant habitat type used by moose throughout the state was regenerating forest created by logging. in central massachusetts moose used areas of forest regeneration intensively in all seasons. while use of regenerating stands in western massachusetts was more variable, moose still concentrated in these sites, especially during summer. early seral stage forest stands provided a concentrated source of abundant browse during the growing season (mcdonald et al. 2008), which allow moose to maximize their forage intake without moving over large areas (belovsky 1981, wickstrom et al. 1984). the use and selection of these sites during summer (≥57% of home range core areas by all groups) suggests that moose relied on regenerating forests to provide the forage required to gain weight at this critical time of year (belovsky and jordan 1978, van ballenberghe and miquelle 1990). the recent pattern of logging in massachusetts appeared to be favorable to moose. harvest sites on state and private lands were widely distributed, with <2% of the forested landscape logged annually (kittredge et al. 2003, mcdonald et al. 2006). this resulted in new patches of early successional habitat within a matrix of mature and maturing forest. the importance of thermal cover for moose in and around forest harvests and burns has been well documented (mcnicol and gilbert 1980, girard and joyal 1984, bangs et al. 1985, masterbrook and cummings 1989, thompson et al. 1995). the small size (mean = 16.5 ha) and moderate harvest intensity (27% of timber volume harvested) of forest harvest units in massachusetts (kittredge et al. 2003) resulted in short distance to edge, which provided both browse and cover in close proximity. shelterwood cuts were commonly applied to harvest units, resulting in cover from solar radiation along with browse, with the added advantage that vegetation growing in shade tends to be more nutritious and has lower secondary compound levels than growth in direct sunlight (hjeljord et al. 1990, schwartz and renecker 2007). the intense use of regenerating forests is similar to habitat use in northern new england and elsewhere (peek et al. 1976, joyal and scherrer 1978, crossly and gilbert 1983, monthey 1984, leptich and gilbert 1989, thompson et al. 1995, scarpitti 2006). however, both leptich and gilbert (1989) and miller and litvaitis (1992) found that only females selected for cut-over areas during summer in northern new hampshire and northern maine, with males selecting upland hardwoods; scarpitti (2006) found selection for regenerating stands only during winter. the high concentration of browse found in regenerating stands mimicked the permanent shrub communities used by moose in other portions of their range, including delta floodplains, tundra and subalpine areas, aspen parklands, and stream valley shrub communities, as well as the 142 habitat in massachusetts – wattles and destefano alces vol. 49, 2013 transitory early successional habitats created by fire and insect outbreaks (phillips et al. 1973, pierce and peek 1984, bangs et al. 1985, mccracken et al. 1997, poole and stuart-smith 2005, peek 2007, gillingham and parker 2008). the clear importance of early successional forest as foraging habitat for moose, however, should not take away from that fact that moose used a mix of cover types and age classes to meet their annual habitat needs in southern new england. mature coniferous, mixed, and deciduous stands were seasonally important foraging sites. additionally, moose used mature forests and a variety of wetlands as thermal shelters during periods of high temperature, and mature coniferous and mixed stands during periods of deep snow. while moose now occupy most suitable habitat in massachusetts and connecticut, additional habitat may exist in unoccupied portions of its historic range in new york and pennsylvania. forest types transition in new york and pennsylvania in a similar way as in massachusetts, from spruce-fir and northern hardwood forests to transitional and central hardwood forests, and suitable habitat likely exists for moose in the forests types of southern new york and pennsylvania. however, different state management goals (wattles and destefano 2011), greater amounts of agriculture, the highly developed mohawk river valley, and high temperatures may prevent or slow the further expansion of moose in this region. management implications the year-round intensive use of regenerating forests by moose in massachusetts underlies the importance of early successional forest. the recent pattern of logging that continually created new patches of young forest seemed to be favorable for moose. however, recently adopted plans by the dcr (the agency that manages the state forest system and public watersheds in massachusetts) that restrict or eliminate logging on some state lands could have a negative impact on moose and other wildlife that use or require early successional forest. moose rely on these sites of high forage density to gain weight for winter and support lactation of calves. a reduction in logging would result in a loss of this cover type over time from some of the largest tracts of conservation land and would force moose to forage for lower density browse in mature forest stands. this could result in higher energy expenditures to obtain the same amount of food, which may be particularly harmful for a species living in an environment at the extremes of its temperature tolerances. that heat stress has been implicated in the recent declines in moose populations elsewhere along the southern edge of the species’ range (murray et al. 2006, lenarz et al. 2009, 2010) demonstrates the importance of energy balance for moose living in these environments. management of moose habitat on a landscape scale in massachusetts should ensure the protection of large blocks of forested habitat that support a mix of age classes and forest cover types, including mature stands of coniferous, mixed-coniferousdeciduous, and deciduous forests, patches of early successional forest, and a variety of wetlands. the mix of cover types, age classes, and wetlands that currently occur in the temperate deciduous forests of massachusetts and southern new england appear to provide suitable habitat for long-term occupation by moose. in general, moose are relatively widely dispersed, actively reproducing, and present at low density in almost all forest types in central and western massachusetts. the absence of major predators and hunting undoubtedly influence the population dynamics of moose in massachusetts. the differences in the distribution, structure, and landscape alces vol. 49, 2013 wattles and destefano – habitat in massachusetts 143 configuration of key habitat components, along with large levels of development and a potentially thermally stressful environment will likely combine to limit the distribution and density of moose in southern new england. acknowledgments the massachusetts division of fisheries and wildlife (mdfw) through the federal aid in wildlife restoration program (w-35-r) provided funding and support for this research. the massachusetts dcr, u.s. geological survey, university of massachusetts-amherst, and safari club international provided additional funding and logistical support. capture of moose would not have been possible without the assistance of field technician k. berger, the massachusetts environmental police and mdfw personnel, and other technicians and volunteers. we thank r. deblinger for logistic support and t. k. fuller and j. mcdonald for reviewing the manuscript. references aebischer, n. j., p. a. robertson, and r.e. kenward. 1993. compositional analysis of habitat use from animal radio-tracking data. ecology 74: 1313–1325. bangs, e. e., s. a. duff, and t. n. bailey. 1985. habitat differences and moose use of two large burns on the kenai peninsula, alaska. alces 21: 17–35. belovsky, g. e. 1981. optimal activity times and habitat choice of moose. oecologia 48: 22–30. ———, and p. a. jordan. 1978. the timeenergy budget of a moose. theoretical population biology 14: 76–104. beyers, h. 2006. hawth's analysis tools for arcgis. version 3.27. <http://www.spa tialecology.com/htools>. boose, e. 2001. fisher meteorological station (since 2001). harvard forest data archive: hf001. petersham, massachusetts, usa. coady, j. w. 1974. influence of snow on behavior of moose. canadian fieldnaturalist 101: 417–436. crossley, a., and j. r. gilbert. 1983. home range and habitat use of moose in northern maine. transactions of the northeast section of the wildlife society 40: 67–75. degraaf, r. m., and m. yamasaki. 2001. new england wildlife: habitat, natural history, and distribution. university press of new england, hanover, new hampshire, usa. destefano, s., r. d. deblinger, and c. miller. 2005. suburban wildlife: lessons, challenges, and opportunities. urban ecosystems 8: 131–137. dussault, c., j-p. ouellet, r. courtois, j. huot, l. breton, and j. larochelle. 2004. behavioural responses of moose to thermal conditions in the boreal forest. ecoscience 3: 321–328. environmental systems research institute, inc. (esri). 2008. arcgis 9.3. redlands, california, usa. faison, e. k., g. motzkin, d. r. foster, and j. e. mcdonald. 2010. moose foraging in the temperate forests of southern new england. northeastern naturalist 17: 1–18. fieberg, j. 2007. kernel density estimators of home range: smoothing and the autocorrelation red herring. ecology 88: 1059–1066. franzmann, a. w., and c. c. schwartz. 2007. ecology and management of the north american moose. second edition. university press of colorado, boulder, colorado, usa. garner, d. l., and w. f. porter. 1990. movement and seasonal home ranges of bull moose in a pioneering adirondack population. alces 26: 80–85. gillingham, m. p., and k. l. parker. 2008. the importance of individual variation in defining habitat selection by moose in northern british columbia. alces 44: 7–20. 144 habitat in massachusetts – wattles and destefano alces vol. 49, 2013 http://www.spatialecology.com/htools http://www.spatialecology.com/htools girard, f., and r. joyal. 1984. l'impact des coupes a blanc mecanisees sur l'original dans le nord-ouest du quebec. alces 20: 3–25. gitzen, r. a., j. t. millspaugh, and b. j. kernohan. 2006. bandwidth selection for fixed-kernel analysis of animal utilization distributions. journal of wildlife management 70: 1334–1344. hall, b., g. motzkin, d. r. foster, m. syfert, and j. burk. 2002. three hundred years of forest and land-use in massachusetts, usa. journal of biogeography 29: 1319–1335. hayne, d.w. 1949. calculation of size of home range. journal of mammalogy 30: 1–18. hemson, g., p. johnson, a. south, r. kenward, r. ripley, and d. macdonald. 2005. are kernels the mustard? data from global positioning system (gps) collars suggests problems for kernel home-range analyses with least-squares cross-validation. journal of animal ecology 74: 455–463. hjeljord, o., n. hovik, and h. b. pedersen. 1990. choice of feeding sites by moose during summer, the influence of forest structure and plant phenology. holarctic ecology 13: 281–292. johnson, d. h. 1980. the comparison of usage and availability measurements for evaluating resource preference. ecology 61: 65–71. jordan, p. a., d. b. botkin, a. s. dominski, h. s. lowendorf, and g. e. belovsky. 1973. sodium as a critical nutrient for the moose of isle royale. proceedings of the north american moose conference and workshop 9: 13–42. joyal, r., and b. scherrer. 1978. summer movements and feeding by moose in western quebec. canadian field-naturalist 92: 252–258. kaufmann, j. h. 1962. ecology and social behavior of the coati, nasua nifica on barro colorado island panama. university of california publications in zoology 60: 95–222. kie, j. g., j. matthiopoulos, j. fieberg, r. a. powell, f. cagnacci, m. s. mitchell, j-m. gaillard, and p. r. moorscroft. 2010. the home-range concept: are traditional estimators still relevant with modern telemetry technology. philosophical transactions of the royal society b 365: 2221–2231. kittredge, d. b., jr., a. o. finley, and d. r. foster. 2003. timber harvesting as ongoing disturbance in a landscape of diverse ownership. forest ecology and management 180: 425–442. lenarz, m. s., j. fieberg, m. w. schrage, and a. j. edwards. 2010. living on the edge: viability of moose in northeastern minnesota. journal of wildlife management 74: 1013–1023. ———, m. e. nelson, m. w. schrage, and a. j. edwards. 2009. temperature mediated moose survival in northeastern minnesota. journal of wildlife management 73: 503–510. leptich, d. j., and j. r. gilbert. 1989. summer home range and habitat use by moose in northern maine. journal of wildlife management 53: 880–885. massachusetts office of geographic information (massgis). 2011. <http:// www.mass.gov/anf/research-and-tech/itserv-and-support/application-serv/officeof-geographic-information-massgis/data layers/layerlist.html> (accessed january 2011). masterbrook, b., and h. g. cummings. 1989. use of residual strips of timber by moose within cutovers in northwestern ontario. alces 25: 146–155. mccracken, j. g., v. van ballenberghe, and j. m. peek. 1997. habitat relationships of moose on the copper river delta in coastal south-central alaska. wildlife monographs 136: 3–52. mcdonald, r. i., g. motzkin, m. s. bank, d. b. kitteridge, j. burke, and d. r. foster. 2006. forest harvesting and alces vol. 49, 2013 wattles and destefano – habitat in massachusetts 145 http://www.mass.gov/anf/research-and-tech/it-serv-and-support/application-serv/office-of-geographic-information-massgis/datalayers/layerlist.html http://www.mass.gov/anf/research-and-tech/it-serv-and-support/application-serv/office-of-geographic-information-massgis/datalayers/layerlist.html http://www.mass.gov/anf/research-and-tech/it-serv-and-support/application-serv/office-of-geographic-information-massgis/datalayers/layerlist.html http://www.mass.gov/anf/research-and-tech/it-serv-and-support/application-serv/office-of-geographic-information-massgis/datalayers/layerlist.html http://www.mass.gov/anf/research-and-tech/it-serv-and-support/application-serv/office-of-geographic-information-massgis/datalayers/layerlist.html land-use conversion over two decades in massachusetts. forest ecology and management 227: 31–41. ———, ———, and d. r. foster. 2008. the effect of logging on vegetation composition in western massachusetts. forest ecology and management 225: 4021–4031. mcnicol, j. g., and f. f. gilbert. 1980. late winter use of upland cutovers by moose. journal of wildlife management 44: 363–371. miller, b. k., and j. a. litvatis. 1992. habitat segregation by moose in a boreal forest ecotone. acta theriologica 37: 41–50. mohr, c. o. 1947. table of equivalent populations of north american small mammals. american midlands naturalist 37: 223–249. monthey, r. w. 1984. effects of timber harvesting on ungulates in northern maine. journal of wildlife management 48: 279–285. morris, k. i. 2002. impact of moose on aquatic vegetation in northern maine. alces 38: 213–218. murray, d. l., e. w. cox, w. b. ballard, h. a. whitlaw, m. s. lenarz, t. w. custer, t. barnett, and t. k. fuller. 2006. pathogens, nutritional deficiency, and climate influences on a declining moose population. wildlife monograph 166: 1–30. peek, j. m. 2007. habitat relationships. pages 351–375 in a.w. franzmann and c.c. schwartz, editors. ecology and management of the north american moose. second edition. university press of colorado, boulder, colorado, usa. ———, d. l. urich, and r. j. mackie. 1976. moose habitat selection and relationships to forest management in northeastern minnesota. wildlife monographs 48: 3–65. phillips, r. l., w. e. berg, and d. b. siniff. 1973. moose movement patterns and range use in northwestern minnesota. journal of wildlife management 37: 266–278. pierce, d. j., and j. m. peek. 1984. moose habitat use and selection patterns in north-central idaho. journal of wildlife management 48: 1335–1343. poole, k. g., r. serrouya, and k. stuartsmith. 2007. moose calving strategies in interior montane ecosystems. journal of mammalogy 88: 139–150. ———, and k. stuart-smith. 2005. finescale winter habitat selection by moose in interior montane forests. alces 41: 1–8. powell, r. a. 2000. animal home ranges and territories. pages 65-110 in l. boitani and t. k. fuller, editors. research techniques in animal ecology. columbia university press, new york, new york, usa. r development core team. 2005. r: a language and environment for statistical computing. r foundation for statistical computing, vienna, austria. <http:// www.r-project.org>. renecker, l. a., and r. j. hudson. 1986. seasonal energy expenditures and thermoregulatory responses of moose. canadian journal of zoology 64: 322–327. ritcey, r. w., and n. a. m. verbeek. 1969. observations of moose feeding on aquatics in bowron lake park, british columbia. canadian field-naturalist 83: 339–343. rogers, a. r., a. p. carr, h. l. beyer, l. smith, and j. g. kie. 2007. hrt: home range tools for arcgis. ontario ministry of natural resources, centre for northern forest ecosystem research, thunder bay, ontario, canada. samuel, m. d., d. j. pierce, and e. o. garton. 1985. identifying areas of concentrated use within home range. journal of animal ecology 54: 711–719. scarpitti, d. l. 2006. seasonal home range, habitat use, and neonatal habitat characteristics of cow moose in northern new hampshire. m.s. thesis, university of 146 habitat in massachusetts – wattles and destefano alces vol. 49, 2013 http://www.r-project.org http://www.r-project.org new hampshire, durham, new hampshire, usa. ———, c. habeck, a. r. musante, and p. j. pekins. 2005. integrating habitat use and population dynamics of moose in northern new hampshire. alces 41: 25–35. schwab, f. e., and m. d. pitt. 1991. moose selection of canopy cover types related to operative temperature, forage, and snow depth. canadian journal of zoology 69: 3071–3077. schwartz, c. c., and l. a. renecker. 2007. nutrition and energetics. pages 441-478 in a.w. franzmann and c.c. schwartz, editors. ecology and management of the north american moose. second edition. university press of colorado, boulder, colorado, usa. telfer, e. s. 1988. habitat use by moose in southwestern alberta. alces 31: 153–166. the weather channel. 2011a. monthly averages for petersham, massachusetts. <http://www.weather.com/weather/wxcli matology/monthly/graph/usma0497> (accessed january 2011). ———. 2011b. monthly averages for windsor, ma. <http://www.weather.com/ weather/wxclimatology/monthly/graph/ usma0497> (accessed january 2011). thompson, m. e., j. r. gilbert, g. j. matula, and k. i. morris. 1995. seasonal habitat use by moose on managed forest lands in northern maine. alces 31: 233–245. u. s. census bureau. n.d. census. 2000. population housing units, area and density for states: 2000. <http://www. census.gov/population/www/censusdata/ density.html.> (accessed june 2010). van ballenberghe, v., and d. g. miquelle. 1990. activity of moose during spring and summer in interior alaska. journal of wildlife management 54: 391–396. vecellio, g. m., r. d. deblinger, and j. e. cardoza. 1993. status and management of moose in massachusetts. alces 29: 1–7. wattles, d. w. 2011. status, movements, and habitat use of moose in massachusetts. m.s. thesis, department of environmental conservation, university of massachusetts, amherst, massachusetts, usa. ———, and s. destefano. 2011. status and management of moose in the northeastern united states. alces 47: 53–68. westveldt, m. r., r. i. ashman, h. i., baldwin, r. p. holdsworth, r. s. johnson, j. h. lambert, h. j. lutz, l. swain, and m. standish. 1956. natural forest vegetation zones of new england. journal of forestry 54: 332–338. westworth, d., l. brusnyk, j. roberts, and h. veldhuzien. 1989. winter habitat use by moose in the vicinity of an open pit copper mine in north-central british columbia. alces 25: 156–166. wickstrom, m. l., c. t. robbins, t. a. hanley, d. e. spalinger, and s. m. parish. 1984. food intake and foraging energetic of elk and mule deer. journal of wildlife management. 48: 1285–1301. worton, b. j. 1989. kernel methods for estimating the utilization distribution in home-range studies. ecology 70: 164–168. alces vol. 49, 2013 wattles and destefano – habitat in massachusetts 147 http://www.weather.com/weather/wxclimatology/monthly/graph/usma0497 http://www.weather.com/weather/wxclimatology/monthly/graph/usma0497 http://www.weather.com/weather/wxclimatology/monthly/graph/usma0497 http://www.weather.com/weather/wxclimatology/monthly/graph/usma0497 http://www.weather.com/weather/wxclimatology/monthly/graph/usma0497 http://www.census.gov/population/www/censusdata/density.html http://www.census.gov/population/www/censusdata/density.html http://www.census.gov/population/www/censusdata/density.html moose habitat in massachusetts: assessing use at the southern edge of the range study area methods study animals and gps telemetry seasons habitat availability and core area habitat use analyses results capture and deployment of gps collars home range core area habitat use discussion management implications acknowledgments references alces20_61.pdf alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 alces vol. 20, 1984 diurnal defecation rate of moose in southwest finland juho matala1 and antti uotila2 1finnish forest research institute metla, p.o. box 68, fi-80101 joensuu, finland; 2university of helsinki hyytiälä forest station, hyytiäläntie 124, fi-35500 korkeakoski, finland. abstract: an accurate measure of defecation rate is essential for application of pellet group counts in moose (alces alces) population estimates. we measured the wintertime, diurnal defecation rate of moose by tracking 7 gps-collared and 22 uncollared moose in southwest finland. the mean defecation rate was 23.5 ± 4.2 pellet groups/d, one of the highest values reported. the mean defecation rate did not differ between the tracking methods (gps vs. uncollared moose); limited sample size precluded conclusions about sex and age differences. the defecation rate was not correlated with calendar week, length of accumulation period, or number of diurnal beds. our results are appropriate for use in southwest finland when using the pellet group method to assess moose population density. alces vol. 49: 155–161 (2013) key words: alces alces, finland, defecation rate, moose, pellet group, tracking. counting fecal pellet groups of moose (alces alces) has been widely used to estimate habitat utilization, feeding behavior, and population trends and density (see franzmann et al. 1976b, forbes and theberge 1993, härkönen and heikkilä 1999, rönnegård et al. 2008, månsson 2009, månsson et al. 2011a, b). reliable moose population estimates are not always realized from pellet group counts (rönnegård et al. 2008), but the method's usefulness has been noted (neff 1968, lautenschlager and jordan 1993, månsson et al. 2011b), despite some uncertainty (neff 1968). however, to successfully estimate moose population density using a pellet group count, it is critical to use an accurate defecation rate in the survey area and time period (timmermann 1974). earlier studies have used 2 main methods to estimate the defecation rates of moose: 1) track moose in snow-covered terrain (joyal and ricard 1986, andersen et al. 1992), and 2) estimate the number of pellet groups in a closed area or island where the number of moose is known (jordan et al. 1993). defecation rates have occasionally been estimated by comparing the aerial censuses and pellet group counts in specific areas (rönnegård et al. 2008). moose enclosures could also be utilized, but results from domestic moose can be affected by food quality and behaviour that are dissimilar to natural conditions. gps radio-collars enable intensive and accurate tracking by identifying specific individuals and the beginning and end points of their specific tracks, providing ideal conditions to measure defecation rates of free-ranging moose. moose defecation rates vary by age, sex, habitat, food quality, season, and year (desmeules 1968, franzmann et al. 1976a, oldemeyer and franzmann 1981, joyal and ricard 1986, andersen et al. 1992, månsson et al. 2011b). large variations in defecation rate have been reported in different areas; for example, in north america variation was 9.6–32.2 pellet groups/d (timmermann 1974), and in northern europe rates varied corresponding author: juho matala, finnish forest research institute metla, p.o.box 68, fi-80101 joensuu, finland. +358 40 801 5275. juho.matala@metla.fi 155 from 14–26.9 pellet groups/d (andersen et al. 1992, remm and luud 2003, rönnegård et al. 2008). these data emphasize the importance of using area-specific rates when using the pellet group method to estimate moose population density. our main objective was to formulate a general estimate of the wintertime, diurnal defecation rate of moose in southwest finland by tracking both gps-collared and uncollared moose in snowy terrain. we also compared these 2 tracking methods and looked for differences in defecation rates between sex and age. study area moose were tracked in 2 separate areas approximately 100 km apart in southwest finland (fig. 1). uncollared moose were tracked in the orivesi-kangasala area (∼wgs84 61°36′ n, 24°22′ e) and gpscollared moose in the loppi-hyvinkää area (∼wgs84 60°38′ n, 24°35′ e). both areas are located in the southern boreal vegetation zone with scots pine (pinus sylvestris) and norway spruce (picea abies) as the dominant tree species. forest cover was 78% of the total land area in orivesi-kangasala and 71% in loppi-hyvinkää (metla 2012). methods tracking of uncollared moose we actively searched for uncollared moose in their known habitats during fresh snow conditions between december and april, 1999–2003. the accumulation period for pellet counts began when moose were seen or flushed, enabling data collectors to accurately time their count by locating fresh resting places, pellet groups, or tracks in fig. 1. study area locations and starting points of moose tracking periods in southern finland. in the orivesi-kangasala area some of the starting point coordinates of uncollared moose are rough estimates and overlap because they could not be separated at the map scale. 156 diurnal defecation rate of moose – matala and uotila alces vol. 49, 2013 the snow. the pellet groups were counted the following day, until moose fled from the counter. the end of the accumulation period was determined in a similar manner as the starting point. the accumulation periods varied from 8–31 h. moose were classified as either adults or calves by visually observing them and their pellets, bedding places, and behavior. their sex could be determined without visual contact by analyzing their urination methods, as bulls urinate in front of the hind hooves. we measured the diurnal defecation rate of 5 bulls, 3 cows, and 3 calves. additionally we measured 3 cows with twin calves and 1 cow with a single calf, without separating the pellet groups of the cows and calves. in total, we were able to record the diurnal defecation rates of 22 uncollared moose and count the number of beds of 7 individuals. tracking of gps-collared moose the finnish game and fisheries research institute (fgfri) provided location data for the gps-collared moose. the fgfri implemented gps-collaring procedures in accordance with finnish legislation, with permission from the national animal experiment board of finland. an individual gps-collared cow was tracked once and another on 5 separate occasions, while 1 cow with twin calves was tracked once, and one cow with a single calf twice; altogether, 7 individuals were tracked. the gps-collared moose were tracked a few days after fresh snowfall in december– march 2010. the counter went to the most recent location of moose tracks which were usually ∼12 h old. the pellet groups were counted along the moose track by following it against the original course of the moose; the coordinates of the pellet groups, beds, and urination sites were located with handheld gps devices. the accumulation period finished when an individual track became mixed with others or sunset precluded tracking. the duration of the accumulation period was determined by identifying the time associated with the closest location of the tracks with gps collar data. the accumulation periods ranged from 6–47 h. data analysis at least 20 pellets were required to make a pellet group. we processed the pellet group data of 29 moose (7 gps-collared, 22 uncollared) to calculate the diurnal defecation rate (number of pellet groups produced per individual in 24 h) from the accumulation periods of individual bulls, cows, and calves. mean values were calculated for the group of cows and calves when it was impossible to identify calf from cow pellet groups; analyses of 25 separate cases were used to calculate the diurnal defecation rate. for comparison, we sampled the mean values of diurnal defecation rates of 3 sex and age classes: 1) bulls, 2) cows, and 3) calves and cow-calves. the last class was required because we were only able to track 3 individual calves, which was insufficient for any reasonable analysis. we also compared the mean defecation rates of the uncollared and gps-collared moose. due to limitations in the linearity and homogeneity of variances in the data, we used the kruskal-wallis test for multiple samples and the mann-whitney test for paired sample comparisons. furthermore, we searched for possible correlations between the diurnal defecation rate and 1) the calendar week of the accumulation period, 2) the duration of the accumulation period, and 3) the number of diurnal beds for part of the samples. all statistical calculations were performed using the statistical package for the social sciences (spss) 17.0 software. data relating to home range of moose during the accumulation period (calculated using the home range tools for arcgis® version 1.1 with the minimum convex polygon method), length of the moose track alces vol. 49, 2013 matala and uotila – diurnal defecation rate of moose 157 during the accumulation period, and the number of diurnal urinations were also collected from the gps-collared moose, but insufficient sample sizes precluded their utilization in the analysis. results the diurnal defecation rate ranged from 12.2–32 pellet groups with an overall mean of 23.5 ± 4.2 (sd; table 1, fig. 2). the mean values of bull, cow, and calf/cow-calf groups were different (kruskal-wallis test: χ2 = 9.9, df = 2, p = 0.007; table 1). the calf and cow-calf group had the highest mean rate, but was statistically different only from the cow group (table 1). the bull group had the lowest mean rate, but also the lowest and highest absolute values (i.e., the widest range; table 1). the cow and bull group rates and the rates of the uncollared and gps-collared moose were not different (table 1). the diurnal defecation rate was not related to the calendar week of the accumulation period (fig. 2, table 2). the gpscollared individual which was tracked 5 times between 17 january and 23 february 2010 showed no trends during this period. the defecation rate did not correlate to the duration of the accumulation period or to the number of diurnal beds (table 2). the mean values for the other variables were: diurnal number of beds = 8.0 ± 4.2 (sd, n = 15), the diurnal urination rate = 1.0 ± 1.0 (sd, n = 9), the area of home range during the accumulation period = 95,267 m2 ± 122,399 (sd, n = 9), and length of track during the accumulation period = 955 m ± 832 (sd, n = 6). discussion the mean defecation rate was considered high (23.5 ± 4.2 pellet groups/d) but similar to the average rate measured in relatively good moose habitat in southern norway (22.9 pellet groups/d; andersen et al. 1992). lower defecation rates were measured in southern sweden (14 pellet groups/d; rönneberg et al. 2008) and on isle royale, north america (20.9; jordan et al. 1993). many studies have reported lower defecation rates in alaska and canada (see desmeules 1968, franzmann et al. 1976a, oldemeyer and franzmann 1981, joyal and ricard 1986). the high values reported in our study area are not unreasonable when comparing the status of the moose population in finland to other nordic countries. the finnish moose population is lower and of a higher productive state compared to those in table 1. mean values of the wintertime, diurnal defecation rate and comparisons between moose groups in southern finland. pellet groups (#/ind/24 h) mann-whitney test mean min max n sd test against u-value p grouping by moose type: bulls 20.1 12.2 32.0 5 7.3 cow 12 0.152 cows 22.3 18.9 24.0 9 2.2 calf + groups 10 0.002 calf and cow-calf groups 25.9 23.0 31.3 11 2.2 bull 11 0.061 grouping by tracking method: gps-collared 23.5 20 26.7 9 1.8 uncollared 64 0.647 uncollared 23.4 12.2 32.0 16 5.2 all moose 23.5 12.2 32.0 25 4.2 158 diurnal defecation rate of moose – matala and uotila alces vol. 49, 2013 sweden and norway (lavsund et al. 2003, tiilikainen et al. 2012). these differences presumably imply better foraging habitat, higher nutritional condition, and a resultant higher defecation rate in finland. because defecation rates are influenced by food quality and availability and show large variation (andersen et al. 1992), it follows that areaspecific defecation rates should be measured when using pellet group counts to estimate population density. the extremes in defecation rate varied largely in our study, but most observations, especially of the gps-collared individuals, were similar to the mean indicating the general reliability of these data and their application to estimate population density. the extreme values could result from local foraging conditions or from longer movements associated with unintended disturbances while tracking the uncollared moose. because the defecation rate can correlate with age and sex, these relationships should be taken into account if a change in population structure occurs (franzmann et al. 1976a). our data indicate that the calf and cow-calf group had slightly higher defecation rates than individual cows as also reported by desmeules (1968), but opposite that of joyal and ricard (1986). bulls and cows were not different in our analyses; however, general conclusions about sex and age differences cannot be made due to the limited sample size. using an accurate defecation rate is critical when applying pellet group counts in moose population density estimates. defecation rates have not been published previously in finland, and given the considerable variation in similar data from northern europe, we consider our results the best available for local use. furthermore, similar defecation rates were obtained for both regions in our study, making the mean values applicable to all of southern finland. the lack of correlation between calendar week and defecation rate indicates a stable accumulation rate throughout winter, suggesting that these rates might be applicable for springtime pellet group counts. variations in the absolute data and means should be taken into account when calculating confidence limits for the final moose density estimate. acknowledgements we thank mr. j. sillanpää for tracking the gps-collared moose and mr. h. mattila 50 fig. 2. the diurnal defecation rate of moose in southern finland relative to the time of the tracking period (calendar weeks). moose types are classified as: 1) individual bulls, calves, and cows, and 2) cow-calf groups (when individuals could not be separated during tracking); cow1 = cow with single calf, cow2 = cow with 2 calves. table 2. pearson correlations between the diurnal defecation rate (pellet groups/moose/24 h) and temporal variables in southern finland; no differences were found. calendar week accumulation period (h) beds/ moose/ 24 h coefficient −0.144 −0.086 −0.403 p-value 0.492 0.682 0.136 n 25 25 15 alces vol. 49, 2013 matala and uotila – diurnal defecation rate of moose 159 for tracking in the kangasala area. dr. j. pusenius from the finnish game and fisheries research institute is greatly acknowledged for organizing the gps-collaring and thus enabling data availability on exact moose locations for our study. we also thank the finnish forest research institute, the university of helsinki, and the ministry of agriculture and forestry for providing financing for our work. m.sc. m. melin is acknowledged for producing the maps in fig. 1 and ms. s.thompson for checking the language of this manuscript. references andersen, r., o. hjeljord, and b-e. saether. 1992. moose defecation rates in relation to habitat quality. alces 28: 95–100. desmeules, p. 1968. détermination du nombre de tas de crottins rejetés et du nombre de reposées établies, par jour, par l'original (alces alces), en hiver. le naturaliste canadien 95: 1153–1157. (in french with english abstract). franzmann, a. w., p. d. arneson, and j. l. oldemeyer. 1976a. daily winter pellet groups and beds of alaskan moose. journal of wildlife management 40: 374–375. ———, j. l. oldemeyer, p. d. arneson, and r. k. seemel. 1976b. pellet-group count evaluation for census and habitat use of alaskan moose. proceedings of north american moose conference workshop 12: 127–142. forbes, g. j., and j. b. theberge. 1993. multiple landscape scales and winter distribution of moose, alces alces, in a forest ecotone. canadian field-naturalist 107: 201–207. härkönen, s., and r. heikkilä. 1999. use of pellet-group counts in determini