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 avail- able 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 regio- nal 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, reach- ing 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 expendi- ture of energy above that required for main- tenance 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 dimor- phic cervids, antlers have significance in sex- ual selection and are correlated with social dominance and mating success (Clutton- Brock and Albon 1989, Bartoš 1990, Solberg and Sæther 1994, Pélabon and Joly 2000, Stewart et al. 2000). These developmental, morphological, and sociobehavioral attri- butes 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 popu- lation (Peterson 1977) should illuminate the poorly understood influence of senescence on antler development (Bubenik 1998). Asymmetry, defined as random devia- tions 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 developmen- tal 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 rela- tive individual fitness regardless of age (Markusson and Folstad 1997, Pélabon and van Breukelen 1998, Bowyer et al. 2001, Ditchkoff et al. 2001). Population wide stres- sors, 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). Nutri- tional 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 preda- tion 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 sta- tus 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 asymme- try will follow similar general patterns measured for other North American popula- tions. 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 pro- tected 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 gener- ally below 1/km2 (Karns 1998). Likewise, moose populations located on the nearest mainland in Southwest Ontario and North- east 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 lar- gest 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 indivi- duals 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 mea- surement 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 deter- mine the degree to which these scores accu- rately 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 sub- populations 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 run- ning 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 deter- mined 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 circumfer- ence, 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 signed- rank 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 Alas- kan 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 avail- able 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 Bube- nik 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 relation- ship 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 exponen- tially 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 para- meters 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 nega- tively 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 differ- ent 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 dia- meter (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 Gas- away 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 compel- ling. 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 indi- viduals. Neither sample is randomly selected; in the case of Isle Royale, indivi- duals were collected after death from natural causes, and so probably include proportio- nately 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 intro- duce 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 Technolo- gical University, unpublished data). This evi- dence 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 asym- metry 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 popu- lation density (Karns 1998, Peterson 1999, R. O. Peterson, unpublished data, Ontario Ministry of Natural Resources, unpub- lished 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 indivi- duals, in many cases ranking larger indivi- duals 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 compara- tive studies on antlers. Moose numbers on Isle Royale are natu- rally regulated with no human interference, which allows individual moose the opportu- nity to reach ages when signs of senescence would be expected. In most cases, the second order polynomials used in regression esti- mated reductions in size for older indivi- duals. 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 drasti- cally 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 nor- mal 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 measur- able restrictions. A. B. Bubenik (pers. com- mun.) 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 pro- nounced 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 govern- ing 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 condi- tion developing the largest and most sym- metric 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 subpopula- tion 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 con- tributing factors, such as genetic founder effects and effects of sampling methodology. This study also supports the contention that antlers are useful indicators for both indivi- dual and population condition (e.g., Markus- son and Folstad 1997, Pélabon and van Breukelen 1998, Strickland and Demarais 2000, Schmidt et al. 2001), although future research should attempt to specifically evalu- ate fitness in relation to measures of antler size and asymmetry for moose. ACKNOWLEDGEMENTS We thank P. DeWitt, E. Parker, C. Peter- son, 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 Ser- vice (Co-op Agreement No. J631005N00- 40003), and ROP, the Robbins Chair at Michigan Technological University in Sus- tainable 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 indica- tors 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, edi- tors. 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 . BOWYER, R. T., K. M. STEWART, J. G. KIE, and W. C. GASAWAY. 2001. Fluctuating asym- metry 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: Evolu- tion, Morphology, Physiology and Social Significance. Springer-Verlag, New York, New York, USA. BUBENIK, A. B. 1990. Epigenetical, morpho- logical, physiological, and behavioral aspects of evolution of horns, prong- horns, and antlers. Pages 3–113 in G. A. Bubenik and A. B. Bubenik, editors. Horns, Pronghorns, and Antlers: Evolu- tion, 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, edi- tors. Ecology and Management of the North American Moose, 1st edition. Smithsonian Institution Press, Washing- ton, D. C., USA. CEDERLUND, G., and G. MARKGREN. 1987. The development of the Swedish moose population, 1970–1983. Swedish Wild- life 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, Washing- ton, D. C., USA. MARKUSSON, E., and I. FOLSTAD. 1997. Rein- deer antlers: visual indicators of indivi- dual 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. Fluc- tuating asymmetry: measurement, analy- sis, 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 indi- vidual 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 Mono- graph Series No. 11. United States Gov- ernment 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 regula- tion? 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 popu- lation density, weather and early life- history. Oecologia 127: 191–197. SCRIBNER, K. T., and M. H. SMITH. 1990. Genetic variability and antler develop- ment. Pages 460–473 in G. A. Bubenik and A. B. Bubenik, editors. Horns, Pronghorns, and Antlers: Evolution, Morphology, Physiology and Social Sig- nificance. 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 asso- ciated 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