91 MOVEMENTS AND RESOURCE USE BY MOOSE IN TRADITIONAL AND NONTRADITIONAL HABITATS IN NORTH DAKOTA James J. Maskey Jr.1,2 and Rick A. Sweitzer1,3 1Department of Biology, 10 Cornell Street, Stop 9019, University of North Dakota, Grand Forks, North Dakota 58202, USA; 2Department of Biology, University of Mary, 7500 University Drive, Bismarck, North Dakota 58504, USA; 3Great Basin Institute,16750 Mt. Rose Hwy, Reno, Nevada 89511, USA ABSTRACT: In the past several decades, moose (Alces alces) have expanded their range in North Dakota from primarily forested areas to the prairie/agriculture mosaic of the state. As a result, moose are now well- established in a large portion of North Dakota, yet little is known about their ecology in the state. We examined the home ranges, habitat selection, and diets of moose in both traditional (forested) and nontraditional ranges (prairie/agricultural) and inferred whether range expansion is the result of agriculture-related landscape changes. From 2004 to 2006, we placed GPS radio-collars on a total of 14 moose in two study areas: Turtle Mountains (forested) and Lonetree (prairie/agricultural). Total and seasonal home ranges were larger for Lonetree moose, and moose in both study areas selected strongly for wooded habitat. In both study areas seasonal diets ranged from 65 to 99% woody browse, with forbs 15% of summer diets. In the Lonetree area row crops made up the second highest consumed forage in fall (12%) and winter (29%) diets. Larger home ranges in the Lonetree area may reflect the low avail- ability and scattered distribution of wooded habitat. Further, the strong selection for planted woodlands and the high proportion of woody browse and row crops in the diet of Lonetree moose suggests that conversion of the native prairie to agriculture has facilitated range expansion by moose in North Dakota. ALCES VOL. 55: 91–104 (2019) Key Words: Alces alces, browse, cropland, diet, habitat selection, home ranges, North Dakota, prairie, woodland INTRODUCTION Moose (Alces alces) are native to North Dakota with their traditional range encom- passing the aspen (Populus tremuloides) and bur oak (Quercus macrocarpa) forests of the Turtle Mountains and Pembina Hills along the northern edge of the state (Knue 1991). While moose were extirpated from North Dakota by the late 1800s, they had begun to re-establish a population in the state by the 1960s. After re-colonizing their historic range, by the 1980s moose had expanded their range to include large expanses of for- mer tall and mixed grass prairie that had been greatly modified by conversion to agri- culture and widespread planting of tree rows to reduce wind erosion subsequent to the Dust Bowl years of the 1930s (Knue 1991, Licht 1997). The colonization and range expansion by moose in North Dakota are likely the result of conversion of the native prairie landscape to an agricultural mosaic that pro- vides suitable cover and forage otherwise absent in unaltered tall or mixed grass prairie habitats. Although moose are known to per- sist in other landscapes modified by humans such as clear-cuts and agricultural areas within forested landscapes (Leptich and Gilbert 1989, Rempel et al. 1997, Schneider and Wasel 2002), the agriculture-dominated landscape of the northern Great Plains MOVEMENTS AND RESOURCE USE BY MOOSE – MASKEY AND SWEITZER ALCES VOL. 55, 2019 92 represents a unique habitat for the species that was not occupied prior to human-induced habitat change. While numerous prior efforts have provided insight into moose move- ments and resource use in traditional habi- tats (Kearney and Gilbert 1976, Leptich and Gilbert 1989, Cederlund and Sand 1994, MacCracken et al. 1997, Labonte et al. 1998), the ecology and behavior of moose in the prairie ecoregions of North America is relatively unknown. The purpose of this project was to inves- tigate the ecology of moose in both tradi- tional woodland habitats and the recently colonized prairie region of North Dakota, including how this species may be taking advantage of landscape alterations to extend its range. The specific objectives were to 1) examine seasonal and annual movements and habitat use of moose in the prairie and woodland regions of North Dakota, 2) inves- tigate the diet of moose in prairie and wood- land regions of North Dakota, and 3) compare movements, habitat use, and diets of moose in these two regions. To meet these objectives, we selected study areas that were representative of traditional and more recently colonized habitats. First, the for- ested Turtle Mountains comprise a major portion of the historic range of moose in North Dakota and was one of the areas first recolonized upon their return (Knue 1991, Seabloom et al. 2011). Second, the Lonetree Wildlife Management Area (WMA) is an agricultural mosaic characteristic of the hab- itats more recently colonized by moose. It is representative of most of the landscape of eastern and central North Dakota, which now comprises much of the primary range of this species in the state. STUDY AREA The Turtle Mountains (48° 57ʹ N, 99° 53ʹ 00” W; Fig. 1) are located along the Canadian border and are characterized by hilly wooded terrain and numerous small lakes and wetlands with interspersed agri- cultural fields, pastureland, and hay fields, especially near the southern edge of the area. The forest of the Turtle Mountains is comprised primarily of aspen and bur oak along with green ash (Fraxinus pennsylvan- ica), paper birch (Betula papyrifera), bal- sam poplar (Populus balsamea) , and box elder (Acer negundo), with an understory of chokecherry (Prunus virginiana), hazel (Corylus cornuta), and several species of willow (Salix spp.). Typical herbaceous species include sarsaparilla (Aralia nudi- caulis), alfalfa (Medicago sativa), brome (Bromus spp.), fescue (Festuca spp.), wheatgrass (Agropyron spp.), sedges (Carex spp.), baneberry (Actea spp.), false lily of the valley (Maianthemum canadensis), wild vetch (Vicia americana), and Virginia anemone (Anemone virginiana) (Stevens 1966, Bakke 1980, ND Forest Service 2003). The Lonetree WMA is large, encom- passing 134 km2 in the central part of the state (47°30ʹ N, 100°15ʹ W; Fig. 1). It con- sists of farmland initially purchased by the U.S. Bureau of Reclamation to be used as part of the Missouri River Garrison Diversion Project (Garrison Diversion Project 2019). Following the cancellation of that portion of the project, management of the land was turned over to the North Dakota Game and Fish Department. Habitats include native mixed grass prairie, corn (Zea mays) and sunflower (Helianthus ann- uus) food plots (range = 6–31 ha), numerous seasonal and semi-permanent wetlands, small impoundments along the Sheyenne River, and planted woodlands in the form of linear tree rows or larger block plantings (Smith et al. 2007). The surrounding area is comprised primarily of pasture and hay land ALCES VOL. 55, 2019 MOVEMENTS AND RESOURCE USE BY MOOSE – MASKEY AND SWEITZER 93 as well as crop fields consisting mostly of small grains. Planted tree rows and wood- lots are present with some natural wood- lands occurring in woody draws along the Missouri Escarpment that marks the border between the Northern Glaciated Plains and Missouri Coteau ecoregions (USEPA 1996). Typical grassland plants found in the area include prairie junegrass (Koeleria macran- tha), indiangrass (Sorghastrum nutans), needle and thread grass (Hesperostipa comata), brome, wheatgrass, and alfalfa. Common tree species in planted and/or native woodlands include green ash, box elder, American elm (Ulnus americana), Russian olive (Elaeagnus angustafolia), plum (Prunus spp.), apple (Malus spp.), chokecherry, fireberry hawthorn (Crataegus chrysocarpa), serviceberry (Amelanchier alnifolia), and willow. METHODS Study animals Global positioning system (GPS) radio- collars (Lotek Wireless Inc. Newmarket, Ontario, Canada) were placed on 14 adult moose (5 cows, 1 bull in the Lonetree WMA; 4 cows, 4 bulls in the Turtle Mountains) in January 2004–2006. Only moose in the Lonetree WMA study area were captured in 2004, with subsequent expansion to the Turtle Mountains in 2005 and 2006. Moose were captured from helicopter with the use of a net gun. Capture operations were per- formed by Leading Edge Aviation (Lewiston, Idaho, USA), and all methods were approved by the University of North Dakota (IACUC Project #0506-3). Collars were set to acquire a location every 4 h, and location data were stored on board. After ~ 52 weeks, radio-col- lars were recovered when moose were Fig. 1. Location of the Lonetree Wildlife Management Area and Turtle Mountains study areas in North Dakota, USA. Boundaries delineate North Dakota counties. MOVEMENTS AND RESOURCE USE BY MOOSE – MASKEY AND SWEITZER ALCES VOL. 55, 2019 94 recaptured or after they dropped off via timed-release mechanisms. Moose were periodically monitored by aerial and ground- based VHF telemetry. Home range estimation Total and seasonal moose home range sizes (km2) were estimated with a 95% fixed-kernel estimator, with seasons defined as winter (1 January–30 April), summer (1 May–31 August), and fall (1 September–31 December) for all analyses. We carried out these calculations using all available loca- tions for non-dispersing moose, with disper- sal defined as locations for a moose that deviated from other grouped relocations for that animal (Dodge et al. 2004). Seasonal home range sizes were estimated for each moose for all seasons for which at least 30 locations were available, and total home ranges were estimated for moose with at least 30 locations in every season (Seaman et al. 1999). Moose were considered to exhibit seasonal migrations if < 25% of their seasonal home ranges overlapped (Dodge et al. 2004). Location data were input into ArcMap 9.2 (ESRI Inc., Redlands, California, USA) and home range calcula- tions were performed using the Home Range Extension (Rodgers and Carr 1998). At the time of our study, least squares cross validation (LSCV) was the most rec- ommended technique to determine the opti- mal smoothing parameter for fixed-kernel home range estimation (Worton 1995, Seaman and Powell 1996, Seaman et al. 1999). However, we experienced similar problems with this method as reported by others (Silverman 1986, Hemson et al. 2005); LSCV was unable to calculate a smoothing parameter for most sets of loca- tions, and if it did, the multi-modal nature of the locations produced home ranges that were dramatically under-smoothed. To deal with these problems, we used biased cross validation (BCV) to calculate the smoothing parameters for all fixed-kernel home range estimations (Wand and Jones 1995, Rodgers and Carr 1998). Although BCV has not been commonly applied to estimate the smooth- ing parameters for home range estimates, the statistical literature has demonstrated its util- ity in selecting kernel bandwidth, as well as its potential superiority to LSCV (Sain et al. 1994, Wand and Jones 1995, Rodgers and Carr 1998). We compared total home range sizes between study sites with a two-sample t-test. For all moose with home range esti- mates for all seasons, we also compared sea- sonal home range sizes among seasons and study sites with repeated-measures ANOVA. When necessary, home range sizes were nat- ural log transformed to meet the assumptions of parametric tests. All statistical compari- sons were performed in the statistical pack- age R 2.6 (R Core Development Team 2007). Habitat selection We first estimated the extent of the area available to moose in each study area by constructing 99% fixed-kernel home ranges for each moose, and then combined the home ranges for each study site into a single poly- gon. We then used land cover data from the United States Geological Survey’s Gap Analysis Program (GAP) compiled from 1992 to 1999 (Strong et al. 2005) as well as National Wetland Inventory 1:24,000 digital quadrangles (USFWS 2000) to determine habitat types available to moose. To make the comparison of habitat use between study sites possible, land cover data were collapsed into 4 habitat types using Spatial Analyst in ArcMap 9.2 (ESRI Inc., Redlands, California, USA). These were defined as woodland (all planted and naturally occur- ring woodlands), wetland (temporary, sea- sonal, permanent, and semi-permanent wetlands), grasslands (planted non-native grasses, hay fields, old fields, and planted ALCES VOL. 55, 2019 MOVEMENTS AND RESOURCE USE BY MOOSE – MASKEY AND SWEITZER 95 or naturally occurring prairie), and crops (all planted row crops or grains). Next, because preliminary analysis indi- cated that the coarse spatial resolution (30- m) of the GAP data were insufficient to detect small areas of habitat, we improved the reso- lution of terrestrial habitat layers by re-digi- tizing data based on 1-m resolution aerial photos of each study site (National Agricultural Imagery Program, USDA-FSA 2005). We did this by overlaying GAP habi- tat layers onto the aerial photos in ArcMap, then manually re-digitizing the GAP layers to conform to the habitat boundaries indicated on the photos. We modified the wetland hab- itat layer by considering all temporary and seasonal wetlands to be part of the terrestrial habitat in which they were imbedded, as these wetlands are typically inundated only during the spring and do not provide a source of emergent or submergent aquatic vegeta- tion (USFWS 2000). We also adjusted wet- land habitat availability to account for the presence of several larger lakes in the 2 study areas. Because the deep-water areas in these lakes were unlikely available to moose, we created a 100 m buffer layer that extended from the shoreline into each lake. This dis- tance was chosen as a conservative estimate of the extent of the littoral zone, where water depth was shallow and emergent and sub- mergent plants would occur. The area of this buffer layer was considered the amount of lake habitat available to moose. We measured the area of each habitat type in each study area using the X-Tools extension for ArcMap 9.2 (ESRI Inc., Redlands, California, USA) and then determined the proportional avail- ability of each habitat (Table 1). All locations for individual moose at each study site were separated into seasons. We then calculated Manly’s standardized selection ratios for each moose with ≥30 locations in a season (Manly et al. 2002, Osko et al. 2004). This method produces selection ratios that represent the probability of a moose using a particular habitat if all habitats were equally available, and the selection ratios for all habitats sum to 1. Because there were four habitat types in this study, a selection ratio of 0.25 indicates non-selection (not different than by chance), while a selection ratio >0.25 represent posi- tive selection for a habitat type. Diet In 2005–2006 we collected 5 samples of fresh moose feces monthly from each study area by searching several locations distrib- uted across each study site, with no more than 2 samples/month collected from a single location (e.g., from the same clear-cut). Samples were combined to generate a series of 2-month composite fecal samples. Samples were sent to the Wildlife Habitat and Nutrition Laboratory at Washington State University (Pullman, Washington, USA) for microhisto- logical determination of plant fragments and estimates of diets to the genus and species level (Van Vuren 1984). Forage plants were classified into 5 categories: woody browse, grasses and sedges, forbs, crops, and other (fruits, nuts, aquatic vegetation). Results of this diet analysis were grouped by season based on the same criteria used for home range and habitat use analyses. RESULTS Home range and movement A high rate of collar failure in 2005 pre- vented the calculation of all seasonal and Table 1. Proportional availability of each of the four major habitat types in the Lonetree and Turtle Mountains study areas in North Dakota, USA. Woodland Wetland Grass Crop Lonetree 0.025 0.053 0.400 0.522 Turtle Mountains 0.451 0.170 0.252 0.127 MOVEMENTS AND RESOURCE USE BY MOOSE – MASKEY AND SWEITZER ALCES VOL. 55, 2019 96 annual home ranges. Estimates of home range size (95% fixed kernel method) ranged from 59.2 to 262.6 km2 (n = 5) in the Lonetree WMA study area which were larger than in the Turtle Mountains (9.6–47.7 km2, n = 4; t5.3= 3.7, P = 0.01; mean number of loca- tions = 2709). Seasonal home range estimates were also larger in Lonetree than the Turtle Mountains (F1,25 = 13.3, P = 0.0012, mean number of locations = 807), ranging from 18.8 to 292.8 km2 and 1.0 to 44.7 km2, respec- tively. Home range size did not differ among seasons (F2, 25 = 0.1, P = 0.91). One moose was excluded from comparisons because its 30 locations occurred in a single month. None of the moose exhibited seasonal migrations. One dispersed from the Lonetree WMA in March 2004, with the 5 others remaining in the general vicinity. Radio- collars were recovered successfully from 7 animals (the 8th failed) in the Turtle Mountains; all remained in the Turtle Mountains throughout the study. Habitat use Moose strongly selected for wooded habitat in all seasons in both study areas; conversely, no selection for cropland or grassland habitats was measured in either study area. Moose in the Turtle Mountains also selected for wetland habitats during the summer (Table 2). Moose diets Moose consumed mostly woody plants in both the Lonetree (≥65%) and Turtle Mountains (≥83%) areas in all seasons of the year (Fig. 2). Consumption of woody browse was particu- larly high in the Turtle Mountains; for exam- ple, moose consumed 99% woody browse during winter, primarily aspen (36%) and wil- low (20%). Willow and aspen were also important components of the diets in the Turtle Mountains during summer (15 and 12%) and fall (19 and 23%). Bur oak stems and leaves were also common forage items in these sea- sons, representing 20% and 23% of summer and fall diets, respectively. In the Lonetree area, Russian olive was the most common woody browse consumed in all seasons and was 50% of the fall diet, followed by willow (10% in summer) and cottonwood (11% in winter). Row crops (primarily corn) were also a major component of the diets in the Lonetree area during fall (12%) and winter (29%; Fig. 2). In contrast, row crops were absent from sam- ples collected in the Turtle Mountains, although alfalfa was an important component in summer and fall diets (13%), representing 90% of forbs consumed in these seasons. Grasses (≤3%) and fruits and nuts (≤1%) were minor components of the diet in both study areas, while emergent and submergent aquatic vegetation were ≤1% of the diet during the open water seasons of summer and fall. Table 2. Mean Manly’s standardized selection ratios (SE) for four habitat types based on data from 13 GPS-collared moose in the Lonetree and Turtle Mountains study areas in North Dakota, USA. Bold numbers indicate positive selection (> 0.25) for a habitat type. Study site Season n Woodland Wetland Crop Grassland Lonetree Winter 6 0.95(0.008) 0.013(0.005) 0.013(0.003) 0.024(0.004) Summer 6 0.89(0.016) 0.048(0.018) 0.024(0.006) 0.034(0.005) Fall 5 0.84(0.033) 0.067(0.017) 0.051(0.017) 0.038(0.009) Turtle Mountains Winter 7 0.76(0.01) 0.16(0.021) 0.031(0.017) 0.048(0.012) Summer 4 0.56(0.06) 0.30(0.039) 0.015(0.007) 0.13(0.068) Fall 4 0.54(0.1) 0.21(0.052) 0.10(0.040) 0.15(0.085) ALCES VOL. 55, 2019 MOVEMENTS AND RESOURCE USE BY MOOSE – MASKEY AND SWEITZER 97 DISCUSSION Sample size and radio-collar performance The low density of moose limited the number of animals that could be studied in the Lonetree WMA. Observations from fixed-wing aircraft indicated that only 5 moose were in the vicinity in 2004, and all were successfully captured and radio- collared that year; likewise, in subsequent years we successfully radio-collared nearly every known moose in the area. Radio- collar failures limited our ability to make comparisons between moose in their tradi- tional range and the prairie habitats. Overall, 9 of 22 radio-collars failed prematurely, Fig. 2. Seasonal diet composition (%) of moose in the Lonetree (a) and Turtle Mountains (b) study areas in North Dakota, USA. MOVEMENTS AND RESOURCE USE BY MOOSE – MASKEY AND SWEITZER ALCES VOL. 55, 2019 98 with 6 of 10 in the Turtle Mountains includ- ing 4 of 5 deployed in 2005. We do not believe that these failures generated any obvious sources of bias within our data. More importantly, we radio-collared a large proportion of the moose in the study areas and our results provide novel information about habitat and forage use in the North Dakota landscape. Migration and dispersal Our results indicate that moose in North Dakota are largely non-migratory. Elsewhere, moose may migrate to avoid deep snows at high elevation or to seek conifer forests that provide thermal cover or reduced snow depth (Pierce and Peek 1984, Ballard et al. 1991, Hundertmark 1998, Thompson and Stewart 1998, Poole and Stuart-Smith 2006). Because moose range in North Dakota lacks these characteristics, elevational differences do not lead to variability in snowfall or depth, and conifer forests are absent. Additionally, habitat selection and diet com- position of moose in both study areas indi- cated that moose selected for wooded habitats and consumed primarily woody browse in all seasons. Other habitats such as croplands realized increased seasonal use, but these habitats were interspersed within the mosaic of woodland patches used year- round by moose. As a result, moose did not need to migrate to gain access to seasonally preferred forage. Home range Detailed comparisons of home range sizes across studies are difficult because of differences in the number of locations col- lected and the variety of estimators used. However, home range size is expected to be a function of the energetic requirements of an animal and the spatial distribution of nec- essary resources (McNab 1963, Elchuk and Weibe 2003, Mitchell and Powell 2004). Thus, where required resources are widely dispersed, home range size will be larger. Mean total home range size (160.5 km2, SE = 38.9) for Lonetree moose was near the upper range of averages reported for non-mi- gratory moose (174–290 km2; Grauvogel 1984, Ballard et al. 1991, Stenhouse et al. 1995). Ballard et al. (1991) attributed large home ranges to the high proportion of unavailable habitat within home ranges. Similarly, 92% of the landscape in the Lonetree WMA consisted of grassland and cropland habitats that moose mostly avoided, while the wooded habitats that moose selected for comprised only 2.5% of the landscape. In contrast, in the Turtle Mountains with a high proportion of wood- land habitat (45.1%), moose had smaller total home ranges (x̄ = 27.7 km2, SE = 10.0) similar to those (2–43 km2) for other pre- dominantly wooded areas in eastern North America (Phillips et al. 1973, Addison et al. 1980, Leptich and Gilbert 1989, Garner and Porter 1990, Dodge et al. 2004). We did not observe significant differ- ences in home range size among seasons, but seasonal home-range size varied con- siderably among moose. Although energy constraints associated with moving through deep snow or predator avoidance may result in smaller home ranges during winter (Phillips et al. 1973, Thompson and Vukelich 1981, Dussault et al. 2005), snow depths considered limiting to moose are rare (> 70 cm; Hundertmark 1998) and large predators are absent in North Dakota. Thus, seasonal home ranges were more likely determined by the distribution of seasonally-important forage resources (Doerr 1983, Lynch and Morgantini 1984, Leptich and Gilbert 1989). As such, the dif- ferences in size of seasonal home ranges among moose likely reflected the spatial pattern of available seasonal resources where moose resided. ALCES VOL. 55, 2019 MOVEMENTS AND RESOURCE USE BY MOOSE – MASKEY AND SWEITZER 99 Habitat selection and diet While the characterization of habitat types was general in order to facilitate com- parisons between the study areas, the habitat selection analyses nevertheless provided important insight into how moose utilized available habitats in North Dakota. Numerous researchers have demonstrated the impor- tance of a variety of types of woody habitats in providing forage and/or cover for moose (e.g., Peek et al. 1976, Peek 1998, Courtois et al. 2002). Therefore, it was not surprising that moose in North Dakota exhibited a strong selection for woody habitats in all seasons in both study areas. In adjacent Minnesota and many areas of Canada, moose inhabited early successional forest created by periodic fire or insect outbreaks and that are now maintained largely by forest harvesting (Peterson 1955, Phillips et al. 1973, Peek et al. 1976). The forests that cover nearly half of the Turtle Mountains study area represent this “typical” moose habitat, and the tree plantings on and around the Lonetree WMA, though more scattered across the landscape, also appear to provide important forest habi- tat for moose. Woody browse dominated the diets of moose in both study areas, similar to prior research demonstrating the importance of woodlands in providing forage for moose (Belovsky 1981, Renecker and Schwartz 1998). Diets in the Turtle Mountains con- sisted in large part of browse species such as aspen, willow, birch, juneberry, and cherry that are typically considered common forage items of moose. In addition to many of these traditional browse species, the Turtle Mountains also had an abundance of bur oak which was a major component of summer and fall diets. In the Lonetree area, the most common woody plants (except willow) are not typically found in traditional moose range, and this was reflected in the local diet. For example, Russian olive is a common shrub in tree plantings and was the most abundant browse item (25% of overall diets), and green ash and box elder, also commonly planted, were ~11% of the winter and sum- mer diets. Woodlands was the only habitat moose selected for in all seasons, but seasonal changes in diet and the use of croplands and wetlands suggest that moose also selectively used seasonally available for- age in other habitat types. Thus, selection ratios may not entirely reflect the impor- tance of habitats other than woodlands. For example, while moose avoided grassland habitats in both study areas, alfalfa was 13% of the summer and fall diets in the Turtle Mountains, indicating that this forb was an important supplemental seasonal forage. Likewise, moose exhibited nega- tive selection for croplands in all seasons, even though its use was greater in fall than in other seasons and corn was an important part of the fall and winter diets of Lonetree moose. This apparent lack of selection may reflect the different composition of crop- lands in the two study areas. Crops in the Turtle Mountains consisted almost entirely of small grains (wheat, barley) that were not expected to serve as moose forage, and in Sheridan and Wells Counties where Lonetree WMA is located, ~26% of the total land area was planted in wheat and barley in 2005 with only 2% corn and 3% sunflowers (USDA 2005). In contrast, if habitat selection analyses were confined to the boundary of the Lonetree WMA where the only croplands were corn and sun- flower food plots, then moose would show an overall positive selection for cropland habitats (Manly’s standardized selection ratio = 0.29). Moose avoided wetlands in most sea- sons, possibly due to avoidance of open areas during warm daytime temperatures (Olson et al. 2016). However, relative to winter, we MOVEMENTS AND RESOURCE USE BY MOOSE – MASKEY AND SWEITZER ALCES VOL. 55, 2019 100 observed an increase in wetland use during summer and fall which was primarily driven by Turtle Mountain moose; use of wetlands was low year-round for Lonetree moose (Table 2). The increased use of wetlands was not reflected in the diets, as aquatic vegeta- tion was a minor component (≤1%) of sum- mer and fall diets in both study areas. However, these plants likely play an import- ant role because it is believed moose con- sume aquatic plants for their critical minerals (De Vos 1958, Belovsky and Jordan 1981) and high digestibility (MacCracken et al. 1997). The latter may represent a potential limitation of this study because it may cause underrepresentation of aquatic plants in fecal samples. Alternatively, the relative increase in wetland use in summer and fall may have been independent of forage requirements and triggered by thermoregulatory behavior or to avoid insects (De Vos 1958, Belovsky and Jordan 1981). The combined results of home range, habitat use, and diet analyses provide insight into the factors influencing space use by moose in both traditional and prairie habitats in North Dakota. While wooded habitats appear to be critical for moose throughout their range in North Dakota, other seasonally available resources such as corn and alfalfa may provide supplemental food sources. Further, the strong selection for planted woodlands and use of crops as a food source in the Lonetree WMA support the hypothesis that range expansion by moose is the direct result of landscape modifications occurring since European settlement. Management implications Moose are a prized big game species in North Dakota, with >15,000 hunters apply- ing annually for a once-in-a-lifetime license (North Dakota Game and Fish Department 2019). This study provides ecological information about the state’s moose popula- tion that will help managers make informed decisions to maintain and enhance this unique wildlife resource. While moose have expanded their range to include areas of North Dakota that were historically prairie, the woodland habitats that they depend on constitute a very small proportion of the overall landscape in these areas, thereby requiring moose to have large home ranges to acquire sufficient resources. As a result, managers should recognize that prairie habi- tats are likely capable of supporting fewer moose than forested areas, and that the con- tinued persistence of prairie populations of moose will be dependent on the maintenance of forest habitat. Additionally, the planted woodlands and food plots of the Lonetree WMA may make this area particularly attractive to prairie moose. The continued management of this and other WMAs to pro- vide food and cover for wildlife should help support the state’s moose population in non-traditional range where availability of preferred habitats is limited. ACKNOWLEDGMENTS We thank North Dakota EPSCoR, the North Dakota Game and Fish Department, the U.S.D.I. Bureau of Reclamation, the U.S. Fish and Wildlife Service, the North Dakota Chapter of The Wildlife Society, the University of North Dakota Biology Department, and the Wheeler Scholarship for funding this project. We would also like to thank W. Jensen, R. Johnson, R. Kreil, and S. Peterson for permitting us to cap- ture moose and conduct work on North Dakota Game and Fish lands. Additionally, we are grateful to R. Newman, B. Rundquist, and J. Vaughan for their helpful input on this manuscript. Finally, gratitude is expressed to J. Smith, E. Pulis, T. Manuwal, J. Faught, and J. Rubbert for their assis- tance in the field. ALCES VOL. 55, 2019 MOVEMENTS AND RESOURCE USE BY MOOSE – MASKEY AND SWEITZER 101 REFERENCES Addison, R. B., J. C. WilliAmson, B. P sAundeRs, and d. FRAseR. 1980. Radio- tracking of moose in the boreal forest of northwestern Ontario. Canadian Field- Naturalist 94: 269–276. BAkke, e. l. 1980. Movements and Habitat Use of Ruffed Grouse in the Turtle Mountains, North Dakota. M. S. Thesis. University of North Dakota, Grand Forks, North Dakota, USA. BAllARd, W. B., J. s. WhitmAn, and d. J. Reed. 1991. Population dynamics of moose in south-central Alaska. Wildlife Monographs 114: 1–49. Belovsky, G. e. 1981. Food plant selection by a generalist herbivore: the moose. Ecology 62: 1020–1030. doi:10.2307/1937001 _____, and P. A. JoRdAn. 1981. Sodium dynamics and adaptations of a moose population. Journal of Mammalogy 62: 613–627. doi:10.2307/1380408 CedeRlund, G. h., and h. sand. 1994. Home-range size in relation to age and sex in moose. Journal of Mammalogy 75: 1005–1012. doi:10.2307/1376618 CouRtois, R., C. dussAult, F. Potvin, and G. dAiGle. 2002. Habitat selection by moose (Alces alces) in clear-cut land- scapes. Alces 38: 117–192. de vos, A. 1958. Summer observations on moose behavior in Ontario. Journal of Mammalogy 39: 128–139. doi:10.2307/ 1376618 dodGe, W. B., JR, s. R. WinteRstein, d. e. BeyeR, JR., and h. CAmPA iii. 2004. Survival, reproduction, and movements of moose in the western Upper Peninsula of Michigan. Alces 40: 71–86. doeRR, J. G. 1983. Home-range size, move- ments, and habitat use in two moose, Alces alces, populations in southeastern Alaska. Canadian Field-Naturalist 97: 79–88. dussAult, C., J. P. ouellet, R. CouRtois, J. huot, l. BReton, and h. JoliCieR. 2005. Linking moose habitat selection to limit- ing factors. Ecography 28: 619–628. doi:10.1111/j.2005.0906-7590.04263.x elChuk, C. l., and k. l. WieBe. 2003. Home range size of northern flickers (Colaptes auratus) in relation to habitat and paren- tal attributes. Canadian Journal of Zoology 81: 954–961. doi:10.1139/z03-077 GARneR, d. l., and W. F. PoRteR. 1990. Movements and seasonal home ranges of bull moose in a pioneering Adirondack population. Alces 26: 80–85. GARRison diveRsion PRoJeCt. 2019. History and federal legislation. (accessed September 2019). GRAuvoGel, C. A. 1984. Seward Peninsula moose population identity study. Federal Aid in Wildlife Restoration Final Report. Alaska Department of Fish and Game, Juneau, Alsaka, USA. hemson, G., P. Johnson, A. south, R. kenWARds, R. RiPley, and d. mACdonAld. 2005. Are kernels the mustard? Data from global positioning system collars suggests problems for kernel home-range analysis with least-squares cross validation. Journal of Animal Ecology 74: 455–463. doi:10.1111/j.1365-2656.2005. 00944.x hundeRtmARk, k. J. 1998. Home range, dis- persal, and migration. Pages 303–335 in A. W. Franzmann and C. C. Schwartz, editors. Ecology and Management of North American Moose. Smithsonian Institution Press, Washington, DC, USA. keARney, s. R., and F. F. GilBeRt. 1976. Habitat use by moose and white-tailed deer on sympatric range. Journal of Wildlife Management 40: 645–657. doi:10.2307/3800559 knue, J. 1991. Big game in North Dakota: a short history. North Dakota Game and Fish Department, Bismarck, North Dakota, USA. lABonte, J., J. P. Quellet, and R. CouRtois. 1998. Moose dispersal and its role in the maintenance of harvested populations. Journal of Wildlife Management 62: 225–235. doi:10.2307/3802282 lePtiCh, d. J., and J. R. GilBeRt. 1989. Summer home range and habitat use by http://www.garrisondiv.org/about/HistoryFederalLegislation/ http://www.garrisondiv.org/about/HistoryFederalLegislation/ http://www.garrisondiv.org/about/HistoryFederalLegislation/ MOVEMENTS AND RESOURCE USE BY MOOSE – MASKEY AND SWEITZER ALCES VOL. 55, 2019 102 moose in northern Maine. Journal of Wildlife Management 53: 880–885. doi:10.2307/3809581 liCht, d. s. 1997. Ecology and Economics of the Great Plains. University of Nebraska Press, Lincoln, Nebraska, USA. lynCh, G. m., and l. e. moRGAntini. 1984. Sex and age differential in seasonal home-range size of moose in northcentral Alberta. Alces 20: 61–78. 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. Journal of Wildlife Management 61: (suppl. 136) 1–52. mAnly, B. F. J., l. l. mCdonAld, d. l. thomAs, t. l. mCdonAld, and W. P. eRiCkson. 2002. Resource Selection by Animals: Statistical Design and Analysis. 2nd edition. Kluwer Academic Publishers, Dordrecht, the Netherlands. mCnAB, B. k. 1963. Bioenergetics and the determination of home range. American Naturalist 97: 130–140. doi:10.1086/ 282264 mitChell, m. s., and R. A. PoWell. 2004. A mechanistic home range model for opti- mal use of spatially distributed resources. Ecological Modelling 177: 209–232. doi:10.1016/j.ecolmodel.2004.01.015 noRth dAkotA FoRest seRviCe. 2003. Legendary forests: State forests guide. North Dakota Forest Service, Bottineau, ND, USA. noRth dAkotA GAme And Fish dePARtment. 2019. Moose hunting. (accessed September 2019). olson, B. t., s. k. Windels, R. A. moen, and n. P. mCCAnn. 2016. Moose modify bed sites in response to high tempera- tures. Alces 52: 153–160. osko, t. J., m. n. hiltz, R. J. hudson, and s. m. WAsel. 2004. Moose habitat preferences in response to changing avail- ability. Journal of Wildlife Management 68: 576–584. doi:10.2193/0022-541X (2004)068[0576:MHPIRT] 2.0.CO;2 Peek, J. m. 1998. Management of moose hab- itat. Pages 377–401 in A. W. Franzmann and C. C. Schwartz, editors. Ecology and Management of North American Moose. Smithsonian Institution Press, Washington, DC, USA. _____, d. l. uRiCh, and R. J. mACkie. 1976. Moose habitat selection and relation- ships to forest management in northeast- ern Minnesota. Wildlife Monographs 48: 1–65. PeteRson, R. l. 1955. North American Moose. University of Toronto Press, Toronto, Ontario, Canada. PhilliPs, R. l., W. e. BeRG, and d. B. siniFF. 1973. Moose movement patterns and range use in northeastern Minnesota. Journal of Wildlife Management 37: 266–278. doi:10.2307/3800117 PieRCe, J. d., and J. m. Peek. 1984. Moose habitat use and selection patterns in north-central Idaho. Journal of Wildlife Management 48: 1335–1343. doi:10.2307/ 3801794 Poole, k. G., and k. stuARt-smith. 2006. Winter habitat selection by female moose in western interior montane for- ests. Canadian Journal of Zoology 84: 1823–1832. doi:10.1139/z06-184 R CoRe develoPment teAm. 2007. R: a language and environment for statisti- cal computing. R Foundation for Statistical Computing, Vienna, Austria. (accessed September 2019). RemPel, R. s., P. C. elkie, A. R. RodGeRs, and m. J. GluCk. 1997. Timber manage- ment and natural disturbance effects on moose habitat: landscape evaluation. Journal of Wildlife Management 61: 517–524. doi:10.2307/3802610 ReneCkeR, l. A., and C. C. sChWARtz. 1998. Food habits and feeding behavior. Pages 403–439 in A. W. Franzmann and C. C. Schwartz, editors. Ecology and Management of North American Moose. Smithsonian Institution Press, Washington, DC, USA. https://gf.nd.gov/hunting/moose https://gf.nd.gov/hunting/moose http://www.R-project.org/ ALCES VOL. 55, 2019 MOVEMENTS AND RESOURCE USE BY MOOSE – MASKEY AND SWEITZER 103 RodGeRs, A. R., and A. P. CARR. 1998. Home Range Extension (HRE) for ArcViewGIS. Ontario Ministry of Natural Resources Centre for Northern Forest Ecosystem Research, Thunder Bay, Ontario, Canada. sAin, s. R., k. A. BAGGeRty, and d. W. sCott. 1994. Cross-validation of multi- variate densities. Journal of the American Statistical Association 89: 807–817. doi: 10.1080/01621459.1994.10476814 sChneideR, R. R., and s. m. WAsel. 2000. The effect of human settlement on the density of moose in northern Alberta. Journal of Wildlife Management 64: 513–520. doi:10.2307/3803249 seABloom, R. W., J. W. hoGAnson, W. F. Jensen. 2011. The mammals of North Dakota. Institute for Regional Studies, North Dakota State University, Fargo, North Dakota, USA. seAmAn, d. e., and R. A. PoWell. 1996. An evaluation of the accuracy of kernel den- sity estimators for home range analysis. Ecology 77: 2075–2085. doi:10.2307/ 2265701 _____, J. J. millsPAuGh, B. J. keRnohAn, G. C. BRundiGe, k. J. RAedeke, and R. A. Gitzen. 1999. Effects of sample size on kernel home range estimates. Journal of Wildlife Management 63: 739–747. doi:10.2307/3802664 silveRmAn, B. W. 1986. Density Estimation for Statistics and Data Analysis. Chapman and Hall, London, United Kingdom. smith, J. R., R. A. sWeitzeR, and W. J. Jensen. 2007. Diets, movements, and conse- quence of providing wildlife food plots for white-tailed deer in central North Dakota. Journal of Wildlife Management 71: 2719–2726. doi:10.2193/2006-379 stenhouse, G. B., P. B. lAtouR, l. klutny, n. mACleAn, and G. GloveR. 1995. Productivity, survival and movement of female moose in a low-density population, Northwest Territories, Canada. Arctic 48: 57–62. doi:10.14430/ arctic1224 stevens, o. A. 1966. Plants of Bottineau County, North Dakota. North Dakota School of Forestry, Bottineau, North Dakota, USA. stRonG, lAuRenCe l., h. thomAs skleBAR, and kevin e. keRmes. 2005. North Dakota GAP analysis project. Northern Prairie Wildlife Research Center, Jamestown, North Dakota. Online. (Version 12JUN2006; accessed September 2019). thomPson, i. d., and R. W. steWARt. 1998. Management of moose habitat. Pages 377–401 in A. W. Franzmann and C. C. Schwartz, editors. Ecology and Management of North American Moose. Smithsonian Institution Press, Washington, DC, USA. _____, i. d., and m. F. vukeliCh. 1981. Use of logged habitats in winter by moose calves in northeastern Ontario. Canadian Journal of Zoology 59: 2103–2114. doi:10.1139/z81-287 united stAtes dePARtment oF AGRiCultuRe. 2005. National Agricultural Statistics Service website. U. S. Department of Agriculture, National Agricultural Statistics Service, North Dakota Field Office, Fargo, ND, USA. (accessed September 2019). _____-Farm Service Agency. 2005. Statewide NAIP photography 2005. USDA-FSA Aerial Photography Field Office, Salt Lake City, Utah, USA. (accessed September 2019). united stAtes enviRonmentAl PRoteCtion AGenCy. 1996. Level III ecoregions of the continental United States. National Health and Environmental Effects Research Laboratory Map M-1, vari- ous scales. United States Environmental Protection Agency, Corvallis, Orgeon, USA. united stAtes Fish And WildliFe seRviCe. 2000. National Wetlands Inventory http://www.npwrc.usgs.gov/projects/ndgap/ http://www.npwrc.usgs.gov/projects/ndgap/ http://www.nass.usda.gov/nd/ http://www.nass.usda.gov/nd/ http://www.apfo.usda.gov MOVEMENTS AND RESOURCE USE BY MOOSE – MASKEY AND SWEITZER ALCES VOL. 55, 2019 104 website. U.S. Department of the Interior, Fish and Wildlife Service, Washington, DC, USA. (accessed September 2019). vAn vuRen, d. h. 1984. Summer diets of bison and cattle in southern Utah. Journal of Range Management 37: 260–261. doi:10.2307/3899151 Wand, m. P., and m. C. Jones. 1995. Kernel Smoothing. Chapman and Hall. London, United Kingdom. WoRton, B. J. 1995. Using Monte Carlo sim- ulation to evaluate kernel-based home range estimators. Journal of Wildlife Management 59: 794–800. doi:10.2307/ 3801959 http://www.fws.gov/nwi/