Spatial ecology of moose in central Alaska: Movement behaviour, home range dynamics, and resource selection

Introduction

Moose (Alces alces) are a keystone species and ecosystem engineer in boreal and subarctic ecosystems, including Alaska, shaping ecosystems through bottom-up and top-down processes. They are prey for large carnivores, alter vegetation structure, nutrients and nitrogen cycles, as well as holding cultural and economic value through harvestings (Molvar, Bowyerm and Ballenberghe, 1993; Edenius et al., 2002; Pastor & Danell, 2003; Jennewein et al., 2020; Bowyer et al., 2026).  Although stable worldwide (IUCN, 2015), southern populations of moose in North America are declining, mainly due to climate and landscape changes (Ditmer et al., 2018; Breithaupt et al., 2025). These changes are altering moose movement and home ranges, so understanding their space and resource use patterns per season can help inform conservation strategies.  Central Akasa is useful to study these patterns as it is relatively undisturbed by human activity and has strong seasonal differences, so offers insight into natural habitat preferences (USDA NRCS, accessed 2026). 

Moose spatial ecology, encompassing home range, movement and resource use, is shaped and varies with sexual dimorphism and seasonal demands (e.g., Dussault et al., 2005a; 2005b; Oehlers et al., 2011). Their high energetic demands mean they must maximise intake in Alaska’s short summers to build fat reserves to survive winter. Males have larger energetic needs due to their size and the rut in autumn, while females incur large winter fat losses due to lactating, while also having to considering their calves’ survival (Schwartz, 1992; Borowik et al., 2021; Thompson et al., 2024).  These differences shape sex specific patterns, driven by energetic needs and resource availability, which vary seasonally across Alaska.

In Alaska, moose occupy a range of habitats, primarily using boreal forests dominated by conifers such as spruce (Picea mariana, Picea glauca), which provide cover from weather and predators. They also use areas with deciduous species, such as birch (Betula), which surround these forests, as well as shrubs like willow (Salix), found in wetland and riparian zones, which provide forage for moose (Shipley, 2010; Alaska Department of Fish and Game, 2015; Johnson & Rea, 2024; Breithaupt et al., 2025). Moose home ranges, traditionally defined as “that area traversed by an individual in its normal activities of food gathering, mating and caring for young” (Burt, 1943), represent resource use and the movement patterns linking resources across the land (Oehlers et al., 2011; Ofstad et al., 2019). As home range reflects a balance of resource uses, effective conservation strategies must consider resource needs and movement processes. 

Plant availability varies seasonally. Summer provides abundant high quality nutritional forage, but is scarce and patchy in winter months, with willow being one of the few high quality plants available year round (Shipley, 2010). This drives high foraging rates in summer, enabling fat accumulation for winter, whilst also high rest rates in shaded areas to avoid heat stress (Jennewein et al., 2020).  In winter, home ranges often expand, as individuals must travel further, but in short bursts due to snow constraints (e.g. Oehlers et al., 2011; Ditmer et al., 2018). Females are more likely to use high quality forage patches, wetlands and canopy cover to gain fat and protect their calves from predation (Dussault et al., 2005a; 2005b; Oehlers et al., 2011; Ofstad et al., 2019). Males are more likely to use lower elevations and prioritise quantity over quality in forage to meet their high energy needs, which reflects larger home ranges, particularly in winter when resources are scarce (Oehlers et al., 2011; Wattles and Destefano, 2013; Dorowik et al., 2021). As climate change raises temperatures, risk of heat stress intensifies so reliance on shade will increase and potentially reduce consumption in the summer. Understanding how sex and season influence movement and habitat selection is therefore needed to predict responses to environmental change (Ditmer et al., 2018; Jennewein et al., 2020).

Moose spatial ecology has been widely studied, yet research primarily focuses on female moose or rarely considers home range, resource use and movement behaviour simultaneously (e.g Dussault et al., 2005a; 2005b; Oehlers, et al., 2011 Thompson et al., 2024). To address this, the present study will investigate the impact of sex and season, integrating a multi-analytical approach across home range, behavioural movement, and resource use to provide a more comprehensive assessment of space use. This study has the following research questions:

1. How do home range sizes vary between winter and summer and between sexes?

2. Which habitats do moose select and avoid over a year?

3. How do behavioural states of moose vary between winter and summer, and does sex influence state use?

Methods

Study site

The moose in this study inhabit the upper Koyukuk River drainage in north central Alaska, within the Central Brook Range, including the Gates of the Arctic National Park and Preserve and the Kanuti National Wildlife Refuge. This region is characterised by boreal forests, wetlands and riparian zones and elevation ranges from lowland valleys to mountainous alpine regions. Human population is minimal apart from the Dalton Highway. This region is highly seasonal, with temperatures ranging from  −40◦ C to 16◦ C (Joly et al., 2015A, 2015b; USDA NRCS accessed 2026).  A full description of the study area is provided in Joly et al. (2015a, 2015b).

Data source

The telemetry data used in this analysis were obtained from Movebank (“Moose in Upper Koyukuk Alaska”), with 35 adult moose (22 female, 13 males) tracked via GPS collar from 17/03/2008 to 01/04/2013, yielding 71,675 fixes. Full details on deployment and data collection can be found in Joly et al. (2015a, 2015b).

Software

All statistical analysis were performed using RStudio with R version 4.5.2 (R Core Team, 2025).

Data processing

Data was cleaned in R Studio. Timestamps were converted to POSIXct format and locations were projected to UTM Zone 5N (EPSG:32605) to ensure spatial consistency. As a result, individuals with fixes outside of this zone were excluded, those with fewer than 11 out of zone fixes were retained once those points removed, as it had minimal impact on temporal structure. Fixes were recorded at 8 hour intervals, with individuals with only one fix a day removed to maintain consistency for the Hidden Markov Model (HMM).  The package “move2” (Kranstauber, Safi, and Scharf, 2024) was used to identify outliers in time lags, speed, and distance. Most outliers were isolated ( <1% of fixes), so were retained to avoid creating long time gaps and subsequently, analytical methods capable of handling irregular data were used (Fleming et al., 2015; Silva et al., 2021). One individual with over 900 hours worth of time lags was removed as the data was not comparable.

The dataset was standardised to a year (15/04/2011 to 14/04/2012), containing 21,502 fixes from 21 individuals (Table 1).  Two seasonal datasets were also created, with “summer” spanning 01/06/2011 to 31/08/2011 and “winter” running from 01/12/2011 to 29/02/2012 (Table 2 ).

Analysis and movement metric

Home range analysis

MCP

Home ranges were estimated using 95% minimum convex polygons (MCPs) per sex and season using “adehabitatHR” (Calenge, 2006). MCPs represent the smallest convex polygon enclosing 95% of GPS locations (Mohr, 1947).  Although simple and widely used in home range research (e.g., Dussault et al., 2005a; Oehlers et al., 2011), they are sensitive to outliers and do not capture core areas. Subsequently, MCPs were used for exploratory comparison with Autocorrelated Kernel Density Estimation (AKDE) estimates.

AKDE

AKDEs 95% home ranges were estimates using  “ctmm” (Calabrese, Fleming, and Gurarie, 2016; Fleming et al., 2019). AKDEs correct for autocorrelation and irregular sampling, so are less biased compared to traditional kernel methods (Fleming et al., 2015; Noonan et al., 2018; Silva et al., 2021). 

Variograms were generated for each individual to assess range residency (Fleming et al., 2015; Silva et al., 2021), three individuals whose variograms indicated non-residency were removed from the analysis, leaving 18 animals investigated (8 female, 10 male). Seasonal AKDEs were conducted using pHREML after ctmm.guess was used to estimate starting movement parameters.  Effective sample size (DOF) was also estimated to assess the reliability of home range estimates.

Resource use

RSF

Resource selection function (RSF) was used to investigate habitat use. RSF investigates locations, so are appropriate to use with coarse 8 hour fixes, which are unsuitable for fine-scale approaches such as step selection function (Thurfjell, Ciuti, Boyce, 2014; Avgar, et al., 2016).

A 30 m land cover raster, resampled to 120m, was used (MRLC, 2011). Within each MCP, ten available points were randomly generated for each used. Covariates included land cover class, patch area, patch isolation ( “landscapmetrics”) (Hesselbarth et al., 2019; Hollister, 2025). Habitats were categorised into ecological classes and variables standardised. An RSF was then fitted as a mixed effects logistic regression using “lme4” ( Millspaugh et al., 2006; Bates et al., 2015) with individual ID included as a random effect. A full year RSF was used here rather to capture annual habitat selection and identify a baseline for which habitats moose select or avoid more strongly.

Movement patterns and behavior

HMM

A two state Hidden Markov Model (HMM) was fitted separately for summer and winter using “momentuHMM” (McClintock and  Michelot, 2018). Fixes over 8 hours capture broader behavioural states, states were classified as low and high movement states. Step length was modelled with a Gamma distribution and turning angle with a von Mises distribution. Sex and time of day were used as covariates to investigate differences in transitional probabilities. From step length and turning angles HMMs infer unobserved states and insight into movement strategies (Patterson et al., 2009; Leos-Barajas et al., 2017; McClintock and Michelot, 2018).

Results

Descriptives

A total of 21 adult moose (9 females, 12 males) were investigated following data cleaning (Table 1 and 2), of which 18 (8 females, 10 males) were investigated in the AKDE home range analysis.

Table 1. Descriptive statistics of moose tracking effort

Sex	Number of Moose	Total Days Tracked	Average Days Tracked	Min Days Tracked	Max Days Tracked	Total Locations	Avg Locations	Min Locations	Max Locations
f	9	365	363.89	362	365	9637	1071.00	1019	1091
m	12	364	361.58	356	364	11865	988.92	891	1085

Table 2. Descriptive statistics of moose tracking effort by season and sex

Season	Sex	Number of Moose	Total Days Tracked	Average Days Tracked	Min Days Tracked	Max Days Tracked	Total Locations	Avg Locations	Min Locations	Max Locations
summer	f	9	92	92.00	92	92	2,413	268.1	246	276
summer	m	12	92	91.17	88	92	2,843	237.0	188	275
winter	f	9	91	90.78	89	91	2,430	270.0	265	273
winter	m	12	91	91.00	91	91	3,075	256.3	238	273
Values represent tracking summaries by season and sex. The same 20 moose are monitored across both summer and winter

Home range analysis

MCP

Table 3: MCP 95% home range summary by sex and season

sex	n	Mean Winter (95% MCP, km²)	SD Winter (km²)	Winter Min (km²)	Winter Max (km²)	Mean Summer (95% MCP, km²)	SD Summer (km²)	Summer Min (km²)	Summer Max (km²)
f	8	59.0	50.7	4.4	163.5	88.9	54.1	20.5	192.5
m	10	114.3	88.2	44.9	325.1	100.1	67.0	25.4	251.1

A preliminary home range analysis was conducted using 95% MCPs. On average, female moose had smaller home ranges than males in both seasons.  In winter females had smaller MCP home ranged (mean= 59 km², SD = 50.7 km² ) compared to males (mean = 114.3 km², SD = 88.2 km²). In summer, home ranges were closer in size, increasing for females to 88.9 km² (SD = 54.1 km² ),  whilst males decreased to 100.1km² (± 67 km²SD).

AKDE

Table 4: 95% AKDE home range summary by sex and season

season	sex	n	Mean 95% AKDE (km²)	Median 95% AKDE (km²)	SD 95% AKDE (km²)	Min 95% AKDE (km²)	Max 95% AKDE (km²)
summer	f	8	258.3	148.9	264.4	31.9	747.0
summer	m	10	363.1	256.5	505.0	55.4	1760.9
winter	f	8	394.0	229.2	505.3	9.8	1535.6
winter	m	10	892.9	398.1	906.9	114.4	2577.6

Figure 1: Seasonal and sex differences in 95% AKDE home‑range area

Figure 2: Winter AKDE outputs

Figure 3: Summer AKDE outputs

Home range size varied between seasons and sexes (Table 4, Figure 1). Mean summer home range estimates using 95% AKDE was 363.1 km² for males, compared to females which were smaller, but less variable (mean = 258.3  km², SD = 264.4  km² ).  Winter home ranges expanded for both sexes, with a mean of 892.9 km² for males and 394 km² for females. There was substantially greater variation amongst winter home ranges, particularly in the males (SD = 906.9 km²), with one home reaching 2577.6 km².  Several winter home ranges had low effective sample sizes (DOF area = 1.2-4.9), indicating uncertainty due to temporal autocorrelation, whilst summer ranges had more constrained estimates (DOF area = 2–30).

Movement patterns and behavior

HMM

The two state HMM identified movement behaviours corresponding to low movement and high movement (Table 5).

Table 5: Seasonal comparison of HMM state dependent parameters

Parameter	Winter	Summer
Low Movement mean step (m)	141.15	1135.19
Low Movement SD (m)	106.55	1151.84
Low Movement zero-mass	0.00	0.00
High Movement mean step (m)	714.17	205.59
High Movement SD (m)	659.82	183.31
High Movement zero-mass	0.00	0.00
Angle concentration (Low Movement)	0.38	0.26
Angle concentration (High Movement)	1.39	0.00

Table 6. Proportion of time spent in each HMM state by season

Season	State	Count	Percent
Winter	1	3888	70.4
Winter	2	1632	29.6
Summer	1	2986	56.4
Summer	2	2309	43.6

Table 7. Sex-specific proportion of locations assigned to each HMM state by season

Season	sex	state	n	percent
Winter	f	1	1829	75.0
Winter	f	2	610	25.0
Winter	m	1	2059	66.8
Winter	m	2	1022	33.2
Summer	f	1	1623	66.4
Summer	f	2	822	33.6
Summer	m	1	1363	47.8
Summer	m	2	1487	52.2

Figure 4: Step length distribution in summer. High movement is displayed in red and low movement in blue.

Figure 5: Step length distribution in winter. High movement is displayed in blue and low movement in red.

The two state HMM identified movement behaviours corresponding to low movement and high movement (Table 5).

In summer, high movement was characterised by longer and more varied step length (mean = 1135.19 m,  SD =  1151.84 ) and weak directionality (κ =0.26),  whilst low movement showed shorter, consistent steps (mean = 205.59 m, SD = 183.21) and near-random turning (κ  ≈ = 0). Moose were more likely to spend time in a high movement state (56.4%), with females more likely to occupy this state (66.4%), compared to males, who spent more time in lower movement states (52.2%). Males, however, showed a higher probability of switching from low to high movement (β =-0.78). Both states were highly persistent (>90% probability of remaining in the same state). Time of day was also included as a covariate, influencing transitioning behaviour with strong diel effects (–2.15 and +0.95). Including time of day and sex as covariates improved model fit (ΔAIC = 82), indicating that sex and the diel cycle influence movement state transitions.

The winter HMM similarly identified two movement states, with high movement characterised by longer step length (714.17 m, SD = 659.82) than low movement (141.15 m, SD = 1.06.55). High movement showed strong directionality (κ = 1.39), whilst low movement (κ = 0.38) showed weak directionality. Moose spent more time in low movement states (70.4%), particularly females (75%). States were persistent (>87% probability of remaining the same). Switching probabilities showed clear diel structure, with strong hour effects (–1.36 and +0.98). Including time of day and sex as covariates improved model fit (ΔAIC = 52.2), indicating that sex and the diel cycle influence movement state transitions.

Resource use

RSF

Table 8. Resource Selection Function (RSF) model results

Predictor	Estimate	Std. Error	z value	Pr(>|z|)
Intercept (grand mean habitat)	-2.330	0.136	-17.159	0.000
Patch area (scaled)	0.203	0.014	14.404	0.000
Isolation (scaled)	0.035	0.006	5.915	0.000
barren land	0.015	0.296	0.051	0.959
deciduous_forest	0.849	0.189	4.493	0.000
developed	-1.266	0.609	-2.080	0.038
evergreen_forest	0.379	0.117	3.248	0.001
herbaceous	-0.155	0.209	-0.741	0.458
mixed_forest	0.603	0.159	3.788	0.000
shrub	-0.220	0.156	-1.417	0.157
Model fitted with random slopes.

Figure 6: Fixed effect coefficient for the mixed effect RSF

Overall, 216,170 used and 21,617 available locations were used in this RSF over 8 habitat categories. Shrub (n=156,985), evergreen (n=42,326), and deciduous forests (n= 13,053) were widely available, whereas the developed (n=215) habitat was the rarest.

A mixed effects RSF with random slopes and random intercepts was conducted; this showed singularity, and the random intercept was essentially zero, suggesting only the random slopes were meaningful. Therefore, a second mixed effects RSF was conducted with only random slopes. This RSF indicated significant differences in habitat selection (Table 8). Moose strongly selected to use deciduous forest (β = 0.849, p < 0.001), evergreen forest (β = 0.379, p = 0.001), and mixed forest habitats (β = 0.603, p < 0.001), whilst developed habitats were avoided (β = –1.266, p = 0.038), but due to limited locational numbers there was more variability here (SE = 0.609). Patch area also selected larger patches (β = 0.203, p < 0.001)  but had a weak positive effect with patch isolation (β = 0.035, p < 0.001). This model also exhibited a singular fit, and model evaluation showed the model had a weak but meaningful discrimination between used and available points (AUC = 0.611).

Discussion

Overall, this study indicated clear differences in the spatial ecology of moose between sexes and seasons.  AKDE estimates revealed sexual dimorphism in space use, as seen in Figure 2, with females consistently exhibiting smaller home ranges than males, and both sexes expanding home ranges in winter. There have been mixed findings regarding seasonal home range size in female moose, with some suggesting winter expansion (e.g., Borowik et al., 2021), whilst others suggest reduction (Oehlers et al., 2011). This is likely to reflect resource quality and distribution, as moose travel further to find desired resources when habitats are poor. Females maintain relatively stable home ranges, expanding marginally in areas with lower quality resources. They continue to prioritise high quality patches that provide a balance of forage and cover, needing to maintain energy and safety, which could account for these varying results (Ofstad et al., 2019).  Females were found to spend 75% of winter in low movement states, but the directionality seen in the HMM suggests searching behaviour, most likely for the deciduous and mixed forest identified in the RSF, which provided high-quality forage. This suggests females balance their energetic needs with high quality forage patches and safety from predation.

Males, in contrast, had larger home ranges and greater variation in winter ranges (SD = 906.9 km²). The HMM showed directionality in winter, and they were more likely to be in the high movement states than females, suggesting that they travel further to meet their high energetic demands and use a broader range of resources (Ofstad et al., 2019; Borowik et al., 2021). Despite this expanded habitat, moose were much more likely to be in low movement (70.4%) in winter, which could reflect slow paced movement to conserve energy when travelling through deep snow to get to spatially distributed forage patches (Oehlers et al., 2011; Ditmer et al., 2018). The strong selection for mixed, evergreen, and deciduous forests highlights a preferred choice of habitats sought out in winter. This pattern contrasts with that of summer, where directionality was weaker or near zero, indicating that moose did not have to travel in directional paths to find resources, but that high quality forage was more abundant. Seasonal ranges reflect movement between preferred resources. Moose spent more time in high movement in summer, particularly females, supporting the theory that females have greater trade offs than males, having to store enough fat in winter, while considering cover and predation for calves (Dussault et al., 2005a; 2005b; Oehlers et al., 2011; Ofstad et al., 2019).

In the AKDE, the degrees of freedom were low in the winter (DOF area = 1.2-4.9), indicating there was low independence of the movement data and strong autocorrelation. This increased uncertainty and reflects the large variability in the home range estimates. As the AKDE models autocorrelation, the home range estimate is still useful but should be interpreted with caution. The MCP, however, still supports these sex differences and follows a similar trend as the AKDE estimates; the 95% home ranges were considerably lower here overall, and females were found to have larger home ranges in summer than winter, unlike in the AKDE. This is expected as the MCP assumes locations are independent, whilst the AKDE understands movement is autocorrelated, so is the better fit here (Fleming et al., 2015; Silva et al., 2021).

The variations in home range sizes shown in Figures 2 and 3 reflect differences in seasonal resource requirements and movement patterns between sexes, highlighting the importance of considering habitat connectivity and heterogeneity in forest management and conservation planning. The strong selection for deciduous, mixed and evergreen forests, and avoidance of developed habitats in the RSF highlights that a mosaic of forest habitats is essential for moose to meet their needs. Males, particularly in winter, had larger home ranges, indicating a greater landscape connectivity is needed, whilst females depend on smaller areas where forage and cover are readily available. Logging practices often prioritise maintaining early successional habitats, creating forage for moose, but in turn reducing needed cover, negatively impacting females in particular who rely on this (e.g Edenius et al., 2002; Oehlers et al., 2011; Ofstad et al., 2019). Keeping strips of forests within a harvested area has been suggested to be maintained in order to keep cover and shelter and reduce fragmentation (Innes, 2010). Loggers need to aim to maintain a connected heterogeneous landscape, incorporating both cover and forage whilst accounting for sex differences in habitat requirements. This is particularly important given increased travel in winter where fragmentation may elevate energetic costs.

These management conditions will become increasingly important with climate change raising temperatures.  In warmer temperatures, moose are more likely to rely on canopy cover to rest and thermoregulate, increasing trade offs between foraging and cooling down  (Jennewein et al., 2020).  This has already been illustrated in southern regions where moose are more likely to rest in the day and to forage nocturnally (Sullender et al., 2025), resulting in trade offs between increased predation risks or limiting energy intake. Further, reliance on shade could increase exposure to biting flies, which thrive in shade and can transfer parasites which are also causing declines in moose (Benedict et al., 2024). Overall, these findings illustrate the need for maintaining connected heterogeneous forests and emphasise the need for long term management strategies to support moose populations under future change.

Despite there being a low proportion of developed habitat availability, the results suggest these areas were used less than expected and were actively avoided. This is consistent with previous research that found moose avoided human disturbances, only using developed land if food was of higher quality in comparison to surrounding areas. Although other studies found males to undertake more risky behaviour here, this development could create fragmented landscapes acting as a barrier, resulting in separate populations, genetic divergence and increasing road collisions (Eldegard, Lyngved and Hjeljord, 2012; Wilson et al., 2015). Subsequently, removing high quality plants here could discourage moose from foraging, reducing collisions, whilst the implementation of wildlife corridors could prevent this disconnect (Rea, 2003; Olsson, Widén, and Larkin, 2008).

References

Alaska Department of Fish and Game., 2015. Featured species-associated forest habitats: boreal forests and coastal temperate forests. Available at:https://www.adfg.alaska.gov/static/species/wildlife_action_plan/appendix5_forest_habitats.pdf (Accessed 29 April 2026).

Avgar, T., et al .2016. Integrated step selection analysis: bridging the gap between resource selection and animal movement. Methods in Ecology and Evolution, 7: 619-630. https://doi.org/10.1111/2041-210X.12528

Bates, D et al., 2015. Fitting Linear Mixed-Effects Models Using lme4. Journal of Statistical Software, 67(1), 1-48.  doi:10.18637/jss.v067.i01.

Benedict, B. M., et al. 2024. Trapped between food, heat, and insects: Movement of moose (Alces alces) and exposure to flies in the boreal forest of Alaska. Ecology and evolution, 14(6), e11625. https://doi.org/10.1002/ece3.11625

Benedict., B, et al., 2024. Trapped between food, heat, and insects: Movement of moose (Alces alces) and exposure to flies in the boreal forest of Alaska. Ecology Evolution. 14(6):e11625. doi: 10.1002/ece3.11625. PMID: 38911494; PMCID: PMC11192998.

Borowik, T., et al. 2021. Annual movement strategy predicts within-season space use by moose. Behavioural Ecology Sociobiology,  75, 119. https://doi.org/10.1007/s00265-021-03059-4

Borowik, T., et al. 2024. A highly variable habitat selection in moose across diel and seasonal scales. Movement Ecology,  12 (69). https://doi.org/10.1186/s40462-024-00508-3

Bowyer, R. et al. 2026. Tradeoffs Among Predator Control, Moose Harvests, and Trophy Antlers: Principles Pertinent to Managing Alaska’s Wildlife. Animals. 16. 472. 10.3390/ani16030472.

Breithaupt, K., et al., 2025.  Using winter diet composition and forage plant availability to determine browse selection and importance for moose (Alces alces) in a landscape modified by industrial forestry, Forestry: An International Journal of Forest Research, 98 (2), 167–180, https://doi.org/10.1093/forestry/cpae019

Burt, W; 1943. Territoriality and Home Range Concepts as Applied to Mammals, Journal of Mammalogy, 24(3), 346–352, https://doi.org/10.2307/1374834

Calabrese, J., Fleming, C and Gurarie. E. 2016. “Ctmm: Anrpackage for Analyzing Animal Relocation Data as a Continuous-Time Stochastic Process.” Edited by Robert Freckleton. Methods in Ecology and Evolution. 7 (9): 1124–32. https://doi.org/10.1111/2041-210x.12559.

Colinandmeg. 2022. Unsplash. Available at URL: https://unsplash.com/photos/a-moose-with-large-antlers-standing-in-a-field-HXBECPqQXaE (Accessed 08/05/2026)

Ditmer M, et al. 2018. Moose movement rates are altered by wolf presence in two ecosystems. Ecology and Evolution, 8 (17) 9017–9033. https://doi.org/10.1002/ece3.4402

Ditmer M, et al., 2018. Moose movement rates are altered by wolf presence in two ecosystems. Ecology and Evolution. 17(8), 9017-9033  https://doi.org/10.1002/ece3.4402

Ditmer, M., et al.,  2017. Moose at their bioclimatic edge alter their behavior based on weather, landscape, and predators. Current Zoology, 64 (4). 419–432. https://doi.org/10.1093/cz/zox047

Dussault, C et al. 2005a. Space use of moose in relation to food availability. Canadian Journal of Zoology. 83(11): 1431-1437. https://doi.org/10.1139/z05-140

Dussault, C et al. 2005b. Linking moose habitat selection to limiting factors. Ecography. 28. 619 - 628. 10.1111/j.2005.0906-7590.04263.x.

Edenius et al., 2002. The role of moose as a disturbance factor in managed boreal forests. Silva Fennica . 36(1). https://doi.org/10.14214/sf.550

Edenius, L., et al. 2002. The role of moose as a disturbance factor in managed boreal forests. Silva Fennica. 36. 57-76. 10.14214/sf.550.

Eldegard, K., Lyngved, J.T. and Hjeljord, O., 2012. Coping in a human-dominated landscape: trade-off between foraging and keeping away from roads by moose (Alces alces). European journal of wildlife research. 58(6), pp.969-979.

Fleming, C, et al. 2015. Rigorous home range estimation with movement data: a new autocorrelated kernel density estimator. Ecology, 96, 1182-1188. https://doi.org/10.1890/14-2010.1

Fleming, C,. et al. 2019. “Overcoming the Challenge of Small Effective Sample Sizes in Home-Range Estimation.” Edited by Jason Matthiopoulos. Methods in Ecology and Evolution. 10 (10): 1679–89. https://doi.org/10.1111/2041-210x.13270.

Hesselbarth, M., et al. 2019. “Landscapemetrics: An Open-Source R Tool to Calculate Landscape Metrics.” Ecography 42 (10): 1648–57. https://doi.org/10.1111/ecog.04617.

Hollister, J.W. (2025). elevatr: Access Elevation Data from Various APIs. R package version 0.99.1. https://CRAN.R-project.org/package=elevatr/

Innes, R. 2010. Alces americanus, moose. In: Fire Effects Information System, [Online]. U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station, Fire Sciences Laboratory (Producer). Available: https://research.fs.usda.gov/feis/species-reviews/alam (Accessed: 07 May 2026)

IUCN (2015) Moose. The IUCN Red List of Threatened Species 2015: ISSN 2307-8235 (online) IUCN 2008: T56003281A22157381. Available at: https://www.iucnredlist.org/species/56003281/22157381 (Accessed: 1 April 2026).

Jennewein, J et al., 2021. Estimating integrated measures of forage quality for herbivores by fusing optical and structural remote sensing data. Environmental Research Letters. 16. 10.1088/1748-9326/ac09af.

Jennewein, J. et al. 2020. Behavioral modifications by a large-northern herbivore to mitigate warming conditions. Movement Ecology, 8(39). https://doi.org/10.1186/s40462-020-00223-9

Johnson, C and Rea, R. 2024. Response of moose to forest harvest and management: a literature review. Canadian Journal of Forest Research. 54(4): 366-388. https://doi.org/10.1139/cjfr-2023-0158

Joly K, et al. 2015. Moose in the Upper Koyukuk Alaska. Movebank Data Repository. Available at: www.movebank.org. Accessed: 01/04/2026.

Joly K., et al., 2015a. Variation in fine-scale movements of moose in the Upper Koyukuk River Drainage, northcentral Alaska. Alces 51: 97-105.

Joly, K., et al.,  2015b. Moose movement patterns in the Upper Koyukuk River Drainage, northcentral Alaska. Alces 51: 87-96.

Kranstauber, B., Safi, K., and Scharf, A. K. 2024. move2: R package for processing movement data. Methods in Ecology and Evolution, 15, 1561–1567. https://doi.org/10.1111/2041-210X.14383

Kumar , K. 2019. Unsplash. Available at URL: https://unsplash.com/photos/black-moose-lying-on-field-during-daytime-MVIqwQvkwG4  (Accessed 08/05/2026)

Leos-Barajas, V., et al. 2016. Analysis of animal accelerometer data using hidden Markov models. Methods in Ecology and Evolution. 8. 10.1111/2041-210X.12657.

McClintock, B., and  Michelot, T. 2018. “momentuHMM: R Package for Generalized Hidden Markov Models of Animal Movement.” Edited by Sarah Goslee. Methods in Ecology and Evolution 9 (6): 1518–30. https://doi.org/10.1111/2041-210x.12995

Millspaugh, J., et al. 2006. Analysis of Resource Selection Using Utilization Distributions. Journal of Wildlife Management. 70. 384-395. 10.2193/0022-541X(2006)70[384:AORSUU]2.0.CO;2.

Mohr, C. 1947. Table of equivalent populations of North American small mammals. The American Midland Naturalist. 37, 223– 249. https://doi.org/10.2307/2421652

Molvar, E, Bowyerm T and Ballenberghe, V. 1993. . Moose herbivory, browse quality, and nutrient cycling in an Alaskan treeline community. Oecologia, 94, 472–479. . https://doi.org/10.1007/BF00566961

Multi-Resolution Land Characteristics (MRLC) Consortium. 2011. MRLC NLCD Viewer. Available at: https://www.mrlc.gov/viewer/.  (Accessed: 1 April 2026).

Nandakumar, H. 2020. Unsplash. Available at URL: https://unsplash.com/photos/brown-moose-on-green-grass-during-daytime-hPQtvIFr7d0 (Accessed 08/05/2026)

Noonan, M., et al., 2018. A comprehensive analysis of autocorrelation and bias in home range estimation. Ecological Monographs 89: 1–21https://doi.org/10.1002/ecm.1344

Oehlers, S et al. 2011. Sex and scale: Implications for habitat selection by Alaskan moose Alces alces gigas. Wildlife Biology. 17. 67-84. 10.2981/10-039.

Ofstad, E et al. 2019. Use, selection, and home range properties: complex patterns of individual habitat utilization. Ecosphere. 10. e02695. 10.1002/ecs2.2695.

Olsson, M, Widén, P and Larkin, J. 2008. Effectiveness of a highway overpass to promote landscape connectivity and movement of moose and roe deer in Sweden. Landscape and Urban Planning. 85(2). 133-139. ISSN 0169-2046. https://doi.org/10.1016/j.landurbplan.2007.10.006.

Pastor, J. and Danell, K. 2003. Moose-vegetation-soil interactions : A dynamic system. Alces. 39.  177–192. https://alcesjournal.org/index.php/alces/article/view/477

Patterson et al., 2009. Classifying movement behaviour in relation to environmental conditions using hidden Markov models. Journal of Aniaml Ecology. 78(6). 1113-1123.   https://doi.org/10.1111/j.1365-2656.2009.01583.x

 R Core Team (2025). _R: A Language and Environment for Statistical Computing_. R Foundation for Statistical Computing, Vienna, Austria. <https://www.R-project.org/>.

Rea, R.V. 2003: Modifying roadside vegetation management practices to reduce vehicular collisions with moose Alces alces. Wildlife Biology. 9: 81-91.

Schwartz, C.C., 1992. Physiological and nutritional adaptations of moose to northern environments. Alces. 139-155. https://alcesjournal.org/index.php/alces/article/view/1801

Shipley, L. 2010. Fifty years of food and foraging in moose: lessons in ecology from a model herbivore. Alces, 46, 1-13. https://alcesjournal.org/index.php/alces/article/view/56

Silva et al., 2021. Autocorrelation-informed home range estimation: A review and practical guide. Methods in Ecology and Evolution. 13:534–544. DOI: 10.1111/2041-210X.13786

Sullender, B., et al. 2025. Warm winters, hot moose: temperature drives activity and habitat trade-offs across a cold-adapted species’ range. Environmental Research Letters. 20. DOI 10.1088/1748-9326/ae263a

Tape K., et al. 2016. Correction: Range Expansion of Moose in Arctic Alaska Linked to Warming and Increased Shrub Habitat. PLOS ONE. 11(7): e0160049. https://doi.org/10.1371/journal.pone.0160049 

Thompson, D, et al., 2024. Seasonal somatic reserves of a northern ungulate influenced by reproduction and a fire-mediated landscape. Frontiers in Ecology and Evolution. 12:1433485. doi: 10.3389/fevo.2024.1433485

Thurfjell, H., Ciuti, S. and Boyce, M.2014. Applications of step-selection functions in ecology and conservation. Movement Ecology. 2, 4 https://doi.org/10.1186/2051-3933-2-4

USDA Natural Resources Conservation Service (NRCS).  Ecological Site Description: MLRA 233X. Available at: https://edit.sc.egov.usda.gov/catalogs/esd/233X (Accessed: 1 May 2026).

Wattles, D and DeStefano, S. 2013. Moose habitat in Massachusetts: assessing use of the southern edge of the range. Alces. 49. 133-147.

Wattles, D.,  and Destefano, S. 2013. Moose habitat in Massachusetts: Assessing use at the southern edge of the range. Alces. 49. 133-147

Wilson, R., et al. 2015. A genetic discontinuity in moose (Alces alces) in Alaska corresponds with fenced transportation infrastructure. Conservation Genetics. 16, 791–800.  https://doi.org/10.1007/s10592-015-0700-x