Seasonal movement and activity of Akia-Maniitsoq caribou cows in West Greenland as determined by satellite. Technical Report No. 99

1 Seasonal movement & activity of Akia-Maniitsoq caribou cows in West Greenland as determined by satellite Technical Report No. 99, 2017 Greenland Institute of Natural Resources 2 Title: Seasonal movement and activity of Akia-Maniitsoq caribou cows in West Greenland as determined by satellite Authors: Christine Cuyler1, John Nagy2 and Karl Zinglersen1 1Greenland Institute of Natural Resources, P.O. Box 570, 3900–Nuuk, Greenland 2CW405 Biological Science Center, University of Alberta, Alberta, Canada Series: Technical Report No. 99, 2017 Date of publication: 19 July, 2017 Publisher/Funding: Pinngortitaleriffik, Greenland Institute of Natural Resources Cover photo: Polled Akia-Maniitsoq caribou cow, August: photo C. Cuyler ISBN: 87-91214-76-9 ISSN: 1397-3657 EAN: 9788791214769 Cited as: Cuyler, C., Nagy, J. & Zinglersen, K. 2017. Seasonal movement and activity of Akia-Maniitsoq caribou cows in West Greenland as determined by satellite. Pinngortitaleriffik – Greenland Institute of Natural Resources. Technical Report No. 99. 94 pp. Contact address: The report is only available in electronic format. You can download a PDF-file of the report at this homepage http://www.natur.gl/publikationer/tekniske rapporter It is possible to obtain a print of the report here: Greenland Institute of Natural Resources P.O. Box 570 DK-3900 Nuuk Greenland Phone: +299 36 12 00 Fax: +299 36 12 E-mail: [email protected] www.natur.gl 3 Seasonal movements & activity of Akia-Maniitsoq caribou cows in West Greenland as determined by satellite By Christine Cuyler1, John Nagy2 & Karl Zinglersen1 1 Greenland Institute of Natural Resources, P.O. Box 570, 3900–Nuuk, Greenland 2 CW405 Biological Science Center, University of Alberta, Alberta, Canada Technical Report No. 99, 2017 Greenland Institute of Natural Resources 4 [Empty page] 5 Table of Contents Summary (English) ................................................................... 8 Eqikkaaneq (Greenlandic) ..................................................... 10 Resume (Danish) ..................................................................... 12 Introduction ............................................................................. 15 Methods .................................................................................... 19 Results ...................................................................................... 27 Discussion ................................................................................ 49 Acknowledgements ................................................................. 64 Literature cited ........................................................................ 65 List of Figures 1. Map of entire Akia-Maniitsoq region. Elevations above 200 meters are light yellow while those below are green. Page 16 2. Akia-Maniitsoq study area, 11,969 km2, previously known calving areas Page 17 3. The 26 mortalities among satellite-collared caribou cows, etc. Page 28 4. Overview of movement paths for satellite-collared Akia-Maniitsoq caribou cows 2008-2010. Page 30 5. Mean daily movement by Julian day for Akia-Maniitsoq caribou cows Page 31 6. Distribution of habitat types present within 150x150 m moving window around each focal 30 m grid cell in the study area, etc. Page 32 7. Categorized probability of occurrence of Akia-Maniitsoq caribou cows in West Greenland during 10 activity periods, etc. Page 41 8. Elevation of sites used during seasonal activity periods by Akia-Maniitsoq caribou cows in West Greenland, 2008-2010. Page 42 9. West Greenland Akia-Maniitsoq satellite-collared caribou, probability of cow occurrence during peak calving, 21 May to 8 June: etc. Page 43 10. Distribution of the 52 calving sites 2008-2010 in relation to substrate, for the satellite-collared Akia-Maniitsoq caribou cows West Greenland, etc. Page 44 11. A: Daily movement pattern of parturient cow: etc. B: Daily movement pattern of non-parturient cow: etc. Page 46 12. Annual variation in the number of caribou cows in various reproductive states, 2008-2010. Page 48 13. Temporal distribution of 52 parturition dates 2008-2010 for satellitecollared Akia-Maniitsoq caribou cows, West Greenland, etc. Page 49 14. Satellite collar deployment locations, 40 caribou cows, 1-7 May 2008, etc. Page 75 15. Caribou movement routes observed, 1-7 May 2008, etc. Page 77 6 List of Tables 1. Labels & definitions for resource selection (RS) habitat and elevation covariates considered in RS models, etc. Page 23 2. Mortality among satellite-collared Akia-Maniitsoq caribou cows, etc. Page 27 3. Mean and total number of locations obtained for tracked females and activity periods, respectively, 2008-2010. Page 27 4. Seasonal activity periods as calculated from daily travel rates, etc. Page 30 5. Availability of habitat types within 0.02 km2 area around focal grid cells Page 31 6. Best-fit activity period resource selection function models, etc. Page 33 7. Parameter estimates & standard errors in best-fit calving and peak of calving (mean calving date ±1 SD) activity period RSF models, etc. Page 34 8. Parameter estimates & standard errors in best-fit post-calving, early/midsummer, and mid/late summer activity period RSF models, etc. Page 35 9. Parameter estimates & standard errors in best-fit fall/pre-breeding, breeding, and post-breeding/late fall activity period RSF models, etc. Page 36 10. Parameter estimates & standard errors in best-fit early/mid-winter, late winter, and pre-calving activity period RSF models, etc. Page 37 11. Habitat selection by seasonal activity period, etc. Page 38 12. Habitat selection during the calving period relative to those selected during all other activity periods, etc. Page 39 13. Percent availability, percent use, & ‘ratio of % use:% availability’ of areas by probability of occurrence category and activity period, etc. Page 40 14. Comparison of calving site conditions chosen by adult cows (n=30) relative to presence or absence of a yearling-at-heel, etc. Page 45 15. Comparison of calving site conditions chosen by adult cows (n =51; sub adult removed) 2008-2010, Akia-Maniitsoq West Greenland Page 45 16. Satellite collar deployment positions on 40 caribou cows, etc. Page 76 17. Calving caribou resource selection models Page 78 18. Peak-calving caribou resource selection models Page 79 19. Post-calving caribou resource selection models Page 80 20. Early summer caribou resource selection models Page 81 21. Mid /late summer caribou resource selection models Page 82 22. Fall /pre-breeding caribou resource selection models Page 83 23. Breeding (rut) caribou resource selection models Page 84 24. Post-breeding /late fall caribou resource selection models Page 85 25. Early/mid-winter activity period caribou resource selection models Page 86 26. Late winter activity period caribou resource selection models Page 87 27. Pre-calving activity period caribou resource selection models Page 88 28. Calving versus post-calving, early summer & mid /late summer, etc Page 89 29. Calving versus fall/pre-breeding, breeding & post-breeding/late fall, etc Page 90 30. Calving versus early/mid-winter, late-winter, and pre-calving, etc. Page 91 31. Reproductive status of collared cows, etc. Page 92 7 List of Appendices 1. Capture of Akia-Maniitsoq caribou cows, 1-7 May 2008. Page 75 2. SBIC (Schwarz-Bayesian information criterion) differences for 10 best-fit resource selection models for each activity period. Page 78 3. Habitat selection by Greenland caribou (Rangifer tarandus groenlandicus) during the CALVING period relative to other activity periods, etc Page 89 4. Reproductive status of collared cows alive at calving Page 92 5. Equipment recommendations for satellite collaring Page 93 *Raw data available at Greenland Institute of Natural Resources: F:/40-59/PaFu/42/Landpattedyr/ 35 ALCOA Rensdyrsporing 8 Summary (English) In May 2008 GPS satellite collars were deployed on 40 cows from the AkiaManiitsoq (AM) caribou (Rangifer tarandus groenlandicus) population. In the period 2008-2010, the satellite-collared cows provided locations, which we analysed for movements, spatial distribution, calving sites, location attributes, seasonal activity periods, and habitat resource selection. The AM caribou are substantially less documented than herds in North America. The results in this report are particularly valuable for management in Greenland because they establish a baseline for AM habitat use in the current absence of significant development and infrastructure in the Central region. However, the short study period and small sample size, exacerbated by high mortality, weaken results. Although obtaining collar data in Greenland is difficult and costly, a longer time series with a stable large number of collared caribou is necessary before sweeping conclusions can be supported. AM cows behave similarly to the mountain ecotype of caribou rather than the barren-ground. Calving was not confined to specific exclusive calving grounds close to the Ice Cap. Birthing cows spaced-away in a continuum across the entire Central region, from seacoast to Ice Cap. There is little relevance to protecting specific exclusive calving grounds in the AM region. Instead, conservation measures would profit from applying a broad-scale habitat management approach to the widespread AM calving habitat. Elevation was a good predictor for probability of cow occurrence at calving. Whether in a xeric zone or not, south facing slopes around 600 m elevation with snow were favoured. Although timing and width of the emergent vegetation period and possibly rain avoidance may be the driving factors behind choice of high elevation for calving by parturient AM cows, these remain to be investigated. AM calving appears to begin earlier than previously assumed, and suggests that the period for protection measures might be shifted forward. Habitat possessing the preferred attributes for birthing comprises 42-45% of the Central region. Thus although calving range is essential for caribou production, at present it is not likely limiting the AM population. However, given the high fidelity by AM cows to previous birthing sites, a warming Arctic or anthropogenic influences could have negative impacts if cows are displaced to habitats less favourable for calf survival. Mortality among satellite-collared cows was high and much appeared due to harvest. If common to the entire population, this would have played a major role in the decline of AM abundance from 2001 to 2010. Patterns of annual movement confirmed a southwest-northeast axis. Distance moved could be short, with individuals typically at the western end of their axis in winter and eastern end in summer. Each cow utilized just a fraction of the available area. This strongly indicates population sub-division within the Central region and suggests that the entire AM population will not be influenced when either stochastic weather events or management actions affect only a portion of the region. Patterns of seasonal movement revealed 10 seasonal activity periods, which were associated with specific habitat attributes. Elevation was the primary habitat 9 attribute that varied significantly across the seasonal activity periods. Breeding occurred at low elevations. If hunting seasons coincide with the rut, then human disturbance may negatively influence breeding and subsequently calf production. Caribou vulnerability to harvest would also likely increase because of their relative accessibility to hunters. Daily movement patterns varied throughout the year. Cows moved least in early March making it optimal for aerial survey. July had maximum movement, suggesting insect harassment as the cause and supporting the necessity of insect relief habitat. During the calving period, a characteristic daily movement pattern was assumed to indicate a birthing event, i.e., rise from normal immediately preceding a sharp drop to near zero and thereafter a gradual rise. In future, birthing could be validated by visually locating satellite-collared cows by airplane or helicopter, or equipping the satellite-collars with video capability. Habitat attributes are not evenly distributed. Relative to the entire Central region, and in contrast to calving habitat, the available area is small for late summer, fall and winter habitats. These could be limiting for the AM population. Specifically, the largest tract of winter habitat, Akia, albeit still small in size, is vulnerable to south westerly storm systems that can render winter forage unavailable or energetically costly to access. Thus, special attention and protection through fine-scale habitat management may be appropriate. AM caribou abundance would likely benefit if for those habitats that are scarce, a) caribou access was preserved, b) anthropogenic disturbance was mitigated, and c) that densities of AM caribou were kept below carrying capacity of these limited ranges. Management is best tailor-made to the population, the seasonal activity under consideration and the amount of habitat available for that activity. Conservation efforts should address not just one several seasonal ranges vital to reproduction, insect relief and survival. Protecting parturient cows and their birthing habitat is not a one-shot cure for ensuring recovery or sustainability of caribou populations. Caribou roam. Globally, caribou range shifts are common, even for ‘sacrosanct’ calving areas. Thus, management must consider conserving currently unused areas for potential future use by caribou. Meanwhile, human influences on the landscape are recognized factors that can exacerbate caribou declines. In North America, the current threshold proposed for preventing caribou decline, is that 65% of the total range remains unexposed to human disturbance. Proactive management and conservation directed towards preserving large undisturbed intact landscapes, relevant for several seasonal activities and their movement corridors, would foster caribou conservation now and for future generations. 16 favorable conditions for calf survival (White 1983, Fancy & Whitten 1991). Calving strategies aim at maximizing survival of new offspring, and may include shifts in latitude or elevation to reduce predation risk, reproductive synchrony to achieve predator satiation, or prioritize access to emergent food resources and green-up (Bergerud & Page 1987, Fryxell et al. 1988, Klein 1990, Albon & Langvatn 1992, Post et al. 2003, Loe et al. 2005). Calving phenology is generally synchronous within populations and closely related to photoperiod as well as the onset and duration of plant growth (Post 2003). It is also affected by length and harshness of the previous winter (Tveraa et al. 2013). Since caribou are ‘capital’ breeders, i.e., parturient cows depend on body reserves rather than forage, peak parturition can precede green-up by several weeks (Durant et al. 2005, Moen et al. 2006, Barboza & Parker 2008, Taillon et al. 2013). Because strategies vary somewhat even among Rangifer, understanding the specific criteria that characterize habitat vital for calving in West Greenland is important for conservation measures to have an influence on abundance. Figure 1. Map of entire Akia-Maniitsoq region, an area of ca. 15,400 km2. Elevations above 200 meters are light yellow while those below are green. 50°W Maniitsoq 66°N 52°W 54°W 64°N Atammik Napasoq NUUK Kapisillit Greenland Ice Cap Akia-Maniitsoq Central region hunting area 3 50 100 kilometres 0 Greenland 17 Although central Alaska and Yukon caribou typically exhibit synchronous and long 300-500 km north-south migrations, with annual distances travelled ranging to 5000 km (Craighead & Craighead 1987, Fancy et al. 1989), West Greenland caribou resemble the mountain/forest dwelling sedentary caribou ecotype, because aggregations are absent and movements short (50-60 km) and non-synchronous (Cuyler & Linnell 2004). Previously, they were assumed to have west-east seasonal migrations between the seacoast (winter) and Greenland Ice Cap (summer) when abundance was high, and shortened migrations or none at all if low (Vibe 1967, Grønnow et al. 1983, Thing 1984). Certainly, Greenland caribou summer range can be inland near the Ice Cap; however, winter range can be intermediate between Ice Cap and seacoast (Cuyler & Linnell 2004). Meanwhile, regionally relevant studies are lacking. Figure. 2. Akia-Maniitsoq study area, 11,969 km2, previously known calving areas, ● Cuyler and Linnell (2004), ● Cuyler unpublished, 2004 brown lines (Aastrup and Nymand 2004), 2015 stippled brown (NunaGis 2015). Elevations below 200 m are green, above 200 m yellow. Although AM caribou are present throughout the Central region, calving locations are not adequately known. Throughout North America, barrenground caribou populations usually have one well identified traditional calving ground, a sub-area of the total range for a specific population (Russell et al. 2002). A calving ground is a specific exclusive area where 80-90% of parturient cows return annually to calve (Gunn & Miller 1986), but may 18 change in spatial location over time (Gunn et al. 2008, Taillon et al. 2012). Identifying calving grounds for AM caribou has been elusive; although Strandgaard et al. (1983) suggested AM calving occurred in the northern inland portion of the range. In the Kangerlussuaq-Sisimiut caribou population to the north, large calving cow aggregations in close proximity to the Greenland Ice Cap were documented by Thing (1984). Subsequent studies for Greenland calving grounds focused on inland areas and observed cows calving (Aastrup 1986; Aastrup & Nymand 2004; J. Nymand, Greenland Institute of Natural Resources, unpublished data). Meanwhile, hunter knowledge for AM indicated a small rugged highland area near the AM seacoast (200-500 m elevation; 65° 23’ N; 51° 40’ W). Further, eight satellitecollared AM cows, 1997-1999, were highly dispersed in the calving season (Cuyler & Linnell 2004), not restricted to areas close to the Ice Cap, were typically at elevations above 300 m (Fig. 2) and exhibited 88% fidelity among sequential birthing sites (Cuyler & Linnell 2004). Site fidelity (philopatry) is the tendency of an animal to stay in, or habitually return to, a specific site or area. These attributes are similar to the mountain caribou ecotype (Bergerud et al. 1984, Skogland 1989). Nevertheless, there was just one small inland calving ground officially recognized for protection (Aastrup & Nymand 2004). Although this calving ground was expanded by 2015 (http://www. nunagis.gl), it remained a specific area associated with the period 20 May to 20 June (Government of Greenland Bureau of Minerals and Petroleum 2000) and human activities have been regulated accordingly. Seasonal habitats have been hitherto poorly understood and knowledge insufficient to provide a firm basis for management recommendations that can meet future challenges e.g., stochastic weather events, infrastructure developments, pathogens, or any combination of these and other factors. Present Study In early May 2008, we deployed GPS satellite collars on 40 caribou cows of the AM population and followed these until July 2010. We delineated mean daily movement over a 1-year cycle and extent of migratory path. We examined the seasonal activity periods of Akia-Maniitsoq caribou cows and describe the habitat attributes associated with those activity periods. We investigated and mapped the location and availability of the habitat attributes and probability of caribou cow occurrence for a given activity period. We delineate movement patterns associated with calving sites, timing of calving period, calving locations, and cow fidelity to calving sites. We determined whether areas 19 used by caribou during calving have characteristics that are uniquely different from those used during rest of the year. Environmental conditions were considered. Since adult female survival is a key determinant of ungulate population dynamics, mortality among the satellite-collared cows is addressed. Since securing sufficient habitat could facilitate conservation of AM caribou, we delineate which seasonal ranges are both essential and limited in availability. Methods Study area The Central region (hunting area 3) of West Greenland is centred at 65° 10’ N; 51° 25’ W north of Nuuk, and is ca. 15,400 km2. Our study area, a sub-set of the total region (Figs. 1, 2), was 11,969 km2 and excluded the heavily glaciated northwest, as well as lakes, glaciers and sand. The surrounding ocean, fjords and glaciers restrict dispersal of caribou from the region, creating a largely "closed" population. The region is primarily rugged uplands and mountains, elevations ≥ 300 m cover 60%, 200-300 m cover 8% and under 200 m 32%. The region is undeveloped and lacks infrastructure. During the study period, 2008-2010, the backcountry was inaccessible to the majority of people. Most were limited to boating along the coast and fjord shorelines, and hiking into the terrain seldom was beyond about 6 km. Other than caribou, mammals present are few and include only the arctic hare (Lepus arcticus) and arctic fox (Alopex lagopus) as well as introduced feral reindeer (1952) (R. t. tarandus) and the recently arrived muskoxen (Ovibos moschatus; natural immigration since ca. 1998 from the Kangerlussuaq population north of Sukkertoppen Ice Cap). Large predators are absent; however, there is an annual caribou harvest. The region is sandwiched between a dominating high pressure over the Greenland Ice Cap to the east and frequent low-pressure oceanic storm systems (wind speeds 22–46 m/sec) from the southwest (Tamstorf 2004, DMI 2014, Gamberg et al. 2016). Precipitation decreases strongly with distance from the seacoast, resulting in a xeric continental climate near the Ice Cap (Tamstorf 2004). The nearest meteorological station is immediately south in the seacoast city of Nuuk, which has an annual mean temperature of -1.4°C, mean July 6.5°C and receives an annual precipitation of 752 mm (Tamstorf et al. 2005, DMI 2014). Vegetation is chiefly low-arctic species. Plant communities vary with elevation, aspect, and proximity to influences of Greenland's wet coastal maritime and dry continental ice cap climates (Lund 20 et al. 2004, Simonsen 2011). The Akia coastal lowlands consist primarily of lichen-rich dwarf shrub heath, which moving inland changes to dwarf shrub heath with increasing grassland. In higher elevations windswept ridges, abrasion plateaus and bare ground dominate (Tamstoft 2004). The dominant plant species in our vegetation classes follows Bay & Simonsen (2009). Detailed descriptions of vegetation are in Lund et al. (2004) and Simonsen (2011). Vegetation and snow maps and digital terrain models are in Tøttrup (2009). Caribou capture & telemetry Cow caribou were captured 1-7 May 2008 with a net-gun fired from a helicopter. Satellite collars were deployed on 40 cows. There were 20 Telonics (Mesa, Arizona, USA and Service Argos, Landover, Maryland, USA) and 20 Iridium System Network (Vectronic Aerospace GmbH, Berlin Germany) global positioning system (GPS) satellite collars. Inter-location intervals for Telonics and Iridium collars were 1 hour (24 per day) and 2 hours (12 per day), respectively. All collars were programmed to release automatically after 108 -118 weeks. Locations from May 2008 to July 2010 were recorded as longitude and latitude coordinates and projected to the UTM Zone 22N Projected Coordinate System, WGS 1984 Geographic Coordinate System, Northern Hemisphere. For GIS analyses, we used ArcMap 10.1 (Environmental Systems Research Institute, Inc., Redlands, California, USA). Caribou capture and collaring locations were spread throughout the Central region south of Sukkertoppen Ice Cap (Appendix 1). We identified cows by presence of vulva patch. Capture was close to calving and in AM cows abdominal swelling from winter rumen digesta/contents is insignificant and late winter fat reserves among pregnant cows are generally zero (Cuyler unpublished data from 1996, 1997, 2008). Viewed from above, cows were assumed pregnant if the sides of the posterior half of their body protruded quite markedly relative to other cows. At capture, 38 of the 40 cows were ascertained pregnant (included a sub adult). Tooth eruption and wear determined age class (sub adult, adult). Cows collared in May 2008 included, 39 adults (age ≥ 36 months) and one sub adult (35 months). The total number of GPS locations obtained in the 2008-2010 study period was 262,137. Daily movements and migratory path length We calculated the direct line distance (km) between all sequential locations per day for each caribou cow. Daily travel rates were standardized to the number of possible locations per day [sum of distances between locations 21 obtained/(no. locations obtained/no. locations possible per day)] and normalized using a log10 transformation. Calculated distances provide a minimum estimate of total individual movement due to 1-2 hour gaps in telemetry. We mapped paths to illustrate movement variation. All paths were created using Hawth’s tools (Beyer 2007). We mapped the annual paths taken by each caribou each year. We summed daily path distances (km) to obtain annual mean total path distance, maximum and minimum. Longitudes / latitudes for the latter two provided the length of a cow’s home range. We generated annual (calculated from the date of capture) minimum convex polygons (MCPs) and paths (straight-line distances between sequential locations) for each caribou using Hawth’s Tools (Beyer 2007). We measured MCP areas and path lengths, standardized these to 365 days (areas or length divided by number of days tracked x 365), and normalized them using a log10 transformation. MCP areas and path lengths are influenced by sample size (Borger et al. 2006). Seasonal activity periods We subdivided the calculated daily travel rate data into 73 5-day periods and used analysis of variance (ANOVA) and Tukey’s honestly significant difference (HSD) pair-wise comparisons (SPSS 11.5, Chicago, Illinois, USA) to identify all sequential 5-day periods when movement rates were not significantly different. These gave the start and end dates for each seasonal activity period. We used the first and last known and estimated parturition and conception dates to define the calving and breeding periods respectively. Habitat selection modelling Herbivore movement rates (Nagy 2011) and habitat requirements for maintenance, growth, reproduction (Ferguson & Elkie, 2004; Gustine et al., 2006; Horn & Rubenstein, 1984) vary seasonally (Maier & White, 1998). Animal habitat affinities are commonly quantified at landscape scales (Johnson et al. 2004) using data for a dependent variable: telemetry animal locations (use) and geographic information system (GIS) generated random locations (available), i.e. 10 random locations per animal location, with each random location sampled from within a circle that was centered on the preceding telemetry location, and having a radius equal to the distance between the preceding and next successive telemetry location. The independent predictor variables are habitat and topographic features thought 22 to influence habitat selected by the caribou (Table 1). The data for dependent and independent variables are used to construct resource selection functions (RSF) of the following form: w(x) = exp (β1x1 + β2x2 + ... + βkxk) Equation 1 (Johnson et al. 2004). The relative probabilities of occurrence of animals on landscapes are commonly calculated from RSFs and mapped using linear stretch transformations of the form: ŵ = [(w(x) - wmin)/( wmax - wmin)] Equation 2 where w(x) is the product of equation 1 and wmin and wmax represent the smallest and largest RSF values, respectively (Johnson et al. 2004). However, relative probabilities of occurrence assessments using this method are sensitive to the value of wmax. Rare large values of wmax may result in underestimation of the relative probabilities of occurrence and affect the reliability of k-fold cross-validation techniques typically used to assess the predictive performance of RSF models (Boyce et al. 2002). The consequences of underestimating habitat values in an area may be significant for conservation and management. We assumed that the attributes (e.g., elevation, vegetation cover types, windexposure) of sites selected by Greenland caribou varied seasonally, to address their changing physiological requirements and behaviors. Therefore, we first identified their activity periods based on significant changes in daily movement rates (Nagy 2011). We used a moving-window GIS approach to assess the presence/absence of 11 vegetation cover types and average elevation within a 0.02 km2 area around each caribou (use) and random (available) location. We used logistic regressions (equation 1) to fit RSF models for each activity period (Johnson et al. 2004, Latham et al. 2011) and selected the best-fitting models using Schwarz information criteria (SBIC) (Cook 2007, Hardin & Hilbe 2012, Schwarz 1978, Strong et al. 1999, Wasser et al. 2011). To calculate relative probabilities of caribou occurrence (w) (equation 3) we modified equation 2. Rather than wmax we used the wmedian RSF values for caribou use locations within the study area for each activity period using the following modified linear stretch transformation: ŵ = [(w(x) - wmin)/(wmedian RSF caribou locations - wmin)]. Equation 3 23 We used wmedian RSF caribou locations under the assumption that at least 50 percent of the caribou locations obtained during any activity period should be in preferred habitats. This method allows for replicable mapping of RSF models and comparisons of relative probability of occurrence among seasonal models. We partitioned ŵ into three equal intervals of relative probability of occurrence categories: A) ≤0.333 (low), B) ˃ 0.333 and ≤ 0.666 (moderate), and C) ˃ 0.666 (high). We calculated the percent of locations that fell within each probability of occurrence category to quantify the known frequency of use of areas we mapped as high, moderate, and low probability of occurrence. We obtained relative probability of occurrence for a specific activity period by pooling data for all years for that activity period. Table 1. Labels & definitions1 for resource selection (RS) habitat and elevation covariates considered in seasonal RS models for West Greenland Akia-Maniitsoq satellite-collared caribou cows. Covariate Definition Elevation Continuous variable; maximum elevation in meters above sea level (m asl) within focal area*. Heath Discrete variable; present (1) or absent (0) within focal area. Vegetation dominated by willow (Betula nana), crowberry (Empetrum nigrum hermaphroditum), blueberry (Vaccinium uliginosum microphyllum), and Labrador tea (Ledum groenlandicum, Ledum palustre decumbens). Open heath Discrete variable; present (1) or absent (0) within focal area. Vegetation dominated by blueberry (Vaccinium uliginosum microphyllum), birch (Betula nana), heather (Cassiope tetragona, Phyllodoce coerulea), and Labrador tea (Ledum palustre decumbens). Copse Discrete variable; present (1) or absent (0) within focal area. Vegetation dominated by willow (Salix glauca). Fen Discrete variable; present (1) or absent (0) within focal area. Vegetation dominated by fireweed (Eriphorum angustifolium hyparcticum) and sedges (Carex rariflora). Grass Discrete variable; present (1) or absent (0) within focal area. Vegetation dominated by grasses (Calamagrostis lapponica/groenlandica), sedges (Carex brunnescens, Carex bigelowii), bunch grasses (Deschampsia flexuosa), and wood rushes (Luzula spicata). Snow-bed Discrete variable; present (1) or absent (0) within focal area. Vegetation dominated by willows (Salix herbacea) and sedges (Carex bigelowii). Wind-exposed Discrete variable; present (1) or absent (0) within focal area. Vegetation dominated by mosses, lichens, dwarf shrubs (Diapensia lapponica lapponica), Labardor tea (Ledum palustre decumbens), campion (Silene acaulis), sweet grass (Hierochloë alpine), sedges(Carex bigelowii), and rhododendrons (Rhododendron lapponica). Soil / Rock Discrete variable; present (1) or absent (0) within focal area. Exposed soil and rock. Sediment Discrete variable; present (1) or absent (0) within focal area. River, estuarine, or coastal flood plains. Water Discrete variable present (1) or absent (0) within focal area. Fresh water lakes, rivers, and streams. Snow / Ice Discrete variable present (1) or absent (0) within focal area. Persistent multiannual snow and ice. 1 Bay & Simonsen 2009; Simonsen 2011 *Focal area is the 5x5 30 m cell moving window around each 30 m cell. 24 Model construction: animal locations & environmental covariates We derived the 12 independent variables (also termed covariates) from Tøttrup’s (2009) 30 m resolution satellite image based digital habitat and elevation models (Table 1), which included heath, open-heath, copse, fen, grass, snow-bed, wind-exposed, soil or rock, sediment, water, snow or ice (Bay & Simonsen 2009; Simonsen 2011) and elevation. We used ArcMap to create a separate 30 m cell-size grid for each independent variable. Habitat grid cells were assigned values of 1 or 0 if habitats were present or absent, respectively. We used a moving window, 5x5 30 m cell, in GIS, to determine the presence (1) or absence (0) of habitats within an area of ≈ 0.02 km2 surrounding each 30 m grid cell (focal cell) of each habitat grid. Focal cells were reclassified as having no data if clouds or shadows occurred in >20% of the cells contained in the moving window. We assigned the maximum elevation within moving windows for the elevation grid focal cells. We used GIS to extract the value for each focal habitat and elevation grid cell to intersect caribou and random locations (generated at a rate of 1 per km2 in the study area). Resources selection analysis Data for each female were sub-sampled randomly to one location per day (of the most accurate locations). 19,572 locations (one location per day per female) were used to calculate resource selection functions. We assessed patterns of habitat selection at the population level (Johnson et al. 2004; Latham et al. 2011), i.e., at the scale of the landscape within the study area, by comparing habitats and elevations at caribou GPS (used) and random (available) locations with used-versus-available design logistic regressions for each seasonal activity period (Johnson et al. 2006, Manly et al. 2002). We used GLM logit STATA 9 (STATCORP, College Station, Texas, USA) to assess the relationship between use and availability of all possible combinations of the 12 explanatory variables and selected the best-fitting models (Cook 2007, Hardin & Hilbe 2012, Schwarz 1978, Strong et al. 1999, Wasser et al. 2011) for each seasonal activity period using Schwarz’s information criteria (SBIC) (STATA 9; Schwarz 1978). We mapped the probability of occurrence of caribou using our modified linear stretch transformation (equation 3). Assessment of habitat selection among activity periods We compared relative selection of habitat types by caribou during calving and all other activity periods using logistic regressions to estimate coefficients for latent selection difference (LSD) functions (Czetwertynski 2007, Latham et al. 25 2011, Mueller et al. 2004). We used ANOVA and Tukey’s honestly significant different (HSD) pairwise comparisons to determine if elevations at caribou locations varied significantly among activity periods. We log10 transformed elevation data prior to conducting ANOVA and back-transformed the resulting statistics for interpretation. Calving Parturition (calving site): movement patterns Cervids commonly exhibit marked (i.e., > 50%) declines in daily movements immediately following parturition (Long et al. 2009). An abrupt lack of movement is evident when caribou birth their calf; cows stop completely (Lent 1966). Following birth of the neonate, directional movement is absent for several hours and thereafter; if it occurs, it is slow with many pauses (Lent 1966). More recently, Ferguson & Elkie (2004) observed that a sedentary period of low movement rate maintained for about 3 days affirmed calving in woodland caribou (Rangifer tarandus caribou). Using this knowledge and daily movement rates in the ±10 days around the date a cow stopped completely Nagy (2011) ascertained date of birthing for 336 cows, which was validated by visual survey. Criteria included daily movement rates rising sharply for several days, then falling precipitously to near zero, with a gradual increase in the days following. We predicted parturition dates using the above. We calculated daily travel rate by each caribou cow during the 15 April - 15 July maximum range of potential calving dates, and assumed that the date of the abrupt fall/cessation in daily movement signaled a birthing event, while the associated GPS location provided the calving site point data. Calving period & cow fidelity to calving site In addition to mean calving date, we defined periods of ‘most’ and ‘peak’ calving to augment assessment of possible temporal variations. The period of ‘most’ calving was estimated to be within ±1.96 standard deviation (SD; 95% CI) of the mean parturition date, and ‘peak’ calving to be within ±1 SD (68% CI) (Nagy 2011). We estimated conception dates by back-dating 229 days from parturition dates (Bergerud 1975, McEwan & Whitehead 1972, Rowell & Shipka 2009). We used the first and last estimated parturition and conception dates to define the start and end dates for the calving and breeding periods, respectively. Although the number of collared cows declined steeply over the study period, we still attempted to examine whether the period of calving 32 Figure 6. Distribution of habitat types present within 150x150 m moving window around each focal 30 m grid cell in the study area. Multi-year permanent snow/ice variable doesn’t include Ice Cap. Use / Available ratios The ratios indicate when caribou use of a habitat was disproportionately less than, equal to, or greater than availability within the study area (<1.0, 1.0, and >1.0). The highest UA ratio was 3.33 during the late fall/early winter season (Table 13). Thus, cows were disproportionately selecting for this range in light of its low availability. The second highest UA was 3.03 and occurred during breeding. The third was fall/pre-breeding at 2.75 followed closely by midwinter at 2.71. Thereafter came late winter and mid/late summer, both with UA’s >2.0. Given that early/mid-summer range was the most available it was not surprising that it also had the lowest UA, 1.49. 33 Table 6. Best-fit activity period resource selection function models for Akia-Maniitsoq caribou cows in the Central region of West Greenland, 2008-2010. Activity period Covariates in best-fit models BIC Calving elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -150298.7 Peak-calving elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -137189.4 Post-calving elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -136853.8 Early/mid-summer elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -140048.1 Mid/late summer elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -140439.1 Fall/pre-breeding heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -137204.7 Breeding elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -143218.3 Late fall/early winter elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -154203.3 Mid-winter elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -144092.0 Late winter elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -139452.9 Pre-calving elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -137443.7 34 Table 7. Parameter estimates and standard errors in the best-fit calving and peak of calving (mean calving date ±1 SD) activity period resource selection function models for Akia-Maniitsoq caribou cows in the Central region of West Greenland, 2008-2010. Habitat covariate Calving Period (mean calving date ±1.96 SD) Peak calving (mean calving date ±1 SD) Parameter estimate SE P-value Parameter estimate SE P-value Elevation 0.001 0.000 0.000 0.002 0.000 0.000 Heath 0.487 0.057 0.000 0.294 0.084 0.000 Open-heath 0.661 0.061 0.000 0.429 0.086 0.000 Copse 0.870 0.061 0.000 0.889 0.090 0.000 Fen -0.103 0.058 0.074 0.051 0.083 0.540 Grass 0.446 0.047 0.000 0.401 0.068 0.000 Snow-bed -0.210 0.104 0.044 -0.010 0.152 0.950 Wind exposed 0.431 0.058 0.000 0.350 0.084 0.000 Soil/rock -0.203 0.049 0.000 -0.223 0.072 0.000 Sediment -0.238 0.070 0.001 -0.383 0.106 0.000 Water -0.605 0.074 0.000 -0.907 0.116 0.000 Snow/ice -1.275 0.329 0.000 -2.295 0.715 0.000 Intercept -3.050 0.091 0.000 -3.734 0.132 0.000 35 Table 8. Parameter estimates and standard errors in the best-fit post-calving, early/mid-summer, and mid/late summer activity period resource selection function models for Akia-Maniitsoq caribou cows in the Central region of West Greenland, 2008-2010. Habitat covariate Activity period Post-calving Early/mid summer Mid/late summer Parameter estimate SE P-value Parameter estimate SE P-value Parameter estimate SE P-value Elevation 0.002 0.000 0.00 0.001 0.000 0.00 0.001 0.000 0.00 Heath 0.907 0.092 0.00 0.558 0.078 0.00 1.385 0.088 0.00 Open-heath -0.002 0.083 0.98 -0.249 0.066 0.00 -0.073 0.069 0.29 Copse 0.884 0.086 0.00 0.974 0.075 0.00 0.794 0.068 0.00 Fen 0.017 0.086 0.85 0.111 0.070 0.11 0.037 0.072 0.60 Grass 0.290 0.071 0.00 0.276 0.063 0.00 0.254 0.063 0.00 Snow-bed 0.235 0.142 0.10 0.380 0.104 0.00 0.289 0.103 0.01 Wind exposed 0.391 0.083 0.00 0.195 0.071 0.01 -0.321 0.070 0.00 Soil/rock -0.324 0.076 0.00 -0.127 0.065 0.05 0.096 0.065 0.14 Sediment -0.102 0.099 0.31 0.115 0.073 0.12 0.006 0.078 0.94 Water -0.936 0.116 0.00 -0.749 0.084 0.00 -0.498 0.085 0.00 Snow/ice -2.219 0.716 0.00 -1.260 0.329 0.00 -15.682 494.905 0.98 Intercept -3.802 0.132 0.00 -2.903 0.104 0.00 -3.384 0.111 0.00 36 Table 9. Parameter estimates and standard errors in the best-fit fall/pre-breeding, breeding, and post-breeding/late fall activity period resource selection function models for Akia-Maniitsoq caribou cows in the Central region of West Greenland, 2008-2010. Habitat covariate Activity period Fall/pre-breeding Breeding Post-breeding/late fall Parameter estimate SE P-value Parameter estimate SE P-value Parameter estimate SE P-value Elevation - - - -0.002 0.000 0.00 -0.004 0.000 0.00 Heath 1.189 0.093 0.00 1.258 0.096 0.00 1.257 0.082 0.00 Open-heath - - - 0.023 0.069 0.74 0.385 0.062 0.00 Copse 0.602 0.074 0.00 0.449 0.062 0.00 0.348 0.052 0.00 Fen -0.155 0.089 0.08 -0.048 0.072 0.50 0.063 0.058 0.28 Grass 0.679 0.068 0.00 0.710 0.060 0.00 0.505 0.055 0.00 Snow-bed -0.034 0.151 0.82 -0.288 0.128 0.03 -0.310 0.101 0.00 Wind exposed -0.065 0.077 0.40 -0.391 0.070 0.00 -0.121 0.059 0.04 Soil/rock -0.172 0.076 0.02 -0.080 0.067 0.23 -0.452 0.060 0.00 Sediment -0.276 0.105 0.01 0.059 0.083 0.48 -0.040 0.069 0.56 Water -0.555 0.112 0.00 -0.961 0.097 0.00 -0.656 0.076 0.00 Snow/ice -14.893 510.544 0.98 -14.467 453.309 0.98 - - - Intercept -3.308 0.101 0.00 -2.253 0.111 0.00 -1.627 0.094 0.00 37 Table 10. Parameter estimates and standard errors in the best-fit early/mid winter, late winter, and pre-calving activity period resource selection models for Akia-Maniitsoq caribou cows in the Central region of West Greenland, 2008-2010. Habitat covariate Activity period Early/mid winter Late winter Pre-calving Parameter estimate SE P-value Parameter estimate SE P-value Parameter estimate SE P-value Elevation -0.003 0.000 0.00 -0.002 0.000 0.00 -0.001 0.000 0.00 Heath 0.901 0.079 0.00 0.950 0.091 0.00 0.274 0.082 0.00 Open-heath 0.653 0.076 0.00 0.304 0.079 0.00 0.529 0.087 0.00 Copse 0.397 0.061 0.00 0.505 0.069 0.00 0.435 0.079 0.00 Fen 0.132 0.063 0.04 -0.079 0.074 0.29 0.074 0.076 0.33 Grass 0.219 0.059 0.00 0.228 0.067 0.00 0.332 0.069 0.00 Snow-bed -0.152 0.108 0.16 -0.408 0.133 0.00 -0.295 0.134 0.03 Wind exposed 0.383 0.063 0.00 0.348 0.073 0.00 0.446 0.077 0.00 Soil/rock -0.305 0.061 0.00 -0.190 0.070 0.01 -0.442 0.071 0.00 Sediment 0.044 0.075 0.56 0.399 0.081 0.00 0.273 0.085 0.00 Water -0.690 0.085 0.00 -0.661 0.095 0.00 -0.489 0.094 0.00 Snow/ice - - - - - - - - - Intercept -2.352 0.103 0.00 -2.631 0.113 0.00 -2.734 0.115 0.00 38 Table 11. Overview of habitat selection by seasonal activity period for Akia-Maniitsoq caribou cows in the Central region of West Greenland, 2008-2010. Seasonal Activity period Resource selection function covariates* elevation heath open-heath copse fen grass snow-bed wind-exposed soil/rock sediment water snow/ice calving + + + + = + - + - - - - post-calving + + = + = + = + - = - - early/mid summer + + - + = + + + = = - - mid/late summer + + = + = + + - = = - = fall/pre-breeding excluded + excluded + = + = = - - - = breeding - + = + = + - - = = - = post-breeding/late fall - + + + = + - - - = - excluded early/mid winter - + + + + + = + - = - excluded late winter - + + + = + - + - + - excluded pre-calving - + + + = + - + - + - excluded *+, =, and - denote covariate coefficients that were significant greater, not significantly different from, and significantly less than zero. 39 Table 12. Habitat selection during the calving period relative to those selected during all other activity periods based on latent selection difference (LSD) function model comparisons for Akia-Maniitsoq caribou cows in the Central region of West Greenland, 2008-2010. Seasonal Activity Period Relative selection* Heath Open-heath Copse Fen Grass Snow-bed Wind-exposed Soil/rock Sediment Water Snow/ice Post-calving 26 >100 = = = 31 = = = >100 = Early summer = >100 = 18 >100 = >100 = 31 = = Mid/late summer 48 >100 = = = 39 >100 30 28 = Fall/pre-breeding 51 >100 = = 20 = >100 = = = Breeding 68 >100 = = 17 = >100 = 43 >100 Post-breeding/late fall 74 = >100 29 = = >100 >100 46 >100 = Early/mid winter 63 30 = 35 >100 = >100 >100 47 >100 = Late winter 62 = = 17 >100 = >100 = 59 >100 62 Pre-calving = = >100 24 >100 = >100 >100 49 = = *Relative selection was calculated for variables with coefficients significantly different from 0 as exp (b) when b>0 and as [1-exp(b) when b<0 (Latham, Latham & Boyce, 2011). Relative selection values <100 indicate that use of the habitat during calving was significantly less by x% than that for the season being compared, values >100 indicate use of habitat during calving was significantly greater than that for the season being compared, and = indicates habitat use during calving and season being compared were not significantly different (Appendix 3). 40 Table 13. Percent availability, percent use, and ratio of percent use: percent availability of areas by probability of occurrence category and activity period for AkiaManiitsoq caribou cows in the Central region of West Greenland, 2008-2010. Seasonal Activity period Probability of occurrence Percent of study area by occurrence category (Available) Percent of caribou locations by occurrence category (Use) U:A ratio for occurrence categories* Low Moderate High Low Moderate High Low Moderate High Calving 29 26 45 5 16 79 0.17 0.62 1.77 Post-calving 30 31 39 6 22 72 0.18 0.69 1.89 Early/mid summer 17 31 52 3 20 77 0.19 0.65 1.49 Mid/late summer 51 15 34 17 11 72 0.33 0.73 2.09 Fall/pre-breeding 57 22 21 16 26 58 0.28 1.18 2.75 Breeding 63 13 24 13 13 74 0.21 1.01 3.03 Late fall/early winter 67 12 21 12 19 69 0.18 1.56 3.33 Mid-winter 57 17 26 13 16 71 0.23 0.94 2.71 Late winter 53 16 31 13 12 75 0.24 0.75 2.44 Pre-calving 33 23 44 8 19 73 0.25 0.82 1.67 *U:A ratios <1.0, 1.0, and ˃1.0 indicate use was disproportionately less than available, proportional to availability, and disproportionately greater than availability. 41 Figure 7. Categorized probability of occurrence of Akia-Maniitsoq caribou cows in West Greenland during 10 activity periods: follow bold black arrows for annual cycle from calving to pre-calving. Maps represent the product of landscape-scale resource selection functions. 48 Mean Julian calving dates for adult cows 2008, 2009, and 2010 were respectively, 152.80 ±6.48 days (range 134-163; n = 30), 151.29 ±10.45 (range 132-165; n = 14) and 142.86 ±5.24 days (range133-149; n = 7). Mean calving dates did not differ for 2008 and 2009, 2009 and 2010 (ANOVA F1 = 0.348, P = 0.558; F1 = 3.98, P = 0.061 respectively), however, a difference was observed between 2008 and 2010 (ANOVA F1 = 14.193, P = 0.006). Employing the nonparametric Mann-Whitney U test on calving dates resulted in no differences between 2008 and 2009, 2009 and 2010 (P > 0.05), but difference between 2008 and 2010 (P ≤ 0.001). 2010 calving was approx. 10 days earlier than in 2008. Mean Julian calving dates for only those cows calving in the two specific years tested also tested non-significant for 2008 and 2009, 2009 and 2010 (P > 0.05). Comparison of cows birthing in both 2008 and 2010 was confounded by the small sample size of three and individual results of 15, 6 and -1 days, where 2010 calving was a mean 7 days earlier than in 2008 (P > 0.05). Calving site fidelity Although sample size was small and the study period short, we noted that calving site fidelity was high. There were 16 sequential calving locations for 14 cows (twelve cows had two locations; two cows had three). Median distance between sequential calving locations was 7.1 km. Mean distance was 12.5 km ± 11.7 SD, minimum was 0.5 km, and maximum 41.5 km. The minimum coincided with the cow that calved in 2008 and again in 2010, after a hiatus of one year. Figure 12. Annual variation in the number of caribou cows in various reproductive states, 2008-2010. 0 5 10 15 20 25 30 35 2008 2009 2010 Number of cows Year Calf Unknown No calf 49 Figure 13. Temporal distribution of 52 parturition dates 2008-2010, West Greenland Akia-Maniitsoq satellite-collared caribou cows (note 28 June sub adult outlier). Discussion The AM caribou are substantially less documented than caribou populations in North America. Therefore, this report could benefit caribou management in Greenland. Identifying the attributes, locations and extent of caribou seasonal habitats, and the timing of use of these, is indispensable to decisions intended to predict and offset negative impacts to caribou habitat, e.g., by climate and anthropogenic factors. The results presented in this report are particularly valuable because they establish a baseline for AM habitat use in the current absence of significant development and infrastructure in the Central region. We do not advise extrapolating the current knowledge from AM to other populations in Greenland, owing to differences in population size, latitude, climate, topography and lack of seasonal habitat use data for the other populations. Further, the short study period and small sample size, which constantly diminished over that period, weaken this study’s findings. Akia-Maniitsoq cow movements Daily Similar to Cuyler & Linnell (2004) this study observed minimum daily movement/activity during a ca. 30-day period concurrent with March, specifically at the beginning of March. Highest cow movement rates occurred in July, and may be associated with insect harassment. The autumn rut 0 1 2 3 4 5 6 12/May 19/May 26/May 2/Jun 9/Jun 16/Jun 23/Jun Number of paturition dates Date 50 (breeding) created a second briefer peak in activity in late September. This knowledge indicates that aerial surveys for caribou abundance in early March will minimize the problem of caribou movement affecting the resulting population estimates. The activity trend observed is not new. A summer activity increase has been observed for caribou / reindeer by several authors (Banfield 1954, Segal 1962, Thomson 1971, Gaare et al. 1975, White et al. 1975, and Roby 1977). The higher activity in summer is ascribed most often to high air temperatures and to biting and parasitic insects (Thing & Thing 1983, Mörschel & Klein 1997, Coleman et al. 2000) and it has been documented that wind speed affects mosquitoes, which have a threshold of 4-6 m/s (Russell et al. 1993, Parrot 2007). Increased caribou movement resulting from insect harassment reduces foraging time and negatively impacts calf growth and cow body condition (Mörschel & Klein 1997, Couturier et al. 2009). Thus insect relief habitat, i.e., windy locations and snow patches (Joly & Klein 2011, Wilson et al. 2012) is an important consideration when managing essential habitats. Following parturition, it is typical among Cervids that there occurrs an immediate and marked decline in daily movement (Long et al. 2009). The sudden cessation of movement at birthing by caribou cows was first described by Lent (1966), who observed that cows always stopped still when labour commenced, regardless if that cow had been with a moving group, which left the birthing cow behind. Ferguson & Elkie (2004) observed the same phenomenon, which was also later support by Nagy’s (2011) movement rates for cows around the time of calving and known to later have calves-at-heel (n=336). Still, an assumption that a cow being stationary equates with birthing requires further substantiation. Video collars, recording daily, on parturient cows would be one solution. Annual There was a distinct seasonal distribution of the collared caribou cows in the Central region. Generally, there was a southwest-northeast axis to their annual movements. Although movements were often short and normally not from seacoast to Ice Cap, each AM cow was typically furthest east in summer and furthest west in winter. Also, each cow utilized only a fraction of the available area. This supports Cuyler & Linnell’s (2004) idea that there is a high degree of population sub-division of the AM caribou in the Central region. This knowledge is relevant for caribou management, e.g., conservation actions 51 applied to only a portion of the region will probably not benefit the entire population. Further, stochastic weather events, which have the potential for influencing caribou abundance negatively, if limited in extent to just a portion of the region, will also not likely affect the entire AM population. Migratory path length AM cow movements between winter and summer ranges were on a relatively small scale. The mean extent, 76 km, of AM cows’ migratory path length was similar to that observed in the 1990’s (Cuyler & Linnell 2004). In contrast, barren-ground caribou (same sub-species as AM cows), and those of northern Alaska, typically make long movements between winter and calving home ranges. These distances can be in the order of 300-500 km for the Central Alaskan or Porcupine herds (Craighead & Craighead 1987, Fancy et al. 1989) or many of the other herds in northern Canada (Hall 1989). On a geographic scale, the AM study area is much smaller than most North American migratory barren-ground caribou home ranges. Thus, our observed short AM migratory paths were not unexpected, and they illustrate the confined areas available to caribou in West Greenland. Also reflected is the diversity of habitats available to the AM caribou across relatively short distances. Since they do not traverse large distances, AM cows could partition any annual energy budget surplus to things other than locomotion, e.g., reproduction. This was supported in the 1990’s when AM cows exhibited an exceptional lifetime fecundity (Cuyler & Østergaard 2005). Movement of collared AM cows between winter and summer ranges was less than 100 km, followed an elevation gradient, and they were spatially independent from each other. Thus, AM cows appear unlike their barrenground cousins and most similar to the Dolphin & Union Island, North American Boreal and Mountain Woodland Caribou R. t. caribou (Oosenburg & Theberge 1980, Bergerud et al. 1984, Skogland 1986, 1989, Edmonds 1988, Hillis et al. 1998, Nagy 2011, Nagy et al. 2011). This also highlights the importance of telemetry studies. Seasonal activity periods Individual caribou select habitat for their own successful survival and reproduction (Bélanger & Côté 2016, Fancy & Whitten 1991, Albon & Langvatn 1992). Patterns of movement revealed 10 seasonal activity periods for the Greenland AM population. These were associated with specific 52 habitat/range attributes. Our classification falls between Russell et al. (1993) who identified 15 activity periods for migratory barren-ground caribou, and Nagy’s (2011) eight periods for both boreal caribou and tundra-wintering barren-ground caribou. Future analyses may reduce the partitioning to fewer than 10 periods, since for example the habitats chosen for early, mid and late winter periods appear similar. Specific elevation use was associated with the different seasons throughout the year. Elevations under 200 m were always used for breeding through to the end of late winter, while elevations above 350 m were invariably used for calving and to the end of mid-summer. Elevation appeared to be a primary attribute defining seasonal ranges. This knowledge could be used for conservation actions, e.g., ending an autumn harvest before the onset of the breeding season, when caribou are at low elevations and accessible to hunters. Late summer and fall ranges provide the basis for a caribou’s body condition build-up that will permit participation in the rut (Cameron et al. 1993), while winter range is vital for survival (Bergerud et al. 2007). Thus these ranges qualify among the essential ranges for caribou, i.e., important for reproduction or survival. Unfortunately, for AM caribou, there is a scarcity of habitat area associated with late summer, fall and winter relative to the other periods. For example, preferred AM winter habitat use is at elevations under 200 m, and while winter range is vital for caribou survival, this habitat is in short supply for AM caribou. Winter range is the smallest area of all the seasonal ranges. AM winter range may therefore need special attention and protections. Increasing the concern, is that the largest tract of winter habitat, the Akia-Nordland lowlands north of Godthåbsfjord, is exposed to southwesterly storm systems. These bring the possibility of several negative weather events, e.g., deep snows, rain-on-snow, and icing. The shortage of late summer, fall and winter ranges, indicates that these could be limiting for the AM caribou population. Seasonal habitat selection Habitat selection varied by season, and seasonal habitats varied considerably in their distribution and area, relative to total study area. Calving habitat had the greatest area and distribution. Tied for the least area were fall/prebreeding and late fall/early winter. In fact, for the activity periods from fall through mid-winter, habitats were relatively few and small in area. The 53 largest was the Akia-Nordland lowlands, followed by the smaller lowlands of the Narssarssuaq and Iluliak valleys. There were also a few scattered smaller valley lowlands. Taken together, they comprised only 21-26% of the total study area. The scarcity of these habitats emphasizes that human influences causing loss of these could have detrimental impacts on the AM population. Interestingly, the fall/pre-breeding activity period evidenced the lowest percentage of cow locations even in the highest occurrence category. An earlier study observed that the lowest probability of occurrence for cows was at locations having the highest values of hunting (Simonsen 2011). Since fall/pre-breeding coincides with the hunting season (1 Aug – 30 Sept), results may reflect avoidance of disturbance / predation risk from hunters using the same areas. From a management perspective, this suggests that the human disturbance associated with a long summer/autumn harvest season may negatively influence breeding and subsequently calf production. Our analyses for probability of cow occurrence during calving and breeding are supported by similar results obtained by Simonsen (2011). Our study, however, was able to identify more areas. As regards a comparison of summer and fall habitats, our probability of occurrence results suggest that fall [September] habitat/range is limited in area/availability, while summer [July] habitat/range is not. The small but important fall area suggests that fall habitat may require protection measures. This would be in addition to July habitat, which is important for among other things selective feeding by cows and insect relief (Tamstorf et al. 2005, Wilson et al. 2012). In the following sections, we will explore linkages among for example nutrition, forage, insect relief, elevation, hunting and winter snow. Calving Mean elevation use, ca. 400 m, in the calving activity period (11 May – 17 June) was significantly higher than for all other activity periods. Cuyler & Linnell (2004) and Simonsen (2011) also observed highest elevation use by AM cows during calving. Strikingly, calving sites themselves were typically at least 200 m above and beyond the mean elevation use for all cows in the calving activity period. This suggests that whatever benefits elevation confers, parturient cows appear to be maximizing these. Calving sites also had a generally southerly aspect, which would take advantage of solar radiation. 54 With the advent of reliable aspect models, future studies may be able to explore the significance of aspect on habitat selection at the landscape scale. Similar to observations in the late 1990’s (Cuyler & Linnell 2004), AM calving sites were dispersed in a continuum across the Central region, where the straight-line distance between seacoast and Greenland Ice Cap is only about 110 km. Certainly 1/3 of births occurred within 25 km of the Greenland Ice Cap, however, several were within 10-20 km of the seacoast. Only 4-5 sites corresponded with the ‘calving ground’ delineated in NunaGis (2015). Considering only inland areas, as suitable for calving, appears to have little relevance for AM cow habitat selection at calving. Cow dispersal at calving is typical for the montane ecotype of caribou (Bergerud et al. 1984, Skogland 1989) and once again, highlights how very unlike the AM caribou are from their barren-ground cousins. In contrast to proximity to the Ice Cap, elevation was the good predictor for probability of AM cow occurrence at calving (and breeding). The AM calving continuum is unsurprising given that 42% of this range possesses the habitat attributes associated with highest probability of occurrence by cows during peak calving: high elevation, heath, open heath, copse, fen, grass, snowbed and wind exposed ridge. Potential xeric habitat for calving is present near the Ice Cap, but the extent of essential calving habitat available elsewhere throughout the region is now recognized. This strengthens capacity to regulate human activities in calving habitat before during and after calving. Based on daily movement rates, the seasonal activity period for calving lasted from 11 May until 17 June. Using actual birthing events, ‘Peak’ calving (68%) was a 19-day period (21 May-8 June), with ‘most’ calving (95%) over a 32-day period, 14 May to 15 June. The period length of AM peak calving is similar to Boreal caribou (R.t. caribou) (Nagy 2011). May births were unexpectedly common, mean 30 May ± 8.2 days. Genetic mixing with feral semi-domestic reindeer (Jepsen et al. 2002), which calf in May, is a plausible explanation for the early calving. Changing weather or plant phenology might also be factors. We note that mean calving date 2010 was 7-10 days earlier than 2008, however, the small 2010 sample size precludes conclusions about a forward shift in calving phenology. 55 Calving site fidelity High fidelity to calving grounds illustrates their important role in the annual cycle of calf production (Gunn & Miller 1986, Fancy & Whitten 1991, Schaefer et al. 2000, Mahoney & Schaefer 2002, Russell et al. 2002, Ferguson & Elkie 2004). The satellite-collared AM cows exhibited high fidelity to previously chosen calving sites. However, our small sample size and short study period do not permit conclusive statements regarding patterns. It is interesting to note that our minimum value of 0.5 km and median value of 7.1 km were similar to the 4 to 11 km observed for caribou of the mountain ecotype in southeastern Labrador (Popp et al. 2011) and the < 5 km for woodland caribou of northern Ontario (Ferguson & Elkie 2004). In contrast, Yukon-Alaskan Porcupine caribou cows, which though faithful to a general area did not calf near same location annually, e.g., minimum 67.1 ± 49.1 km (Fancy & Whitten 1991). Given the high fidelity shown by AM cows, anthropogenic disturbance could have strong negative impacts if cows are forced to reduce fidelity to preferred sites (Faille et al. 2010). Calving site choice Spring migration of parturient Rangifer cows is often northward in latitude, or towards increasing elevations. This is thought to be an evolutionary response to temporal and spatial resource variation that will maximize calf survival, e.g., increase available forage quantity and quality, or reduce the risk of neonate predation (Bergerud & Page 1987, Fryxell et al. 1988, Klein 1990, Albon & Langvatn 1992, Griffith et al. 2002, Russell et al. 2002, Post et al. 2003, Loe et al. 2005). There is no doubt that calving areas are essential to productivity of a caribou population, and that access to these has direct consequences at the population level (Carroll et al. 2005). Carroll et al. (2005) found that calving sites were primarily related to the timing and presence of snowmelt. Parturition and subsequent lactation represent enormous energy expenditures for cows (Boertje 1985, Bergerud et al. 2007), yet parturient caribou employ a capital reproductive strategy, i.e., nitrogen demands throughout gestation and early lactation are chiefly covered by maternal body reserves, which were deposited late in the previous growing season (Moen et al. 2006, Barboza & Parker 2008, Taillon et al. 2013). Thus, peak parturition can precede green-up by several weeks (Durant et al. 2005, Moen et al. 2006, Barboza & Parker 2008, Taillon et al. 2013), and in fact, parturient cows normally arrive at calving areas before green-up has occurred (Whitten & Cameron 1980, Kellyhouse 2001). Once green-up begins, calving 56 areas are associated with high rates of biomass increase (Kellyhouse 2001), and the emergent plants are high in soluble carbohydrates, nitrogen and phosphorus, which deteriorate as summer advances (White et al. 1975, Whitten & Cameron 1980, Jorgensen et al. 2002). Thus, emergent forage at calving sites may well coincide with the birthing cows’ period of greatest energy needs, which is about three weeks post-calving (Parker et al. 1990). As regards available forage quantity and quality at calving, rugged terrain, even in lowlands, widens the period of emergent forage availability in areas of patchy snowmelt (Nelleman & Thomsen 1994). Thus, high elevation would not be a prerequisite to emergent forage on the rough-featured AM region. AM lowlands can have patchy snowmelt with emergent vegetation underway, yet most parturient adult AM cows birthed at relatively high elevations of ca. 600 m or more. Given mean calving date, 30 May, and the ca. 65°N latitude of the Central region, emergent vegetation is unlikely at the elevations typically chosen for birthing by AM cows. AM calving appears to occur where food resources are limited and vegetation green-up many weeks in the waiting, e.g., maximum NDVI’s (normalized difference vegetation index, ‘density of green’) across the Central region generally do not occur until July (Tøttrup 2009). Why would AM cows choose high elevations or xeric areas? Why not birth in the snowmelt occurring in the Central region’s lowlands? The latter would permit access to emergent forage, which as spring progressed could be pursued up the readily available elevation gradients throughout the region. Since insect harassment first begins several weeks post-calving, escape from these is not likely a factor in birthing site choice. Insect relief could, however, keep cows at high elevations, which are often windy, in the post-calving period. Decreasing the neonate predation risk is a mote point; because large predators have been absent in West Greenland since at least the 1870’s (Vibe 1981). The elevations chosen by AM cows suggest the possibility of greater nutrient value contained in emergent vegetation associated with snowmelt in highlands, and that it is available over a longer period. Future investigation of the nutrient value and period length of emergent vegetation at high elevation versus lowlands may cast light on this conundrum. Weather might also influence calving site choice. The Akia-Maniitsoq region is exposed to south westerly storm systems. Birthing cows choosing high elevations or dry xeric areas close to the Greenland Ice Cap may be reducing 57 the risk of unfavourable weather conditions for neonate survival, e.g., rain, which could increase neonate mortality through excessive heat loss associated with cold stress hypothermia. Heavy rainfall and wind can drastically increase neonate mortality in Arctic animal populations (Blix & Steen 1979, Mallory et al. 2009, Anctil et al. 2014, Yannic et al. 2014) including caribou neonates (Kelsall 1968, Blix 1980). Even without wind, wet neonate heat loss increases 5-fold (Lentz & Hart 1960, Hart et al. 1961, Markussen et al. 1985). Although rain coupled with wind increases neonate mortality, its role on parturient caribou migration and calving site choice has been ignored (Miller & Gunn 1986). Reducing the likelihood of exposing neonates to rain and wet substrates would benefit offspring survival in large ungulates (Azzam et al. 1993). This hypothesis that high elevation (combined with 65°N lat.) or xeric areas promotes neonate survival by reducing the risk of rain, needs further study. If true, this may clarify what options are open to parturient cows on ranges with low relief landscapes where the only rain avoidance choices may be locations of driest climate or the most northerly latitudes possible. Interestingly, having a yearling-at-heel (calf from the previous spring) appeared to affect a cow’s elevation choice (median 186 m lower) for parturition. The difference approached significance (P = 0.058, df = 10, t = 2.2281) and may be biologically significant. While cows are capital breeders, i.e., rely on body reserves for birthing and initial lactation, her yearling-at-heel may have fewer body reserves remaining at winter’s end, making high elevations, snow-bound, with reduced forage availability less than ideal. Parturient cows with yearling-at-heel from the previous spring may choose birthing sites at lower elevations to balance their yearling’s forage needs against enhancing survival of the coming neonate. To date, there are no studies addressing parturient caribou calving site choice in the presence or absence of a yearling-at-heel. Aerial observation of parturient cows precalving combined with investigations of preand post-calving relocations, might ascertain if this trade-off is occurring. Management Implications Mortality 2008-2010 Adult female and calf survival are the two key demographic factors that determine population growth rates (Hatter & Bergerud 1991). The 2008-2010 mortality among the AM satellite-collared cows was disturbingly high: 26 of 64 penalties for illegal harvest and the financial resources wasted when a collared cow is removed untimely from telemetry studies. Hunter harvest exacerbated collared-cow mortality in the current study. An information campaign could increase public understanding of the significance of collar data for effective management of habitat and caribou to the benefit of all. Acknowledgements The Arctic Environment Program of the Danish Ministry of Environment and Energy, the aluminium smelter company ALCOA, the Danish Assistance to the Arctic Program (DANCEA), the Greenland Government and the Greenland Institute of Natural Resources provided funding. The former were obtained by Peter Aastrup. Satellite-based vegetation maps were provided by ALCOA. The Greenland government’s Asiaq aided us with the digital elevation model. We thank Air Greenland Charter and their pilots for safe flying during caribou capture. We thank Barry Minor for efficient net-gunning and capture of caribou and also Hans Mølgaard for his valuable assistance. A. Brooker advised on unsuitability of linking passive microwave variables to point data. We thank Wendy Loya for extensive review of the manuscript and Joseph McCullough for proofreading. 65 Literature cited Anctil, A., A. Franke, and J. Bêty. 2014. Heavy rainfall increases nestling mortality of an Arctic top predator: experimental evidence and long-term trend in peregrine falcons. Oecologia 174: 1033–1043. Aastrup, P. 1986. Rensdyrundersøgelser ved Vandkraftprojekt Kangerluarsunnguaq/ Buksefjord, Nuuk/Godthåb 1984-85. Grønlands Fiskeriog Miljøundersøgelser, Roskilde, Denmark. [In Danish.] Aastrup, P. 2004. Rensdyrs adfærd ved forstyrrelser. In: Samspillet mellem rensdyr, vegetation og menneskelige aktiviteter i Vestgrønland (ed.) Aastrup P. Pinngortitaleriffik – Greenland Institute of Natural Resources. Technical report No. 49. 320 pp. [In Danish] Aastrup, P., and J. Nymand. 2004. Bestandsforhold i kælvningsområder i Vestgrønland 1995, 1997 og 1998. In: Samspillet mellem rensdyr, vegetation og menneskelige aktiviteter i Vestgrønland (ed.) Aastrup P. Pinngortitaleriffik – Greenland Institute of Natural Resources. Technical report No. 49, Nuuk, Greenland. [In Danish] Albon, S. D, and R. Langvatn. 1992. Plant phenology and the benefits of migration in a temperate ungulate. Oikos 65:502–513. (doi: 10.2307/3545568) Anon. 1999. Rules and regulations. Department of the Interior, Fish and Wildlife Service 50 CFR Part 17. Federal Register 64(234):68508–68544. FR Doc. 99–31357 Filed 12–1–99. USA. Azzam, S. M., J. E. Kinder, M. K. Nielsen, L. A. Werth, K. E. Gregory, L. V. Cundiff, and R. Koch. 1993. Environmental effects on neonatal mortality of beef calves. Journal of Animal Science 71: 282–290. Banfield, A.W.F. 1954. Preliminary investigation of the Barren-ground caribou. Canadian Wildlife Service Wildlife Management Bulletin, Serial 1, No. 10 A. Ottawa: Canadian Wildlife Service. Barboza, P. S., and K. L. Parker. 2008. Allocating protein to reproduction in Arctic reindeer and caribou. Physiological and Biochemical Zoology 81(6): 835–855. (doi:10.1086/590414). Bay, C., and C. E. Simonsen. 2009. Botanical investigations in relation to the Alcoa project in West Greenland, 2009. Technical report. National Environmental Research Institute. Aarhus University, Aarhus, Denmark. Bélanger, E., and S. Côté. 2016. Habitat selection and survival of the declining Torngat Mountains caribou herd. Technical sessions. 16th North American Caribou Workshop, 16-20 May, 2016, Thunder Bay, Ontario, Canada. Bergerud, A. T. 1974. The role of the environment in the aggregations, movement, and disturbance behavior of caribou. Pages 552–548 in V. Geist and F. Walther, editors. The behavior of ungulates and its relation to management. Vol. 2 IUCN New Series Publication 24. Morges, Switzerland. Bergerud, A. T. 1975. Reproductive season of Newfoundland caribou. Canadian Journal of Zoology, 53(9): 1213-1221. Bergerud, A. T., Butler, H. E., and D. R. Miller. 1984. Antipredator tactics of calving caribou: dispersion in mountains. Canadian Journal of Zoology 62: 1566–1575. Bergerud, A. T., and R. E. Page. 1987. Displacement and dispersion of parturient caribou at calving as antipredator tactics. Canadian Journal of Zoology 65: 1597–1606. 66 Bergerud, A. T., S. N. Luttich, and L. Camps. 2007. The return of caribou to Ungava. McGillQueen’s University Press, Montreal and Kingston/London/Ithica, Canada. Beyer, H. L. 2007. Hawth's analysis tools for ArcGIS (Version 3.27). Available at http://www.spatialecology.com/htools Blix, A. S. 1980. Temperature regulation and climatic influence on mortality in newborn arctic mammals in E. Reimers, E. Gaare and S. Skjenneberg, editors. Proceedings of the Second International Reindeer / Caribou symposium. 17. –21. September, Røros, Norway 1979. Direktoratet for vilt og ferskvannsfisk, Trondheim, Norway. Blix, A. S., and J. B. Steen. 1979. Temperature regulation in newborn polar homeotherms. Physiological Reviews 59: 285–304. Boertje, R. D. 1985. An energy model for adult female caribou of the Denali herd, Alaska. Journal Range Management 38(5): 468–473. Bolger, D.T., Newmark, W.D., Morrison, T.A., and Doak, D.F. 2008. The need for integrative approaches to understand and conserve migratory ungulates. Ecology Letter 11: 63-77. Borger, L., Franconi, N., DeMichele, G., Gantz, A., Meschi, F., Manica, A., Lovari, S. and Coulson, T. 2006. Effects of sampling regime on the mean and variance of home range size estimates. Journal of Animal Ecology 75: 1493-1405. Boyce, M. S., Vernier, P. R., Nielsen, S. E. and Schmiegelow, F. K. A. 2002. Evaluating resource selection functions. Ecological Modelling, 157: 281-300. Bradshaw, C.J.A., Boutin, S. and Hebert, D.M. 1997. Effects of petroleum exploration on woodland caribou in northeastern Alberta. J. Wildl. Manage. 61 (4): 1127-1133. Cameron, R. D., D. J. Reed, J. R. Dau, and W. T. Smith. 1992. Redistribution of calving caribou in response to oil-field development on the Arctic slope of Alaska. Arctic 45: 338–342. Cameron, R.D., Smith, W.T., Fancy, S.G., Gerhart, K.L. & R.G. White. 1993. Calving success of female caribou in relation to body weight. Canadian Journal of Zoology 71: 480-486. Cameron, R. D., W. T. Smith, R. G. White, and B. Griffith. 2005. Central Arctic caribou and petroleum development: distributional, nutritional and reproductive implications. Arctic 58(1): 1–9. Cameron, R.D. & Whitten, K.R. 1979. Seasonal movements and sexual segregation of caribou determined by aerial survey. Journal of Wildlife Management 43: 626-633. Carrol, G. M., Parrett, L. S., George, J. C. & Yokel, D. A. 2005. Calving distribution of the Teshekpuk caribou herd, 1994-2003. – In: Haugerud, R. E. (ed.). Proceedings of the 10th North American Caribou Workshop, Girdwood, Alaska, USA 2004. Rangifer Special Issue No. 16: 27-35. Coleman, J., Pedersen, C., Reimers, E., Holand, Ø. & Moe, S. 2000. 24-h activity pattern of wild reindeer in summer with emphasis on behaviour compensation at night due to limited grazing during the day. Proceedings of the 8th North American Caribou Workshop, Whitehorse, Yukon, Canada, 20-24 April, 1998. Rangifer Special Issue 12: 143. Committee, P.C.T. 1993. Sensitive habitats of the Porcupine caribou herd. Report accepted by the International Porcupine Caribou Board from the Porcupine Caribou Technical Committee January 1993. 67 Cook, N.R. 2007. Use and misuse of the receiver operating characteristic curve in risk prediction. Circulation, 115, 928-35. Couturier, S., Cote, S.D., Otto, R.D., Weladji, R.B. & Huot. J. 2009. Variation in calf body mass in migratory caribou: the role of habitat, climate, and movements. Journal of Mammalogy 90: 442-452. Craighead, D.J. and Craighead, J.J. 1987. Tracking caribou using satellite telemetry. National Geographic Research. 3: 462-479. Cuyler, C., and J. D. C. Linnell. 2004. Årlig vandringsmønster hos satellitmærkede rensdyr I Vestgrønland. In: Samspillet mellem rensdyr, vegetation og menneskelige aktiviteter i Vestgrønland (ed.) Aastrup P. Pinngortitaleriffik – Greenland Institute of Natural Resources. Technical report No. 49, Nuuk, Greenland. [In Danish.] Cuyler C., Nymand J., Jensen A., Mølgaard H.S. 2016. 2012 status of two West Greenland caribou populations, 1) Ameralik, 2) Qeqertarsuatsiaat. Greenland Institute of Natural Resources Technical Report No. 98, 179 pp. Cuyler, C., M. Rosing, J.D.C. Linnell, P.M. Lund, P. Jordhøy, A. Loison & A. Landa. 2003. Status of 3 West Greenland caribou populations; 1) Akia-Maniitsoq, 2) Ameralik & 3) Qeqertarsuatsiaat. Pinngortitaleriffik – Greenland Institute of Natural Resources. Technical report No. 46. 74 pp. Cuyler, C., M. Rosing, H. Mølgaard, R. Heinrich, and K. Raundrup. 2011. Status of two west Greenland caribou populations 2010; 1) Kangerlussuaq-Sisimiut and 2) Akia-Maniitsoq. Pinngortitaleriffik – Greenland Institute of Natural Resources. Technical Report No. 78, Nuuk, Greenland. Cuyler L.C. & Østergaard J. 2005. Fertility in two West Greenland caribou populations 1996/97: Potential for rapid growth. Wildlife Biology. 11(3): 31-37. Czetwertynski, S. M. 2007. Effects of hunting on the demographics, movement, and habitat selection of American black bears (Ursus americanus). Ph.D. thesis, Department of Renewable Resources, University of Alberta, Edmonton. Dau, J. R., and R. D. Cameron. 1986. Effects of a road system on caribou distribution during calving. Rangifer Special. Issue 1: 95–101. Durant, J. M., D. O. Hjermann, T. Anker-Nilssen, G. Beaugrand, A. Mysterud, N. Pettorelli and N. C. Stenseth. 2005. Timing and abundance as key mechanisms affecting trophic interactions in variable environments. Ecology Letters 8(9): 952–958. (doi: 10.1111/j.14610248.2005.00798.x). Dyer, S. J., J. P. O’Neill, S. M. Wasel, and S. Boutin. 2001. Avoidance of industrial development by woodland caribou. Journal of Wildlife Management 65(3): 531–542. Edmonds, E.J. 1988. Population status, distribution, and movements of woodland caribou in west central Alberta. Canadian Journal of Zoology 66: 817-826. Environment Canada. 2011. Scientific review for the identification of critical habitat for woodland caribou (Rangifer tarandus caribou), boreal population, in Canada. Environment Canada, Ottawa, Ontario, Canada. Fancy, S. G., Pank, L.F., Whitten, K. R. and Regeln W.L. 1989. Seasonal movements of caribou in arctic Alaska as determined by satellite. Canadian Journal of Zoology 67(3): 644–650. 68 Fancy, S. G. & Whitten, K. R. 1991. Selection of calving sites by Porcupine herd caribou. Canadian Journal of Zoology 69: 1736–1743. Faille, G., Dussault, Chr., Ouellet, J.-P., Fortin, D., Courtois, R., St-Laurent, M.-H. & Dussault, C. 2010. Range fidelity: The missing link between caribou decline and habitat alteration? Biological Conservation. 143: 2840-2850. Ferguson, S. H. and Elkie, P. C. 2004. Habitat requirements of boreal forest caribou during the travel seasons. Basic and Applied Ecology, 5(5): 465-74. Festa-Bianchet, M., Ray, J. R., Boutin, S., Côté, S. D. & Gunn, A. 2011. Conservation of caribou in Canada: an uncertain future. Canadian Journal of Zoology. 89: 419-434. Fryxell, J. M., J. Greever, and A. R. E. Sinclair. 1988. Why are migratory ungulates so abundant? American Naturalist 131: 781–798. Gaare, E., Thomson, B.R., & Kjos-Hanssen, O. 1975. Reindeer activity on Hardangervidda. In Fennoscadian tundra ecosystems. Part 2. Animals and system analysis, ed. F.E. Wielgolaske, pp. 206-215. Berlin, Heidelberg & New York: Springer-Verlag. (Ecological Studies, 17.) Gaillard, J. M., M. Festa-Bianchet, and N. G. Yoccoz. 1998. Population dynamics of large herbivores: variable recruitment with constant adult survival. Trends Ecological Evolution 13: 58–63 and 170. Gerhart, K.L. 1995. Nutritional and ecological determinants of growth and reproduction in caribou. Ph. D. Thesis, University of Alaska Fairbanks, Fairbanks, Alaska. Government of Greenland Bureau of Minerals and Petroleum. 2000. Rules for field work and reporting regarding mineral resources in Greenland. No. 69.03.20+01. November 2000. 38 pp. www.bmp.gl Denmark Centre for Environment website maps: http://gis.au.dk/ RDImportantAreas/Default.aspx Grant, L., Johnson, C.J. and Thiessen, C. B. 2016. Examining survival in a declining Woodland caribou herd: a consideration of management history and disparities with relatively stable, neighbouring herds. Technical sessions. 16th North American Caribou Workshop, 16-20 May, 2016, Thunder Bay, Ontario, Canada. Griffith, B., Douglas, D. C., Walsh, N. E., Young, D. D., McCabe, T. R., Russell, D. E., White, R. G., Cameron, R. D. & Whitten, K. R. 2002. The Porcupine caribou herd. – In: Douglas, D.C., Reynolds, P.E. & Rhode, E.B. (eds.). Arctic Refuge coastal plain terrestrial wildlife research summaries. U.S. Geological Survey, Biological Resources Division. Biological Science Report USGS/BRD/BSR – 2002-0001, pp. 8-37. Grønnow, B., Meldgaard, M. and Nielsen, J. B. 1983. Aasivissuit - the great summer camp archaeological, ethnographical and zoo-archaeological studies of a caribou-hunting site in West Greenland. Meddelelser om Grønland, Man & Society. 5: 96pp Gunn, A. & Miller, F. 1986. Traditional behavior and fidelity to caribou calving grounds by barren-ground caribou. Rangifer, Special Issue No. 1: 151–158. Gunn, A., Poole, K.G. and Wierzchowski, J. 2008. A geostatistical analysis for the patterns of caribou occupancy on the Bathurst calving grounds 1966–2007. Indian and Northern Affairs Canada, Yellowknife, NWT. Gustine, D. D., Parker, K. L., Lay, R. J., Gillingham, M. P. and Heard, D. C. 2006. Calf survival of woodland caribou in a multi-predator ecosystem. Wildlife Monographs(165): 1-32. 69 Hall, E. 1989. People / Caribou in the Northwest Territories: Northwest Territories, Department of Renewable Resources. Hall, D. K., G. A. Riggs, and V. V. Salomonson. 2014. MODIS/Terra Snow Cover 5-min L2 Swath 500m V005, May-August 2008, 2009 and 2010, Boulder, Colorado USA: National Snow and Ice Data Center. Hanski, I. 1998. Connecting the Parameters of Local Extinction and Metapopulation Dynamics. Oikos 83: 390-396. Hardin, J. W., and Hilbe, J. M. 2012. Generalized linear models and extensions. STATA Press, College Station, Texas, USA. Harris, G., Thirgood, S., Hapcraft, J.G.C., Cromsigt, J.P.G.M. and Berger, J. 2009. Global decline in aggregated migrations of large terrestrial mammals. Endangered Species Research 7: 5576. Hart, J. S., O. Heroux, W. H. Cottle, and Mills, C. A. 1961. The influence of climate on metabolic and thermal responses of infant caribou. Canadian Journal of Zoology 39: 845–856. Hatter, I.W. and Bergerud, W.A. 1991. Moose recruitment, adult mortality and rate of change. Alces 27: 65-73. Hillis, T.L., Mallory, F.F., Dalton, W.J. & Smiegielski, A.J. 1998. Preliminary analysis of habitat utilization by woodland caribou in northwestern Ontario using satellite telemetry. Rangifer Special Issue 10: 195-202. Horn, H .S. and Rubenstein, D. J. 1984. Behavioural adaptations and life history. In Behavioural ecology: an evolutionary approach. Edited by J.R. Krebbs and N.B. Davies. pp. 279-298. Hornseth, M., and Rempel, R. 2016. Seasonal resource selection functions of Woodland caribou (Rangifer tarandus caribou) across a gradient of anthropogenic disturbance: a land-use planning tool. Presentation at 16th North American Caribou Workshop, Thunder Bay, Ontario, Canada, 16-20 May 2016. Howat, I. M., A. Negrete, and Smith, B. E. 2014. The Greenland Ice Mapping Project (GIMP) land classification and surface elevation datasets The Cryosphere 8: 1509–1518, doi:10.5194/tc-8-1509-2014. Hughes, J., Albon, S. D. & Irvine, R. J. 2009. Is there a cost of parasites to caribou? Parasitology 136: 253–265. Hunt, L.M., J. Shuter, A.R. Rodgers, R. Kushneriuk, D.E.B. Reid, G.S. Brown, B.R. Patterson, I.D., Thompson, and J.M. Fryxell. 2016. Behavioural responses of Woodland caribou to vehicle traffic: a case of Pickle Lake, Ontario, Canada. Presentation at 16th North American Caribou Workshop, Thunder Bay, Ontario, Canada, 16-20 May 2016. Jakimchuk, R.D., Ferguson, S.H. & Sopuck, L.G. 1987. Differential habitat use and sexual segregation in the Central Arctic caribou herd. Canadian Journal of Zoology 65: 534-541. Jepsen, B. I., H. R. Siegismund, and Fredholm, M. 2002. Population genetics of the native caribou (Rangifer tarandus groenlandicus) and the semi-domestic reindeer (Rangifer tarandus tarandus) in Southwestern Greenland: Evidence for introgression. Conservation Genetics 3: 401–409. 70 Johnson, C. J., S. E. Nielsen, E. H. Merrill, T. L. Mcdonald, and Boyce, M. S. 2006. Resource selection functions based on use-availability data: theoretical motivation and evaluation methods. Journal of Wildlife Management 70(2): 347–57. Johnson, C. A., C. Nielsen, I. Naujokaitis-Lewis, E. Neave, J. Pasher, F. Schmiegelow, C. Malouin, D. Hervieux, M. Vance and Virc, S. 2016. Presentation at 16th North American Caribou Workshop, Thunder Bay, Ontario, Canada, 16-20 May 2016. Johnson, C. J., Seip, D. R. and Boyce, M. S. 2004. A quantitative approach to conservation planning: using resource selection functions to map the distribution of mountain caribou at multiple spatial scales. Journal of Applied Ecology, 41(2): 238-51. Joly, K. & Klein, D. R. 2011. Complexity of caribou population dynmaics in a changing climate. Alaska Park Science, 10(1): 26-31. Jorgensen, J. C., Udevitz, M. S. & Felix, N. A. 2002. Forage quality and quantity – In: Douglas, D.C., Reynolds, P.E. & Rhode, E.B. (eds.). Arctic Refuge coastal plain terrestrial wildlife research summaries. U.S. Geological Survey, Biological Resources Division. Biological Science Report USGS/BRD/BSR – 2002-0001, pp. 46-50. Kansas, J., Vargas, J., Skatter, H.G., Balicki, B. and McCallum, K. 2016. Using Landsat imagery to backcast fire and post-fire residuals in the Boreal Shield of Saskatchewan – implications for Woodland caribou management. Presentation at 16th North American Caribou Workshop, Thunder Bay, Ontario, Canada, 16-20 May 2016. Kellyhouse, R. A. 2001. Calving ground selection and fidelity: Teshekpuk and Western Arctic caribou herds. M.S. Thesis. University of Alaska, Fairbanks, Alaska, USA. 124 pp. Kelsall, J. P. 1968. The migratory barren-ground caribou of Canada. Queens’s Printer, Ottawa, Canada. Klein, D. R. 1990. Variation in quality of caribou and reindeer forage plants associated with season, plant part, and phenology. Rangifer Special Issue 3: 123–130. Latham, A. D. M., M. C. Latham, and M. S. Boyce. 2011. Habitat selection and spatial relationships of black bears (Ursus americanus) with woodland caribou (Rangifer tarandus caribou) in northeastern Alberta. Canadian Journal of Zoology 89(4): 267–77. Lent, P. C. 1966. Calving and related social behavior in the barren-ground caribou. University of Basutoland, Bechuanaland, and Swaziland, Roma, Basutoland. Zeitschrift für Tierpsychologie 6: 701–756. Lentz, C. P., and J. S. Hart. 1960. The effect of wind and moisture of heat loss through the fur of newborn caribou. Canadian Journal of Zoology 38: 679–688. Loe, L. E., C. Bonenfant, A. Mysterud, J. M. Gaillard, R. Langvatn, F. Klein, C. Calenge, T. Ergon, N. Pettorelli, and N. C. Stenseth. 2005. Climate predictability and breeding phenology in red deer: timing and synchrony of rutting and calving in Norway and France. Journal of Animal Ecology 74: 579–588. Long, R. A., J. G. Kie, R. T. Bowyer, and M. A. Hurley. 2009. Resource selection and movements by female mule deer Odocoileus hemionus: effects of reproductive stage. Wildlife Biology 15(3): 288–98. Lund, P. M., Bay, C. and Motzfeldt, K. G. 2004. Vegetation beskrivelse. In: Samspillet mellem rensdyr, vegetation og menneskelige aktiviteter i Vestgrønland (ed.) Aastrup P. 71 Pinngortitaleriffik – Greenland Institute of Natural Resources, Technical report No. 49. 320 pp. [in Danish] MacNearney, E., K. Pigeon, G. Stenhouse, W. Nijland, N.C. Coops, and L. Finnegan. 2016. Evaluating spatial distribution of Woodland caribou in response to a growing anthropogenic disturbance footprint. Presentation at 16th North American Caribou Workshop, Thunder Bay, Ontario, Canada, 16-20 May 2016. Mahoney, S. P. & Schaefer, J. A. 2002. Hydroelectric development and the disruption of migration in caribou. Biological Conservation, 107(2): 147-153. Maier, J. A. K. and White, R. G. 1998. Timing and synchrony of activity in caribou. Canadian Journal of Zoology, 76(11): 1999-2009. Mallory, M. L., A. J. Gaston, M. R. Forbes, and H. G. Gilchrist. 2009. Influence of weather on reproductive success of northern fulmars in the Canadian High Arctic. Polar Biology 32: 529–538. 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 for field studies. Second edition. Kluwer Academic Publisher, Dordrecht, Netherlands. 221 pp. Markussen, K. A., A. Rognmo, and A. S. Blix. 1985. Some aspects of thermoregulation in newborn reindeer calves (Rangifer tarandus tarandus). Acta Physiologica Scandinavica 123(2): 215–220. McEwan, E. H. and Whitehead, P.E. 1972. Reproduction in female reindeer and caribou. Canadian Journal of Zoology, 50(1): 43-46. Miller, F. L., and A. Gunn. 1986. Effect of adverse weather on neonatal caribou survival – a review. Rangifer, Special Issue No. 1: 211–217. Moen, J., R. Anderson, and A. Illius. 2006. Living in a seasonal environment. pp. 50–71 in K. Danell, R. Bergstrom, P. Duncan, and J. Pastor, editors. Large herbivore ecology, ecosystem dynamics and conservation. Cambridge University Press, Cambridge, United Kingdom. Mörschell, F. M. & Klein, D. R. 1997. Effects of weather and parasitic insects on behavior and group dynamics of caribou of the Delta Herd, Alaska. Canadian Journal of Zoology. 75: 16591670. Mueller, C., Herrero, S. and Gibeau, M. L. 2004. Distribution of subadult grizzly bears in relation to human development in the Bow River watershed, Alberta. Ursus, 15(1): 35–47. Nagy, J. A. S. 2011. Use of space by caribou in northern Canada. PhD Dissertation, University of Alberta, Edmonton, Alberta, Canada. Nagy, J. A. and Campbell, M.W. 2012. Herd structure, movement, calving grounds, activity periods, home range similarity, and behaviours of migratory and tundra-wintering barren-ground caribou on mainland Nunavut and eastern mainland Northwest Territories, Canada. Report prepared for the Beverly and Qamanirjuaq caribou management board. Canada. 155pp. Nagy, J. A., Johnson, D.L., Larter, N.C., Campbell, M.W., Derocher, A.E., Dumond, M., Allaire, D. and Croft, B. 2011. Subpopulation structure of cribou Rangifer tarandus L.. in Arctic and sub-Arctic Canada. Ecological Applications. 21 6.: 2334-48. 72 Nellemann, C., and R. Cameron. 1996. Effects of petroleum development on terrain preferences of calving caribou. Arctic 49: 23–28. Nellemann, C., and R. D. Cameron. 1998. Cumulative impacts of an evolving oil-field complex on the distribution of calving caribou. Canadian Journal of Zoology 76: 1425–1430. Nellemann, C., and M. G. Thomsen. 1994. Terrain ruggedness and caribou forage availability during snowmelt on the arctic coastal plain, Alaska. Arctic 47(4): 361–367. Nellemann, C., I. Vistnes, P. Jordhøy, O. Strand, and A. Newton. 2003. Progressive impact of piecemeal infrastructure development on wild reindeer. Biological Conservation 113: 307– 317. Oosenburg, S.M. & Theberge, J.B. 1980. Altitudinal movements and summer habitat preferences of woodland caribou in the Kluane Ranges, Yukon Territory. Arctic 33: 59-72. Parker, K. L., R. G. White, M. P. Gillingham, and D. F. Holleman. 1990. Comparison of energy metabolism in relation to daily activity and milk consumption by caribou and muskoxen neonates. Canadian Journal of Zoology 68(1): 106–114. Parrett, L. S. 2007. Summer ecology of the Teschekpuk caribou herd. MS thesis, University of Alaska Fairbanks, Alaska, USA. Popp, J. N., J. A. Schaefer, and F. F. Mallory. 2011. Female site fidelity of the Mealy Mountain caribou herd. Rangifer tarandus caribou in Labrador. Rangifer Special Issue 19: 87–95. Post, E. 2003. Timing of reproduction in large mammals. pp. 437-449 in M. D. Schwartz, editor. Phenology: An Integrative Environmental Science. Springer Netherlands, Netherlands. Post, E., Bøving, P. S., Pedersen, C. & MacArthur, M. 2003. Synchrony between caribou calving and plant phenology in depredated and non-depredated populations. Canadian Journal of Zoology 81: 1709–1714. Roby, D. 1977. Behavioral patterns of barren-ground caribou in the central arctic herd within the trans-Alaska pipeline corridor. MSc thesis, University of Alaska, Fairbanks. Ronson, A. 2016. Caribou range planning in Alberta: how caribou and protected areas planning intersects. Presentation at 16th North American Caribou Workshop, Thunder Bay, Ontario, Canada, 16-20 May 2016. Rowell, J. E. and Shipka, M. P. 2009. Variation in gestation length among captive reindeer (Rangifer tarandus tarandus). Theriogenology, 72(2) : 190-97. Russell, D. E, G. Kofinas, and B. Griffith. 2002. Barren-ground caribou calving grounds workshop report of proceedings. Environment Canada Technical Report Series No. 390. Ottawa, Canada. Russell D. E., Martell A. M. and Nixon W. A. C. 1993. Range ecology of the Porcupine caribou herd in Canada. Rangifer Special Issue 8: 168 pp. Schaefer, J.A., Bergman, C. M. & Luttich, S. N. 2000. Site fidelity of female caribou at multiple scales. Landscape Ecology 15: 731-739. Schwarz, G. 1978. Estimating the dimension of a model. Annals of Statistics 6: 461–64. Segal, A.N. 1962. The periodicity of pasture and physiological functions of reindeer. In Reindeer in the Karelian ASSR, pp. 130-150. Petrozavodsk: Akademie Nauka (Translated by Department of the Secretary of State Bureau for Translation, Ottawa, Canada.) 73 Simonsen, C. E. 2011. Seasonal distribution patterns of female caribou (Rangifer tarandus) in West Greenland – insights from species distribution models using MAXENT. MSc thesis, University of Copenhagen, Copenhagen, Denmark, 84 pp. Singh, N.J., Grachev, I.A., Bekenov, A.B. and Milner-Gulland, E.J. 2010. Saiga antelope calving site selection is increasingly driven by human disturbance. Biological Conservation 143: 1770-1779. Skogland, T. 1986. Movements of tagged and radio-instrumented wild reindeer in relation to habitat alteration in the Snøhetta region, Norway. Rangifer Special Issue 1: 267-272. Skogland, T. 1989. Comparative social organisation of wild reindeer in relation to food, mates and predator avoidance. Advances in Ethology 29: 1-74. Smith, W. T. & Cameron, R. D. 1983. Responses of caribou to industrial development on Alaska’s Arctic Slope. Acta Zoologica Fennica 175: 43–45. Strandgaard, H., Holthe, V., Lassen, P. & Thing, H. 1983. Rensdyrundersøgelser I Vestgrønland 1977-82. Report. Vildt biologisk Station, Kalø, DK-8410 Rønde, Denmark. [In Danish.] Strong, D. R., Whipple, A. V., Child, A. L. & Dennis, B. 1999. Model selection for a subterranean trophic cascade: root-feeding caterpillars and entomopathogenic nematodes. Ecology, 80(8): 2750-61. Taillon, J., Barboza, P. S. & Cote, S. D. 2013. Nitrogen allocation to offspring and milk production in a capital breeder. Ecology 94(8): 1815–1827. Taillon, J., Festa-Bianchet, M. & Côté, S. 2012. Shifting targets in the tundra: Protection of migratory caribou calving grounds must account for spatial changes over time. Biological Conservation. 147: 163–173. Tamstorf, M. P., Aastrup, P. & Cuyler, L. C. 2005. Modeling critical caribou summer ranges in West Greenland based on ARGOS and Landsat data. Polar Biology 28:714–724. DOI: 10.1007/s00300-005-0731-8 Tamstorf, M. P. 2004. Satellitbaseret vegetationskortlægning I Vestgrønlnad. In: Samspillet mellem rensdyr, vegetation og menneskelige aktiviteter i Vestgrønland (ed.) Aastrup P. Pinngortitaloriffik – Greenland Institute of Natural Resources. Technical Report No. 49, Nuuk, Greenland. 320 pp. [In Danish] Thing, H. 1984. Feeding ecology of the West Greenland caribou (Rangifer tarandus groenlandicus) in the Sisimiut-Kangerlussuaq Region. Danish Review of Game Biology 12(3):53 pp. Thing, E. M. & Thing, H. 1983. Cow-calf behaviour in West Greenland caribou on the Sdr. Strømfjord summer range. Acta Zool. Fennica 175: 113-115. Thomson, B.R. 1971. Wild reindeer activity. In Report from the grazing project of the Norwegian IBP committee, July-December 1970, Hardangervidda. Viltundersøkelser, Trondheim, Norway. Tveraa, T., A. Stien, B. J. Bardsen, & P. Fauchald. 2013. Population densities, vegetation greenup, and plant productivity: impacts on reproductive success and juvenile body mass in reindeer. Plos One 8(2), e56450. (doi:10.1371/journal.pone.0056450). Tøttrup, C. 2009. Vegetation and snow mapping of area between Maniitsoq Ice Cap and Godthåbsfjorden. Geographic Resource Analysis and Science Ltd, Institute of Geography, University of Copenhagen Technical report to NERI/Alcoa. 27pp. www.gras.ku.dk 80 Table 19. Post-calving caribou resource selection models Ten best-fit post-calving activity period resource selection models SBIC ∆SBIC Relative preference for best-fit model based on ∆SBIC* elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -136853.8 0 Best-fit model elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -136842.2 12 Very strong elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock -136783.2 71 Very strong elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment -136780.9 73 Very strong heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -136779.9 74 Very strong heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -136778.0 76 Very strong elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed -136775.6 78 Very strong heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -136770.6 83 Very strong heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -136768.5 85 Very strong elevation + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -136766.0 88 Very strong *Relative preference for best-fit model based on Hardin & Hilbe (2012). 81 Table 20. Early summer caribou resource selection models Ten best-fit early summer activity period resource selection models SBIC ∆SBIC Relative preference for best-fit model based on ∆SBIC* elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -140048.1 0 Best-fit model elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -140036.7 11 Very strong elevation + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -140012.2 36 Very strong elevation + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -140005.7 42 Very strong elevation + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -140000.1 48 Very strong heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -139994.3 54 Very strong elevation + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -139992.6 56 Very strong heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -139992.2 56 Very strong copse + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -139987.6 61 Very strong heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -139985.1 63 Very strong *Relative preference for best-fit model based on Hardin & Hilbe (2012). 82 Table 21. Mid /late summer caribou resource selection models Ten best-fit mid/late summer activity period resource selection models SBIC ∆SBIC Relative preference for best-fit model based on ∆SBIC* elevation + heath + open + heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -140439.1 0 Best-fit model heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -140410.6 29 Very strong elevation + heath + open + heath + copse -140408.5 31 Very strong elevation + heath + open + heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -140406.6 33 Very strong elevation + heath + open + heath + copse + fen + grass -140406.2 33 Very strong heath + open + heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -140402.4 37 Very strong elevation + heath + open + heath + copse + fen -140401.9 37 Very strong elevation + heath + open + heath + copse + fen + grass + snow-bed -140398.6 41 Very strong elevation + heath + open + heath + copse + fen + grass + snow-bed + wind-exposed -140398.0 41 Very strong heath + copse + fen + grass -140391.4 48 Very strong *Relative preference for best-fit model based on Hardin & Hilbe (2012). 83 Table 22. Fall /pre-breeding caribou resource selection models Ten best-fit fall/pre-breeding activity period resource selection models SBIC ∆SBIC Relative preference for best-fit model based on ∆SBIC* heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -137204.7 0 Best-fit model heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -137198.2 7 Strong heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -137196.2 9 Strong heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -137190.5 14 Very strong elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -137188.8 16 Very strong heath + copse + fen + grass -137188.4 16 Very strong heath + open-heath + copse + fen + grass -137187.4 17 Very strong heath + copse + fen + grass + snow-bed -137184.4 20 Very strong heath + open-heath + copse + fen + grass + snow-bed -137183.2 22 Very strong elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -137180.9 24 Very strong *Relative preference for best-fit model based on Hardin & Hilbe (2012). 84 Table 23. Breeding (rut) caribou resource selection models Ten best-fit breeding activity period resource selection models SBIC ∆SBIC Relative preference for best-fit model based on ∆SBIC* elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -143218.3 0 Best-fit model elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -143213.1 5 Positive elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed -143126.4 92 Very strong elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock -143118.2 100 Very strong elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment -143117.2 101 Very strong elevation + heath + open-heath + copse + fen + grass + snow-bed -143111.7 107 Very strong elevation + heath + open-heath + copse + fen + grass -143093.9 124 Very strong heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -143043.9 174 Very strong heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -143034.5 184 Very strong elevation + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -143034.0 184 Very strong *Relative preference for best-fit model based on Hardin & Hilbe (2012). 85 Table 24. Post-breeding /late fall caribou resource selection models Ten best-fit post-breeding/late fall activity period resource selection models SBIC ∆SBIC Relative preference for best-fit model based on ∆SBIC* elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -154203.3 0 Best-fit model elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -154196.2 7 Strong elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment -154135.9 67 Very strong elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock -154128.7 75 Very strong elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed -154081.3 122 Very strong elevation + heath + open-heath + copse + fen + grass + snow-bed -154079.6 124 Very strong elevation + heath + open-heath + copse + fen + grass -154048.3 155 Very strong elevation + heath + open-heath + copse + fen -153983.9 219 Very strong elevation + heath + open-heath + copse -153957.5 246 Very strong elevation + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -153946.3 257 Very strong *Relative preference for best-fit model based on Hardin & Hilbe (2012). 86 Table 25. Early/mid-winter activity period caribou resource selection models Ten best-fit early/mid-winter activity period resource selection models SBIC ∆SBIC Relative preference for best-fit model based on ∆SBIC* elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -144092.0 0 Best-fit model elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -144085.8 6 Positive elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock -144036.5 56 Very strong elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment -144031.9 60 Very strong elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed -144021.2 71 Very strong elevation + heath + open-heath + copse + fen + grass + snow-bed -143994.7 97 Very strong elevation + heath + open-heath + copse + fen + grass -143983.1 109 Very strong elevation + heath + open-heath + copse -143973.9 118 Very strong elevation + heath + open-heath + copse + fen -143971.0 121 Very strong elevation + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -143962.3 130 Very strong *Relative preference for best-fit model based on Hardin & Hilbe (2012). 87 Table 26. Late winter activity period caribou resource selection models Ten best-fit late winter activity period resource selection models SBIC ∆SBIC Relative preference for best-fit model based on ∆SBIC* elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -139452.9 0 Best-fit model elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -139451.2 2 Weak elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed -139417.2 36 Very strong elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock -139413.4 40 Very strong elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment -139411.0 42 Very strong elevation + heath + open-heath + copse + fen + grass + snow-bed -139402.4 51 Very strong elevation + heath + open-heath + copse + fen + grass -139387.0 66 Very strong elevation + heath + open-heath + copse + fen -139377.5 75 Very strong elevation + heath + open-heath + copse -139370.1 83 Very strong elevation + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -139345.7 107 Very strong *Relative preference for best-fit model based on Hardin & Hilbe (2012). 88 Table 27. Pre-calving activity period caribou resource selection models Ten best-fit pre-calving activity period resource selection models SBIC ∆SBIC Relative preference for best-fit model based on ∆SBIC* elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -137443.7 0 Best-fit model elevation + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water -137442.2 2 Weak elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -137440.7 3 Positive elevation + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -137439.6 4 Positive elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock -137431.4 12 Very strong elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment -137425.4 18 Very strong elevation + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock -137424.8 19 Very strong elevation + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment -137418.6 25 Very strong elevation + heath + open-heath + copse + fen + grass + snow-bed + wind-exposed -137406.5 37 Very strong heath + open-heath + copse + fen + grass + snow-bed + wind-exposed + soil/rock + sediment + water + snow/ice -137406.3 37 Very strong *Relative preference for best-fit model based on Hardin & Hilbe (2012). 89 Appendix 3 Habitat selection by Greenland caribou (Rangifer tarandus groenlandicus) during the CALVING period relative to those selected during the other activity periods, based on latent selection difference (LSD) function model comparisons, 2008-2010. Table 28. Calving VERSUS post-calving, early summer, and mid /late summer activity periods Habitat types Calving versus post-calving Calving versus early summer Calving versus mid /late summer Coefficient S.E. P-value RS* Coefficient S.E. P-value RS Coefficient S.E. P-value RS heath -0.301 0.094 0.001 26 0.029 0.084 0.732 = -0.661 0.094 0.000 48 openheath 0.520 0.098 0.000 >100 0.858 0.084 0.000 >100 0.478 0.085 0.000 >100 copse 0.059 0.103 0.567 = 0.112 0.094 0.233 = -0.004 0.085 0.964 = fen -0.130 0.100 0.192 = -0.200 0.087 0.022 18 -0.093 0.090 0.302 = grass 0.120 0.076 0.113 = 0.145 0.070 0.039 >100 0.134 0.071 0.059 = snowbed -0.366 0.170 0.032 31 -0.429 0.143 0.003 = -0.488 0.141 0.001 39 windexp -0.003 0.097 0.974 = 0.378 0.087 0.000 >100 0.842 0.085 0.000 >100 soilrock 0.042 0.084 0.616 = -0.102 0.077 0.188 = -0.355 0.081 0.000 30 sediment -0.170 0.121 0.158 = -0.372 0.099 0.000 31 -0.328 0.103 0.001 28 water 0.310 0.139 0.026 >100 0.065 0.115 0.572 = -0.092 0.116 0.428 = snowice 0.889 0.784 0.257 = 0.297 0.466 0.524 = - - - - _cons 0.700 0.133 0.000 -0.230 0.117 0.050 0.483 0.121 0.000 *Relative selection (RS) was calculated for variables with coefficients significantly different from 0 as exp (b) when b>0 and as [1-exp(b) when b<0 (Latham, Latham & Boyce, 2011b). Relative selection values <100 indicate that use of the habitat during calving was significantly less by x% than that for the season being compared, values >100 indicate use of habitat during calving was significantly greater than that for the season being compared, and = indicates habitat use during calving and season being compared were not significantly different.