Northeastern Naturalist C. Goulet, J.A. Litvaitis, and M.N. Marchand 2015 Vol. 22, No. 3 530 2015 NORTHEASTERN NATURALIST 22(3):530–540 Habitat Associations of the Eastern Hognose Snake at the Northern Edge of its Geographic Distribution: Should a Remnant Population Guide Restoration? Celine Goulet1, John A. Litvaitis1,*, and Michael N. Marchand2 Abstract - Heterodon platirhinos (Eastern Hognose Snake [EHS]) is a species of conservation concern in the northeastern United States. As an initial step toward potential restoration, we examined habitat associations of a peripheral population of EHS in New Hampshire. At the landscape scale, transmitter-equipped snakes were found most often in developed lands, followed by, in order of frequency, mixed forest, Pinus strobus/P. resinosa (Eastern White/ Red Pine) stands, and 7 other cover types. Within individual home ranges, snakes selected Tsuga canadensis (Eastern Hemlock) stands most often, followed by, in order of frequency, White/Red Pine stands, mixed forest, Fagus grandifolia/Quercus spp. (American Beech/ oak) stands, and 6 other cover types. Compared to random locations, microhabitat features at snake activity sites included higher ground-surface temperatures, closer proximity to wetlands, less canopy closure, and more abundant shrubs, ground debris, and rock cover. When combined with a previous study conducted in the same area, we found that cover-type associations (mesic forest) of this population differed from known affinities (open, xeric habitats) of EHS throughout much of its geographic distribution. Home ranges were also larger than those reported in most studies. We suspect our population persists because it occurs in a large parcel of land with limited human alteration and use. Habitat there is suitable but likely is not optimal. Such limitations should be considered when selecting sites to establish new populations of EHS in northern regions. Introduction Heterodon platirhinos Latreille (Eastern Hognose Snake [EHS]) is widely distributed throughout the eastern United States and southern Ontario, Canada (Ernst and Ernst 2003). However, populations are declining and are of regional conservation concern (Therres 1999). These declines may be especially evident along the periphery of its distribution. In New Hampshire, EHS is at the northeastern extent of its geographic distribution (Michener and Lazell 1989) and is listed as endangered (New Hampshire Endangered Species Conservation Act RSA 212-A, New Hampshire Fish and Game Threatened and Endangered Wildlife List Administrative Rule FIS 1000). Throughout much of its distribution, EHS is typically associated with dry, open areas, often with sandy soils (Michener and Lazell 1989, Platt 1969, Plummer and Mills 2000, Robson 2011). Selection of xeric habitats may be driven by availability of critical resources, such as prey, especially Anaxyrus spp. (toads), or refugia. Alternatively, the thermal environment may be the primary determinant that influences 1Department of Natural Resources and the Environment, University of New Hampshire, Durham, NH 03824. 2Nongame and Endangered Wildlife Program, New Hampshire Fish and Game Department, Concord, NH 03301. *Corresponding author - john@unh.edu. Manuscript Editor: Rudolf G. Arndt Northeastern Naturalist 531 C. Goulet, J.A. Litvaitis, and M.N. Marchand 2015 Vol. 22, No. 3 habitat selection because the snakes seek to maintain a body temperature within the range of 29.0–32.0 °C (Platt 1969). Preferential selection of habitats in response to thermal quality is well documented among reptile populations, especially at their temperature extremes (Blouin-Demers and Weatherhead 2001, Diaz 1997, Harvey and Weatherhead 2006). To understand the factors limiting EHS at its northern distributional extent, LaGory et al. (2009) examined habitat selection in a remnant population in southcentral New Hampshire. Using radio telemetry, they found that old field and forest edges with sandy loam soils were preferred at the landscape scale (second-order selection; Johnson 1980). Snakes did not, however, exhibit selection within home ranges (third-order selection), apparently because preferred macrohabitat features (cover type and soil characteristics) were well distributed throughout the study area. Building upon the work of LaGory et al. (2009), we conducted a more detailed evaluation of habitat associations by EHS in the same study area. Our aim was to identify additional features that could be used to manage remaining populations and to identify sites on which to establish new populations through reintroduction. Thus, we examined microhabitat features associated with individual snake locations (activity site, or fourth–order selection) in addition to the characteristics of habitat selection at the landscape and home-range scales. Study Area Our study was conducted on the 1144-ha New Boston Air Force Station (NBAFS) in Hillsborough County, south-central New Hampshire—the site used by LaGory et al. (2009). Dominant land-cover types include mixed (40.2%) and deciduous (29.3%) forests managed with selective harvests and small clearcuts (LaGory et al. 2009). Topography is hilly, and elevations range from 104 to 389 m above mean sea level. Canton series fine sandy loam is dominant throughout much of the site. The regional climate is humid continental, with an average annual temperature of 8.0 ºC and monthly averages of -5.2 ºC in January to 21.0 ºC in July (LaGory et al. 2009). Methods Capture, monitoring, and home-range estimation We captured snakes by hand and monitored them from mid-April to October 2008. We did not continue our surveys into the hibernation season due to site-access restrictions. Snakes were weighed and assigned to an age class: <10 g = hatchlings, 10–100 g = juveniles, and >100 g = adults). A veterinarian surgically implanted temperature-sensitive transmitters (Advanced Telemetry Systems, Model F1820t) into all adult, non-gravid snakes captured (procedures in Reinert and Cundall 1982). To minimize the risk of infection arising from an additional surgery (Sperry et al. 2009), we did not recover transmitters, especially since it would have been performed when snakes were about to enter hibernacula. Any reduction in body condition would certainly decrease their chance of over-winter survival. Snakes were handled in accordance with rules of the University of New Hampshire Institutional Animal Care and Use Committee (Protocol 070403). Northeastern Naturalist C. Goulet, J.A. Litvaitis, and M.N. Marchand 2015 Vol. 22, No. 3 532 We located radio-tagged snakes at approximately 2-day intervals using a portable receiver (Communications Specialists [CS], Model R-1000) equipped with a threeelement Yagi antenna (CS, Model RA-150). Locations were determined by homing, a non-triangulation technique where the transmitted signal is followed until the animal is observed (White and Garrott 1990). This procedure facilitated measurements of surface temperature and solar radiation at capture sites. We recorded behavior (active = on ground surface, or inactive = in retreat or not visible after extensive searching of ground surface) and geographic coordinates at each location. We used minimum-convex polygons (Mohr 1947) and 95% isopleths of fixedkernel density estimates with least-squares cross validation (Seaman and Powell 1996) to delineate home ranges with Hawth’s Tools Extension in ArcGis 9.0 (Environmental Systems Research Institute, Inc., Redlands, CA). Consistent with LaGory et al. (2009), estimates were limited to snakes that were tracked for >50 days and that had >15 locations. Habitat selection Landscape and home-range scales. We characterized habitats at the landscape and home-range scales using the digital New Hampshire Land Cover Assessment data layer (available from http://www.granit.unh.edu) and reclassified land-cover classes to produce a layer consisting of 9 cover types: (1) developed lands (e.g., sites with buildings and paved roads), (2) mowed fields, (3) Fagus grandifolia Erhr. (American Beech)/Quercus spp. (oak) hardwoods, (4) Betula spp. (birch)/Populus spp. (aspen) stands, (5) Pinus strobus L. (Eastern White Pine)/P. resinosa Sol. ex Aiton (Red Pine) stands, (6) Tsuga canadensis (L.) Carrière (Eastern Hemlock) stands, (7) Picea spp. (spruce)/Abies spp. (fir) stands, (8) mixed forest (even mix of conifer and deciduous trees), and (9) cleared lands (early-successional areas dominated by grasses and forbs). We maintained species assemblages, rather than combine cover types as did LaGory et al. (2009), to facilitate management recommendations. The final land-cover layer was then used in habitat-selection analyses. We used compositional analysis to analyze cover-type use versus availability (Aebischer et al. 1993). The advantage of this approach is that individual snakes are the sample units, and it has been widely applied in studies of snake-habitat associations (e.g., DeGregorio et al. 2011, Moore and Gillingham 2006, Waldron et al. 2008). At the landscape scale, available habitat was the same for all snakes, and it was delineated as the study area (because snakes were essentially found throughout the area) plus any portion of a 95% fixed-kernel home range that extended beyond the perimeter of NBAFS. Use was delineated by the composition of individual home ranges. At the home-range scale, availability was measured as the proportion of each cover type within an individual’s 95% fixed-kernel home range, and use was the percentage of radio locations within each cover type (McLoughlin et al. 2004). Available cover types that had no evidence of use were assigned a small, non-zero value (0.003; Bingham and Brennan 2004). We performed log-ratio transformations separately for the matrices of used and available data to remove linear dependency (Pendleton et al. 1998). Nonrandom habitat use was evaluated using a Wilks’ lambda (Λ) test statistic. With the occurrence of habitat selection (nonrandom use), we averaged differences between log-ratios across animals to obtain Northeastern Naturalist 533 C. Goulet, J.A. Litvaitis, and M.N. Marchand 2015 Vol. 22, No. 3 Table 1. Variables sampled within 5-m radius plots centered on Eastern Hognose Snake activity sites and an equivalent number of random locations at New Boston Air Force Station, NH, 2008. Variable Description Method of measurement ASPECT Aspect Calculated using Arc View Spatial Analyst CANOPY Percent canopy closure Visually estimated closure utilizing paper cylinder DEBRIS Percent downed woody debris Visually estimated coverage ELEVLAND Elevation at 5-m resolution Calculated using Arc View Spatial Analyst HERBDENS Percent herbaceous cover Visually estimated LAI Leaf Area Index Hemispherical photos taken at center of plot and analyzed using GLA software (Frazer et al. 1999) LEAFDENS Percent leaf litter coverage Visually estimated LEAFDEPTH Leaf litter depth Measured depth at center of plot OVERDBH DBH of nearest overstory tree Diameter at breast height (DBH) measured with diameter tape OVERDIST Distance to nearest overstory Measured from center of plot to closest point of tree tree up to 20 m POND Distance to nearest pond Calculated using Arc View Spatial Analyst RETREAT Distance to nearest retreat Measured from center of plot to closest point of retreat site up to 20 m ROCKDENS Percent rock coverage Visually estimated SHRUBIST Distance to nearest shrub Measured from center of plot to closest point of shrub up to 20 m SLOPE Degree of slope Visually estimated SOILDENS Percent bare soil coverage Visually estimated SOLRAD Solar radiation Hemispherical photos taken at center of plot and analyzed using GLA software (Frazer et al. 1999) STUMPDENS Percent stumps coverage Visually estimated SURFTEMP Surface temperature Measured snake locations with digital hygrometer UNDERDBH Diameter at breast height of Measured using diameter tape nearest understory tree UNDERDIST Distance to nearest understory Measured from center of plot to closest point of tree tree up to 20 m 1Overstory is defined as vegetation with a DBH of ≥7.5 cm. 2Understory is defined as vegetation <2 m in height with a DBH of <7.5 cm a mean for each cover type. We then created ranking matrices to assess relative preferences (Johnson 1980). We considered cover types with use-availability ratios significantly greater than 0 as preferred and those with significantly less than 0 as avoided. Paired t-tests were then used to rank habitats by relative use (Pendleton et al.1998). We limited selection analyses to snakes that were tracked for >50 days and having >15 locations. Microhabitat selection. Habitat at snake locations (activity sites) was described using 21 structural, physical, and climatic features (Table 1). We conducted sampling in 5-m radius plots centered on locations occupied by a snake within two Northeastern Naturalist C. Goulet, J.A. Litvaitis, and M.N. Marchand 2015 Vol. 22, No. 3 534 weeks of the snake vacating the site. For each location, we identified an associated random site by moving in a pre-determined random direction and distance up to 50 m (95% confidence interval of the mean daily movement). Within 15 min of locating a snake, we surveyed the associated random site and measured the same variables there. Using individual locations as the sample unit at this scale creates the problem of pseudoreplication. However, it is difficult to avoid this approach when data sets are limited (e.g., Robson 2011, Steen et al. 2010). Prior to analysis, we examined Pearson correlation coefficients to minimize multicolinearity among variables and removed from consideration 1 variable of highly correlated pairs (r > 0.7). We employed multivariate analysis of variance (MANOVA) to test if used and random sites differed. We then conducted stepwise discriminant analysis (DA) to examine the differences among group centroids between occupied and random locations (Blouin-Demers and Weatherhead 2001). Variables were included in the DA model based on univariate ANOVAs, with order of entry corresponding to the relative discriminating ability of individual covariates. Significance of the final model was determined by a Wilks’ lambda (Λ) test statistic. We repeated analyses using all combinations of removed variables to compare discriminate-function significance and model performance and examined linear correlations between variables and resultant discriminant function for biological meaning. Results Capture, monitoring, and home-range estimation We recorded 179 locations for 7 snakes (3M, 4F). Monitoring periods ranged from 9 to 231 days (Table 2). Of these snakes, 5 were first caught during this study period, and 2 were recaptures from a pilot study conducted the previous year (Goulet 2010). Sample size is admittedly small, and this limitation was exaggerated by the occurrence of a transmitter failure or predation for 2 of the snakes (Table 2). We calculated home-range size for 5 snakes. Minimum-convex polygon home ranges averaged 72.7 ha (range = 28.7–128.6 ha ± 35.25) and 95% fixed-kernel home ranges averaged 282.8 ha (range = 179.2–588.3 ha ± 172.26). Table 2. Description of transmitter-equipped adult Eastern Hognose Snakes captured at New Boston Air Force Station, NH, 2008. Monitoring period refers to the da te of the first and last location. Snake Sex Weight (g) Monitoring period Locations Fate H015 F 440 24 Apr–5 May 3 Transmitter failure H026 M 470 24 Apr–11 Sep 45 Monitored full access period1 H040 F 390 22 May–8 Oct 42 Monitored full access period H041 F 450 26 May–7 Jul 14 Lost to predation H042 F 700 1 Jun–2 Sep 15 Monitored full access period H043 M 360 17 Jun–23 Sep 34 Monitored full access period H045 M 200 15 Jul–11 Sep 26 Monitored full access period 1Access to study area was restricted during some periods. Northeastern Naturalist 535 C. Goulet, J.A. Litvaitis, and M.N. Marchand 2015 Vol. 22, No. 3 Habitat selection Habitat use at the landscape scale was non-random (Λ = 3.21-17, P < 0.01), with snakes found most often in developed lands, followed by mixed forest and then Eastern White/Red Pine stands (Table 3), with occurrences in the other habitats to a lesser extent. Habitat use at the home-range scale also was non-random (Λ = 0.23, P < 0.01). The most preferred habitats at this spatial scale were Eastern Hemlock stands, followed in order of preference by Eastern White/Red Pine stands, mixed forest, and beech/oak stands (Table 4). Among activity sites (n = 179), aspect, percent canopy closure, leaf-area index, distance to nearest retreat, and distance to nearest understory tree were highly correlated with other variables (elevation [r = 0.97], solar radiation [r = 0.97], and distance to nearest overstory tree [r = 0.90]) and thus were removed from the analyses. Occupied and random sites differed (F = 3.62, P < 0.01). The DA defined a single discriminant function (eigenvalue = 0.17, Λ = 0.85, P less than 0.01) that accounted for variation between occupied and random sites. The most Table 4. Habitat selection at the home-range scale by transmitter-equipped adult Eastern Hognose Snakes at New Boston Air Force Station, NH, 2008. Selection patterns were based on compositional analysis and cover types not sharing a common letter are considered to have a difference in preference rank (P < 0.05). The lower the rank number, the more preferred the cover type. Habitat Rank Rank-order difference Hemlock 1 A White/red pine 2 A B Mixed forest 3 A B C Beech/oak/hardwoods 4 A B C Cleared lands 5 B C Disturbed lands 6 B C Spruce/fir 7 B C Developed lands 8 B C Fields 9 B C Birch/aspen 10 C Table 3. Habitat selection at the landscape scale by transmitter-equipped adult Eastern Hognose Snakes at New Boston Air Force Station, NH, 2008. Selection patterns were based on compositional analysis and cover types not sharing a common letter are considered to have a difference in preference rank (P < 0.05). The lower the rank number, the more preferred the cover type. Habitat Rank Rank-order difference Developed lands 1 A Mixed forest 2 A B White/red pine 3 A B C Hemlock 4 A B C D Fields 5 A B C D Birch/ aspen 6 A B C D Spruce/ fir 7 B C D Cleared lands 8 C D Disturbed lands 9 C D Beech/oak/hardwoods 10 D Northeastern Naturalist C. Goulet, J.A. Litvaitis, and M.N. Marchand 2015 Vol. 22, No. 3 536 parsimonious model (correct classification rate = 64.0%) included solar radiation, distance to nearest shrub, rock and debris cover, ground-surface temperature, and distance to nearest wetland (Fig. 1). Positive correlations between original variables and the discriminant function were detected for rock cover (r = 0.26), debris cover (r = 0.22), solar radiation (r = 0.10), and ground-surface temperature (r = 0.10), whereas distance to nearest shrub (r = -0.50) and distance to nearest wetland (r = -0.10) contributed negatively to group separation. Discussion Snakes in our study demonstrated nonrandom habitat selection across spatial scales. Use of cover types varied between landscape and home-range scales. At the landscape scale, developed lands were used disproportionately to their availability, whereas Eastern Hemlock stands were preferred at the home-range scale. These seemingly divergent results, however, may have been a selective response to common structural components within habitats at each scale, such as the availability of thermal resources and refugia, rather than to the vegetation composition by itself. For example, developed lands on the NBAFS are characterized by open and edge habitat as well as a high frequency of surfaces with greater thermal Figure 1. Group centroids on a single discriminant axis describing activity sites occupied by H. platirhinos (Eastern Hognose Snake) versus random locations at New Boston Air Station, NH, 2008. The illustrated gradient depicts habitat structure progressing from random sites (e.g., dense canopy closure, low ground cover, and reduced solar radiation) towards occupied sites (e.g., low canopy closure; dense shrub, debris, and rock cover; and high surface temperature). Northeastern Naturalist 537 C. Goulet, J.A. Litvaitis, and M.N. Marchand 2015 Vol. 22, No. 3 conductivity (e.g., pavement, concrete, rock, gravel, wood and mulch piles, and surface debris). Similarly, the hemlock forests contain small-scale timber harvests and canopy gaps associated with rocky outcrops. Such features within both habitat types thus provide a range of thermoregulatory opportunities in addition to proximity to protective cover. This pattern of selection is consistent with populations of EHS in Ontario, Canada (Robson 2011), as well as in the southeastern United States (Steen et al. 2012) where EHS preferred a range of human-modified habitats in response to the thermal environment. Our results are in contrast to those of LaGory et al. (2009) despite both studies having been conducted in the same population. They found snakes to prefer old fields at the landscape scale and detected no differential selection at the home-range scale. Conflicting results could be due to variation in land-cover characterization or home-range estimation. For example, snakes in our study had larger home ranges than snakes monitored by LaGory et al. (2009), and that size difference could have altered ratios of used vs. available habitats, thereby affecting the cover types identified as preferred. Furthermore, our GIS land-cover datasets were classified into 10 categories whereas LaGory et al. (2009) used 5. Determining habitat types used based on fine versus broad land-cover characterizations may have masked similar patterns of habitat selection in that EHS may have been utilizing the same habitat types or responding to the same variables (e.g., open habitat or thermoregulatory resources), but the community-type assignment as such differed. Alternatively, the divergent selective patterns reported here could instead be real and have been driven by annual variation in climate or management practices resulting in variation in the distribution of critical resources (e.g., basking and shelter sites) or by sampling error. Selection based on thermoregulatory and refugia resources is supported by the variables which were found to dictate selection of microhabitats. Used sites had both small-scale canopy openings that allowed much of the transmitted light to penetrate to the ground surface, as well as a high availability of shelter in the form of shrubs, leaf litter, rocks, and debris (Fig. 1). Combined, these features enabled snakes to either increase body temperature through basking or to avoid overheating or predation by utilizing a retreat (Cunnington et al. 2008). Additionally, activity sites were in proximity to wetlands and thus to prey (e.g., toads and frogs), suggesting that prey distribution may also have influenced habitat selection at this scale. Our results indicate that the habitat-selection process of EHS is influenced by the availability of basking and shelter sites. Evolutionarily, these requirements were fulfilled by occupying xeric habitats (Platt 1969, Plummer and Mills 2000, Robson 2011). However, with the modification and degradation of these communities, availability of natural basking and shelter resources has become increasingly rare, perhaps forcing EHS populations to utilize anthropomorphic surrogates. This interpretation is supported by the association of EHS with human-modified habitats (e.g., clear-cuts, road-sides, tree plantations) that have limited canopy closure and relatively high ground-surface temperatures. In utilizing such habitats, snakes may suffer higher rates of mortality not only by way of predation by pets and road Northeastern Naturalist C. Goulet, J.A. Litvaitis, and M.N. Marchand 2015 Vol. 22, No. 3 538 mortality but also through intentional killing by humans who view them as either threats or pests (Robson 2011). As the distribution of a species is reduced, information on habitat requirements often is restricted to remnant, disjunct, or peripheral populations (e.g., Tash and Litvaitis 2007). The present distribution of EHS in New Hampshire likely reflects the contraction and fragmentation of a more continuous historic range (Michener and Lazell 1989). The factors contributing to that decline (e.g., habitat degradation, fire suppression, and killing by humans) probably did not spread uniformly through the region. Instead landscapes more conducive to development (e.g., sandy soils), as in the Merrimack Valley, have been substantially more affected than the rural landscape that contains NBAFS. As predicted by the contagion hypothesis that states the last population to be affected by extinction forces will persist the longest (Channell and Lomolino 2000), we suggest that the abundance of EHS at NBAFS compared to other historically occupied habitats in New Hampshire may be a consequence of limited development and restricted public access more than of high-quality habitats present at NBAFS. Supporting that notion is the utilization of habitat types (e.g., hemlock forest) that are atypical for EHS and thus may reflect less-than-optimal conditions for it. Fortunately, given that snakes were indeed capable of locating appropriate environments in a forest-dominated landscape by using clearings and canopy gaps, small-scale modifications to existing habitat may enable the NBAFS population, and potentially other populations, to persist and expand. In contrast to Waldron et al. (2008), we recommend caution in using information from a remnant population such as this one as a basis for reintroductions. Revitalizing historically occupied habitats in the Merrimack Valley may prove to be more productive. This approach may also enhance connection with populations to the south. However, existing land uses, and road networks in particular (Rouse et al. 2011, Thomasson 2012, Xuereb 2012), in the Merrimack Valley will present challenges. Thus, if historically occupied sites are not conducive to restoration, our findings could prove useful to generate suitable habitats in other mesic forest landscapes. Acknowledgments We thank personnel of the New Boston Air Force Station, especially Stephen Najjar, for assistance in all phases of this study. Dr. Michael Dutton and his staff performed transmitter implantations. Tom Lee provided constructive input during project development. S. Buchanan and 3 anonymous reviewers provided helpful comments on drafts of this paper. New Hampshire Fish and Game Department and the College of Life Sciences and Agriculture, University of New Hampshire provided financial support. Literature Cited Aebischer, N.J., P.A. Robertson, and R.E. Kenward. 1993. Compositional analysis of habitat use from animal radio-tracking data. Ecology 74:1313–1325. Bingham, R.L., and L.A. Brennan. 2004. Comparison of type I error rates for statistical analyses of resource selection. Journal of Wildlife Management 68:206–212. Northeastern Naturalist 539 C. Goulet, J.A. Litvaitis, and M.N. Marchand 2015 Vol. 22, No. 3 Blouin-Demers, G., and P.J. Weatherhead. 2001. Habitat use by Black Rat Snakes (Elaphe obsoleta obsoleta) in fragmented forests. Ecology 82:2882–2896. Channell, R., and M.V. Lomolino. 2000. Dynamic biogeography and conservation of endangered species. Nature 403:84–86. Cunnington, G.M, J. Schaefer, J.E. Cebak, and D. Murray. 2008. Correlations of biotic and abiotic variables with ground surface temperature: An ectothermic perspective. Ecoscience 15:472–477. DeGregorio, B.A., B.J. Putman, and B.A. Kinsbury. 2011. Which habitat selection method is most applicable to snakes? Case studies of the Eastern Massasauga (Sistrurus catenatus) and Eastern Fox Snake (Pantherophis gloydi). Herpetological Conservation and Biology 6:372–282. Diaz, J.A. 1997. Ecological correlates of the thermal quality of an ectotherm’s habitat: A comparison between two temperate lizard populations. Functional Ecology 11:79–89. Ernst, C.H., and E.M. Ernst. 2003. Snakes of the United States and Canada. Smithsonian Institute Press, Washington, DC. 668 pp. Frazer, G.W., C.D. Canham, and K.P. Lertzman. 1999. Gap Light Analyzer (GLA), Version 2.0: Imaging software to extract canopy structure and gap light transmission indices from true-colour fisheye photographs. Users manual and program documentation. Simon Fraser University, Burnaby, British Columbia, and the Institute of Ecosystem Studies, Millbrook, New York. Available online at: http://www.rem.sfu.ca/forestry/downloads/ Files/GLAV2UsersManual.pdf. Accessed 27 December 2014. Goulet, C. 2010. A multi-scale evaluation of Eastern Hognose Snake (Heterodon platirhinos) habitat at the northern extent of its range. M.Sc. Thesis. University of New Hampshire, Durham, NH. 94 pp. Harvey, D.S., and P.J. Weatherhead. 2006. Hibernation site selection by Eastern Massasauga Rattlesnakes (Sistrurus catenatus catenatus) near their northern range limit. Journal of Herpetology 40:66–73. Johnson, D.H. 1980. The comparison of usage and availability measurements for evaluating resource preference. Ecology 61:65–71. LaGory, K.E., L.J. Walston, C. Goulet, R.A. van Lonkhuyzen, S. Najjar, and C.P. Andrews. 2009. An examination of scale-dependent resource use by Eastern Hognose Snakes in south-central New Hampshire. Journal of Wildlife Management 73:1387–1393. McLoughlin, P.D., L.R. Walton, H.D. Cluff, P.C. Paquet, and M.A. Ramsay. 2004. Hierarchical habitat selection by Tundra Wolves. Journal of Mammalogy 85:576–580. Michener, M.C., and J.D. Lazell. 1989. Distribution and relative abundance of the hognose snake Heterodon platirhinos in eastern New England. Journal of Herpetology 23:35–40. Mohr, C.O. 1947. Table of equivalent populations of North American small mammals. The American Midland Naturalist 37:223–249. Moore, J.A., and J.C. Gillingham. 2006. Spatial ecology and multi-scale habitat selection by a threatened rattlesnake: The Eastern Massasauga (Sistrurus catenatus catenatus). Copeia 2006:742–751. Pendleton, G.W., K. Titus, E. DeGayner, C.J. Flatten, and R.E. Lowel. 1998. Compositional analysis and GIS for study of habitat selection by Goshawks in southeast Alaska. Journal of Agricultural, Biological, and Environmental Statistics 3:280–295. Platt, D.R. 1969. Natural history of the hognose snakes Heterodon platirhinos and Heterodon nasicus. University of Kansas Publication of the Museum Natural History 18:253–420. Northeastern Naturalist C. Goulet, J.A. Litvaitis, and M.N. Marchand 2015 Vol. 22, No. 3 540 Plummer, M.V., and N.E. Mills. 2000. Spatial ecology and survivorship of resident and translocated Hognose Snakes (Heterodon platirhinos). Journal of Herpetology 34:565–575. Reinert, H.K., and D. Cundall. 1982. An improved surgical implantation method for radiotracking snakes. Copeia 1982:702–705. Robson, L.E. 2011. The spatial ecology of Eastern Hognose Snakes (Heterodon platirhinos): Habitat selection, home-range size, and the effect of roads on movement patterns. M.Sc. Thesis. University of Ottawa, Ottawa, ON, Canada. 55 pp. Rouse, J.D., R.J. Willson, R. Black, and R.J. Brooks. 2011. Movement and spatial dispersion of Sistrurus catenatus and Heterodon platirhinos: Implications for interactions with roads. Copeia 2011:443–456. Seaman, D.E., and R.A. Powell. 1996. An evaluation of the accuracy of kernel density estimators for home-range analysis. Ecology 77:2075–2085. Sperry, L.K. Butler, L.M. Romero, and P.J. Weatherhead. 2009. Effects of parasitic infection and radio-transmitters on condition, hematological characteristics, and corticosterone concentrations in Texas ratsnakes. Journal of Zoology 278:100–107. Steen, D.A., J.M. Linehan, and L.L. Smith. 2010. Multiscale habitat selection and refuge use of Common Kingsnakes, Lampropeltis getula, in southwestern Georgia. Copeia 2010:227–231. Steen, D.A., C.J.W. McClure, J.C. Brock, D.C. Rudolph, J.B. Pierce, J.R. Lee, W.J. Humphries, B.B. Gregory, W.B. Sutton, L.L. Smith, D.L. Baxley, D.J. Stevenson, and C. Guyer. 2012. Landscape-level influences of terrestrial snake occupancy within the southeastern United States. Ecological Applications 22:1084–1097. Tash, J.P., and J.A. Litvaitis. 2007. Characteristics of occupied habitats and identification of sites for restoration and translocation of New England cottontail populations. Biological Conservation 137:584–598. Therres, G.D. 1999. Wildlife species of regional conservation concern in northeastern United States. Northeastern Wildlife 54:93–100. Thomasson, V. 2012. Habitat suitability modeling for the Eastern Hog-nosed Snake, Heterodon platirhinos, in Ontario. M.Sc. Thesis. University of Ottawa, Ottawa, ON, Canada. 117 pp. Waldron, J.L., S.M. Welch, and S.H. Bennett. 2008. Vegetation structure and the habitat specificity of a declining North American reptile: A remnant of former landscapes. Biological Conservation 141:2477–2482. White, G.C., and R.A. Garrott. 1990. Analysis of Radio-tracking Data. Academic Press, San Diego, CA. 383 pp. Xuereb, A.T.J. 2012. Characterizing population genetic structure and inferring the influence of landscape features on gene flow in a temperate snake species. M.Sc. Thesis. Queen’s University, Kingston, ON, Canada. 88 pp.