CAVITY-NEST WEBS IN A LONGLEAF PINE ECOSYSTEM/Dinmica de Redes en la Nidificacion en Cavidades en un Ecosistema de Pinus palustris Lori A Blanc, Jeffrey R Walters.

The Condor. Lawrence: Feb 2008. Vol. 110, Iss. 1; pg. 80 Abstract (Summary) Cavity-nesting communities can be viewed as interconnected webs that interact through the creation of and competition for cavities as nest sites. Using a web approach, we depicted the flow of cavity creation and use in the cavity-nesting bird community of a Florida longleaf pine ( Pinus palustris ) ecosystem to examine the relationship between cavity-nesting bird abundance and cavity resources, and to identify species with potential to respond to cavity management for the endangered Red-cockaded Woodpecker ( Picoides borealis ). We identified two groups into which most cavity-nesting species could be placed: 1) six species associated with pine snags and Red-cockaded Woodpecker cavities, and 2) five species associated with hardwood snags. We found the majority of nests (60%) in large pine snags. Red-cockaded Woodpeckers and Northern Flickers ( Colaptes auratus ) excavated most cavities used by other cavity nesting birds. The Northern Flicker was the primary creator of large nest cavities, through its excavation in snags and enlargement of cavities originally excavated by the Red-cockaded Woodpecker in live pine. Large secondary cavity nesting birds were the primary users of Redcockaded Woodpecker cavities, and we identified three cavity-nesting species with potential to respond to Red-cockaded Woodpecker cavity management. The cavitynesting web dynamics documented in this study, including the role of large pine snags, Red-cockaded Woodpecker cavities in live pine, and excavating species within the community, can serve as a baseline for comparison to other southern pine forests. INTRODUCTION Cavity-nesting species represent a consideral portion of forest bird communities and rely upon the availability of cavity resources for nesting and roosting. Cavities can occur naturally through fungal decay at broken tree limbs, or through the process of woodpecker cavity excavation. Woodpeckers (primary excavators) typically excavate new cavities each year for nesting, and abandoned woodpecker cavities are regularly used by nonexcavating (secondary cavity-nesting) species, which rely upon existing cavities for nesting and roosting. Weak excavators use existing cavities or excavate their own cavities in highly decayed wood. These cavity-nesting guilds interact through the creation of and competition for cavities as nesting and roosting sites, and interactions among species can have a strong influence on forest community structure (Martin and Eadie 1999 ). Examining cavity-nesting birds at the community level can aid in the identification of important cavity-making agents, tree resources, and interspecific interactions (Dobkin et al. 1995 , Martin and Eadie 1999 , Bednarz et al. 2004 , Martin et al. 2004 , Saab et al. 2004 ). One useful way of examining connections and interdependencies among cavity-nesting species is to visually depict the community as a hierarchy of interconnected tree resources, cavity excavators, and secondary cavity users (i.e., a nest web; Martin and Eadie 1999 ). In this study, we apply the nest web concept to the long- leaf pine ( Pinus palustris ) ecosystem of the southeastern United States to identify cavity-nesting bird community

structure in relation to the federally endangered Red-cockaded Woodpecker ( Picoides borealis ). Fire-maintained longleaf pine forests, which are dominated by living pine, can be a relatively cavity-poor environment due to a low availability of snags (i.e., dead trees) and hardwoods, and the low tree density resulting from regularly occurring fire (Conner et al. 2001 ). The Red-cockaded Woodpecker is the only species in this ecosystem that excavates cavities in living pine trees, and does so exclusively (Ligon 1970 ). Redcockaded Woodpeckers live in family groups, and each group typically maintains multiple cavity trees (i.e., a cluster) within its territory (Walters 1990 ). METHODS STUDY AREA We conducted fieldwork from April through July in 2002 and 2003 on Eglin Air Force Base, located in northwestg Florida (Fig. 1 ). Eglin is the largest forested military reservation in the U.S., encompassing 187 555 ha, and is the largest, least fragmented longleaf pine ownership in the Southeast (McWhite et al. 1999 ). Approximately 78% of the Eglin reservation consists of longleaf pine sandhills (McWhite et al. 1999 ), characterized by scattered longleaf pine, with an open to sparse midstory of turkey oak ( Quercus laevis ) and a ground cover of various fire-adapted forbs and grasses (Kindell et al. 1997 ). Red-cockaded Woodpeckers inhabit sandhills habitat over the entire reservation, with over 6000 Red-cockaded Woodpecker cavity trees and 300 occupied clusters in 2002 when this study commenced (K. Gault, Eglin Air Force Base, pers. comm.). Thus, the Eglin reservation is well suited for large-scale studies that involve replicate plots within longleaf communities. Eglin contains a large amount of old-growth longleaf pine (Varner and Kush 2001 ), over 1700 dead standing Red-cockaded Woodpecker cavity trees (Williams et al. 2006 ), and an abundance of old-growth pine snags due to excessive tree mortality after the reintroduction of fire following years of suppression (Varner et al. 2005 ). As part of their habitat restoration process, Eglin land managers have also conducted hardwood midstory reduction through mechanical and herbicidal methods (McWhite et al. 1999 ). As a result, both hardwood and pine snags were abundant on parts of the reservation during the course of this study, but in other areas at Eglin, few snags were available, as is typical of other fire-maintained southeastern pine forests. Additionally, Eglin land managers do not harvest dead trees, and thus large, old-growth snags remain on the landscape as a resource for cavity nesting birds. This provided us with the opportunity to observe heterospecific use of Red-cockaded Woodpecker cavities in areas of varying snag densities. DATA COLLECTION We recorded the relative abundance (detections per survey) of birds at all sampling stations within each plot twice annually between 8 April and 21 June 2002 and 2003 using a point-transect count sampling technique adapted for the open habitat (Provencher et al. 2002 , 2003 ). We sampled two plots (totaling 24 sampling stations) per morning within the first 3 hr after sunrise. We conducted an 8-min point count at each sampling station, during which we recorded species, time detected, vocalization type, location, and movement within a 100-m fixed radius of the point center (Hutto et al. 1986 ). Two observers sampled simultaneously at point-count stations on alternating transects situated 200 m apart and synchronized start times for the point counts. Birds detected by the two

observers at the same time were not double-counted. The same two observers conducted the point counts during both years. In counts at each sampling station, we included birds detected as the observers moved between adjacent stations. This transect-count addition to the point-count sampling enabled us to account for birds fleeing ahead of the observers and may be appropriate for open habitat (Bibby et al. 1992 ). For each species and each round of sampling, we calculated relative abundance (detections per survey) by summing the total number of individuals detected at all 12 sampling stations in a plot. We then used the mean number of detections for the plot (across the two sampling rounds) as the measure of abundance for that season. In a previous study at Eglin, Provencher et al. (2002) found no substantial difference between a similar modified point-transect count method and standard point-count methodology, except for reduced variability in bird detection rates using the former method. Because all study plots were in the same habitat type, observer bias was minimized by using the same two observers both years and double counts were minimized, we believe that our sampling methodology provides a reliable measure of the relative abundance of birds occurring in all study plots. Nonetheless, we emphasize that our measurement was used as an index of abundance rather than an attempt to quantify breeding densities. STATISTICAL ANALYSES Finally, we incorporated nest data into two nest web diagrams (Martin and Eadie 1999 ), which reflect the flow of cavity creation and use between levels of primary cavity excavators, weak excavators, secondary cavity nesting birds, and trees (cavity resources). Nest webs included all nest cavities for which we could identify the excavator, either through known use of the cavity or reasonable estimation based on cavity characteristics. Nests used in the web include reuse of the same cavity across years but not renesting attempts by the same species in the same cavity within the same year. Abundance webs and nest webs serve as alternative depictions of cavity-nesting bird community structure at Eglin. When sampling for bird abundance, it is possible to detect birds that are foraging but not nesting in the plot. Basing a web on nest data provides a more definitive measure of which species actually use the plot for nesting, although it is possible to miss nests of species that are shy or stealthy or for which our nest searching methodology is not well suited. Thus, the abundance and nest webs should complement one another. RESULTS Most species detections were reflective of nesting effort within each plot; for 11 of the 14 species, the abundance of a species was positively correlated with the number of nests found for that species (Table 1 ). Correlations could not be computed for two species: the Carolina Chickadee was often detected, but we found no nests, whereas we found many Eastern Screech-Owl nests, but rarely detected them in point counts. Neither of the noncavity-nesting bird groups (i.e., the open savanna group or developing hardwood midstory group) was significantly associated with live Red-cockaded Woodpecker cavity trees ( r = 0.03, P = 0.79 and r = -0.17, P = 0.16, respectively; n = 72 plots), hardwood snags ( r = -0.16, P = 0.17 and r = 0.02, P = 0.90, respectively; n = 72 plots), or pine snags ( r = 0.09, P = 0.45 and r = 0.10, P = 0.43, respectively; n = 72 plots), indicating that all three types of cavity resources were available to cavity-nesters regardless of habitat condition. Therefore, the associations depicted in our abundance web (Fig. 2 ) likely reflect community structure as influenced by cavity resource type, rather than habitat condition alone.

DISCUSSION Our nest data for three species, the Brown-headed Nuthatch, Carolina Chickadee, and Eastern Bluebird, should be interpreted with caution for several reasons. First, the phenology of these species may have been too early for our searching efforts to have produced a representative sample of nests (Mostrom et al. 2002 , Withgott and Smith 1998 , Gowaty and Plissner 1998 ). Second, given our large plot size and widely spaced transects, our nest-searching methodology may not have been well suited for locating nests in small, short snags, typical of these species. Therefore, we may have underestimated the nesting effort of these species. CONSERVATION AND MANAGEMENT IMPLICATIONS Our results suggest that the Red-cockaded Woodpecker, through its cavity-excavation in living pine, may positively affect American Kestrel, Northern Flicker, and Eastern Screech-Owl abundance, and cavity restrictor plates have potential to negatively affect these larger species. In addition, the frequent use of pine snags by other species in this study, despite the abundance of Red-cockaded Woodpecker cavities, suggests that retaining snags may benefit Red-cockaded Woodpeckers by reducing heterospecific occupation of their cavities. Given the large amount of management directed toward preventing heterospecific use of Red-cockaded Woodpecker cavities, future experimental studies are needed to clarify the relationship between snag availability and Red-cockaded Woodpecker nesting success (Kappes and Harris 1995 , U.S. Fish and Wildlife Service 2003 ). [Reference] Aitken, K., and K. Martin. 2007. The importance of excavators in hole-nesting communities: availability and use of natural tree holes in old mixed forests of western Canada. Journal of Ornithology (Online First). doi: 10.1007/s10336-007-0166-9. Allen, D. H. 1991. An insert technique for constructing artificial Red-cockaded Woodpecker cavities. USDA Forest Service Technical Report SE-73. Allen, J. C., S. M. Krieger, J. R. Walters, and J. A. Collazo. 2006. Associations of breeding birds with fire-influenced and riparian-upland gradients in a longleaf pine ecosystem. Auk 123:1110--92. Aubry, K. B., and C. M. Raley. 2002. The Pileated Woodpecker as a keystone habitat modifier in the Pacific Northwest. USDA Forest Service General Technical Report PSWGTR- 181. Bai, M-L. 2005. Tree cavity abundance and nest site selection of cavity nesting birds in a natural boreal forest of West Khentey, Mongolia. Ph.D. dissertation. Georg-AugustUniversitt zu Gttingen, Gttingen, Germany. Bednarz, J. C., D. Ripper, and P. M. Radley. 2004. Emerging concepts and research directions in the study of cavity-nesting birds: keystone ecological processes. Condor 106:1--92. Belson, M. S. 1998. Red-headed Woodpecker ( Melanerpes erythrocephalus ) use of habitat at Wekiwa Springs State Park, Florida. M.Sc. thesis, University of Central Florida, Orlando, FL. Benjamini, Y., and Y. Hochberg. 1995. Controlling the false discovery rate--a practical and powerful approach to multiple testing. Journal of the Royal Statistical Society 57:289--92. Bibby, C. J., N. D. Burgess, and D. A. Hill. 1992. Bird census techniques. Academic

Press, San Diego, CA. Blanc, L., and J. R. Walters. 2007. Cavity-nesting community webs as predictive tools: where do we go from here?. Journal of Ornithology 148:417--92. Blanc, L. A., and J. R. Walters. 2008. Cavity excavation and enlargement as mechanisms for indirect interactions in an avian community. Ecology 89, in press. Bonar, R. L. 2000. Availability of Pileated Woodpecker cavities and use by other species. Journal of Wildlife Management 64:52--92. Bull, E. L., C. G. Parks, and T. R. Torgersen. 1997. Trees and logs important to wildlife in the interior Columbia River Basin. USDA Forest Service General Technical Report PNW-391. Bull, E. L., S. R. Peterson, and J. W. Thomas. 1986. Resource partitioning among woodpeckers in northeast Oregon. USDA Forest Service Research Note PNW-444. Carter, J. H. III, J. R. Walters, S. H. Everhart, and P. D. Doerr. 1989. Restrictors for Redcockaded Woodpecker cavities. Wildlife Society Bulletin 17:68--92. Conner, R. N., and D. C. Rudolph. 1995. Excavation dynamics and use patterns of Redcockaded Woodpecker cavities: relationships with cooperative breeding, p. 343-352. In D. L. Kulhavy, R. G. Hooper, and R. Costa [EDS. ], Red-cockaded Woodpecker: recovery, ecology and management. Center for Applied Studies in Forestry, College of Forestry, Stephen F. Austin State University, Nacogdoches, TX. Conner, R. N., D. C. Rudolph, and J. R. Walters. 2001. The Red-cockaded Woodpecker: surviving in a fire-maintained ecosystem. University of Texas Press, Austin, TX. Conner, R. N., and D. Saenz. 2005. The longevity of large pine snags in eastern Texas. Wildlife Society Bulletin 33:700--92.

Intraspecific and interspecific variation in use of forest-edge habitat by snakes Gerardo L F Carfagno, Patrick J Weatherhead. Canadian Journal of Zoology. Ottawa: Oct 2006. Vol. 84, Iss. 10; pg. 1440, 13 pgs Abstract (Summary) Variation in use of edge habitat among populations and species of snakes should reflect underlying causes (e.g., thermal ecology, prey availability) and consequences (e.g., predation on birds' nests) of habitat selection. We compared the habitat use of ratsnakes, Elaphe obsoleta (Say in James, 1823), in Illinois and Ontario and compared habitat use by ratsnakes and racers, Coluber constrictor (L., 1758), in Illinois. Ratsnakes in Illinois used upland forest more and forest edges less than ratsnakes in Ontario. Female ratsnakes in Illinois used edges less than males, regardless of their reproductive status. Relative to ratsnakes, racers preferred forest edges and avoided forest interior. Female racers used edges more than males, especially while gravid. These results, and most of the seasonal patterns in habitat use, were broadly consistent with variation expected from differences in thermoregulatory needs, although other factors potentially influencing habitat use cannot be ruled out. Although it has been proposed that some forest fragmentation is likely to be beneficial for ratsnakes in Canada, such fragmentation may be detrimental to ratsnakes in Illinois but beneficial to racers. Thus, relative to forest-interior species, edgenesting birds should be more vulnerable to predation by ratsnakes in Ontario, and fragmentation should increase the vulnerability of forest birds to nest predation by racers. INTRODUCTION Habitat loss and fragmentation imperil natural communities throughout the world (Meffe and Carroll 1997; Wilson 2001). Fragmentation increases the relative amount of edge habitat, which can have important consequences for community interactions (Donovan et al. 1995; Robinson et al. 1995). Generalist predators that use edge habitat are of special concern, especially for avian populations (Yahner and Scott 1988; Andren 1992; Harrison and Bruna 1999). Nest predation is the greatest cause of nest failure in birds (Ricklefs 1969; Martin 1988; Martin 1993) and is especially significant in some edge habitats (Andren and Angelstam 1988; Burkey 1993; King et al. 1998; Batary and Baldi 2004). Therefore, avian conservation requires better understanding of habitat use by nest predators (Chalfoun et al. 2002a, 2002B; Larivire 2003; Stephens et al. 2003). Snakes have been identified as important nest predators (e.g., Thompson et al. 1999; Morrison and Bolger 2002; Thompson and Burhans 2003). Therefore, an essential step toward identifying links between habitat structure and nest predation by snakes is the determination of how snakes select and use habitat, particularly edges (Weatherhead and Blouin-Demers 2004). Among the ecological similarities between ratsnakes and racers, both prey extensively on birds and mammals (Klimstra 1959; Fitch 1963a, 1963b; Weatherhead et al. 2003; Carfagno et al. 2006), which should reduce the importance of prey abundance for habitat use. Ratsnakes and racers occur together in Illinois in areas characterized by a mixture of forests and fields (Keller and Heske 2000). Thus, both species would have access to the edges between those habitats. The species are also similar in size and color as adults, which makes their physical properties associated with heating and cooling similar.

Despite these similarities, there are several reasons to assume that racers and ratsnakes should differ in their habitat use, particularly in their use of edges. Habitat use by snakes may shift with changing environmental conditions (Webb and Shine 1998). As in Ontario, for example, forest edges in Illinois may be the optimal thermal habitat when the need to shuttle between shade and sun is great (Heatwole and Johnson 1979; Blouin-Demers and Weatherhead 2001a, 2001c). During warmer months, however, forested sites could allow ratsnakes to achieve preferred temperatures solely by using sun flecks or canopy gaps (Lillywhite and Henderson 1993), whereas old fields could generally become too warm for racers. Therefore, we predicted that ratsnakes in Illinois would preferentially use forest edges in the spring and fall, but would move to forested interior habitat in the summer. Contrary to the pattern expected for ratsnakes, we predicted that racers would show an increasing preference for edges in the hottest parts of the summer as open habitats became too warm. MATERIALS AND METHODS Study site and species We conducted our research from 2002 to 2004 at the Cache River State Natural Area in Johnson County, southern Illinois. Genetic and morphological analyses have suggested that E. obsoleta consists of four species, with Ontario snakes part of the eastern species and Illinois snakes part of the central species (Burbrink et al. 2000; Burbrink 2001). More recently, however, an analysis of the Ontario population studied by Blouin-Demers and Weatherhead showed that it is a mix of eastern and central species (Gibbs et al. 2006). Thus, for the purpose of this study we consider ratsnakes in Ontario and Illinois to be different populations of the same species, which we refer to simply as ratsnakes. Radiotelemetry We captured most ratsnakes and racers as they emerged from shared communal hibernacula each spring, but also opportunistically throughout the season. Hibernacula were located by radio-tracking snakes caught opportunistically during the first field season. Upon capture, snakes were sexed, measured to the nearest millimetre, and weighed to the nearest gram. Reproductive condition of females was assessed by palpation of the abdomen periodically throughout the field season and by observation of mating behavior. We classified a female as not gravid that season when we failed to observe mating behavior or the presence of eggs. However, many females (especially ratsnakes) were inaccessible to us (i.e., in tree canopies) most of the time. Therefore, we were unable to assess reproductive condition for all female ratsnakes. Habitat characterization We assessed macro- and micro-habitat at randomly selected points for every other snake location to quantify habitat availability. Each point was selected at a random direction and distance (10-200 paces) from the snake location (Blouin-Demers and Weatherhead 2001a). Thus, this approach did not sample available habitat in the study area in a strictly random sense. Rather, it was a random sample of habitat actually available to each snake.

At the landscape scale, habitat at our study site consisted of a mixture of forest and open areas (e.g., fields, wetlands) with occasional old barns and other man-made structures, similar to the habitat in the study site used by Blouin-Demers and Weatherhead (2001a) in Ontario. We classified habitat into five categories. Upland forest was dominated by oaks and hickory and was located along a ridge adjacent to the Cache River floodplain. Bottomland forest was dominated by cypress and tupelo and occurred below the ridge and directly adjacent to the river, where flooding was common, especially in spring. Grasses and forbs dominated old fields. Later successional fields were dominated by woody species including shrubs, vines, and dense saplings. Finally, we defined edge habitat as the interface between closed and open canopy habitats (including forest-field and forestriver interfaces). We used a distance of 15 m from the interface to define edge habitat (Blouin-Demers and Weatherhead 2001a). DATA ANALYSIS All analyses were performed using SYSTAT Version 10.0 (Systat Software Inc., Richmond, California). Using individual animals as replicates, we used compositional analysis to determine landscape-scale habitat preferences (Aebischer et al. 1993), which allowed us to include all locations in the landscape-scale analysis, weighted by the amount of time individual snakes spent at each location. We calculated a composition for each snake, pooling data for snakes that were tracked in multiple years. For females that changed reproductive condition, we calculated separate compositions for the snake when gravid and when not. We calculated differences between log-ratio transformed proportions of habitat use and availability. To determine whether use differed from availability, we used MANOVA to determine whether the deviations between habitat use and availability points differed from a normal distribution centered on a mean of zero. We used t tests for painvise comparisons among the five habitat types. The analysis was repeated with observations categorized by sex and reproductive condition. We also calculated compositions separately by month to test for differences across the season. Blouin-Demers and Weatherhead (2001a) analyzed macrohabitat selection without weighting locations by the amount of time a snake had spent there, and by pooling data across all individuals. To allow direct comparisons between studies, we reanalyzed the raw telemetry data from their Ontario study using compositional analysis. RESULTS Landscape-scale habitat selection by Ontario ratsnakes Blouin-Demers and Weatherhead (2001a) had sufficient (>25) telemetry locations for 46 ratsnakes (15 males and 31 females), for a total of 2990 telemetry points between May and August. Sixteen females were not gravid when tracked, nine were gravid, and six were gravid one year and not the other. Habitat availability was recalculated with the exclusion of aquatic habitat because we considered this habitat unavailable to terrestrial ratsnakes. Qualitatively, compositional analysis did not change overall landscape-scale habitat use by ratsnakes in Ontario. Consistent with the results of the original analyses (Blouin-Demers and Weatherhead 2001a), edge habitat was used significantly more, and forests and open habitats less, relative to their availability. Therefore, we do not present the results of those reanalyses here and instead use Blouin-Demers and Weatherhead's (2001a) original results for comparison with our results from Illinois.

Landscape-scale habitat selection in Illinois We tracked 22 ratsnakes and 16 racers over the course of the 3-year study. Eighteen of the ratsnakes (nine males and nine females) and 15 of the racers (seven males and eight females) had sufficient locations (>25) to be used in the compositional analysis, for a total of 1727 ratsnake and 1419 racer telemetry points between April and September. One of the female ratsnakes was gravid while tracked and three were gravid one year and not the other; we were unable to determine the reproductive status of the other five. We recalculated compositions with data for all females combined because gravid and nongravid females used habitat similarly (see below). Because we know females are not reproductive every year in this population and elsewhere at this latitude (e.g., Fitch 1999), we assume the unknown group comprised both gravid and non-gravid females. Racer compositions were also calculated using all females, but in this case because all female racers were gravid each year of the study. The mean number of observations per individual averaged 82.2 for ratsnakes and 94.6 for racers. Discussion Ratsnake latitudinal comparison Our results for edge use differ from results previously reported from our study site (Carfagno et al. 2006). In addition to those analyses being restricted to 3 months of one active season, habitats were classified differently. Because bottomland (avoided) and upland (preferred) forest types were combined in those analyses, preference for edges was increased relative to preference for forest habitat. The more inclusive current analyses still showed forest edge habitat to be important for ratsnakes in southern Illinois, but not more than upland forest interior. The lack of an overall edge preference that we documented is consistent with Keller and Heske's (2000) results from Illinois but differs from Durner and Gates' (1993) study of ratsnakes in Maryland. These results are not directly comparable, however, because neither of those studies quantified edge as a specific habitat type. This is especially problematic in fragmented landscapes, where the extent of fragmentation might influence the degree to which snake locations in forest are proximate to edges. [Reference] References Adolph, S.C. 1990. Influence of behavioral thermoregulation on microhabitat use by two Sceloporus lizards. Ecology, 71: 315-327. doi: 10.2307/1940271. Aebischer, N.J., Robertson, P.A., and Kenward, R.E. 1993. Compositional analysis of habitat use from animal radio-tracking data. Ecology, 74: 1313-1325. doi: 10.2307/1940062. Andren, H. 1992. Corvid density and nest predation in relation to forest fragmentation: a landscape perspective. Ecology, 73: 794-804. doi: 10.2307/1940158. Andren, H., and Angelstam, P. 1988. Elevated predation rates as an edge effect in habitat islands: experimental evidence. Ecology, 69: 544-547. doi: 10.2307/1940455. Batary, P., and Baldi, A. 2004. Evidence of an edge effect on avian nest success. Conserv. Biol. 18: 389-400. doi: 10.1111/j.15231739.2004.00184.x. Bauwens, D., Garland, T., Castilla, A.M., and Van Damme, R. 1995. Evolution of sprint

speed in lacertid lizards: morphological, physiological, and behavioral covariation. Evolution, 49: 848-863. doi: 10.2307/2410408. Blouin-Demers, G., and Weatherhead, P.J. 2001a. Habitat use by black rat snakes (Elaphe obsoleta obsoleta) in fragmented forest. Ecology, 82: 2882-2896. doi: 10.2307/2679968. Blouin-Demers, G., and Weatherhead, P.J. 2001b. Thermal ecology of black rat snakes (Elaphe obsoleta) in a thermally challenging environment. Ecology, 82: 3025-3043. doi: 10.2307/2679832.