2009 NORTHEASTERN NATURALIST 16(1):13–26 Distribution of Red-backed Salamander (Plethodon cinereus) with Respect to Cover-object Characteristics in the Green Mountains of Vermont Laurie S. Richmond1,2,* and Stephen C. Trombulak1 Abstract - We examined the distribution of Plethodon cinereus (Red-backed Salamander) with respect to cover-object type in the Green Mountains of Vermont by exploring their distribution under cover objects relative to the total availability of cover objects on the forest floor. We conducted cover-object searches in sixteen 50-m transects in forest stands >50 years old to explore the distribution of large (>3.47 cm snout–vent length) and small (<3.47 cm snout–vent length) salamanders with respect to object material, size, and texture. There were more salamanders than would be expected by chance under rocks and fewer under woody objects (branches and logs). Salamander counts were higher than would be expected under large cover objects and lower under small ones. Our results also indicate that salamanders were more common than would be expected under fibrous woody objects and less common under solid ones. Finally, we found that large salamanders were more common than would be expected under rocks, while small salamanders were more common under woody objects. These results could have important implications for improving the recovery of salamanders following forest management applications. Introduction Plethodon cinereus Green (Red-backed Salamander) is an important component of forested ecosystems in the northern Appalachian Mountains. It is one of the most abundant vertebrates both in terms of numbers and biomass in New England (Burton and Likens 1975a). Burton and Likens (1975b) reported that in New Hampshire, salamanders—of which the Redbacked Salamander makes up 95% of the total salamander biomass—are responsible for approximately 20% of the energy flow through mammalian and avian populations. Welsh and Droege (2001) argued that plethodontid salamanders (Caudata: Plethodontidae) should be used to monitor ecosystem integrity in North American forests because they possess a tight physiological link with forest processes, are efficient to count, and are so abundant that statistically meaningful samples are easy to obtain. Much past research has shown that forest harvesting can cause significant declines in salamander populations in the eastern United States (Ash and Bruce 1994, deMaynadier and Hunter 1998, Petranka et al. 1993). Many of these studies trace the process by which salamander populations 1Department of Biology, Middlebury College, Middlebury, VT 05753. 2Current address - Department of Fisheries, Wildlife, and Conservation Biology, 200 Hodson Hall, 1980 Folwell Avenue, University of Minnesota, Saint Paul, MN 55108. *Corresponding author - richm051@umn.edu. 14 Northeastern Naturalist Vol. 16, No. 1 recover following forest management events. These studies show that the abundance of plethodontid salamanders changes with forest age, suggesting that salamanders can serve as indicators of the ecological responses to forest management practices and rates of ecological recovery following disturbance (deMaynadier and Hunter 1998, Petranka et al. 1993, Welsh and Droege 2001). In order to understand the recovery patterns of Red-backed Salamanders following forest harvesting, it is important to understand their behavior and life-history characteristics in forest environments. Red-backed Salamanders avoid exposed surfaces and are most often found under cover objects such as logs, bark, and rocks (Moore et al. 2001). Such objects are advantageous because they offer salamanders a cool, moist refuge from predators and contain a large supply of invertebrate prey (Gergits and Jaeger 1990). Many adult Red-backed Salamanders establish territories that include cover objects (Gergits and Jaeger 1990). Both males and females chemically mark these territories and act aggressively towards intruders (Jaeger and Gergits 1979; Jaeger et al. 1982, 1995). Red-backed Salamanders often associate with only one cover object and exhibit high rates of site fidelity (Gergits and Jaeger 1990). They have relatively small home ranges, estimated in Michigan to have an upper limit of about 24.3 m2 (Kleeberger and Werner 1982). Studies indicate that Redbacked Salamanders prefer large rather than small cover objects (Mathis 1990). Moore et al. (2001) observed that salamanders have similar abundances under rocks and objects made of wood. However, few studies have examined salamander cover object preference by dividing woody objects into more specific categories of size and texture. Wood objects can have a variety of textures depending upon level and type of decay. Heatwole (1962) divided wood into four texture categories— solid, fibrous, chunky, and crumbly—and observed that the texture of logs influenced whether salamanders would be inside, finding the most salamanders within chunky logs and almost none within solid logs. However, Heatwole (1962) did not explore whether wood texture, which affects its ability to house salamanders inside, increases the chance that a salamander is associated with that object overall, either inside or underneath. Several studies have found that in territorial disputes, larger salamanders are dominant over smaller ones (Mathis 1990, Townsend and Jaeger 1998). Faragher and Jaeger (1997) did not find a significant correlation between size of salamanders and size of cover objects in Virginia. However, they did not explore the relationship between salamander size and other attributes of cover objects, such as material type and texture. Much of the research on salamander territoriality and cover-object preference has involved the placement and manipulation of artificial cover objects on the forest floor or in the laboratory. However, Marsh and Goicochea (2003) suggested that the placement of artificial cover objects can lead to 2009 L.S. Richmond and S.C. Trombulak 15 biases in data on Red-backed Salamander populations, particularly when monitoring changes in population size. In this study, we examined salamander distribution under natural cover objects relative to cover object abundance, size, type, and texture. We conducted intensive cover-object searches in the forests of the Green Mountains in central Vermont. We collected detailed information about cover objects and associated salamanders to answer questions about salamander distribution: (1) Do Red-backed Salamanders exhibit non-random relationships with cover objects based on the material, size, or texture of cover objects? (2) Do large and small salamanders show different distributions with respect to cover object types? Some salamander research suggests that salamander recovery following forest management is linked to the salamander’s relationship with cover objects present on the forest floor (deMaynadier and Hunter 1995). Therefore, information about the association of salamanders with different cover objects not only provides insight into the natural history of these animals, but can aid in the development of forest management strategies that will increase the rate at which Red-backed Salamander populations recover following management activities. Methods Study sites We explored the distribution of Red-backed Salamanders under cover objects by conducting surveys of cover objects in eight forest stands in the Middlebury District of the Green Mountain National Forest in Vermont (Table 1, Fig. 1). All stands were in deciduous forest >50 years old and were separated by 0.25–20 miles in distance. Table 1. Description of the forest stands studied in the Middlebury District of the Green Mountain National Forest, VT. The soil type symbols are: BsE = Berkshire and Marlow extremely stony loams, 20 to 50% slopes; BsC = Berkshire and Marlow extremely stony loams, 3 to 20% slopes; LxC = Lyman-Berkshire very rocky complex, 5 to 20% slope; PsC = Peru extremely stony loam, 0 to 20% slopes. Stand ID Stand region Age (yrs) Altitude (m) Soil type # of transects Dates assessed 2218 Ripton A 145 579–609 BsE 1 3 July 1825 Ripton B 58 396–427 LxC 3 20, 27 September 1866 Ripton B 66 396–427 PsC 3 13, 17 September 2200 Ripton B 140 549–609 BsC 1 1 August 729 Lincoln A 80 488 BsE 1 6 July 165 Lincoln B 105 549–609 BsE 5 13, 14, 18, 19 June 601 Bristol 73 488–518 PsC 1 12 July 3318 Leicester 103 366–396 BsC 1 11 July 16 Northeastern Naturalist Vol. 16, No. 1 The dominant tree species on each stand were Acer saccharum Marsh. (Sugar Maple), A. rubrum Linnaeus (Red Maple), Betula alleghaniensis Britton (Yellow Birch), and Fagus grandifolia Ehrhart (American Beech). We obtained spatially explicit digital information on stand locations and ages from the District Forest Office. Data on stand elevation and soil type were derived from digital elevation and soil data sets available from the Vermont Center for Geographic Information. All analysis of spatial information was done with ArcView 3.2. Information on soil characteristics was derived from Ferguson (1998). Transects We examined salamander distributions by conducting cover-object and salamander surveys along linear transects placed throughout the study sites. During spring and summer 2001, we searched from one to five transects in each of the eight forest stands, for a total of 16 different transects. Stand areas ranged from 6.6 to 27.3 hectares, and transects were spaced evenly throughout the forest stands. The location of preliminary transect starting points differed according to the number of transects in a stand. When a stand only had one transect (n = 5), the preliminary starting point was based on the stand’s approximate center, identified using the digital stand data and a GPS unit. When a stand had three transects (n = 2), the preliminary starting points were evenly Figure 1. Map of the study region in Vermont. The black area on the state map represents the boundary of the northern portion of the Green Mountain National Forest. Boxes on the map of the northern half of the forest boundary show the approximate location of the forest stand regions from which we selected 1 or 3 (Ripton B) stands. 2009 L.S. Richmond and S.C. Trombulak 17 distributed within a stand, also identified using the digital stand data and a GPS unit. When a stand had five transects (n = 1), the first preliminary starting point was positioned near one edge of the stand, and subsequent points were positioned after walking in the direction of the stand’s center for 5, 10, 15, and 20 minutes. After the preliminary starting points were established, the same method was used to establish all transects. From each preliminary starting point, we threw a stick over a shoulder in a haphazard direction; the point where the stick landed became the transect’s true starting point. We staked down a measuring tape at the starting point and extended it 50 m in the direction opposite from where the stick had been thrown. We searched the area 5 m to each side of the transect, resulting in transects with a 500-m2 search area. Between 13 June and 27 September 2001, we searched for salamanders under every potential cover object—branch, log, bark strip, and rock—we were able to turn over. We did not search the leaf litter. Whenever possible, we carefully broke apart and sifted through woody cover objects to search for salamanders residing within them. The search time varied between 2 and 4.5 h for each transect, depending on the number of salamanders and cover objects present. We recorded information on each potential cover object searched and each salamander found. Locations of all items of interest— cover objects and salamanders—were recorded by transect. Each transect was searched only once. Cover objects We recorded the number and type of each potential cover object found on each transect regardless of whether it covered or contained a salamander. We defined a potential cover object to be any solid object large enough to cover a small Red-backed Salamander (snout–vent length = 2.4 cm). Objects smaller than 2.4 cm were not searched or counted. In addition, we did not assess cover objects, particularly rocks, which were too heavy to lift or highly embedded. All cover-object searches were conducted by one researcher (L.S. Richmond) over the course of the field season, increasing the consistency of cover-object categorization. We categorized cover objects by material, size, and texture. We recognized three types of material: cylindrical woody object, bark, and rock. Cylindrical woody objects consisted of both branches and logs. For the purposes of this paper, we will refer to the category as “woody object.” For rocks and woody objects, we recognized the following size categories based on the observed range of sizes: small (rocks <256 cm2, woody objects <15 cm), medium (rocks 256–702 cm2, woody objects 15–30 cm), and large (rocks >702 cm2, logs >30 cm). Woody object measurements represent diameter of branch or log where it was touching the ground, and rock measurements represent surface area covering the ground 18 Northeastern Naturalist Vol. 16, No. 1 (greatest length x greatest width). We did not record the relative size of bark strips because bark was not very common and it was difficult to determine the range of possible sizes. We categorized the texture of woody objects following Heatwole (1962): solid (branches and logs that do not break apart and have little decay), chunky (wood as rectangular blocks that can be separated by hand), crumbly (small particles that fall to dust when pulled apart), and fibrous (loose strands). Salamanders We handled salamanders following the guidelines recommended by the American Society of Ichthyologists and Herpetologists (Herpetological Animal Care and Use Committee 2004). For each salamander, we recorded its size, location on the transect, and associated cover object. Size was measured as snout–vent length (SVL) to the nearest 0.01 cm using dial calipers. We recorded the characteristics of the cover object associated with the salamander and noted whether the salamander was found underneath or inside the cover object. We returned all salamanders and cover objects to their original positions immediately after searching each object. For textured objects that we pulled apart during searching, we took care to rebuild the object in its original shape and we placed the salamander in approximately the same location as where it was found within the object. Statistical analysis Data analysis was conducted with the statistical software “R” (The R Foundation for Statistical Computing 2005, http://www.r-project.org). We conducted chi-square analysis using the Pearson’s chi-squared test. In the cases where the samples sizes were too small to accurately run a Pearson’s chi-squared test, we conducted analysis using Fisher’s exact test, which reports a single p-value with no test statistic. Both Pearson’s chi-squared test and Fisher’s exact test determine if the proportions in the table are different from what would be expected under the null hypothesis of random distribution. In our analyses, the null hypothesis was that the salamanders were randomly distributed under cover objects and did not select among cover objects based on material, size, or texture. For examinations of salamander distribution based on salamander size, the null hypothesis was that salamanders were randomly distributed under cover objects with respect to salamander size. Results We encountered a total of 240 salamanders, 41 of which escaped before measurement. We found an average of 15 Red-backed Salamanders per 500-m2 transect (0.03 salamanders/m2). Only 7.4% (240/3251) of the 2009 L.S. Richmond and S.C. Trombulak 19 searched cover objects were occupied by salamanders. Sixty-eight (28.3%) of the 240 salamanders were within rather than underneath their associated cover object. We observed significant relationships between Red-backed Salamander presence and cover-object material, size, and texture. In each of these analyses, the distribution of salamanders deviated significantly from the null hypothesis of random distribution (Tables 2–4). In the case of material type, there were more salamanders than expected under rocks and fewer than expected under woody objects (Table 2). For tests involving cover-object size of rocks and woody objects, we found more salamanders than would be expected under large objects and fewer than expected under small objects (Table 3). In the case of woody objects, salamander counts were higher than would be expected under objects of medium size as well. Objects made of bark were excluded from this analysis because we did not collect information on the size of bark in the field (see Methods). Table 2. The observed (and expected in parentheses) number of cover objects with (present) and without (absent) Plethodon cinereus (Red-backed Salamander) by type of material (woody [branches and logs], bark, or rock) in Green Mountain National Forest, VT, 2001 (χ2 = 33.69, df = 2, P < 0.001). Woody Bark Rock Present 187 (206.9) 8 (12.3) 45 (20.9) Absent 2615 (2595.1) 158 (153.7) 238 (262.1) Table 3. The observed (and expected in parentheses) number of cover objects with (present) and without (absent) Plethodon cinereus (Red-backed Salamander) by object size for both rock and woody (branch or log) objects: Small (rocks < 256 cm2, woody objects < 15 cm), medium (rocks 256–702 cm2, woody objects 15–30 cm), and large (rocks > 702 cm2, logs > 30 cm). Woody object measurements represent diameter of branch or log where touching the ground and rock measurements represent surface area covering the ground. Green Mountain National Forest, VT, 2001 (rocks: (χ2 = 53.6, df = 2, P < 0.001; woody objects: (χ2 = 203.71, df = 2, P < 0.001). Small Medium Large Rocks only*** Present 4 (8.7) 21 (30.4) 20 (5.4) Absent 51 (46.3) 173 (163.2) 14 (28.6) Woody objects*** (logs and branches) Present 66 (144.6) 43 (13.7) 78 (28.6) Absent 2101 (2022.4) 163 (192.3) 351 (400.4) *** indicates significance P < 0.001. Table 4. The observed (and expected in parentheses) number of cover objects with (present) and without (absent) Plethodon cinereus (Red-backed Salamander) by woody object texture: solid, fibrous, chunky, or crumbly. Green Mountain National Forest, VT, 2001 (Fisher’s Exact Test: P < 0.001). Solid Fibrous Chunky Crumbly Present 119 (163.8) 52 (17.0) 11 (1.8) 5 (0.9) Absent 2336 (2291.2) 254 (285.6) 16 (25.2) 9 (13.1) 20 Northeastern Naturalist Vol. 16, No. 1 When examining cover-object texture, we found more salamanders than expected associated with fibrous, chunky, and crumbly objects and fewer than expected with solid ones (Table 4). In addition, texture had an influence on location of salamanders. There was a significantly nonrandom distribution of salamanders with respect to their location (within or underneath the object) and object texture (Table 5). Salamander presence was more frequent than expected underneath rather than within solid objects, while salamander presence within the cover object was higher than expected for decaying woody objects of fibrous or chunky texture (Table 5). Salamanders that were found within solid logs were either located in the space underneath peeling bark, or in cases where the center of the logs were slightly decayed, in holes or crevices within the log that could be pried open. Based on the median size of salamanders caught over the course of the study, we divided salamanders into two size categories: small (SVL ≤ 3.47 cm) and large (SVL > 3.47 cm). We examined the distribution of large and Table 6. Observed (and expected in parentheses) number of small (SVL ≤ 3.47 cm) and large (SVL > 3.47 cm) Plethodon cinereus (Red-backed Salamander) found associated with cover objects based on object material, size, and texture. Object category definitions same as those used in previous analyses. (Object material type: Fisher’s exact test P = 0.0037; object size [rocks only]: Fisher’s exact test P = 0.53; object size [woody objects only]: χ2 = 4.90, df = 2, P = 0.086; object texture: Fisher’s exact test P = 0.56). Object material** Woody Rock Bark Small 94 (84.4) 13 (18.8) 0 (3.8) Large 63 (72.6) 22 (16.2) 7 (3.2) Object size (rocks only) Small Medium Large Small 2 (1.5) 8 (7.1) 3 (4.5) Large 2 (2.5) 11 (11.9) 9 (7.5) Object size (woody objects only) Small Medium Large Small 39 (33.5) 23 (22.3) 32 (38.3) Large 17 (22.5) 14 (14.9) 32 (25.7) Object texture Solid Fibrous Chunky Crumbly Small 59 (58.1) 24 (26.9) 7 (6.0) 4 (3.0) Large 38 (38.9) 21 (18.1) 3 (4.0) 1 (2.0) **Indicates significance, P < 0.01. Table 5. The observed (and expected in parentheses) number of Plethodon cinereus (Redbacked Salamander) found either within or underneath woody cover objects categorized by texture, Green Mountain National Forest, Vermont, 2001 (χ2 = 44.7, df = 3, P < 0.001). Solid Fibrous Chunky Crumbly Within 24 (43.3) 33 (18.9) 10 (4) 1 (1.8) Underneath 95 (75.7) 19 (33.1) 1 (7) 4 (3.2) 2009 L.S. Richmond and S.C. Trombulak 21 small salamanders under cover objects in four categories: object material, object size for rocks only, object size for woody objects only, and object texture (Table 6). The only significant effect was for material type. Large salamanders were found more often than expected under rocks and less frequently than expected under woody objects. Discussion The density of salamanders in our study area (0.03 m-2) is lower than those reported in other studies (e.g., 0.258 m-2; Burton and Likens 1975a). This is not surprising given the fact that we limited our searches to cover objects and did not thoroughly search the leaf litter. For this reason, it is difficult to compare our density results with those reported from other regions. Studies indicate that occurrence of salamanders under cover objects can vary with weather (specifically relative wetness) and season (Faragher and Jaeger 1997, Moore et al. 2001). Precipitation data indicates that there were significant precipitation events during the sampling season. However, since these data were not recorded hourly, we are unable to determine if daily precipitation occurred before or after sampling. In addition, salamanders might respond in particular ways to precipitation events a day or even days before sampling. For the purposes of this study, we averaged all of the cover object data irrespective of weather. An area of further study would be to explore how precipitation events affect patterns of salamander distribution within and under cover objects. Our study supports many of published conclusions regarding Red-backed Salamander preference for cover objects. Our results show that salamanders are associated with only a small percentage of cover objects (7.4%), suggesting that, in the Green Mountains at least, individual salamanders have ample opportunity to select cover objects and are not forced to choose cover objects based on limited availability. With respect to cover-object material, Red-backed Salamanders were more common than would be expected under rocks. This result is consistent with the findings of Moore et al. (2001). In addition, Red-backed Salamanders were more common than expected under large cover objects (for both rocks and woody objects), supporting Mathis’ (1990) laboratory findings. Salamander prevalence under rocks could be explained by a number of factors. A single rock does not provide an opportunity for salamanders to reside inside of it, but we often found many rocks in one area, stacked next to and on top of each other, creating cool and moist crevices that the salamanders could inhabit. Also, of all object types, rocks seemed to form the tightest seal with the ground; they were often embedded in the soil rather than simply lying on the leaf litter. This tight seal might provide salamanders with increased protection from predators and increased moisture, which could be especially important during dry spells. 22 Northeastern Naturalist Vol. 16, No. 1 The relationship between salamander presence and cover object size is likely due to different microclimatic conditions under objects as a function of their size. Mathis (1990) found that in August in Virginia, soil temperature under large cover objects was significantly lower than under small objects, suggesting that larger objects provided more of a temperature buffer. Keen (1984) observed that larger objects are more likely to maintain adequate moisture; thus, salamanders might preferentially shelter under larger objects to protect themselves from desiccation. Fraser (1976) found that invertebrate prey increased as the moisture of the substrate increased, which might suggest that larger objects have higher prey densities. However, Gabor (1995) did not find a correlation between object size and amount of invertebrate prey. Therefore, prey availability might not fully explain increased salamander presence under large objects. In our study, Red-backed Salamanders exhibited a significantly nonrandom distribution with respect to cover-object texture. Salamanders were more common than expected under (and within) fibrous objects and less common that expected under solid objects. These results are consistent with observation of other studies, which found increased salamander presence in decaying logs compared to solid ones (Heatwole 1962). Fibrous woody objects offer additional spaces for salamanders to reside inside and places in which they can lay eggs (Heatwole 1962). In addition, the spaces within the fibrous logs could provide additional places for invertebrate prey to reside. The relationship between invertebrate prey occurrence and object texture could be an important area for further study, as we did not encounter any literature to support or deny this possibility. We were concerned about the results of the texture data because the sample sizes for chunky and crumbly texture were so low. In addition, object size could have influenced these results. Almost all small woody objects (less than 15 cm) were of solid texture. Since salamander counts were much lower than expected under small woody objects, this could skew the texture results. However, an analysis of texture that only looked at logs (woody objects consisting of the trunk of a tree generally greater than 15 cm in diameter) also showed a significant relationship between salamander presence and object texture (Fisher’s Exact Test: P = 0.0017), suggesting that wood texture plays a role in salamander distribution. When we explored the relationship between salamander size and coverobject characteristics, only object material type was significantly correlated. Large salamanders were more common than would be expected under rocks and small salamanders were more common than would be expected under woody objects. Previous analyses indicated overall salamander prevalence under rocks, suggesting that rocks could be a more favorable cover-object material. The increased presence of large salamanders under rocks could be the result of territorial behavior, where large salamanders out-compete small salamanders for favorable cover objects. Territoriality in Red-backed 2009 L.S. Richmond and S.C. Trombulak 23 Salamanders has been well documented (Mathis 1990, Townsend and Jaeger 1998). While our study was not designed to investigate the behavioral basis for differential distribution of cover objects by large and small salamanders, the pattern we observed in this field study is consistent with Mathis’ (1990) and Townsend and Jeager’s (1998) conclusion that large salamanders can out-compete small salamanders for resources, in this case for access to preferred cover objects. However, given the overall low densities of salamanders compared to available cover objects, territoriality may not fully explain the differential distribution of large and small salamanders. Heatewole’s (1960) descriptions of Red-backed Salamander burrowing suggest that increased head size and body strength might improve burrowing ability. Therefore, an alternative explanation for this differential distribution of large and small salamanders is that small salamanders are less physically able than large salamanders to work their way under rocks, perhaps due to the tighter seal rocks have with the ground. Thus, the relative paucity of small salamanders under rocks may be less a result of behavioral exclusion by large salamanders and more a result of active selection for objects that require less effort to crawl under. Distinguishing these two explanations would require direct observation of salamander behavior under conditions where access by small salamanders to rocks is controlled for both physical barriers and the presence of large salamanders. We did not observe a significant relationship between salamander size and cover-object size. This result is consistent with the work of Faragher and Jaeger (1997), who did not find a significant correlation between cover object size and salamander size on wet or dry days, as well as of Moore et al. (2001), who only found a correlation between salamander size and coverobject size during autumn. Thus, the size of a cover object may not generally correlate with characteristics used by salamanders to assess territory quality, such as prey availability (Gabor 1995, Gabor and Jaeger 1995). However, overall, salamanders were disproportionately common under large cover objects, suggesting that those objects have preferred characteristics. There may be a difference between cover objects that salamanders prefer and those that they compete for in territorial disputes. Also, it is possible that given the overall low densities of salamanders, large cover objects were not so limiting; thus, territorial competition may not have come into play. Our results indicate that Red-backed Salamanders are more common than would be expected under large rocks, or woody objects with fibrous texture; however, it is possible that territory selection may be based primarily on other characteristics, such as soil or leaf litter moisture (Grover 1998, 2000) or invertebrate prey abundance (Gabor and Jaeger 1995), which may only imperfectly correlate with object type, size, or texture. The results of this study have implications for the management of Redbacked Salamanders following forest clearing and for designing monitoring programs of forest health using salamanders as indicators. Much research 24 Northeastern Naturalist Vol. 16, No. 1 has documented the decline and slow recovery of salamander populations following forest management activities (Ash 1997, Ash and Bruce 1994, deMaynadier and Hunter 1998, Petranka et al. 1993). Some have suggested that these declines are linked to a decrease in soil and leaf-litter moisture following the removal of canopy (Ash 1997, Seastedt and Crossley 1981); however, Brooks and Kyker-Snowman (2008) found that forest clearing had little lasting impact on the forest-floor microclimate. In either case, the presence of favorable microhabitat is likely important to aid the recovery of salamanders following forest clearing. We found that salamanders were more common than expected under large cover objects, rocks, and fibrous wood, which could indicate that these are favorable object qualities. While forest harvesting activities can involve the leaving behind of coarse woody debris (McCarthy and Bailey 1994), Goodburn and Lorimer (1998) found that old-growth forests in Michigan and Wisconsin had significantly higher levels of coarse woody debris—compared with managed stands—both in biomass and average diameter. Our research on Red-backed Salamander distribution under cover objects suggests that the type of cover objects plays an important role in salamander distribution. If this is the case, the quality of the cover objects left behind may be just as important as the quantity. For example, in our research, salamander counts were higher under large woody objects, which are often less common in managed forest stands (Goodburn and Lorimer 1998). Managers looking to promote the reestablishment of Red-backed Salamander populations following forest clearing should consider protecting and possibly even placing more of the preferred types of cover objects in the cleared areas. Acknowledgments We would like to thank C. Casey and M. Burbank of the US Forest Service for providing access to study sites and digital maps of stand data. We would also like to thank N. Kieves and S. Perry for assistance in the field. In addition, we would like to thank the editor and two anonymous reviewers for their thoughtful comments on this manuscript. This research was supported through a Middlebury College Summer Research Fellowship from the Mellon Foundation. In this research, we complied with all applicable institutional Animal Care guidelines, and all required state and federal permits were obtained. Literature Cited Ash, A.N. 1997. Disappearance and return of plethodontid salamanders to clearcut plots in the southern Blue Ridge Mountains. Conservation Biology 11:983–989. Ash, A.N., and R.C. Bruce. 1994. Impacts of timber harvesting on salamanders. Conservation Biology 8:300–301. Brooks, R.T., and T.D. Kyker-Snowman. 2008. Forest floor temperature and relative humidity following timber harvesting in southern New England, USA. Forest Ecology and Management 254:65–73. Burton, T.M., and G.E. Likens. 1975a. Salamander populations and biomass in the Hubbard Brook Experimental Forest, New Hampshire. Copeia 1975:541–546. 2009 L.S. Richmond and S.C. Trombulak 25 Burton, T.M., and G.E. Likens. 1975b. Energy flow and nutrient cycling in salamander populations in the Hubbard Brook Experimental Forest, New Hampshire. Ecology 56:1068–1080. deMaynadier, P.G., and M.L. Hunter, Jr. 1995. The relationship between forest management and amphibian ecology: A review of the North American literature. Environmental Review 3:230–261. deMaynadier, P.G., and M.L. Hunter, Jr. 1998. Effects of silvicultural edges on the distribution and abundance of amphibians in Maine. Conservation Biology 12:340–352. Faragher, S.G., and R.G. Jaeger. 1997. Distributions of adult and juvenile Redbacked Salamanders: Testing new hypotheses regarding territoriality. Copeia 1997:410–414. Ferguson, H.J. 1998. Soil survey of Rutland County, Vermont. United States Department of Agriculture, Natural Resources Conservation Service and US Forest Service in cooperation with Vermont Agency of Natural Resources and Vermont Agricultural Experiment Station, Washington, DC. Fraser, D.F. 1976. Empirical evaluation of the hypothesis of food competition in salamanders of the genus Plethodon. Ecology 57:238–251. Gabor, C.R. 1995. Correlational test of Mathis’ hypothesis that bigger salamanders have better territories. Copeia 1995:729–735. Gabor, C.R., and R.G. Jaeger. 1995. Resource quality affects the agonistic behaviour of territorial salamanders. Animal Behaviour 49:71–79. Gergits, W.F., and R.G. Jaeger. 1990. Site attachment by the Red-backed Salamander, Plethodon cinereus. Journal of Herpetology 24:91–93. Goodburn, J.M., and C.G. Lorimer. 1998. Cavity trees and coarse woody debris in old-growth and managed northern hardwood forests in Wisconsin and Michigan. Canadian Journal of Forest Resources 28:427– 437. Grover, M.C. 1998. Influence of cover and moisture on abundances of the terrestrial salamanders Plethodon cinereus and Plethodon glutinosus. Journal of Herpetology 32:489–497. Grover, M.C. 2000. Determinants of salamander distributions along moisture gradients. Copeia 2000:156–168. Heatwole, H. 1960. Burrowing ability and behavioral responses to desiccation of the salamander, Plethodon cinereus. Ecology 41:661–668. Heatwole, H. 1962. Environmental factors influencing local distribution and activity of the salamander, Plethodon cinereus. Ecology 43:460–472. Herpetological Animal Care and Use Committee. 2004. Guidelines for Use of Live Amphibians and Reptiles in Field and Laboratory Research (2nd Edition). American Society of Ichthyologists and Herpetologists, Lawrence, KS. Jaeger, R.G., and W.F. Gergits. 1979. Intra- and interspecific communication in salamanders through chemical signals on the substrate. Animal Behaviour 27:150–156. Jaeger, R.G., D. Kalvarsky, and N. Shimizu. 1982. Territorial behavior of the Redbacked Salamander: Expulsion of intruders. Animal Behaviour 30:490–496. Jaeger, R.G., J.A. Wicknick, M.R. Griffis, and C.D. Anthony. 1995. Socioecology of a terrestrial salamander: Juveniles enter adult territories during stressful foraging periods. Ecology 76:533–543. 26 Northeastern Naturalist Vol. 16, No. 1 Keen, W.H. 1984. Influence of moisture on the activity of a plethodontid salamander. Copeia 1984:684–688. Kleeberger, S.R., and J.K. Werner. 1982. Home range and homing behavior of Plethodon cinereus in northern Michigan. Copeia 1982:409–415. Marsh, D.M., and M.A. Goicochea. 2003. Monitoring terrestrial salamanders: Biases caused by intense sampling and choice of cover objects. Journal of Herpetology 37:460–466. Mathis, A. 1990. Territoriality in a terrestrial salamander: The influence of resource quality and body size. Behaviour 112:162–175. McCarthy, B.C., and R.R. Bailey. 1994. Distribution and abundance of coarse woody debris in a managed forest landscape of the central Appalachians. Canadian Journal of Forest Research 24:1317–1329. Moore, A.L., C.E. Williams, T.H. Martin, and W.J. Moriarity. 2001. Influence of season, geomorphic surface, and cover item on capture, size, and weight of Desmognathus ochrophaeus and Plethodon cinereus in Allegheny Plateau riparian forests. American Midland Naturalist 145:39–45. Petranka, J.W., M.E. Eldridge, and K.E. Haley. 1993. Effects of timber harvesting on southern Appalachian salamanders. Conservation Biology 7:363–370. The R Foundation for Statistical Computing. 2005. R. Version 2.1.1. Vienna, Austria. Seastedt, T., and D. Crossley, Jr. 1981. Microarthropod response following cable logging and clear-cutting in the southern Appalachians. Ecology 62:126–135. Townsend, V.R., Jr., and R.G. Jaeger. 1998. Territorial conflicts over prey: Domination by large male salamanders. Copeia 1998:725–729. Welsh, H.H., and S. Droege. 2001. A case for using plethodontid salamanders for monitoring biodiversity and ecosystem integrity of North American forests. Conservation Biology 15:558–569.