Nest Success and Attributes of Brood Crevices Selected by Green Salamanders (Aneides aeneus) on the Blue Ridge Escarpment.
The green salamander (Aneides aeneus) is a rare species associated with rock outcrop habitats of the Appalachian Plateau and Blue Ridge Province (Petranka, 1998). As a result of its close association with rock habitat, this species has a patchy distribution and ranges from southwestern Pennsylvania to northern Alabama and extreme northeastern Mississippi (Hammerson, 2004; Pauley and Watson, 2005). There is also a group of disjunct populations on the Blue Ridge Escarpment in southwestern North Carolina and adjacent areas in South Carolina and Georgia (Hammerson, 2004; Pauley and Watson, 2005).
The green salamander is listed as either Rare or Endangered in most of the states in which it occurs (Brodman, 2004). In North Carolina it is listed as Threatened (North Carolina Administrative Code, 2017) and considered a Species of Greatest Conservation Need in the North Carolina Wildlife Action Plan (North Carolina Wildlife Resources Commission, 2015). The green salamander is listed as a Federal Species of Concern (LeGrand et al., 2014), and is currently under status review because of its continual decline throughout its range (Federal Register, 2015). Threats to the species include loss and alteration of habitat, over-collecting, epidemic disease, and climate change (Corser, 2001; Wilson, 2001).
The Blue Ridge Escarpment is thought to be a refugium for green salamanders given its stable microclimates during periods of widespread climate change (Bruce, 1968). Prior to the mid-1970s, green salamanders occurred at relatively high densities on the escarpment (Gordon, 1952; Bruce, 1968; Snyder, 1971). However, a dramatic decline in populations occurred in the mid-late 1970s (Snyder, 1991), and continued at least through the 1990s (Corser, 2001; Wilson, 2001), with no definitive causes known (Corser, 2001). Since Corser (2001), no quantitative studies have been published regarding the reproductive success of this species.
Green salamanders are habitat specialists that require exposed rock with crevices that provide specific microhabitats (Gordon and Smith, 1949; Gordon, 1952; Snyder, 1971; Canterbury, 1991; Petranka, 1998). They also use large trees that are in close proximity to rock outcrops for foraging as well as nesting (Waldron and Humphries 2005; Smith et al., 2017). Green salamanders use two types of rock crevices: breeding (herein referred to as brood) and transitory (Gordon, 1952). These specialized crevices are considered the basis of a green salamander's territory, as they are defended by an individual throughout the breeding season (Cupp, 1980). Gravid females use brood crevices for oviposition and brooding eggs and hatchlings (Gordon, 1952; Snyder, 1971; Canterbury, 1991; Cupp, 1991). Brood crevices are often used perennially and located in either the male's crevice where mating occurred or more commonly in a crevice in close proximity to the male's crevice (Gordon, 1952; Snyder, 1971; Canterbury, 1991). Generally, there is one female per brood crevice, although occasionally two females will nest in the same crevice (Gordon and Smith, 1949; Gordon, 1952). Because green salamanders breed biennially (Canterbury and Pauley, 1994; Petranka, 1998), different females occupy brood crevices every other year (Snyder, 1971; Canterbury, 1991). Transitory crevices are less specialized than brood crevices and are used by juveniles, males, and non-breeding females (Gordon, 1952; Canterbury, 1991). Transitory crevices serve as daytime refuges that provide stable microclimatic conditions (Gordon, 1952; Canterbury, 1991; Cupp, 1991). The use of transitory crevices can vary from one day to an entire breeding season (Gordon, 1952; Canterbury, 1991), depending on the crevice attributes (Rossell et al., 2009; Smith et al., 2017) and the sex and age of the individual occupying the crevice (Canterbury, 1991).
Understanding the relationship between microhabitat preference and reproductive success is important for developing effective conservation strategies (Morrison et al., 1992). Although numerous studies have reported microhabitat characteristics of brood crevices, including dimensions (e.g., Gordon and Smith, 1949; Gordon, 1952; Cupp, 1971; Snyder, 1971; Canterbury, 1991; Davis, 2004), no studies have examined the microhabitat characteristics of brood crevices in relation to nest success, or compared those characteristics to what is available in the surrounding environment. Our primary objectives in this study were to 1) investigate the nest success of green salamander populations during six nesting seasons on the Blue Ridge Escarpment, 2) investigate the attributes of brood crevices in relation to nest success, and 3) compare the attributes of brood crevices to those randomly available in the surrounding environment.
Our study area encompassed breeding locations of green salamanders in and around DuPont State Recreational Forest (35.19173[degrees]N, 83.62245[degrees]W; 4189 ha), in Henderson and Transylvania Counties, North Carolina (Fig. 1). This area provides habitat for the highest estimated densities of green salamanders known in North Carolina (Williams, pers. comm.) and is located on the southwest portion of the Blue Ridge Escarpment in the Blue Ridge Physiographic Province. Hardwoods dominate the forest communities of the area and are comprised mostly of second-growth Montane Oak ((hiercus)-Hickory (Carya) and Acidic Cove Forests (Schneider, 2011). Topography is diverse and includes broad rolling hills, with some exposed granitic domes at higher elevations, and gorges formed by the Little River and its tributaries at lower elevations (elevation range: 679-1097 m; Schneider, 2011). The area contains an abundance of exposed rock in the form of various-sized outcrops and rock faces and is part of the Granite gneiss unit of the Chauga Belt (North Carolina Geologic Survey, 1985). Ashe-Edneyville complex soil and Ashe stony sandy loam soil are the primary upland soils (Schneider, 2011). Temperatures of the region are considered mild, with average annual temperatures ranging from 9 to 22 C, and annual precipitation ranging from 152 to 178 cm (Schneider, 2011).
We used data from six green salamander nesting seasons from June 2010 to November 2015. Data were collected as part of an on-going monitoring program by the North Carolina Wildlife Resources Commission. As part of the program, each nest site was georeferenced using Global Positioning System information (GPS; Garmin GPSMAP 62S, Olathe, KS) and the general dimensions of the brood crevice and its position on the rock were recorded and photographed to assist in long-term tracking. We used data only from nests where success or failure was determined, following Snyder's (1971) definition of nest success as evidence of at least one egg hatching in a clutch. This allowed a direct comparison of our results with Snyder (1971), the only study to date to quantify green salamander nest success. We considered a nest successful when we directly witnessed eggs hatching, observed hatchlings in the brood crevice, or documented fresh remnants of hatched eggs at the end of August or early September, the typical hatching period in North Carolina. Frequency of monitoring varied by nest, and depended on seasonal time constraints of field staff and access of nest sites. Nests were monitored using a flashlight to look into crevices to observe nests, brooding females and hatchlings a minimum of three times during each nesting season. We also noted the presence of potential predators at rock outcrops during monitoring.
We measured attributes of all brood crevices that were documented from 2010 to 2015 following methods of Rossell et al. (2009). For each brood crevice, we measured the total length (measured using a string placed along the bottom edge of the crevice opening on the rock face), maximum depth (measured using a flattened wooden probe), maximum opening width (measured using a tape measure), drip-edge (horizontal distance of cover provided by the closest overhanging ledge above the crevice, measured using a plumb bob and tape measure), crevice angle (angle of crevice above horizontal estimated using a protractor and level), and height above the ground (measured from the lowest point of the crevice opening). We also noted the presence of any internal fissures that would allow green salamanders deeper access into the rock (herein referred to as anastomosing fissures; Gordon, 1952).
To investigate whether attributes of brood crevices differed from randomly available crevices, we identified a subset of rock outcrops where all crevices could be surveyed. We further limited our sample to only those outcrops where we were 100% confident of the location of every brood crevice used during the six-year study period based on our monitoring (>2500 total visits to rock outcrops). For each outcrop that fit these criteria, we paired each brood crevice with a randomly selected crevice on the same outcrop for comparison. We located a random crevice using a random numbers table to determine the side of the brood crevice the random crevice was to be located, and to determine the height above the ground measured along a vertical line 1 m from the brood crevice. We considered a random crevice suitable if its opening was at least 0.6 cm wide (i.e., the minimum opening size of brood crevices recorded during the study period), its length or depth was at least 8 cm (the minimum size to accommodate an adult green salamander; total length of adult green salamander = 8-14 cm; Petranka, 1998), and it was never used as a brood crevice during the study period. If we were unable to locate a suitable crevice at the random point, we used a random numbers table to generate a direction (0-360[degrees]) to follow until a suitable crevice was found. In the event there was no suitable crevice in that direction, we repeated the procedure until one was located. Once a suitable random crevice was selected, we recorded its attributes as described above.
We calculated straight-line distances between closest outcrops with nesting green salamanders using the measure tool in ArcGIS (ESRI, Redlands, CA, version 10.5). We summarized attributes of all brood and random crevices using median, quartiles (Q3 = third quartile and Q1 = first quartile), and interquartile range (IQR) given several of the variables exhibited positive skewness (i.e., length, maximum depth, maximum opening width, and drip edge). We used logarithmic transformation on all numeric variables except height above the ground and crevice angle to improve their normality. To investigate the association of nest success with crevice attributes and straight-line distance (SLD) to closest outcrop with nesting green salamanders, we used a generalized linear mixed model with nest success as a binary response variable and crevice attributes and SLD to closest outcrop as predictor variables. We included nest as a random effect. We used a stepwise selection procedure, selecting the variable with the lowest p-value, to develop the final model. We used the Type III F-test to calculate p-values for the fixed predictor variables, and the likelihood ratio test to examine the significance of the random effect of nest.
For the paired crevice data, we used a conditional logistic regression model using a stepwise procedure to determine whether green salamanders used any attribute differently than was randomly available. We set an alpha of P < 0.05 to retain variables in the model. The p-value for assessing the significance of the relationship was based on the Wald chisquare test. We considered all statistical comparisons significant at alpha < 0.05. An attribute was determined selected if its use differed from its random availability. We used Statistical Analysis System (SAS; SAS Institute, Cary, NC, version 9.4) programs for all analyses.
We conducted a total of 2578 rock outcrop surveys of 168 green salamander nests during 2010-2015. Each outcrop was visited on average 14 times per season (range: 1-54). Nest success ranged from 73% to 92% (mean = 83.3%, Table 1), and there was no significant effect of year on success (F= 0.05, df = 5, P = 0.82). Of the 140 successful nests, we observed at least one hatchling at 138 nests and fresh remnants of hatched eggs at two nests. A total of 28 nests failed. Of those, 36% (n = 10) did not have a brooding female present (or was present and then disappeared), and the crevices (n = 10) all contained potential predators of green salamander eggs. These included six brood crevices with cave crickets (Ceuthophilis spp.), three with slugs (Philomycidae), and one with a mouse (Peromyscus spp.) nest. We were unable to determine whether any missing females were predated themselves or if they abandoned nests or consumed their own eggs. Of the remaining nests that failed, brooding females were absent in all cases, and brood crevices were either totally devoid of eggs (n = 14) or contained eggs that were severely desiccated (n = 2), or covered in fungus (n = 2).
Nests occurred in 88 crevices from 74 individual rock outcrops. Seventy-one outcrops (96%) contained one brood crevice, while two outcrops contained two brood crevices (3%), and one outcrop contained four brood crevices (1%). We recorded 160 occurrences of potential predators at the rock outcrops over the course of the study, including seven species of salamanders and two species of snakes (Appendix 1). Summary statistics of brood crevice attributes are provided in Table 2. Of the 88 brood crevices, 53% (n = 47) had anastomosing fissures and 47% (n = 41) did not. The first iteration of the stepwise linear mixed model showed no significant relationships between any of the crevice attributes and nest success (all P > 0.05; Table 3), indicating the independent effect of each variable.
Straight-line distance to closest outcrop with nesting green salamanders ranged from 3 to 891 m (median = 70 m, Table 2). There was a significant negative relationship between nest success and SLD to closest outcrop in the first iteration of the linear mixed model (P = 0.043, Table 3, Fig. 2). The second iteration of the mixed model indicated no significant relationships between any of the crevice attributes and nest success after adjusting for SLD to closest outcrop, which had the smallest p-value in the first iteration (Table 3). The odds ratio estimate for log base 2 of SLD to closest outcrop was 0.812 (95% CI = 0.664, 0.994), indicating that the odds of a successful nest decrease by a factor of 0.812 as the SLD to closest outcrop with nesting green salamanders doubles. There was also a significant nest effect after adjusting for SLD to closest outcrop, indicating that crevices that had one successful nest were more likely to have a successful nest in another year, irrespective of any crevice attribute or distance to closest outcrop with nesting green salamanders ([chi square] = 4.65, df = 1, P = 0.016).
Twenty-nine brood crevices, representing 29 outcrops, were used to compare attributes of occupied and random crevices. Table 4 provides summary statistics of occupied and random brood crevices. Presence of anastomosing fissures was similar for both occupied and random crevices; 45% (n = 13) contained fissures and 55% (n = 16) did not. The first iteration of the stepwise logistic regression model indicated that length of occupied crevices was significantly less than random crevices (P = 0.033), and height above the ground of occupied crevices was significantly greater than random crevices (P = 0.023, Table 5). The second iteration of the regression model indicated no significant differences between any of the crevice attributes after adjusting for height above the ground, which had the smallest p-value in the first iteration (Table 5). The odds ratio estimate for this attribute was 1.027 (95% CI = 1.004, 1.051); thus, the odds of a crevice used for brooding increases by a factor of 1.027 as the height above the ground increases by 1 cm.
Nest success of green salamanders during 2010-2015 (range 73-92%, mean = 83.3%, Table 1) was comparable, if not higher, than success reported by Snyder (1971) for the same region in 1970 (range 60-80%, mean = 75.6%; Snyder, 1971), a time when green salamander populations on the Blue Ridge Escarpment were considered robust (Snyder, 1991; Corser, 2001). We were unable to determine the cause of nest failure with certainty. However, we suspect predation played a major role, as 36% of brood crevices with failed nests (n = 10) contained species known to consume green salamander eggs and hatchlings (Snyder, 1971). We also documented 160 observations of other potential predators at the outcrops over the course of the study (Appendix 1). Snyder (1971) reported that predation also was the likely cause of most nest failures in his study, although he indicated that some nests disappeared as a result of brooding females consuming their eggs.
The lack of a year effect on nest success suggests that weather did not influence reproduction of green salamanders during our study. Total annual precipitation during our study fell within the normal range of the region, ranging from 150 to 179 cm, except for 2013, which was an exceptionally wet year with a total annual precipitation of 263 cm (U.S. Climate Data, 2017). Adverse weather conditions have been suggested as the proximal cause of reproductive failure in several populations of green salamanders on the Blue Ridge Escarpment (Snyder, 1971; Snyder, 1991; Corser, 2001). Snyder (1971) and Corser (2001) reported severe drought during the brooding period as the likely cause in several cases of nest abandonment. In addition, Snyder (1991) and Corser (2001) speculated that bouts of extreme cold weather during late winter-early spring were responsible for several population crashes on the Blue Ridge Escarpment in the late 1970s and 1990s, respectively.
Nest success of green salamanders is directly related to the survival and fitness of adult females (Snyder, 1971). In our study, nest success significantly decreased as straight-line distance to closest outcrop with nesting green salamanders increased. We are unsure of the reasons for this and can only speculate without additional data. However, one possibility is that fitness of females decreases the more isolated productive rock outcrops are from one another, because of a lack of gene flow. The extent of dispersal by green salamanders is poorly understood. Green salamanders are considered sedentary, and long-distance migrations to outcrops are generally less than 50 m (Gordon, 1952; Woods, 1969; Canterbury, 1991; Pauley and Watson, 2005). The longest distance moved by a green salamander was documented by Gordon (1952), who recovered an immature male 107 m from its original capture site. In our study area, the straight-line distance to closest outcrop with nesting green salamanders ranged from 3 to 891 m (median = 70 m). Therefore, based on the above information it is conceivable that green salamanders from the most isolated outcrops have higher levels of inbreeding depression and lower rates of nest success compared to those from outcrops in closer proximity to each other because of lack of gene flow between populations. This speculation is supported by Cabe et al. (2007) who reported small, but significant genetic differentiation among populations of red-backed salamanders (Plethodon anerus) separated by as little as 200 m of continuous habitat. In addition, decreased fitness was associated with inbreeding depression in isolated populations of the rare cliff-dwelling species Centaurea corimbosa (Colas et al, 1997). In both these studies, the authors attributed their findings to lack of gene flow and limited dispersal (Colas et al., 1997; Cabe et al, 2007). However, additional studies on genetics and reproductive success of green salamander populations are needed to help confirm our supposition.
Dimensions of brood crevices in our study were comparable to those previously reported on the Blue Ridge Escarpment by Gordon (1952) and Snyder (1971). Brood crevices were relatively short horizontal crevices with small opening widths located on the upper portion of a rock face (Table 2). Our data generally agree with descriptions reported in the literature (e.g., Gordon and Smith, 1949; Gordon, 1952; Snyder, 1971; Petranka, 1998), and highlights the specialized habitat requirements of brood crevices. Attributes of brood crevices presumably help provide optimal conditions for nest success (Gordon, 1952; Snyder, 1971; Cupp, 1991). We did not find any crevice attribute significantly related to nest success. However, there was a significant nest effect, which suggests that there were attributes not quantified (e.g., microclimate of crevices) that give some nests a higher probability of success than others. Because nest effects were confounded with crevice attributes, it is possible that there were attributes (e.g., height above the ground; Table 3) that had an effect on success, but were not significant in the model. Therefore, additional data are needed to further explore these relationships.
We found crevices that contained one successful nest were more likely to have a successful nest in another year, regardless of their attributes or the distance to closest outcrop with nesting green salamanders. It is possible that larger, more fecund green salamanders choose brood crevices where they had a prior nest success. Previous researchers have reported that individual females have an affinity for using the same brood crevice in successive years (Gordon, 1952; Cupp, 1991). Egg and clutch size are both critical determinants for reproductive success in salamanders (Reinhard et al., 2015). Although positive correlations between body size and egg and clutch size have not yet been demonstrated for green salamanders, they have been reported for several other plethodontid species, including arboreal salamander (A. lugubris; Anderson, 1960), California slender salamander (Batrachoseps attenuatus; Anderson, 1960), Allegheny mountain dusky salamander (Desmogriathus ochrophaeus; Martof and Rose, 1963), long-nosed salamander (Bolitoglossa rostrata-, Houck, 1977), and red-backed salamander (Fraser, 1980). Another possible explanation is that certain crevices provide more suitable conditions for nesting than others, and that factors responsible for those conditions were not measured in our study (e.g., microclimate of crevices).
In our study green salamanders nested in crevices that were significantly higher above the ground than those available in the surrounding habitat. Height above the ground was the only attribute significantly different from random crevices in the final regression model (Table 5), highlighting its importance. Females likely prefer brood crevices higher on a rock because they reduce competition and potential nest predation from other crevice-dwelling plethodontid salamanders and from other predators in general (Cliburn and Porter, 1986, 1987; Pauley and Watson, 2005). Similarly, field observations and laboratory tests by Cliburn and Porter (1987) found that green salamanders in northeastern Mississippi stratified vertically in the presence of northern slimy salamander (P. glutinosus), with green salamanders selecting higher crevices. Although we did not observe any other species of salamander inhabiting brood crevices during our study, there were numerous observations of other salamander species occupying the same rock outcrops where green salamanders were nesting (Appendix 1).
We found crevice length also an important attribute of brood crevices, as it was statistically significant in the first iteration of the regression model and nearly significant (P = 0.057) in the second iteration (Table 5). Females chose crevices that were substantially shorter than those available (Table 4), suggesting that shorter crevices may be easier to defend, or possibly provide better microclimatic conditions for brooding young. Rossell et al. (2009) and Smith et al. (2017) both reported that the length of transitory crevices did not differ from those available on rock outcrops at their study sites in North Carolina and Virginia, respectively. These contrasting results between brood and transitory crevices provide evidence that the habitat requirements for brood crevices are more specialized than transitory crevices.
Several attributes (i.e., width, depth, angle, drip edge, anastomosing fissures) did not differ from those available in the surrounding habitat (Table 5), suggesting these attributes were not limiting factors when females choose a brood crevice. Numerous authors have postulated that narrow crevices are a requisite of green salamander habitat (e.g., Gordon and Smith, 1949; Gordon, 1952; Bruce, 1968; Snyder, 1972; Cupp, 1980), and our data support this premise, as maximum opening widths of occupied and available brood crevices were small (Table 4) and indicative of outcrops containing reproductive animals. Green salamanders are uniquely adapted for inhabiting narrow crevices given their flattened bodies (Conant and Collins, 1998; Petranka, 1998). Thus, small opening width is likely an important attribute of a brood crevice, because it would reduce direct competition and potential nest predation from larger woodland salamanders (Gordon, 1952) or other predators. Crevices with small openings also may be more efficient at maintaining high relative humidity levels and constant temperatures (Rossell et al., 2009), conditions that are important for maintaining the viability of eggs (Snyder, 1971). Anastomosing fissures are another attribute considered to be important to green salamanders, as they provide access to deeper crevices for hibernation (Gordon, 1952; Cupp, 1991). Our results suggest that anastomosing fissures are not necessarily an important component of brood crevices, as only 53% of all brood crevices (N = 88) in our study contained this type of fissure.
In summary our study investigated the nest success and microhabitat characteristic of brood crevices selected by green salamanders on the Blue Ridge Escarpment over a 6 y period. We found nest success in our study population was comparable to historic levels reported in the 1970s for the same region. Nest success was not related to year or any brood crevice attribute. However, it was inversely related to distance between rock outcrops with reproducing green salamanders. In addition, brood crevices that contained one successful nest were more likely to have a successful nest in another year, irrespective of crevice attribute or distance to closest outcrop with nesting green salamanders. Females selected specialized crevices for brooding that were higher above the ground and shorter in length than those available in the surrounding habitat, attributes that likely minimize competition and potential nest predation from other species of salamanders and other predators in general. Based on these results, we suggest future research focus on the genetics and reproductive success of green salamanders, the macrohabitat of the forest surrounding outcrops that contained breeding populations, and the microclimatic conditions that produce
productive brood crevices.
Acknowledgments.--We thank the North Carolina Forest Service for project support, B. Miller for assisting with geological information, and T. Forrest for information on invertebrates. We also thank K. Weeks and A. Boynton for administrative and programmatic support.
Anderson, P. 1960. Ecology and evolution in island populations of salamanders in the San Francisco Bay region. Ecol. Monogr., 30:359-385.
Brodman, R. 2004. R9 species conservation assessment for the green salamander, Aneides aeneus (Cope and Packard). U.S. Forest Service, Eastern Region of the Forest Service Threatened and Endangered Species Program, Milwaukee, Wl. 17 p.
Bruce, R. C. 1968. The role of the Blue Ridge Embayment in the zoogeography of the green salamander, Aneides aeneus. Herpetologica, 24:185-194.
Cabe, P. R., R. B. Page, T.J. Hanlon, M. E. Aldrich, L. Connors, and D. M. Marsh. 2007. Fine-scale population differentiation and gene flow in a terrestrial salamander (Plethodon cinerus) living in a continuous habitat. Heredity, 98:53-60.
Canterbury, R. A. 1991. Ecology of the green salamander, Aneides aeneus, (Cope and Packard), in West Virginia. M.S. Thesis, Marshall University, Huntington, WV. 186 p.
--AND T. K. Pauley. 1994. Timing of mating and egg deposition of West Virginia populations of the salamander Aneides aeneus. J. Herpetology, 28:431-434.
Colas, B., I. Oliyieri, and M. Riba. 1997. Centaurea corymbosa, a cliff-dwelling species tottering on the brink of extinction: a demographic and genetic study. Proc. Natl. Acad. Sci., 94:3471-3476.
Conant, R. and J. T. Collins. 1998. A field guide to reptiles and amphibians of eastern and central North America, 3rd ed. Houghton Mifflin Company, New York, NY. 616 p.
Corser, J. 2001. Decline of disjunct green salamander (Aneides aeneus) populations in the southern Appalachians. Biol. Consent., 97:119-126.
Cuburn, J. W. and B. Porter. 1986. Comparative climbing abilities of the salamanders Aneides aeneus and Plethodon glutinosus (Caudata, Plelhodontidae). J. Mississippi Acad. Sci., 31:91-96.
--and-- 1987. Vertical stratification of lhe salamanders Aneides aeneus and Plethodon glutinosus (Caudata, Plelhodontidae)./ Alabama Acad. Sci., 58:18-22.
Cupp, P. V., Jr. 1971. Fall courtship of the green salamander Anedies aeneus. Herpetologica, 27:38-310.
--1980. Territoriality in the green salamander, Aneides aeneus. Copeia, 1980:463-468.
--1991. Aspects of the life history and ecology of the green salamander, Aneides aeneus, in Kentucky. J. Tennessee Acad. Sci., 66:171-174.
Dams, A. 2004. Aneides aeneus (green salamander) and Plethodon glutinosus (slimy salamander). Nesting observations. Heipetological Rev.. 35:51-52.
Federal Register. 2015. Status review green salamander. Federal Register, 80:37569. http://www.gpo. gov/fdsys/pkg/FR-2015-07-01/pdf/2015-16001.pdf. Accessed 2 November 2016.
Fraser, D. 1980. On the environmental control of oocyte maturation in a plethodontid salamander. Oecologia, 146:302-307.
Gordon, R. E. 1952. A contribution to the life history and ecology of the plethodontid salamander Aneides aeneus (Cope and Packard). Amer. Midi. Nat., 47:666-701.
--. and R. L. Smith. 1949. Notes on the life history of the salamander Aneides aeneus. Copeia, 1949:173-175.
Hammerson, G. 2004. Desmognathus wrighti. The IUCN red list of threatened species, http://dx.doi.org/ 10.2305/IUCN.UK. 2004. RLTS.T59259A11907504.en. Accessed 18 April 2016.
Houck, L. D. 1977. The reproductive biology of a neotropical salamander, Bolitoglossa rostrate (Urodela: Plethodontidae). Copeia, 1977:70-83.
Legrand, H. E., Jr., J. A. Ratcliffe, and J. T. Finnegan. 2014. Natural Heritage Program List of the Rare Animal Species of North Carolina. North Carolina Natural Heritage Program, North Carolina Department of Environment and Natural Resources, Raleigh.
Martof, B. and F. Rose. 1963. Geographic variation in southern populations of desmognathus ochrophaeus. Amer. Midi. Nat., 69:376-425.
Morrison, M. L., B. G. Marcot, and R. W. Mannan. 1992. Wildlife-habitat relationships: concepts and applications. University of Wisconsin Press, Madison. 364 p.
North Carolina Administrative Code. 2017. Threatened species list, http://reports.oah.state.nc.us/ncac. Accessed 12 February 2018.
North Carolina Geologic Survey. 1985. Geologic map of North Carolina. Department of Natural Resource and Communication Development, Division of Land Resource, Raleigh, North Carolina.
North Carolina Wildlife Resources Commission. 2015. North Carolina wildlife action plan. North Carolina Wildlife Resources Commission, Raleigh.
Pauley, T. K. and M. B. Watson. 2005. Aneides aneus, p. 656-658. In: M. Lannoo (ed.), Amphibian declines, the conservation status of United States species. University of California Press, Berkeley.
Petranka, J. W. 1998. Salamanders of die United States and Canada. Smithsonian Institution Press, Washington, DC. 587 p.
Reinhard, S., S. Renner, and A. Kupper. 2015. Age and fecundity in Salamandra algira (Caudata: Salmandridae). Salamandra, 51:19-24.
Rossell, C. R., Jr., J. Hicks, L. A. Williams, and S. C. Patch. 2009. Attributes of rock crevices selected by green salamanders, Aneides aeneus, on the Blue Ridge escarpment. Herpetological Rev., 40:151-153.
Schafale, M. P. 2012. Guide to the natural communities of North Carolina, Fourth Approximation. North Carolina Natural Heritage Program, Department of Environment and Natural Resources, Raleigh. 208 p.
Schneider, B. J. 2011. DuPont State Forest land and resource management plan. North Carolina Division of Forest Resources, Raleigh. 63 p.
Smith, W. H., S. L. Slemp, C. D. Stanley, M. N. Blyckburn, and J. Wayland. 2017. Rock crevice morphology and forest contexts drive microhabitat preferences in the green salamander (Anades aeneus). Can. J. Zool., 95:353-358.
Snyder, D. H. 1971. The function of brooding behavior in the Plethodontid salamander, Aneides aeneus: a field study. Ph.D. Dissertation, University of Notre Dame, South Bend, IN. 140 p.
--1991. The green salamander (Aneide aeneus) in Tennessee and Kentucky, with comments on the Carolinas' Blue Ridge population. J. Tennessee Acad. Sa., 66:165-169.
United States Climate Data. 2017. http:/www.usclimatedata.com. Accessed 1 August 2017.
Waldron, J. L. and W. J. Humphries. 2005. Arboreal habitat use by the green salamander, Aneides aeneus, in South Carolina. J. Herpetology, 39:486-492.
Wilson, C. R. 2001. Status of the green salamander (Aneides aeneus) in the Blue Ridge Mountains. M.S. Thesis, Appalachian State University, Boone, North Carolina. 73 p.
Wood, J. E. 1969. The ecology and natural history of Mississippi populations of Aneides aeneus and associated salamanders (abstract). Dissertation Abstracts International, 29B:3554.
Submitted 5 March 2018
Accepted 11 September 2018
Appendix 1. Number of potential predators observed during surveys (N = 2578) of 74 rock outcrops containing green salamanders on the Blue Ridge Escarpment in North Carolina, 2010-2015. Surveys occurred from June to November during each nesting season, and none of the potential predators were observed in green salamander brood crevices Species Scientific name Southern Gray-cheeked Salamander Plethodon metcalfi White-spotted Slimy Salamander Plethodon cylindraceus Seal Salamander Desmogn ath us mon ticola Northern Dusky Salamander Desirwgnathus fuscus Unidentified Dusky Salamander Desmognathus spp. Blue Ridge Two-lined Salamander Euiycea wilderae Three-lined Salamander Eurycea guttolineata Red-spotted Newt Notophthalmus v. viridescens Ringneck Snake Diadophis punctatus Eastern Ratsnake Pantherophis alleghaniensis Species No. of observations Southern Gray-cheeked Salamander 132 White-spotted Slimy Salamander 1 Seal Salamander 5 Northern Dusky Salamander 1 Unidentified Dusky Salamander 1 Blue Ridge Two-lined Salamander 9 Three-lined Salamander 1 Red-spotted Newt 5 Ringneck Snake 1 Eastern Ratsnake 2
C. REED ROSSELL, JR. (1)
Department of Environmental Studies, University of North Carolina, Asheville 28804
LORI A. WILLIAMS
North Carolina Wildlife Resources Commission, 1722 Mail Service Center, Raleigh 27699
ALAN D. CAMERON
1030 West Blue Ridge Road, Flat Rock, North Carolina 28731
CHARLES R. LAWSON
North Carolina Wildlife Resources Commission, 1722 Mail Service Center, Raleigh 27699
STEVEN C. PATCH
Department of Mathematics, University of North Carolina, Asheville 28804
(1) Corresponding author: e-mail: firstname.lastname@example.org
Caption: Fig. 1.--Study area of green salamander nest sites on the Blue Ridge Escarpment in North Carolina, 2010-2015
Caption: Fig. 2.--Box-plots of successful (n = 140) and failed nests (n = 28) of green salamanders in relation to straight-line distance (SLD) to closest outcrop with nesting green salamanders on the Blue Ridge Escarpment in North Carolina, 2010-2015. Nest success was defined as evidence of al least one egg hatching in a clutch, and SLD was estimated using ArcGIS
Table 1.--Percent of green salamander nests determined successful on the Blue Ridge Escarpment in North Carolina, 2010--2015. Nest success was defined as evidence of at least one egg hatching in a clutch, n = total number of nests monitored Year n % successful 2010 12 91.7 2011 33 87.9 2012 42 76.2 2013 30 73.3 2014 25 92.0 2015 26 88.5 Table 2.--Median, first quartile (Ql), third quartile (Q3), and interquartile range (IQR) of attributes of brood crevices (N = 88) and straight-line distances to closest outcrops with nesting green salamanders (N = 74) on the Blue Ridge Escarpment in North Carolina, 2010-2015. Straight-line distance (SLD) was estimated using ArcGIS Attribute Median Q1 Q3 IQR Length (cm) 53.0 32.2 77.8 45.6 Maximum opening width (cm) 2.3 1.8 4.1 2.4 Maximum depth (cm) 13.6 10.4 20.9 10.6 Height above the ground (cm) 61.5 49.0 90.0 41.0 Angle above horizontal 14.5 8.3 24.0 15.8 ([degrees]) Drip edge (cm) 9.9 6.2 27.1 21.0 SLD (m) 70.0 16.0 283.5 267.5 Table 3.--P-values and odds ratio estimates (ORE) of the stepwise procedure using a generalized linear mixed model to examine the relationship between nest success and attributes of brood crevices and straight-line distance (SLD) to closest outcrop with nesting green salamanders on the Blue Ridge Escarpment in North Carolina, 2010-2015. Odds ratio estimates less than one indicate a possible negative association between predictor and nest success, and ORE greater than one indicate a possible positive association between predictor and nest success Attribute P-value ORE P-value adjusted for SL.D Length (cm) 0.208 1.496 0.716 Maximum opening width (cm) 0.131 1.859 0.219 Maximum depth (cm) 0.425 1.514 0.705 Height above the ground (cm) 0.107 1.015 0.134 Angle above horizontal 0.136 0.973 0.160 ([degrees]) Drip edge (cm) 0.817 0.945 0.577 Presence of anastomosing 0.575 0.753 0.846 fissures SLD (m) 0.043 0.704 -- Attribute ORE Length (cm) 1.137 Maximum opening width (cm) 1.645 Maximum depth (cm) 1.223 Height above the ground (cm) 1.013 Angle above horizontal 0.974 ([degrees]) Drip edge (cm) 0.872 Presence of anastomosing 0.904 fissures SLD (m) -- Table 4.--Median, first quartile (Q1), third quartile (Q3), and interquartile range (IQR) of attributes of occupied and random brood crevices of green salamanders on the Blue Ridge Escarpment in North Carolina, 2010-2015 Attribute Actual (N = 29) Median Q1 Q3 IQR Length (cm) 47.7 27.5 66.7 39.2 Maximum opening width (cm) 2.7 1.5 4.1 2.6 Maximum depth (cm) 11.8 10.0 16.9 6.9 Height above the ground (cm) 82.8 56.9 104.5 47.6 Angle above horizontal ([degrees]) 10.0 6.0 18.0 12.0 Drip edge (cm) 8.5 3.0 18.0 15.0 Attribute Random (N = 29) Median Q1 Q3 IQR Length (cm) 75.1 51.1 108.1 57.0 Maximum opening width (cm) 3.1 1.8 4.3 2.5 Maximum depth (cm) 13.5 8.9 16.1 7.2 Height above the ground (cm) 53.0 43.4 83.3 39.9 Angle above horizontal ([degrees]) 15.0 5.0 18.0 13.0 Drip edge (cm) 8.3 3.0 18.4 15.4 Table 5.--P-values and odds ratio estimates (ORE) of the stepwise procedure using a conditional logistic regression models comparing attributes of occupied (N = 29) and random (N = 29) brood crevices of green salamanders on the Blue Ridge Escarpment in North Carolina, 2010-2015. Odds ratio estimates less than one indicate a possible negative association between predictor and nest success, and ORE greater than one indicate a possible positive association between predictor and nest success Attribute P-value ORE Length (cm) 0.033 0.473 Maximum opening width (cm) 0.140 0.625 Maximum depth (cm) 0.670 0.863 Height above the ground (cm) 0.023 1.027 Angle above horizontal 0.426 0.981 ([degrees]) Drip edge (cm) 0.562 1.175 Presence of anastomosing 1.00 1.00 fissures Attribute P-value adjusted for ORE height above the ground Length (cm) 0.057 0.478 Maximum opening width (cm) 0.215 0.654 Maximum depth (cm) 0.560 0.789 Height above the ground (cm) -- -- Angle above horizontal 0.632 0.986 ([degrees]) Drip edge (cm) 0.643 1.172 Presence of anastomosing 0.899 0.931 fissures
|Printer friendly Cite/link Email Feedback|
|Author:||Rossell, C. Reed, Jr.; Williams, Lori A.; Cameron, Alan D.; Lawson, Charles R.; Patch, Steven C.|
|Publication:||The American Midland Naturalist|
|Date:||Jan 1, 2019|
|Previous Article:||Common Nighthawks (Chordeiles minor) in Agricultural Landscapes: Genetic Structure of Populations Restricted to Urban Rooftop Nesting.|
|Next Article:||Space Use of Predatory Larval Dragonflies and Tadpole Prey in Response to Chemical Cues of Predation.|