Abstract Title of Document: ENVIRONMENTAL VARIABLES ASSOCIATED WITH POPULATION CHANGES OF PLETHODONTID SALAMANDERS IN THE GREAT SMOKY MOUNTAINS NATIONAL PARK, USA Nicholas Caruso, Master of Science, 2011 Directed By: Dr. Karen R. Lips Department of Biology I used a long term collection database to compare 72 current populations of six species and three hybrids of Plethodon salamanders in the Great Smoky Mountains National Park (GRSM). I analyzed population abundance and adjusted for detection probabilities, over time for each species, with respect to null models. I also examined biotic and abiotic factors as potential causes for changes in population abundance. Population response varied among species, Plethodon glutinosus and P. teyahalee declined while P. jordani x metcalfi and P. ventralis increased at a greater rate than what was expected by historic variation in abundance. Declines of GRSM salamanders most likely began in the late 1960?s to early 1970?s and were associated with cooler and moister habitats. I conclude that species? biology may explain the variation in population responses and propose future research to determine the cause. ENVIRONMENTAL VARIABLES ASSOCIATED WITH POPULATION CHANGES OF PLETHODONTID SALAMANDERS IN THE GREAT SMOKY MOUNTAINS NATIONAL PARK, USA By Nicholas Caruso Thesis submitted to the Faculty of the Graduate School of the University of Maryland, College Park, in partial fulfillment of the requirements for the degree of Master of Science 2011 Advisory Committee: Professor Karen R. Lips, Chair Professor Daniel S. Gruner Professor James M. Dietz ? Copyright by Nicholas Caruso 2011 ii Acknowledgements This study was approved by the UMD Institutional Animal Care and Use Committee (R?09?09); research permits were issued from the Great Smoky Mountains National Park through the National Park Service. I would like to express my gratitude to my advisor Dr. Karen Lips, for her immense support and guidance through my graduate degree. I would also like to thank the members of my committee, Dr. Daniel Gruner and Dr. James Dietz, for their assistance. I thank Dr. Jake Kerby for providing me the opportunity to learn DNA extraction and PCR techniques. Special thanks go to Dr. Richard Highton, whose incredible knowledge is a continued inspiration and the Smithsonian National Museum of Natural History: Addison Wynn, Roy McDiarmid, and Steve Gotte. I would also like to thank Paul Super for his assistance in getting around and living within the Great Smoky Mountains. I am very appreciative to my loving parents, Frank and Maria, as well as my brother and sister, Matt and Jess, who have always been there for me and offered their love and support. I would like to thank Dot Lundberg for her support and enduring many stressful late nights of writing with me. I would like to thank the SIU and UMD herp labs: Sarah, Sandra, Brooke, Erin, Toren, Kyle, Cora, Stephanie, and Rebecca for their helpful suggestions and comments. I would also like to thank all of my volunteers for assistance in the field. iii Table of Contents Acknowledgements ...................................................................................................ii Table of Contents .....................................................................................................iii List of Tables............................................................................................................iv List of Figures ...........................................................................................................v Introduction...............................................................................................................1 Methods ..................................................................................................................12 Results.....................................................................................................................27 Discussion ...............................................................................................................34 Conservation Implications .......................................................................................43 Tables......................................................................................................................48 Figures ....................................................................................................................61 Literature Cited .......................................................................................................87 iv List of Tables Table Page 1. Number of populations surveyed for each species ..............................................48 2. Number of sites based on the number of historic species present ........................49 3. Biological information for GRSM Plethodon salamanders .................................50 4. Prevalence and intensity of Bd infection for all species of Plethodon salamanders sampled during my 2009 surveys ...................................................52 5. Detection probabilities for each species.............................................................. 53 6. Number of declined, stable and increased populations for each species...............................................................................54 7. AICc scores and GLMM estimates for each model.............................................55 8. Temperature and precipitation ranges among the three elevational groups ....................................................................................59 9. Results of High/Low clustering analysis (Getis-Ord General G).........................60 v List of Figures Figure Page 1. Map showing the widespread declines in Plethodon salamander populations..................................................................................... 61 2. Map showing the location of the resurveyed sites in the GRSM with respect to elevation.................................................................. 62 3. Frequency graph showing the distribution in historic sampling............................................................................................... 63 4. Relative population abundance over time at selected sites for Plethodon glutinosus.............................................................. 64 5. Relative population abundance over time at selected sites for Plethodon jordani x metcalfi .................................................. 65 6. Relative population abundance over time at selected sites for Plethodon jordani x teyahalee ................................................ 66 7. Relative population abundance over time at selected sites for Plethodon jordani .................................................................. 67 8. Relative population abundance over time at selected sites for Plethodon metcalfi ................................................................. 68 9. Relative population abundance over time at selected sites for Plethodon serratus ................................................................. 69 10. Relative population abundance over time at selected sites for Plethodon teyahalee ............................................................... 70 11. Relative population abundance over time at selected sites for Plethodon ventralis ................................................................ 71 12. Observed and expected population responses for Plethodon glutinosus ................................................................................... 72 13. Observed and expected population responses for Plethodon teyahalee .................................................................................... 73 14. Observed and expected population responses for Plethodon jordani........................................................................................ 74 15. Observed and expected population responses vi for Plethodon serratus ...................................................................................... 75 16. Observed and expected population responses for Plethodon jordani x metcalfi........................................................................ 76 17. Observed and expected population responses for Plethodon ventralis ..................................................................................... 77 18. Variable contributions to training gain for the trimmed model of Bd in the GRSM................................................................... 78 19. Variable contributions to AUC for the trimmed model of Bd in the GRSM................................................................... 79 20. Predicted maximum suitability of Bd in the GRSM........................................... 80 21. Predicted average suitability of Bd in the GRSM............................................... 81 22. Average suitability of each population response................................................ 82 23. Proportion of population response based elevational bands ............................... 83 24. Population response based on habitat disturbance type...................................... 84 25. Number of populations found during historic (1960 ? 1979) and current (2009) surveys based on community size. ........................................................................................... 85 26. Graph showing the adjusted relative abundance for Plethodon jordani from 1962 ? 2009........................................................... 86 1 Introduction Plethodon salamanders were once immensely abundant throughout the United States. In the Great Smoky Mountains National Park (GRSM), densities of Plethodon jordani and P. teyahalee were estimated at 8,600 and 2,300 individuals/hectare respectively (Merchant 1972). In the Southern Appalachian region, salamander density was estimated at 5,961?9,935/ha (Hairston 1987), and salamander density in western North Carolina was estimated at 10,000/ha (Petranka et al. 1993). In a New Hampshire forest, Burton and Likens (1975a) found that P. cinereus occurred in such large densities that their estimated biomass exceeded the combined biomass of all small vertebrates. Because of their naturally high abundance and density, Plethodon salamanders are important components of forest ecosystems (Davic and Welsh 2004). These salamanders are considered the dominant vertebrate predators of leaf litter arthropods (Hairston 1987) and function as keystone predators of leaf litter invertebrates (Davic 1983; Wyman 1998). Woodland salamanders also function as important prey species for birds and small mammals and form an important link in the food web. Plethodon salamanders are able to efficiently convert small prey (e.g., forest invertebrates) into available biomass for larger predators (e.g., small mammals; Burton and Likens 1975b). In addition to their role in forest food webs, Plethodon salamanders are important to forest ecosystem function. Woodland salamanders are considered ecosystem engineers, because they are able to modify resource availability for other species by causing either physical or chemical changes in habitats (Jones et al. 1994). The mechanism by which salamanders alter their habitat is through the creation or use of underground burrows (Brooks 1946; Heatwole 1960), which may lead to translocation of 2 nutrients underground from the forest floor (for plants), deposition of excretory nutrients (for bacteria and fungi), and increased dispersion of gases through the soil (Davic and Welsh 2004). Through control of forest floor invertebrates, salamanders may have an indirect effect on leaf litter decomposition and nutrient recycling (Hairston 1987; Wyman 1998; Wyman 2003). Woodland salamanders are excellent study organisms for monitoring abiotic and biotic components of forest ecosystems because fluctuations in their populations can be associated with changes in leaf litter, air and water pollution, invertebrates, pH, and moisture (Welsh and Droege 2001). Moreover, Welsh and Droege (2001), reported that salamander population trends can be detected more quickly and with fewer years of monitoring effort than with other vertebrates. Therefore, successive surveys of terrestrial salamanders over a long period of time can provide reliable estimates of population changes. These changes, therefore, will be important to assess the changes in the forest ecosystem. Studies of Plethodon population abundance over time have shown opposing trends for different species, times, and geographic locations (Wyman 2003). From 1976 to 1990 at two different sites, Hairston and Wiley (1993) found no changes in abundance for P. glutinosus (teyahalee; Highton 1984) and P. jordani. Elsewhere, Highton (2005) described declines of both P. jordani and P. teyahalee over 40 years. Globally, at least 21 of 350 species within the family Plethodontidae have become more threatened in the last 20 years (Stuart et al. 2004). Declines of 13 of these species were attributed to habitat loss, while 8 were classified as enigmatic, declining despite residing in suitable habitat (Stuart et al. 2004). Currently, 22 species of 3 Plethodon salamanders have an IUCN listing as ?Near Threatened? or higher, including P. jordani (IUCN 2010). Synchronous, widespread declines occurred in 180 populations of 38 species of eastern U.S. Plethodon salamanders (Highton 2005; Figure 1); of these species only P. welleri is currently listed as ?Endangered? (IUCN 2010). Reductions in abundance were first noticed in the late 1980?s throughout these populations, which suggests that the same factor or factors affected these populations in a similar manner (Highton 2005). Species with populations that declined occurred at sites where other species did not decline (Highton 2005; Figure 1), which suggest that species? biology is an important factor in determining which species declined and which did not decline. Extensive habitat destruction was observed at 16 sites (22 populations; Highton 2005), but it did not explain the other 158 population declines documented by Highton (2005), including populations in protected areas, in which habitat loss, land?use change and overexploitation cannot explain these losses. Plethodon salamanders face many different threats. Six identified main threats are likely the cause of most amphibian population declines: overexploitation, land use change, alien species, emerging infectious diseases, pollution (e.g., habitat acidification), and global climate change (Collins and Storfer 2003). Within Plethodon, the main threat to population abundance appears to be timber harvesting (Wyman 2003), although it has not caused any documented species extinctions (IUCN 2007); however, habitat acidification may be another important threat (Wyman 2003). Additionally, global climate change has been implicated in declines in Central American plethodontid 4 populations (Rovito et al. 2009) and Plethodon species are susceptible to emerging infectious diseases (Vazquez et al. 2009; Weinstein 2009) Timber harvesting can create unsuitable habitat for terrestrial plethodontids (Ash 1988) by reducing leaf litter, shade soil moisture and increasing surface temperature (Bury 1983; Ash 1988). These conditions, therefore, create drier habitats that limit surface activity and may cause mortality due to increased physiological stress (Petranka et al. 1993). Additionally, Petranka et al. (1993, 1994) reported that Plethodon abundance may not recover until 50 ? 70 years after harvesting because younger tree stands create less favorable (i.e., drier and warmer) microhabitats than mature stands. Although protected areas (e.g., National Parks) preserve salamander habitat from harvesting, changes in populations may be lagged effects of this past logging. Because Plethodon salamanders are lungless and rely on moist skin for cutaneous respiration, it is likely that all species are affected similarly to harvested forests (see Petranka et al. 1993); however, some evidence suggests that P. teyahalee may be less sensitive to drier forests, as a result of timber harvesting, than P. jordani (Ash 1997). Acidic habitat can affect salamanders as well as their ecosystem. These acidic environments may create unsuitable habitat for terrestrial salamander populations (Wyman and Jancola 1992; Frisbie and Wyman 1995). Individuals raised in low pH (< 3.8) aquatic and terrestrial environments can experience direct mortality, increased deformity rates and reduced rates of growth and development (Rowe and Freda 2000; Wyman 2003). Additionally, salamanders living on acidic substrates (< 3.8) lose sodium, water, and overall body mass more readily than salamanders living on less acidic substrates (Frisbie and Wyman 1991). This disruption in sodium balance may create 5 unfavorable habitats to terrestrial salamander populations on acid soils or in areas receiving increased acid deposition (Wyman 2003). Infectious diseases can cause amphibian morbidity and mortality. Chytridiomycosis is an emerging infectious disease of amphibians caused by the fungal pathogen Batrachochytrium dendrobatidis (Bd; Longcore et al. 1999). Bd attacks the keratinized epidermis of adult amphibians and the keratinized mouthparts of tadpoles (Berger et al. 1998). Susceptible amphibian species with high intensity infections experience electrolyte depletion and osmotic imbalance, which leads to cardiac arrest and death (Voyles et al. 2007; 2009). Disease occurrence varies among different hosts and different habitats and is determined by the interaction between three factors: the host, the pathogen and the environment (Wobeser 2006). Some amphibians can clear an infection or can persist without signs of disease (Daszak et al. 2004). In terrestrial salamanders, cutaneous antifungal bacteria (Harris et al. 2006) produce antifungal metabolites (Brucker et al. 2008), which can reduce the level of Bd infection (Harris et al. 2009a, b). Bd thrives under cooler temperatures (Piotrowski et al. 2004) and moist environments (Johnson et al. 2003) and can be cleared in warmer (Woodhams et al. 2003) and drier conditions (Johnson et al. 2003). Optimum growth of Bd occurs between 17 ? 25?C (Piotrowski et al. 2004) and the critical thermal maximum of Bd is 30?C (Piotrowski et al. 2004). Consistent with Bd?s environmental tolerances, infection prevalence, in several empirical studies, increased during cooler months compared to warmer months (Woodhams et al. 2003; Berger et al. 2004; Kriger and Hero 2006; Weinstein 2009). Variation in species response to Bd has also been attributed to differences in 6 habitat, microhabitat (Lips et al. 2003; Rowley and Alford 2007; Brem and Lips 2008), thermoregulation (Richards?Zawacki 2010), amphibian immune response (Ramsey et al. 2010), skin peptides (Harris et al. 2006), and amphibian behavior (Longo et al. 2009), amphibian density (Briggs et al. 2010). Arrival of Bd into a susceptible population is followed by amphibian mortality, increases in prevalence and intensity of disease, and population declines (Lips et al. 2006; Briggs et al. 2010). Following Bd?s arrival, amphibian communities experience a loss of regional and local scale diversity, in which many endemic species are lost (Crawford et al. 2009; Smith et al. 2009). Declines attributed to Bd have occurred in amphibian populations throughout the world (Berger et al. 1998; Lips et al. 2006; Rachowicz et al. 2006). Within the United States, confirmed cases of amphibian population declines attributed to Bd have occurred in California (Rachowicz et al. 2006) and Colorado (Muths et al. 2003) and mortality events associated with Bd have occurred in adult anurans in California, Colorado, North Dakota, North Carolina, and Wyoming (Green et al. 2002). Amphibian response to Bd throughout the US varies. Populations in the Western US experience epizootic outbreaks consistent with a na?ve population, while Eastern US amphibians do not show the same patterns (Green et al. 2002). Bd has caused salamander population declines worldwide (Berger et al. 1998; Collins and Storfer 2003; Lips et al. 2006) and is present on every continent in which amphibians occur (Berger and Speare 2000; Goka et al. 2009). In Pe?alara National Park, Spain, Salamandra salamandra have experienced population declines (Mart?nez? Solano et al. 2003), in which heavy Bd infections were confirmed in several dead 7 individuals (Bosch and Mart?nez?Solano 2006). Although Bd?related declines and mortalities have rarely been reported for salamanders, Bd infections have been confirmed in many genera, including: Ambystoma (Davidson et al. 2003; Ouellet et al. 2005), Notophthalmus (Ouellet et al. 2005), Necturus (Speare and Berger 2000), Siren (Speare and Berger 2000), Pseudotriton (Speare and Berger 2000), Desmognathus (Grant et al. 2008), Eurycea (Grant et al. 2008), Bolitoglossa (Pasmans et al. 2004; Lips et al. 2006; Rovito et al. 2009; Crawford et al. 2010), and Taricha (Padgett?Flohr and Longcore 2007). Within the genus Plethodon, reports of infected species include: Plethodon neomexicanus (Cummer et al. 2005), P. cinereus (www.spatialepidemiology.net/bd/), P. glutinosus (Chinnadurai et al. 2009), P. yonahlossee (Chinnadurai et al. 2009) and P. metcalfi (Vazquez et al. 2009). Although population declines attributed to Bd have occurred within Plethodontidae in Central America (Lips 1998; Lips and Donnelly 2005; Lips et al. 2004; Lips et al. 2006; Rovito et al. 2009; Crawford et al. 2010), there are no confirmed cases of Plethodon population declines directly caused by Bd in the US. Ranavirus and has been implicated in amphibian declines in various species (e.g., Collins et al. 1988; Chinchar 2002; Ariel et al. 2009). In the US, ranaviruses have been associated with some amphibian mortalities (Green et al. 2002) especially in Ambystoma spp. (Green et al. 2002; Greer et al. 2009). Dead amphibians with ranaviruses have been found in New Hampshire, Colorado, North Dakota, Minnesota, Maine, Utah, Idaho, Tennessee, Massachusetts, Wyoming, and North Carolina (Green et al. 2002). Higher ranavirus prevalence estimates are associated with higher elevations (Gahl and Calhoun 2008) and its virulence is likely increased through 8 natural (e.g., metamorphosis) and anthropogenic stressors (e.g., nitrogenous compounds; Forson and Storfer 2006; Gray et al. 2007; Gray et al. 2009a). Within Plethodon, reports of susceptible species include: P. jordani, P. metcalfi and P. jordani x metcalfi (Gray et al. 2009b; M.J. Gray and N.M. Caruso unpublished data). Climate change can impact protected amphibian populations (Carey et al. 2001; Carey and Alexander 2003; Rohr et al. 2008; Rodenhouse et al. 2009; Rohr and Raffel 2010). Warmer and drier climates are likely to negatively impact many amphibian populations (Blaustein et al. 2003; Pounds 2001; Rohr and Madison 2003). In the eastern US, predictive models predict warmer and wetter climates (Girvetz et al. 2009). These changes may be problematic for terrestrial lungless salamanders (i.e., plethodontid salamanders), which rely on moist, permeable skin to allow gas exchange (Whitford and Hutchison 1965). Because of their permeable skin, plethodontid salamanders are prone to dehydration, are restricted to cool, moist microhabitats, and have limited surface activity, even during favorable conditions (Feder 1983; Bernardo and Spotila 2006). Amphibian species most at risk of global climate change include those that are already at their physiological limits of temperature and/or moisture (e.g., montane- adapted terrestrial plethodontids; Bernardo and Spotila 2006), those that depend on ephemeral wetlands, or those that cannot disperse (Reaser and Blaustein 2005). Based on the degree of metabolic depression associated with reduced temperature, current high elevation species of plethodontid salamanders with restricted ranges (e.g., P. jordani) are most likely distributed near the upper limits of their thermal tolerances (Bernardo and Spotila 2006; Bernardo et al. 2007). Global circulation 9 models predict that rising CO2 levels will lead to climate changes in the northern hemisphere, including the GRSM (IPCC 2007), in which average temperatures will continue to increase (Girvetz et al. 2009). Because of these warming temperatures, lower elevation forest habitats may become unsuitable for southern Appalachian salamanders (Milanovich et al. 2010). Changes in climate can also affect disease dynamics by shifting the host- pathogen relationship in favor of the pathogen (Wobeser 2006; Rohr et al. 2008). Decreases in temperature may reduce the host?s ability to fight infection (Vazquez et al. 2009) while increased precipitation may create more suitable habitat for disease (Rohr et al. 2008; Weinstein 2009). Natural changes in climate, such as El Ni?o ? Southern Oscillation (ENSO) events may create more favorable conditions for pathogen persistence and transmission (Weinstein 2009), in producing higher than average rainfall in the Eastern United States (NOAA 2010; NCDC 2010). Therefore, ENSO years may allow a pathogen that normally exists at low levels, which do not cause decreases in population abundance (Vredenburg et al. 2010), to increase transmission and cause declines. Objectives I analyzed the change in population abundance over 49 years for six species and three hybrids of Plethodon salamanders in the GRSM at 35 historic collecting sites. I used a collection dataset to determine historic population abundance and field surveys to determine current population abundance. My goal for this project was four?fold. (1) To determine if changes in abundance of Plethodon salamanders differed from expected 10 based on historic population abundance and adjust count indices based on detection probabilities; (2) to determine the relationship among changes in population abundance and species, sites, and elevations as well as determine the approximate timing of population declines; (3) to estimate the change in diversity within each site as a function of species presence/absence and abundance changes; (4) to search for associations between these population changes and potential threats (e.g., Bd?s predicted and actual occurrence, forest disturbance, current temperature, precipitation patterns, and changes in precipitation and temperature from 1961 ? 2006). These analyses will indicate potential causes for declines in Plethodon salamander populations the GRSM. If populations declined because of forest stand age or disturbance, I expected declines to have occurred in young forests, with stable or increasing populations in old growth forests. Furthermore, I would expect populations of P. teyahalee to remain stable because this species is more tolerant of timber harvesting (Ash 1997). If acid deposition has caused Plethodon salamander declines in the GRSM then I expected an elevational pattern, with all high elevation, which is correlated with increased acid deposition in the GRSM (Robinson et al. 2008), population to declined, while all low elevation populations would remain stable or increase. I searched for associations between declines and environmental variables (Lips et al. 2003; Gray et al. 2009a) and spatial relationships. If Bd were a potential cause for declines, declines should be associated with areas of higher Bd suitability [as predicted by Maxent models (Phillips et al. 2006)], I further expected that population declines would be spatially clustered in areas of high Bd suitability. Furthermore, I expected populations which inhabited cooler temperatures and higher precipitation, which may promote Bd or 11 ranavirus persistence, will have declined, while populations which inhabit warmer temperatures and lower precipitation will not have declined. Additionally, I expected to find these ranavirus or Bd within populations and species that have declined and at a lower prevalence in populations and species which have remained stable or increased. If changes in temperature and precipitation were a potential direct cause for declines, I would expect that populations will have declined in areas that have become warmer and drier over the last 55 years. I expected populations to have remained stable or increased in areas that have become cooler and wetter since 1951. Additionally, I expected the timing of declines to be synchronous with extreme weather events (e.g., El Ni?o ? Southern Oscillations; Pounds et al. 1999). 12 Methods Study Area The GRSM, which straddles the Tennessee?North Carolina border, encompasses 205,665 ha of contiguous forest, categorized into five major forest types: Cove Hardwood, Spruce?Fir, Northern Hardwood, Hemlock, and Oak?Pine (Jenkins 2007). Approximately 95% of the park is forested and 25% of this is old? growth forest (Davis 1993). Forest stands, especially mature stands, are critical habitat for Plethodon salamanders (Dupuis et al. 1995) because they provide extensive canopy coverage, which inhibits sunlight penetration (Phillips and Shure 1990) that can otherwise lead to higher temperatures and drier leaf litter (Ash 1995). Moreover, mature stands provide abundant coarse woody debris, which serves as moist retreat sites for terrestrial salamanders (Feder 1983). Elevations in the GRSM range from 267 ? 2,025 meters with highly variable topography, which include gentle and steep slopes, level valleys, talus slopes, incised drainages and rocky summits (Jenkins 2007). The GRSM has a temperate climate; annual temperatures range from -2 to 31 oC in the low elevations (~265 meters), while at the highest elevations (2,025 meters) annual temperatures range from -18 to 19 oC (NPS 2010). Precipitation also varies with elevation; currently, the highest elevations receive an average of 2,160 mm of precipitation per year while the lower elevations average about 1,397 mm annually (NPS 2010). The GRSM is home to nearly 10% of the world?s salamander species (Petranka 1998); at least 30 species (5 families) reside within the GRSM (Tilley and Huheey 2004). Because of the high salamander diversity the GRSM is 13 known as ?The Salamander Capital of the World? (NPS 2010) and is an international temperate forest biodiversity refuge (NPS 2010). Despite its protected status, GRSM salamander diversity may be threatened by factors such as historic timber harvesting, acid precipitation, emerging infectious diseases, and climate change. Although the GRSM is currently protected, historically, forest harvesting has occurred until 1939 (NPS 2010), while precipitation pH at the highest elevations (NPS 2010) is currently low enough to cause reduced growth or mortality of salamanders (Wyman 2003). Bd has been present in the southeastern US since 1978 (Daszak et al. 2005) and both Bd and ranavirus are present within the GRSM (Chatfield et al. 2009; Gray et al. 2009b; Todd-Thompson et al. 2009). Additionally, projections of future climate trends using species distribution models (SDMs) show range contractions for southern Appalachian salamanders as early as 2020 (Milanovich et al. 2010). Site Selection From the 1960 to 2001, Richard Highton collected over 17,000 plethodontid salamanders at over 300 sites throughout the GRSM. Detailed field notes were taken for every visit, including: geographic coordinates, date of collection, species, number of individuals encountered, and number of collectors. These collected specimens as well as the research notes are currently housed at the US National Museum of Natural History (USNM). These sites were sampled for over 50 years in the same manner, which allowed for unique opportunity to study multiple species throughout the GRSM for over a half century (Shaffer et al. 1997). 14 I selected 35 of Highton?s collecting sites that occurred within the GRSM (Figure 2) and which had been sampled during multiple years. I sampled 2 plots within each site to the increase power for detecting changes in abundance (Smith and Petranka 2000). I selected sites that encompassed a wide geographic and elevation range (488 to 1,972 meters) to represent a variety of climate space and include the distribution of the 6 species of Plethodon salamanders within the GRSM. I collected sampled field sites during March, May through July and November of 2009 (Figure 2). I sampled seven low elevation sites (~500 m ? 1,000 m), 19 mid elevation sites (~1,000 m ? 1,500 m) and nine high elevation sites (~1,500 m ? 2,000 m). Study Species Six species of Plethodon salamanders occur within the GRSM (Tilley and Huheey 2001). All six species are lungless, are completely terrestrial, and utilize natural cover objects as aboveground shelter on the forest floor (Wells 2007). One to three species occupied the 35 study sites for a total of 72 populations of Plethodon salamanders (Tables 2 and 3). Susceptibility to Bd varies among species of terrestrial plethodontids. Plethodon glutinosus has been found to be infected with Bd in the wild (Chinnadurai et al. 2009; Table 3) while Plethodon metcalfi shows clinical signs of chytridiomycosis and mortality when infected with high doses of Bd in the lab (Vazquez et al. 2009). Susceptibility to Bd for the other species of GRSM salamanders is unknown, and no fully terrestrial plethodontid salamander has been found to be infected within the GRSM (Chatfield et al. 2009). 15 High elevation populations of terrestrial salamanders are most likely distributed near the upper limits of their thermal tolerances (Bernardo and Spotila 2006; Bernardo et al. 2007). High elevation species experience high rates of metabolic depression with increasing temperatures, which can limit their ability to disperse through lower valleys or adapt to warming conditions (Bernardo and Spotila 2006). P. serratus, P. metcalfi, P. teyahalee, and P. jordani inhabit all three elevational bands (Table 1), while P. ventralis is the only species that is not found at elevations over 1,000 m (Table 1) and P. glutinosus inhabits elevations below 1,500 m (Table 1). Species and populations inhabiting the warmer, low elevation habitats may not be able deal with the physiological stress associated with warming trends (Bernardo and Spotila 2006). Field Methods From 1960 ? 2001, R. Highton and colleagues surveyed these 35 sites at least once and not more than seventeen times (Figure 3). Each site is characterized as an area marked by one geographic coordinate, in which Highton and colleagues searched natural cover objects (i.e., rocks and logs) for approximately one hour. All salamanders were captured by hand, identified and usually collected (R. Highton personal communication). During a site visit, Highton used topographic and quadrat maps to georeference the site location, which allowed him to return to the same site over many years. I located using a combination of a handheld GPS Unit and topographic maps. I sampled each site two to four times and searched two different per visit. I surveyed each site multiple times to limit the chances that the number of salamanders encountered was not biased towards good or bad weather events but rather was a more accurate index of 16 abundance for that site during over time. I sampled each site during optimal conditions (e.g., early morning or during cooler, moister weather conditions) to maximize the chance of encountering salamanders, consistent with the historic sampling (R. Highton personal communication). Using red flagging, I marked off two 50 m x 3 m plots within each site. I combined relative abundance indices for the two plots for each site visit, which increases the power to detect population abundance changes in terrestrial plethodontids (Smith and Petranka 2000). Estimated home range sizes for Plethodon glutinosus, P. jordani, P. metcalfi, and P. teyahalee are less than 15 m2 (Table 3; Merchant 1972; Nishikawa 1990; Marvin 1998), so I maintained at least 15 meters between plots within a site visit and between visits. To allow comparison among Highton?s surveys and my surveys, I used the number of salamanders accounted for effort (the number of surveyors multiplied by the amount of time surveying) as abundance indices. In each plot, I turned all natural cover objects and captured all salamanders by hand. Upon capture, I placed each salamander in a new plastic bag to prevent cross? contamination in case of infected individuals (Hyatt et al. 2007). I recorded the species identification, snout?to?vent length (SVL), tail?to?vent length (TVL), and mass for each individual. Additionally, I measured the length and width of each natural cover object that I searched under and indicated the type of cover (i.e., rock, log, surface, tree bark). To test for Bd, I used a skin swab, using a cotton?tipped plastic applicator (Medical Wire and Equipment, M113), I swabbed the ventral surface ten times and each limb five times for a total of 30 swabs per individual (Hyatt et al. 2007). I stored all swabs in vials (Fisher, 02?681?374) of 75% ethanol at 17 room temperature until quantitative PCR analysis, which I performed in January of 2010. After processing, I released all salamanders at their point of capture and replaced the natural cover object to its original position to minimize habitat disturbance (Smith and Petranka 2000). I quantified the total amount of time searching for salamanders at a site. I did not include time spent handling and processing salamanders or measuring cover objects because these actions were not performed during historic surveys. Between captures, I used new powder?free latex gloves to prevent disease transmission. Swab Analysis I performed PCR analysis on all swabs at the University of South Dakota under the supervision of Dr. Jake Kerby. DNA from all swabs was extracted using a QIAGEN DNeasy tissue and blood purification kit (QIAGEN, 69581). DNA was amplified using a FAST PCR machine using the primers and probes described in Boyle et al. (2004). I pooled swabs into groups of three and ran those pooled swabs in triplicates to reduce the likelihood of obtaining a false negative (Hyatt et al. 2007). When the pooled swabs returned a positive result, we reran the three swabs individually in triplicates, which identified the positive sample and allowed for estimates of infection intensity. I estimated the prevalence of Bd, defined as the number infected out of the total sample, and calculated 95% Clopper?Pearson binomial confidence intervals using the Hmisc package (Harrell Jr. et al. 2010) in Program R (R Development Core Team 2009). Additionally, I estimated the intensity 18 of infection by multiplying the number of zoospores detected in each sample by 100 (the dilution of the final sample) to determine the level of infection. This value was used as the estimate of the number of Bd zoospores in each swab, or genomic equivalents (GE). Infection intensity is an important predictor of amphibian population responses to Bd (Vredenburg et al. 2010). High average infection intensity (> 10,000 GE) in a population is associated with individual mortality, population declines and extirpations (Vredenburg et al. 2010; V.T. Vredenburg personal communication). Data Analysis Detection Probabilities I estimated detection probabilities for each species during survey decades to account for variation in amphibian detection and surface activity (Schmidt 2009). I used the software program Presence version 3.0 (Hines 2006) to determine detection probabilities for each species using presence/absence data for each survey visit and a multi-season model to estimate detection probabilities during each decade of survey. I used these detection probabilities to adjust count indices used in subsequent analyses (Schmidt 2009; Bailey et al 2004). Patterns in Abundance and the Environment I used R (R Development Core Team 2009) and ArcMap version 10.0 (ESRI, Redlands, California, USA) for all statistical analyses. I used a generalized linear mixed effects model (GLMM; lme4; Bates and Maechler 2010) to measure the 19 change in abundance over time and analyzed the relationship between abundance and changes in the environment (i.e., temperature and precipitation patterns) for each species. I ran five models for each species (Table 7). First, I analyzed the relationship between population abundance and the year of survey (YEAR; Model 1) to determine if population abundance has changed. Next, I included environmental variables (average mean temperature, annual precipitation, change in annual temperature from 1951 ? 2006, and change in annual precipitation from 1951 ? 2006) as site covariates for the six taxa with more than one population. To evaluate Bd and ranavirus as a potential cause for declines I used annual mean temperature (AMT; Model 2) and annual precipitation (AMP; Model 3) from WorldClim (1 km resolution; Hijmans et al. 2005). These variables were chosen because they are important in determining patterns of Bd abundance using climatic envelope models (Puschendorf et al. 2009; Murray et al. 2011). These data were interpolated from a compilation of five different global weather stations (Hijmans et al. 2005). These variables demonstrate the magnitude of the annual temperature and precipitation at all sites, which allows comparison among sites based on their climate (e.g., warmer or cooler). Higher values for AMT or AMP denote areas that are currently warmer or wetter than the lower values. To evaluate climate change as a potential cause for declines, I used two variables: the change in annual temperature (? Temp; Model 4) from 1951 ? 2006 and change in annual precipitation (? Precip; Model 5) from 1951 ? 2006 (Girvetz et al. 2009). These variables explain the direction (i.e., increase or decrease) and 20 magnitude that the annual temperature and precipitation over the 55 years of surveys at these sites. A positive value for ? TEMP or ? PRECIP indicates areas that have become warmer or wetter over the last 55 years respectively, while negative values for ? TEMP or ? PRECIP denotes areas that have become cooler or drier over the last 55 years respectively. I obtained estimates for of the climate change variables from Climate Wizard (4 km resolution; Girvetz et al. 2009), an online tool that provides an ensemble average of 16 global circulation models for future climate predictions, as well as provides a compilation of the last 50 years of climate trends. For the US, the past 55 years of climate were derived from 8,000 climate monitoring stations (Girvetz et al. 2009). To extract climate data for each site, I used ArcGIS and Geospatial Modeling Environment (GME; Beyer 2010). Using ArcGIS, version 9.3.1 (ESRI, Redlands, California, USA), I imported my sites as a shapefile and imported the climate data as raster layers. I used the insectpntrst command in GME to extract the climate data for each site based on geographic location (Beyer 2010). I modeled the relationship between the adjusted counts and the survey date and nested the site within the survey year as the random variable (Shaffer et al. 1997). Therefore I was able to evaluate each species as a composite of all populations, which allowed me to determine the overall change in relative abundance for that species, as well as compare each surveyed population at a given site. I standardized the effort by the number of person hours and used a Poisson distribution (e.g., Roulin and Bersier 2007) and fit each model using maximum likelihood. The fixed effects varied in units, therefore, I standardized (i.e., subtracted the mean and divided by the standard 21 deviation) the year of survey, AMT, AMP, ? TEMP, and ? PRECIP (Marquardt et al. 1980). All models were selected in the following manner. First, I ran each model with standardized year as the only fixed effect. Next, I ran each model with the standardized environmental site covariates (AMT, AMP, ? TEMP, and ? PRECIP) and selected the best fit models based on the lowest Akaike's information criterion with second order correction for small sample sizes (AICc; Sugiura 1978). Determining Population Response I used a conservative approach for classifying each population?s change in abundance. I calculated the model coefficients and standard error for each population using the results from the overall GLMM (Qian et al. 2010) and assigned each population as ?Decline?, ?Stable?, or ?Increasing? based on the coefficients (?1) and standard error. If the upper and lower bounds of a population?s standard error were negative, I considered this population to have declined. If the upper bound of a population?s standard error was positive and lower bound was negative (i.e., the error bars crossed zero), I considered this population to be stable. If both of the upper and lower bounds of a population?s standard error were positive, I considered this population to have increased. Additionally, I performed pairwise comparisons on all species to determine if changes in relative abundance varied among species. All pairwise comparisons were analyzed at the 95% confidence level. For each species, I graphed population changes over time from sites that had been sampled at least 8 times. I chose this number because 8 years of data is sufficient to detect changes in 22 abundance for Plethodon salamanders when surveying six sites (Smith and Petranka 2000), which was the average number of sites per species of my study. Null Models I created null models of changes in population abundance for each species through time based on the mean and variance of the historic relative abundance of each species from 1960 to 1979. I chose this time period because I wanted to know if my estimates for relative abundance change were within the normal fluctuations of relative abundance among historic sampling (Pounds et al. 1997) and during the ?pre- decline? period reported by Highton (2005). First, I estimated the mean and variance of adjusted relative abundance indices for all species with greater than one sampled populations between 1960 and 1979. Next, for each taxon, I generated 10,000 datasets of relative abundance during the total survey period (1960 ? 2009). To generate these data, I used a quasi-Poisson distribution because the historic data were over dispersed (variance was greater than the mean; Ver Hoef and Boveng 2007). Next, I ran a generalized linear model (GLM), with a quasi-Poisson distribution (Ver Hoef and Boveng 2007) on each of run of the 10,000 datasets to determine the change in relative abundance over time. From these results, I determined if a population declined (i.e., standard error of ?1 were below 0), was stable (i.e., standard error of ?1 were above and below 0) or increased (standard error of ?1 were above 0; see Determining Population Response). 23 I performed a chi square test between the null model and my estimates, to determine if a species had changed more or less than expected, based on the null models. Maxent Predictions for Bd in the GRSM I used Maxent (version 3.3.0; Phillips et al. 2006) to model the suitability of Bd in the GRSM based on its presence in the United States and the climate. Maxent is a powerful presence-only SDM (Elith et al. 2006) that models the predicted environmental suitability (ES) based on the environmental conditions of the inputted presence localities for a given organism (Phillips et al. 2006). Maxent generates output ES maps for the target organism, in which higher values indicate areas that are predicted to be more suitable (Phillips et al. 2006). I obtained Bd localities from the Global Bd?mapping project (http://www.bd?maps.net/), an online resource providing information on Bd?s past and current range. Additionally, I used the locations of Bd occurrence in Appalachian plethodontids from previous research (Chatfield et al. 2009; Chinnadurai et al. 2009) and this study. I used 75 US locations with documented Bd presence to model predicted range. I used the 19 continuous bioclimatic variables (~5 km resolution; Hijmans et al. 2005) and followed the methods of Murray et al. (2009), in which all 19 variables were included in the full model and then selected those variables that contributed 90% of the information to the full model run. Next, I re?ran the trimmed model with the most important variables. I projected results of the trimmed model, using the same bioclimatic variables, onto the full range of the six GRSM Plethodon salamanders. This region 24 was obtained by clipping the trimmed bioclim variables by the shapefile of those species? distribution maps. Distribution maps were obtained from International Union for Conservation of Nature (IUCN 2011). I used the area under the curve (AUC) of the receiver operator characteristic (ROC) to assess model accuracy (Fielding and Bell 1997). I used bootstrapping (N=100) on unique training and testing datasets (75%; 25% respectively) on the trimmed model to determine average and standard deviation of Bd?s suitability in the projected region. I determined the relationship between responses for each population (i.e., declined, stable, and increased; see Determining Population Response) and the average predicted suitability of Bd. I used a multinomial logistic regression (mlogit; Croissant 2010) with the population responses as the response variable and average predicted suitability as the explanatory variable. Elevation and Forestry Practices Patterns in Abundance To obtain site elevation data, I used USGS 1?arc second National Elevation Dataset (NED; Gesch et al. 2002). To obtain logging history data, I used a forest disturbance GIS polygon (Griggs 2009). This shapefile shows the extent of historic logging in the GRSM (Griggs 2009). Although the actual amount of logging cannot be derived, these data show relative amounts of logging from the most (?Heavy Cut?) to the least (?Undisturbed?). I used the model coefficients and standard errors from the overall GLMM (see Determining Population Response) to determine population status and related the status to forestry practices and elevation. I approached these 25 analyses using R (R Core Development Team 2009) and a two?sided Kendall?s Correlation Test (Hollander and Wolfe 1973). Community Size Patterns I compared pre-decline (1960 ? 1979) to post-decline (2009) species richness (based on presence/absence of a species at a site) and determined the number of declined species (as determined by GLMMs; see Determining Population Response) at each site. Temporal Patterns in Abundance I used piecewise?linear relationships on generalized linear models to determine the approximate timing of population changes. First, I analyzed the linear relationship between the relative abundance of each species and the year of each survey using a generalized linear model (GLM) and fitted the data using a quasipoisson distribution, which fits the data to the Poisson distribution while accounting for overdispersion (Ver Hoef and Boveng 2007). The GLM for each species was then used to find ?breakpoints? in the regression, which are periods in which the slope changes (e.g., the beginning of a decline in population abundance would be signified by a change from a positive slope to a negative slope). I approached this analysis using the segmented package in Program R (Muggeo 2003; Muggeo 2008). I used 10,000 iterations for each segmented model. Spatial Patterns in Abundance 26 I used ArcMap to measure spatial clustering of population responses. First, I used the model coefficients and standard error for each population using the results from the overall GLMM (see Determining Population Response; Qian et al. 2010). Next, I gave each population a numeric value based on the model coefficients; ?1? for declined, ?2? for stable, and ?3? for increased. Next, I imported these coded model coefficients into an ArcMap shapefile using the geographic coordinates for each site. I projected the shapefile using ?NAD_1983_HARN_StatePlane_Tennessee?. Similar to the GLMMs, I did not analyze the 3 taxa (P. glutinosus x teyahalee, P. metcalfi, and P. jordani x teyahalee) for which I sampled one population. I assigned the coded coefficients as the Input Field, and used inverse distance and Euclidean distance as the Conceptualization of Spatial Relationships and Distance Method respectively. I analyzed each species using the High/Low Clustering (Getis-Ord General G) in the Spatial Statistic tools, which measures concentrations of high (increased populations) or low (decreased populations) values. High Z scores indicate clustering of populations that increased while low Z scores indicate clusters of decreased populations (Getis and Ord 1992). 27 Results Bd Surveys I swabbed 665 plethodontid salamanders, which included four genera; 12 species (Desmognathus imitator, D. ocoee, D. santeetlah, D. wrighti, Eurycea wilderae, Gyrinophilus porphyriticus danielsi, Plethodon glutinosus, P. jordani, P. metcalfi, P. serratus, P. teyahalee and P. ventralis), and 2 hybrids (P. jordani x metcalfi and P. jordani x teyahalee). I swabbed 485 Plethodon species (Table 4) and none were found to be infected within the GRSM (95% C.I. = 0 ? 0.79). I found one D. santeetlah with a low intensity of Bd infection (29 GE) near a small stream at a high elevation site (1504 meters) on July 10, 2009. Prevalence for all plethodontid salamanders in the GRSM was 0.15% (95% C.I. = 0.007 ? 0.847). Detection Probabilities All GRSM Plethodon salamanders were detected imperfectly (i.e., p < 1) during at least one survey decade (Table 5). Therefore, for each species, I corrected the relative abundance indices by the formula N = C/p (where N the true parameter value, C is a count index and p is a detection probability during the survey decade; Schmidt 2009). I used the adjusted count indices for all subsequent analyses. Species Abundance Patterns Changes in population abundance varied with species (Table 6); some species and populations declined, some fluctuated but remained stable, while others increased (Figures 4 ? 11). Plethodon glutinosus (z = -0.2977; p = 0.0087), P. teyahalee (z = - 28 3.1655; p < 0.0001), P. jordani x teyahalee (z = -0.3753; p = 0.0004) and P. glutinosus x teyahalee (z = -3.2135; p = 0.0281) decreased in abundance overall (Table 7). Plethodon glutinosus declined at all 5 of the resurveyed populations (Table 6) and P. teyahalee declined at 13 sites, increased at one site, and remained stable at two of the 16 resurveyed populations (Table 6). Plethodon jordani x teyahalee (N = 1) and P. glutinosus x teyahalee (N = 1) declined at their resurveyed populations (Table 6). Plethodon jordani (z = -0.0287.; p = 0.8100), P. metcalfi (z = -0.1461; p = 0.4170), and P. serratus (z = 0.0409; p = 0.8640) remained stable overall (Table 7) although individual populations showed mixed responses (Table 6). Plethodon serratus (N = 18) and P. jordani (N = 18) showed the most variation; both species had populations that declined, remained stable, and increased (Table 6). Plethodon metcalfi was stable at the one resurveyed population (Table 6). Plethodon ventralis (z = 0.3560; p < 0.0001) and the hybrid P. jordani x metcalfi (z = 0.6195; p < 0.0001) increased in abundance overall (Table 7). Plethodon ventralis increased at all four of the resurveyed populations while P. jordani x metcalfi increased at 87% (N = 8) of the resurveyed populations and remained stable at 13% of the resurveyed populations (Table 6). Null Models For all species, null models predicted that some populations declined, some increased, while others remained stable (Figures 12 ? 17). Plethodon glutinosus (?2 = 25.7975; df = 2; p < 0.0001) and P. teyahalee (?2 = 50.1073; p < 0.0001) declined at a 29 greater number of populations than expected (Figures 12 and 13). Population responses for Plethodon jordani (?2 = 1.9088; df = 2; p = 0.385) and P. serratus (?2 = 0.9981 = 2; p = 0.6071) were not significantly different than expected from the null models (Figures 14 and 15). Plethodon jordani x metcalfi (?2 = 26.5368; df = 2; p < 0.0001) and P. ventralis (?2 = 20.1869; df = 2; p < 0.0001) increased at a greater number of populations than expected (Figures 16 and 17). Maxent Predictions for Bd in the GRSM The trimmed model included seven Bioclim variables, which contributed ~94% information to the full model. The mean test AUC for the trimmed model was 0.993). BIO 2 (Mean Diurnal Temperature Range) contributed the most unique information to the trimmed model, while the most model training was lost when removing BIO 7 (Precipitation of Driest Quarter) from the trimmed model (Figures 18 and 19). The model predicts that the western portion of the GRSM is more suitable for Bd than the eastern portion of the GRSM (Figures 21 and 22). Based on the 100 models, the maximum suitability was 0.90 (Figure 20), while average suitability in the GRSM of Bd was 0.51 (? 0.12 standard deviation; Figure 21). The median suitability was highest in the declined populations compared to the stable or increased populations (Figure 22). However, the likelihood of population responses (i.e., declined, stable, increased) were not significantly different (?2 = 0.8816; df = 2; p = 0.6435) based on predicted average Bd suitability (Figure 22). 30 Elevation and Forestry Practices Patterns in Abundance All three elevational bands included populations that exhibited all three responses (Figure 23). The mid (1000 ? 1,500 m) and low (500 ? 1000 m) elevation sites had the highest percentage of declined populations, while the low elevation sites also had the highest percentage of increased populations (Figure 23). There was no correlation between the percentage of declined species and elevation (z = -0.9371, p = 0.3487). For the two species that had the largest elevational range, Plethodon jordani (1130 m) and P. serratus (996 m), declined populations only occurred at elevations greater than 1,300 m. Declined, stable and increased populations were found in forests of all ages (Figure 24) and there was no correlation between the population response (i.e., declined, stable, increased) and the type of disturbance at each site (z = 0.4761; p = 0.634). Temperature and Precipitation Patterns in the GRSM Populations in the lower elevations experienced the highest temperatures and the lowest precipitation (Hijmans et al. 2005; Table 8), which may be less favorable environment for Bd and salamanders than the cooler and moister higher elevations (Lips et al. 2003; Bernardo and Spotila 2006). Populations at the higher elevations experienced decreases in annual temperatures over the last 55 years (Girvetz et al. 2009; Table 8); these cooler temperatures may promote more suitable habitat for terrestrial salamanders and Bd (Lips et al. 2003; Bernardo and Spotila 2006) than the 31 lower and mid elevations, which have populations that have increased in temperature since 1961 (Girvetz 2009; Table 8). Acid Precipitation History in the GRSM Currently, the precipitation pH at low elevation (640 meters) ranges from 4.08 ? 6.05 (NADP 2010), while the highest elevations experience rainfall pH range of 3.92 ? 5.7, with an average of 4.62 (MACTEC 2010). Diversity Patterns Historic salamander community composition at the 35 sites ranged from one to three species (Table 2; Figure 25) and averaged 1.91 species per site. Current community composition ranged from zero to three species (Figure 25) and averaged 1.23 species per site. During my resurveys of those same populations, I encountered species for 41 of the 69 historic populations. However, at 3 sites, I encountered a species during my surveys that was not found during any historic visit; therefore, adding three new populations to the 69 historic populations. At 66% of sites (23 out of 35) had at least one declined species and 39% of populations (28 out of the 72) declined (Table 2). Temporal Patterns in Abundance Breakpoints, which indicate changes in the slope of population abundance over time (e.g., Lips et al. 2006) that fall within the temporal bounds of this study (i.e., 1960 ? 2009), could be determined for one species (Plethodon jordani; Figure 32 26). Plethodon jordani began to change in 1970 (95% C.I. = 1965 ? 1975; Figure 26). The pattern in abundance change for Plethodon jordani appears to be gradual over time (Figure 26). Spatial Patterns in Abundance Declined populations co-occurred with stable and increased populations for GRSM Plethodon salamanders. For all species, the 3 population responses (?declined?, ?stable? or ?increased?) did not show any significant clustering and were randomly distributed throughout the GRSM (Table 9). For a given species, populations that had declined or increased were not spatially clustered around other declined or increased populations of the same species. Patterns in Abundance and the Environment Based on AICc, adding the four environmental variables (AMT, AMP, ? EMP and ? PRECIP) to the separate species GLMMs, did not increase model fit (p > 0.05) for Plethodon glutinosus, P. teyahalee, P. jordani x metcalfi, and P. ventralis (Table 7). I did not have sufficient sample size to run GLMMs with environmental variables for the singlet populations of P. glutinosus x teyahalee, P. jordani x teyahalee, and P. metcalfi. P. jordani and P. serratus showed an increase in model fit compared to the Year model when adding one of the four environmental variables (Table 7). The fit of the P. jordani model significantly increased when I added the average annual temperature (AMT; ?2 = 7.1735; df = 1; p = 0.0074) and average 33 annual precipitation (AMP; ?2 = 7.6863; df = 1; p = 0.0056; Table 7). Populations of P. jordani that decreased in relative abundance during the survey period, occupied cooler areas (z = 3.0830; p = 0.0021) and received a greater amount of annual precipitation (z = ?3.1630; p = 0.0016; Table 7). Based on AICc, the AMP and AMT models performed better than Year model (Table 7). Plethodon serratus also increased in model fit when I added AMP (?2 = 6.1449; df = 1; p = 0.0132) and AMT (?2 = 4.4015; df = 1; p = 0.0359; Table 7). Populations of P. serratus that have declined in relative population abundance since 1960 are cooler (z = 2.1920; p = 0.0284) and receive higher annual average precipitation (z = ?2.6900; p = 0.0071; Table 7). Based on AICc, the AMP and AMT models performed better than Year model (Table 7). Adding AMT to the Plethodon ventralis model increased the model performance based on AICc; however, this increase was not significant (?2 = 2.9104; df = 1; p = 0.0880; Table 7). Although all populations for this species increased, populations of P. ventralis increased most in areas that are cooler (z = -2.0050; p = 0.0450; Table 7). 34 Discussion Some Plethodon populations in the GRSM are declining since the 1960?s and I found no consistent pattern in declines among species sites or populations. Plethodon glutinosus and P. teyahalee have declined at a rate greater than expected, while P. jordani, P. metcalfi, and P. serratus have remained stable overall, and P. jordani x metcalfi and P. ventralis have increased. I found no association among declining populations, elevation, forestry practices, Bd suitability, or spatial relationship, but I found evidence that declines began in the late 1960?s to early 1970?s. Furthermore, I concluded that salamanders in cooler and moister habitats may be more susceptible to the causative agent (e.g., Bd), while salamanders in warmer habitats are thriving. Detection Probabilities and Null Models Count indices can only be used as valid representations of population abundance when detection probability is constant (Schmidt 2009). During my study as well as throughout historic collections, terrestrial salamander detection probability varied over time and among the six species (Table 5). These findings are consistent with other studies, in which temporary subterranean emigration reduces and varies the proportion of salamanders available for capture on the surface among survey seasons (Bailey et al. 2004). Because of the difference in survey effort between my study and Highton?s work as well as the decrease in salamander detection, detection probabilities were consistently lowest during ?2000?s? surveys (Table 5). I adjusted the relative abundance indices during my surveys as well as the historic collections to 35 reduce the likelihood of finding artificial variation in population abundance as well as account for differences (e.g., amount of effort) among surveys (Link and Nichols 1994; Shenk et al. 1998; Schmidt 2009). Future studies should use consistent sampling methods to reduce the variation in detection; however, because species and populations vary in detection (Schmidt 2009), future studies should also account for detection probabilities when using count data in addition to resurveying multiple plots within sites (Smith and Petranka 2000). Furthermore, count indices unadjusted for detection can lead to misinterpretation of data (Schmidt 2009). For example, if I used the unadjusted counts for Plethodon glutinosus, I would have determined that only two of the five populations declined, while the other three remained stable and I would have underestimated the extent to which this species had declined (Table 6). Null models of population abundance are essential for understanding how the observed populations have changed with respect to natural fluctuations (Pounds et al. 1997) and can provide evidence if changes in abundance are a natural occurrence (Alford and Richards 1999). Although salamanders display natural variation in their populations due to the differences in surface and subterranean activity (Feder 1983; Bernardo and Spotila 2006) as well as detection (Bailey et al. 2004), terrestrial plethodontids exhibit one of the lowest amounts of variation in population abundance among amphibian families (Green 2003). My results indicated that the observed changes in abundance for Plethodon glutinosus, P. teyahalee, P. jordani x metcalfi and P. ventralis are beyond the normal variations for those populations. These models examined, for each species, individual populations as well as the species as whole. Previous reports (Hairston and Wiley 1993) indicate no 36 declines in P teyahalee (glutinosus; Hairston 1992) and P. jordani at similar sites to my study from the 1970?s to 1990?s. However, the changes I observed during the course of this study highlight the importance of surveying multiple populations. For example, if I had chosen to survey the three sites in the GRSM at which Plethodon teyahalee had not declined in addition to any three other sites, I most likely would have reached the conclusion that this species had not declined beyond what is expected from natural variation. The power to determine if abundance has changed increases with the number of survey years and the number of surveyed populations (Smith and Petranka 2000); based on my results, I recommend that future studies should examine no less than six populations for each species of terrestrial plethodontids. Maxent Predictions for Bd in the GRSM Species distribution models (SDM) based on climatic envelope estimates can provide insight into the potential distribution of an organism (Phillips et al. 2006; Ron 2005; Lawler 2009; Lawler et al 2010; Milanovich et al. 2010) and have also been used to predict the impact of an infectious pathogen (Bd) on its amphibian hosts (Murray et al. 2011) and have shown very high probability for Bd to occur throughout the GRSM (Ron 2005). Similar to previous studies of Bd?s potential distribution using SDMs (Puschendorf et al. 2009; Murray et al. 2011), my results show that Mean Diurnal Range (BIO 2) and Precipitation during the Driest Quarter (BIO 7) are important predictors of Bd?s potential distribution in the GRSM, which is consistent with Bd?s biology (Johnson et al. 2003; Piotrowski et al. 2004). My results indicate 37 that suitability for Bd varies throughout the GRSM, and that the one locality where I found Bd had a slightly below average suitability of 0.429. Although I did not find Bd in other habitats, areas with higher predicted suitability (e.g., the western portion of the GRSM; Figure 21) than this locality may have a high probability for Bd occurrence because these habitats are generally cooler and receive more precipitation (Hijmans et al. 2005). Results from Maxent modeling neither supported nor rejected the hypothesis that Bd is associated with declines in GRSM Plethodon salamanders. Although the median suitability was highest in declined populations, areas with the highest Bd suitability showed all three population responses, while areas with the lowest Bd suitability were associated with declined and increased populations (Figure 22). Murray et al. (2011), cautioned against using predicted suitability as an exact measure of disease risk, as other factors related to species? biology (e.g., amphibian immune response or skin peptides; Harris et al. 2006; Ramsey et al. 2010) may explain transmission and infection dynamics. Future studies should examine species- specific differences in microbial communities (Harris et al. 2006; Brucker et al. 2008a, b) and whether the effectiveness of these microbial communities are affected by temperature and moisture. Species Patterns in Abundance I used a conservative approach when determining population responses for all species (see Determining Population Response). Using this approach, only populations that showed a distinct decline pattern (i.e., upper and lower standard error of the slope estimate were below zero) were considered declined populations, which 38 may have led to an underestimation of declines. However, this approach provided confidence in Changes in population abundance varied among species (Table 6). P. glutinosus and P. teyahalee, declined while P. jordani, P. metcalfi and P. serratus remained stable and P. jordani x metcalfi and P. ventralis increased (Tables 7 and 8). These results may indicate that these species vary in their susceptibility to the agent for declines. The observed variation in population response among species does not support the idea that historic park forestry practices caused the observed changes. If these changes were the cause of immature forest stands (Petranka et al. 1993) then I expected Plethodon teyahalee, which tolerates drier habitats (Ash 1997), would have remained stable despite the negative impacts of immature forest stands on salamander habitat (Ash 1997; Petranka et al. 1993, 1994). Susceptibility to chytridiomycosis and ranavirus varies among amphibians (Berger et al. 2004; Gray et al. 2009a; Vazquez et al. 2009; Crawford et al. 2010). Within GRSM terrestrial plethodontids, P.metcalfi and P. glutinosus are susceptible to chytridiomycosis in the laboratory, while infected Plethodon glutinosus have been recovered from the wild (Chinnadurai et al. 2009; Vazquez et al. 2009). With respect to ranavirus, species-specific susceptibilities from laboratory tests are not known although P. jordani, P. metcalfi, P. jordani x metcalfi have been found to be infected in the wild (Gray et al. 2009b; M.J.G and N.M.C unpublished data). Based on the observed variation in population responses alone, it is unlikely that ranavirus or Bd has caused the observed declines as Plethodon jordani x metcalfi has not shown 39 reductions in population abundance. If either of these disease-causing agents has contributed to population declines in the GRSM, then I would expect that interactions between Bd and other factors, such as natural (e.g., host density; Gray et al. 2009a; Briggs et al. 2011) and anthropogenic stressors (e.g., pesticides; Gray et al. 2009a; Carey et al. 1999) or differences in host immunity (Ramsey et al. 2010) or skin microbes (Harris et al. 2006) would explain the variation in response in these populations because of the variation among populations within each species (Table 6). Future research is needed to understand the actual variation in species ability to cope with the stress associated with a changing climate (Bernardo and Spotila 2006; Bernardo et al. 2007) beyond SDMs and environmental suitability (Lawler 2009; Lawler et al. 2010; Milanovich et al. 2010). To determine which species are most at risk, studies should measure the stress (e.g., metabolic depression; Bernardo and Spotila 2006) associated with warming temperatures. Specifically, to determine climate?s role in the declines of GRSM salamanders, future studies should determine if Plethodon glutinosus and P. teyahalee are more sensitive to changes in temperature and precipitation than P. ventralis and P. jordani x metcalfi given that these species co-occur but have opposite changes in population abundance. Elevation and Forestry Practices Patterns in Abundance Population declines occurred at all elevations and no elevation was more or less affected than others (Figure 23). This pattern is most likely a consequence of the elevational range of these species rather than from the environmental effects 40 associated with elevation changes. The only species that increased at the high elevations (> 1,500 m) was Plethodon jordani x metcalfi and 36% of the resurveyed high elevation populations consisted of this hybrid (Table 1). Because declines were not limited to specific elevations, these data may suggest a biological/evolutionary link (e.g., sensitivity to changes in climate or presence of microbial symbionts) for declines rather than an environmental cause because not all species declined in the same habitat. Therefore, species which are closely related to Plethodon glutinosus or P. teyahalee (e.g., P. aureolus; Kozak et al. 2006) may be similarly sensitive or susceptible. If acidic precipitation has caused the observed declines, I expected that higher elevation populations would have been the most negatively impacted given that the precipitation pH decreases with increasing elevation (Robinson et al. 2008). However, elevation patterns may not explain the actual variation in acidic habitats in the GRSM. Future studies should examine the exact range of acidic conditions experienced by GRSM salamanders to determine if they experience pH levels low enough to cause mortality or impact growth (Rowe and Freda 2000). If habitat acidification is not a cause for the declines of GRSM salamanders then I expect the soil pH to be above 3.8. Population responses were random with respect to their spatial relationship (Table 9) and declined populations co-occurred with stable and increased populations. These data are consistent with the earlier finding of Highton (2005) for Eastern US Plethodon species. These results indicate that the cause for these declines is species- specific and is not limited to specific areas, which may further suggest a biological or 41 evolutionary cause for declines rather than environmental or habitat acidification. Biological factors such as density (Briggs et al. 2010), skin peptides (Harris et al. 2006), ecology (Lips et al. 2003), phylogenic clumping (Corey and Waite 2008), or variation in susceptibility to disease (Smith et al. 2009; Crawford et al. 2010) may explain these patterns. Forestry practices cannot explain patterns in GRSM Plethodon declines as declined, increased, and stable populations occurred in all levels of historic logging in similar proportions (Figure 24). Furthermore, the lowest proportion of declined populations occurred in the ?heavy cut? sites which were expected to have the highest percentage of declined populations if forestry practices were the cause of declines. Given that the last timber harvest in the GRSM took place in 1939 (NPS 2010), it is unlikely that past timber harvesting has affected present-day salamanders in the GRSM. Diversity Patterns Although all six species were present during current surveys, my results suggest that the GRSM has lost some of its historic diversity as 23 sites have had at least one species decline (Table 2). Moreover, the majority of the declined populations (20 out of 28) were the large-bodied salamanders, P. glutinosus and P. teyahalee and their hybrids. These results may indicate that the declines of GRSM Plethodon salamanders are non-random and may further suggest that the cause of declines is linked to these species? biology. Furthermore, these results may suggest a 42 loss of a functional role as these large salamanders generally eat larger prey items that may be too large for other Plethodon species (Adams and Rolf 2003). Temporal Patterns in Abundance Plethodon jordani declines most likely began in the late 1960?s or early 1970?s. The timing of declines may indicate a climate-link to the observed changes. During the early and mid-1970?s, the GRSM experienced a strong ENSO event (1972; NOAA 2010). This ENSO corresponded to periods of cool, moist conditions (Ropelewski and Halpert, 1986), in which monthly precipitation remained well-above average throughout the year (NCDC 2010). Terrestrial salamanders are expected to thrive in cool, moist conditions (Grover 1998); however, these conditions may also promote pathogen (e.g., Bd) persistence in the soil (Johnson and Speare 2005), and increase transmission in dense salamander populations (Briggs et al. 2010). If moist conditions increase pathogen intensity and prevalence and dry conditions decrease it (see Weinstein 2009) then further strong ENSO years may be associated with increases in Bd or ranavirus transmission and further decreases in salamander abundance. If these declines are caused by a climate-disease link then I expect high pathogen presence will be found in museum specimens collected at the beginning of decline declines (late 1960?s to early 1970?s) as well as during periods of prolonged and above average rainfall in the GRSM (1972 ? 1975; NCDC 2010). Environmental Patterns 43 Precipitation and temperature patterns were important in explaining the variation in relative abundance changes among salamander populations (Table 7), which are consistent with both Bd and ranavirus biology (Chinchar 2002; Piotrowski et al. 2004). Forest soils that receive a greater amount of precipitation and are cooler may create more favorable conditions for Bd and ranavirus to persist and infect amphibians (Chinchar 2002; Piotrowski et al. 2004). During periods of high or continuous precipitation pathogen prevalence may increase. High soil moisture can allow for pathogen survival in the soil (Johnson and Speare 2005; Brunner et al. 2007), which can increase the likelihood of infections. Wet conditions can also promote increases in salamander activities and densities (Grover 1998), which lead to increases in contact with infected individuals (Briggs et al. 2010). Dry conditions, on the other hand, have increased survivorship of infected terrestrial salamanders in empirical studies (Weinstein 2009). Because surveys for Bd and ranavirus in the GRSM during 2009 (this study; M.J.G. and N.M.C unpublished data) and 2006 ? 2007 (Chatfield et al. 2009) were made during months that received lower than average rainfall (NCDC 2010), prevalence estimates for terrestrial plethodontids in this region may be underestimated. This study provides further evidence of declines in Plethodon salamanders (Highton 2005). Consistent with Highton (2005), this study shows widespread declines with no clear spatial pattern in population response. However, Highton (2005) reported declines in 88% of Eastern US Plethodon populations, while this study shows that 39% GRSM Plethodon populations have declined. This difference 44 may be as a result of the difference in spatial scale or because of the differences in species surveyed. Building upon the reported declines by Highton (2005), I analyzed patterns in population responses to determine potential causes for declines. This study provides evidence that historic logging in the GRSM was not the cause for population declines in Plethodon salamanders while acid precipitation, disease, and climate change may still be potential causes. Because of the complexity in these changes in population abundance (e.g., variation in species? response and variation in species within a site) it is likely declines have been caused by multiple factors (e.g., disease epizootic caused by fluctuations in the climate). 45 Conservation Implications This study highlights the need for future studies to examine multiple populations when considering if a species has declined. Using a power analysis, Smith and Petranka (2000) show that the number of plots or sites needed to detect real changes in abundance varies with the number of survey years. Future studies should conduct similar analyses to determine if the number of sites is sufficient to detect changes in abundance. For similar studies, I recommend at least six sites per species. Furthermore, future studies should search multiple plots within each of these sites, which have also shown to increase power to detect changes in abundance (Smith and Petranka 2000). Currently, both Bd and ranavirus are present in few locations within the GRSM and at a low prevalence in terrestrial salamander populations (Chatfield et al. 2009; Gray et al. 2009b; this study; N.M.C. and M.J.G. unpublished data); however, abiotic or biotic factors (e.g., increased rainfall or increased amphibian density) may favor pathogen transmission (Weinstein 2009; Briggs et al. 2011). Therefore, it is important to reduce the spread of these pathogens into na?ve areas. I recommend installing boot-wash stations at GRSM trailheads and educating the public (e.g., trail signs or posting information on websites) about these pathogens to prevent anthropogenic spread. More data are needed in order to determine the role of Bd in the declines of GRSM Plethodon salamanders. First, future studies are needed to determine the arrival of Bd in the GRSM by examining histology of museum specimens. Additionally, studies should examine museum specimens for an increase in Bd 46 prevalence following its initial arrival. These data may suggest an epizootic event, which has been indicative of Bd as a potential cause (Lips et al. 2006). Future studies should also examine susceptibility of current populations of Plethodon salamanders as well as for the presence of cutaneous bacteria or Bd-inhibiting metabolites (Harris et al. 2006; Brucker 2008a, b), which may suggest a co-evolutionary history between these salamanders and Bd. Presence of pathogen-defenses (e.g., bacteria; Harris et al. 2006) in population or species that have not declined (e.g., Plethodon jordani x metcalfi) and absence of these defenses in populations that have declined (e.g., P. glutinosus) may suggest Bd as a potential cause for declines. Further research is also needed to determine the role of climate change in the declines of GRSM Plethodon salamanders. First, studies are needed to determine the relationship between the soil conditions (i.e., temperature and moisture) experienced by terrestrial salamanders and air temperature and precipitation. Additionally, studies should examine the effects of changes in air temperature and precipitation on the overall fitness of salamanders (e.g., metabolic stress; Bernardo and Spotila 2006) and to determine which populations are at risk based on the magnitude of temperature and precipitation changes. If variation in species response, as determined by these studies, suggest climate change as a potential cause (e.g., Plethodon glutinosus and P. teyahalee are more sensitive than other species), then future monitoring is needed for these populations and potentially relocating the high risk populations to habitats that are more suitable for terrestrial salamanders. This study documents two species of concern, Plethodon glutinosus and P. teyahalee. Although these species are still present in the GRSM, they have declined 47 throughout their resurveyed range and show decline patterns in other studies (Highton 2005). Future studies should examine these species, as well as other closely related glutinosus group species (Kozak et al. 2006) throughout their range to determine where they have declined. Currently, both Plethodon glutinosus and P. teyahalee are listed as ?least concern? and have populations that are considered ?stable? (IUCN 2010). Because of the population abundance trends for both of these species throughout their range (Highton 2005; this study); I recommend reevaluation of the IUCN red list category and criteria for both Plethodon glutinosus and P. teyahalee. 48 Tables Species Total Populations Low Elevation Populations Mid Elevation Populations High Elevation Populations P. glutinosus 5 4 1 0 P. glutinosus x teyahalee 1 1 0 0 P. teyahalee 16 2 12 2 P. jordani x teyahalee 1 0 1 0 P. jordani 18 1 13 4 P. metcalfi 1 0 0 1 P. jordani x metcalfi 8 0 4 4 P. serratus 18 5 13 0 P. ventralis 4 4 0 0 Total 72 17 44 11 Table 1: Number of populations surveyed for each species. Low elevations are between 500 ? 1,000 m, Mid elevations are between 1,000 ? 1,500 m and High elevations are above 1,500 m. 49 Historic Number of Species Number of Sites Sites with ? 1 Declined Species 1 2 2 2 1 1 Low Elevation 3 4 4 1 2 0 2 11 7 Mid Elevation 3 6 6 1 7 2 2 2 1 High Elevation 3 ? ? Table 2: Number of sites based on the number of historic species present. Low elevations are between 500 ? 1,000 m, Mid elevations are between 1,000 ? 1,500 m and High elevations are above 1,500 m. The number of declined species is based off of GLMM results. ??? indicates that no sites with three species historically were sampled at high elevations 50 Species Groups (Kozak et al. 2006) Elevational Range Home Range Distribution Bd/Ranavirus P. glutinosus glutinosus group Up to 1,500 m (Petranka 1998) 4 ? 14 m2 (Merchant 1972; Marvin 1998) Throughout the Eastern US (Petranka, 1998). +/0 P. teyahalee glutinosus group Up to 1,550 m (Petranka 1998) 6.5 ? 14.3 m 2 (Nishikawa 1990) Blue Ridge Province of southeastern TN, southwestern NC, northwestern SC and northeastern GA (Highton 1987) ?/? P. jordani glutinosus group 213 ? 1951 m (Grobman 1944) 1.7 ? 11.4 m 2 (Merchant 1972) TN?NC border; disjunct population in extreme northeastern GA (Petranka 1998) ?/+ Table 3: Biological information for GRSM Plethodon salamanders. For Bd/Ranavirus, the ?+? denotes that a species has been found infected in the field with the respected pathogen, the ?-?denotes a species that has not been found to be infected with the respected pathogen, and ?0? denotes species that have not been surveyed for that pathogen (Rothermel et al. 2008; Gray et al. 2009b; Chinnadurai et al. 2009; this study; M.J. Gray and N.M. Caruso unpublished data. 51 P. metcalfi glutinosus group Above 750 m (Highton and Peabody 2000). 1.87 ? 5.04 m2 (Nishikawa, 1990). Southern Blue Ridge Mountains in NC, SC, and GA (Highton and Peabody 2000) ?/+ P. serratus cinereus group Up to 1,686 m (Huheey and Stupka 1967) Unknown Four disjunct isolates: southeastern MO and western IL; northwestern GA, eastern AL, eastern TN, and western NC; central LA; southeastern OK and western AR (Petranka, 1998) ?/? P. ventralis welleri group Up to 579 m (King 1939; Highton 1972) Unknown Scattered from northern MS to southeastern VA, including populations in northern AL and GA, eastern TN, and western NC (Petranka 1998) ?/0 Table 3: continued 52 Species Number Swabbed Number Infected Prevalence (%) 95% CI P. glutinosus 7 0 0 0.00?35.43 P. teyahalee 8 0 0 0.00?32.44 P. jordani x teyahalee 29 0 0 0.00?11.70 P. jordani 228 0 0 0.00?1.66 P. metcalfi 28 0 0 0.00?12.06 P. jordani x metcalfi 110 0 0 0.00?3.37 P. serratus 19 0 0 0.00?16.82 P. ventralis 56 0 0 0.00?6.42 Total 485 0 0 0.00?0.79 Table 4: Prevalence and intensity of Bd infection for all species of Plethodon salamanders sampled during my 2009 surveys. The number of total animals swabbed, number of infected individuals, prevalence estimates, 95% Clopper-Pearson binomial confidence intervals, and the average intensity are summarized. 53 Species p [1960's] p [1970's] p [1980's] p [1990's] p [2000's] P. glutinosus 0.8303 1.0000 1.0000 1.0000 0.3077 P. glutinosus x teyahalee ? 1.0000 1.0000 ? 0.0000 P. teyahalee 0.8786 0.5521 0.6176 ? 0.0722 P. jordani x teyahalee 1.0000 1.0000 1.0000 ? 1.0000 P. jordani 0.9643 1.0000 1.0000 1.0000 0.9425 P. metcalfi ? 1.0000 1.0000 ? 1.0000 P. jordani x metcalfi 0.8750 1.0000 ? ? 1.0000 P. serratus 0.6030 0.7599 0.3000 0.5000 0.2195 P. ventralis 0.8000 0.7698 ? 1.0000 0.3969 Table 5: Detection probabilities for each species. Detection probabilities (p[ ]) were estimated for each decade. ?? ? denotes decades in detection probabilities could not be estimated. 54 Species Pairwise Comparisons Total Populations Decline Stable Increasing P. glutinosus AB 5 5 (100%) 0 0 P. glutinosus x teyahalee A 1 1 (100%) 0 0 P. teyahalee BCG 16 13 (81%) 2 (12%) 1 (6%) P. jordani x teyahalee ABD 1 1 (100%) 0 0 P. jordani AE 18 5 (28%) 9 (50%) 4 (22%) P. metcalfi EF 1 0 1 (100%) 0 P. jordani x metcalfi BC 8 0 1 (13%) 7 (87%) P. serratus ABCDEFG 18 3 (17%) 10 (55%) 5 (28%) P. ventralis BCF 4 0 0 4 (100%) Total 72 28 (39%) 23 (32%) 21 (29%) Table 6: Number of declined, stable and increased populations for each species. Pairwise comparisons with unlike letters show differences among species (p < 0.05). Numbers in parentheses show percentages of total populations. 55 Species Model K AICc ?AICc Estimate Standard Error Z value p value Year 5 211.5 0.47 -0.2977 0.1135 -2.6240 0.0087 Year + ? TEMP 6 213.1 0.15 -0.0896 0.1100 -0.8150 0.4153 Year + AMT 6 214.1 0.13 -0.0220 0.0741 -0.2970 0.7662 Year + AMP 6 214.1 0.13 0.0228 0.0885 0.2580 0.7963 P. glutinosus Year + ? PRECIP 6 214.2 0.12 -0.0022 0.1494 -0.0150 0.9883 P. glutinosus x teyahalee Year 5 84.2 1.00 -3.2135 1.4634 -2.1960 0.0281 Year 5 200.6 0.31 -3.1655 0.7546 -4.1950 < 0.0001 Year + AMP 6 200.8 0.27 -0.3822 0.2649 -1.4430 0.1490 Year + AMT 6 201.3 0.21 0.3143 0.2572 1.2220 0.2217 Year + ? TEMP 6 202.5 0.12 -0.2534 0.3177 -0.7980 0.4252 P. teyahalee Year + ? PRECIP 6 202.8 0.10 0.0330 0.2789 0.1180 0.9026 Table 7: AICc scores and GLMM estimates for each model. Bolded numbers indicate significant variable for each model (p < 0.05). Models are ranked in descending order based on AICc score. K indicates the number of parameters in each model, while ?AICc is the model weight. 56 P. jordani x teyahalee Year 5 59.0 1.00 -0.3753 0.1061 -3.5370 0.0004 Year + AMP 6 820.4 0.53 -0.2740 0.0866 -3.1630 0.0016 Year + AMT 6 820.9 0.41 0.2670 0.0866 3.0830 0.0021 Year 5 825.8 0.03 0.0287 0.1194 0.2410 0.8100 Year + ? TEMP 6 828.1 0.01 0.0163 0.1038 0.1570 0.8760 P. jordani Year + ? PRECIP 6 828.1 0.01 -0.0021 0.1086 -0.0190 0.9850 Table 7: continued 57 P. metcalfi Year 5 66.7 1.00 -0.1491 0.1836 -0.8120 0.4170 Year 5 204.0 0.46 0.5484 0.1079 5.0820 < 0.0001 Year + ? PRECIP 6 205.8 0.19 -0.0745 0.0725 0.1028 0.3040 Year + ? TEMP 6 206.6 0.12 -0.0356 0.0691 -0.5160 0.6060 Year + AMP 6 206.7 0.12 -0.0237 0.0869 -0.2720 0.7850 P. jordani x metcalfi Year + AMT 6 206.8 0.11 0.0166 0.0875 0.1900 0.8500 Year + AMP 6 462.3 0.60 -0.6023 0.2239 -2.6900 0.0071 Year + AMT 6 464.1 0.25 0.5055 0.2306 2.1920 0.0284 Year 5 466.3 0.08 0.0409 0.2394 0.1710 0.8640 Year + ? TEMP 6 467.8 0.04 0.2849 0.3031 0.9400 0.3470 P. serratus Year + ? PRECIP 6 468.4 0.03 -0.0906 0.2614 -0.3470 0.7290 Table 7: continued 58 Year + AMT 6 612.6 0.35 -0.4247 0.2118 -2.0050 0.0450 Year 5 612.8 0.34 0.3560 0.0551 6.4650 < 0.0001 Year + ? TEMP 6 614.8 0.12 0.1571 0.1270 1.2370 0.2160 Year + AMP 6 615.2 0.10 0.1020 0.1305 0.7820 0.4344 P. ventralis Year + ? PRECIP 6 615.3 0.09 0.0860 0.1734 0.4960 0.6200 P. metcalfi Year 5 66.7 1.00 -0.1491 0.1836 -0.8120 0.4170 Table 7: continued 59 Temperature Range (oC) Annual Precipitation Range (mm) Annual Temperature Change (oC) Annual Precipitation Change (%/year) Low Elevations -5 to 29 1,323 to 1,511 -0.067 to 0.023 -0.067 to 0.104 Mid Elevations -7 to 26 1,485 to 1,845 -0.077 to 0.034 -0.148 to 0.151 High Elevations -8 to 24 1,770 to 1,946 -0.079 to -0.007 -0.138 to 0.274 Table 8: Temperature and precipitation ranges among the three elevational groups. Temperature range and annual precipitation were derived from Hijmans et al. (2005). Annual temperature change and annual precipitation change reflect the magnitude of change from 1951 ? 2006 (Girvetz et al. 2009). 60 Species Observed General G Expected General G Z score p value P. glutinosus ? ? ? ? P. teyahalee 0.058 0.0667 -1.6504 0.7207 P. jordani 0.0001 0.0002 -0.6525 0.5141 P. jordani x metcalfi 0.1414 0.1429 -0.3575 0.117 P. serratus 0.0543 0.0588 -1.5674 0.0989 P. ventralis ? ? ? ? Table 9: Results of High/Low clustering analysis (Getis-Ord General G). Only species with at least 2 surveyed populations were analyzed. Species without variation in population responses could not be analyzed. 61 Figures Figure 1: Map showing the widespread declines in Plethodon salamander populations. At each resurveyed sites; these declines occurred during the 1980s (Highton 2005). ?Decline? indicates populations where the 1990s average abundance was lower than the average abundance from 1960s ? 1980s. ?Non-Decline? indicates populations where the 1990s average abundance was greater than or equal to the 1960s ? 1980s average abundance. This map was created using the data in Highton (2005). 62 Figure 2: Map showing the location of the resurveyed sites in the GRSM with respect to elevation. 63 Figure 3: Frequency graph showing the distribution in historic sampling. Numbers along the x-axis show the number of visits at a particular site. 64 Figure 4: Relative population abundance over time at selected sites for Plethodon glutinosus. Filled-in circles represent one sampling occasion. Numbers in the gray-shaded area represent the site?s elevation. The solid line represents the linear relationship between the adjusted abundance index and the survey year. 65 Figure 5: Relative population abundance over time at selected sites for Plethodon jordani x metcalfi. Filled-in circles represent one sampling occasion. Numbers in the gray-shaded area represent the site?s elevation. The solid line represents the linear relationship between the adjusted abundance index and the survey year. 66 Figure 6: Relative population abundance over time at selected sites for Plethodon jordani x teyahalee. Filled-in circles represent one sampling occasion. Numbers in the gray-shaded area represent the site?s elevation. The solid line represents the linear relationship between the adjusted abundance index and the survey year. 67 Figure 7: Relative population abundance over time at selected sites for Plethodon jordani. Filled-in circles represent one sampling occasion. Numbers in the gray-shaded area represent the site?s elevation. The solid line represents the linear relationship between the adjusted abundance index and the survey year. 68 Figure 8: Relative population abundance over time at selected sites for Plethodon metcalfi. Filled-in circles represent one sampling occasion. Numbers in the gray-shaded area represent the site?s elevation. The solid line represents the linear relationship between the adjusted abundance index and the survey year. 69 Figure 9: Relative population abundance over time at selected sites for Plethodon serratus. Filled-in circles represent one sampling occasion. Numbers in the gray-shaded area represent the site?s elevation. The solid line represents the linear relationship between the adjusted abundance index and the survey year. 70 Figure 10: Relative population abundance over time at selected sites for Plethodon teyahalee. Filled-in circles represent one sampling occasion. Numbers in the gray-shaded area represent the site?s elevation. The solid line represents the linear relationship between the adjusted abundance index and the survey year. 71 Figure 11: Relative population abundance over time at selected sites for Plethodon ventralis. Filled-in circles represent one sampling occasion. Numbers in the gray-shaded area represent the site?s elevation. The solid line represents the linear relationship between the adjusted abundance index and the survey year. 72 Figure 12: Observed and expected population responses for Plethodon glutinosus. The ?2 statistic and p value are displayed at the top of the graph. 73 Figure 13: Observed and expected population responses for Plethodon teyahalee. The ?2 statistic and p value are displayed at the top of the graph. 74 Figure 14: Observed and expected population responses for Plethodon jordani. The ?2 statistic and p value are displayed at the top of the graph. 75 Figure 15: Observed and expected population responses for Plethodon serratus. The ?2 statistic and p value are displayed at the top of the graph. 76 Figure 16: Observed and expected population responses for Plethodon jordani x metcalfi. The ?2 statistic and p value are displayed at the top of the graph. 77 Figure 17: Figure 15: Observed and expected population responses for Plethodon ventralis. The ?2 statistic and p value are displayed at the top of the graph. 78 Figure 18: Variable contributions to training gain for the trimmed model of Bd in the GRSM. ?Only variable? indicates the training gain when a single variable is run in isolation; ?without variable? denotes the effect of removing a single variable from the full model (jack-knife). Values are means from 100 replicates 79 Figure 19: Variable contributions to area under the curve (AUC) for the trimmed model for Bd in the GRSM. ?Only variable? indicates the AUC when a single variable is run in isolation; ?without variable? denotes the effect of removing a single variable from the full model AUC (jack-knife). Values are means from 100 replicates 80 Figure 20: Predicted maximum suitability of Bd in the GRSM. Higher values denote areas that are more suitability for Bd occurrence. Maximum suitability was determined by the model with the highest suitability for the GRSM. This Figure shows that very few areas in the GRSM reach the maximum suitability level (0.9 ? 1), while the maximum model predicts some areas to have low suitability (0.33 ? 0.4). 81 Figure 21: Predicted average suitability of Bd in the GRSM. Higher values denote areas that are more suitability for Bd occurrence. Average is based on all 100 models. This Figure shows that the western portion of the GRSM has higher suitability for Bd than the eastern half of the GRSM. 82 Figure 22: Average suitability of each population response. The filled in circles represent the median suitability for each population response. This Figure shows that the average suitability does not vary with population response for GRSM Plethodon salamanders. 83 Figure 23: Proportion of population response based elevational bands. For population responses, ?1? denotes declined populations, ?2? is stable populations and ?3? is increased populations, while the numbers on the right denote the proportion of the specific population response in each of the elevational bands (Low: 500 ? 1,000 m; Mid: 1,000 ? 1,500 m; High: > 1,500 m) 84 Figure 24: Proportion of population response based on type of historic logging (NPS 2007). For population responses, ?1? denotes declined populations, ?2? is stable populations and ?3? are increased populations, while the numbers on the right denote the proportion of the specific population response in each of the logging categories. 85 Figure 25: Number of populations found during historic (1960 ? 1979) and current (2009) surveys based on community size. 86 Figure 26: Graph showing the adjusted relative abundance for Plethodon jordani from 1962 ? 2009. Solid lines indicate the piecewise linear regressions. The solid point with error bars represents the estimated point of change and 95% confidence intervals. This Figure shows that declines for this species most likely began during the late 1960?s to early 1970?s. 87 Literature Cited Adams, D.C. 2007. Organization of Plethodon salamander communities: guild- based community assembly. Ecology 88:1292?1299. Adams, D.C., and F.J. Rohlf. 2000. Ecological character displacement in Plethodon: biomechanical differences found from a geometric morphometric study. Proceedings of the National Academy of Sciences of the United States of America 97:4106?4111. Alford, R.A. and S.J. Richards. 1999. Global amphibian declines: a problem in applied ecology. Annual Review of Ecology, Evolution, and Systematics 30:113?165. Ariel E., J. Kielgast, H.E. Svart, K. Larsen, H. Tapiovaara, B. Bang Jensen, R. Holopainen. 2009. Ranavirus in wild edible frogs Pelophylax kl. esculentus in Denmark. Diseases Aquatic Organisms 85:7?14. Ash, A.N. 1988. Disappearance of salamanders from clearcut plots. Journal of the Elisha Mitchell Scientific Society 104:116?122. Ash, A.N. 1995. Effects of clear?cutting on litter parameters in the southern Blue Ridge mountains. Castanea 60:89?97. Bailey, J.E. and T.K. Pauley. 1993. Aspects of the natural history of the Cumberland Plateau salamander, Plethodon kentucki, in West Virginia. Association of Southeastern Biologists Bulletin 40:133. Bailey, L.L., T.R. Simons, and K.H. Pollock. 2004. Estimating detection probability parameters for Plethodon salamanders using the robust capture-recapture design. Journal of Wildlife Management 68:1?13. Bates, D., and M. Maechler. 2010. lme4: Linear mixed-effects models using S4 classes. R package version 0.999375-35. http://CRAN.R- project.org/package=lme4 Becker, M.H. and R.N. Harris. 2010. Cutaneous bacteria of the redback salamander prevent morbidity associated with a lethal disease. PLoS One 5:e10957. Becker, M.H., R.M. Brucker, C.R. Schwantes, R.N. Harris, and K.P.C. Minbiole. 2009. The bacterially produced metabolite Violacein is associated with survival of amphibians infected with a lethal fungus. Applied and Environmental Microbiology 75:6635?6638. 88 Bennett, S.H., J.W. Gibbons and J. Glanville. 1980. Terrestrial activity, abundance and diversity of amphibians in differently managed forest types. American Midland Naturalist 103:412?416. Berger, L., R. Speare, P. Daszak, D.E. Green, A.A. Cunningham, C.L. Goggin, R. Slocombe, M. A. Ragan, A.D. Hyatt, K.R. McDonald, H. B. Hines, K.R. Lips, G. Marantelli, and H. Parkes. 1998. Chytridiomycosis causes amphibian mortality associated with population declines in the rain forests of Australia and Central America. Proceedings of the National Academy of Sciences of the United States of America 95:9031?9036. Berger, L., R. Speare, H.B. Hines, G. Marantelli, A.D. Hyatt, K.R. McDonald, L.F. Skerratt, V. Olsen, J.M. Clarke, G. Gillespie, M. Mahoney, N. Sheppard, C. Williams, and M.J. Tyler. 2004. Effect of season and temperature on mortality in amphibians due to chytridiomycosis. Australian Veterinary Journal 82:434?439. Bernardo, J., and J.R. Spotila. 2006. Physiological constraints on organismal response to global warming: mechanistic insights from clinally varying populations and implications for assessing endangerment. Biology Letters 2:135?139. Bernardo J., R.J. Ossola, J. Spotila, and K.A. Crandall. 2007. Interspecies physiological variation as a tool for cross?species assessments of global warming?induced endangerment: validation of an intrinsic determinant of macroecological and phylogeographic structure. Biology Letters 3: 695? 698 Beyer, H.L. 2010. Geospatial Modeling Environment. Available from Bielby, J., N. Cooper, A. Cunningham, T. Garner, and A. Purvis. 2008. Predicting susceptibility to future declines in the world?s frogs. Conservation Letters 1:82?90. Bishop, S.C. 1941b. The Salamanders of New York. New York State Museum Bulletin, Number 324, Albany, New York. Blaustein, A.R., T.L. Root, J.M. Kiesecker, L.K. Belden, D.H. Olson, and D.M. Green. 2003. Amphibian breeding and climate change: Reply to Corn. Conservation Biology 17:626?627. Bosch, J., and I. Mart?nez?Solano. 2006. Chytrid fungus infection related to unusual mortalities of Salamandra salamandra and Bufo bufo in the Pe?alara Natural Park, Spain. Oryx 40:84?89. 89 Boyle, D.G., D.B. Boyle, V. Olsen, J.A.T. Morgan, and A.D. Hyatt. 2004. Rapid quantitative detection of chytridiomycosis (Batrachochytrium dendrobatidis) in amphibian samples using real?time Taqman PCR assay. Diseases of Aquatic Organisms 60:141?148. Brem, F.M.R. and K.R. Lips. 2008. Batrachochytrium dendrobatidis infection patterns among Panamanian amphibian species, habitats and elevations during epizootic and enzootic. Diseases of Aquatic Organisms 81:189? 202. Briggs, C.J., R.A. Knapp, and V.T. Vredenburg. 2010. Enzootic and epizootic dynamics of the chytrid fungal pathogen of amphibians. Proceedings of the National Academy of Sciences of the United States of America 107:9695?9700. Brooks, M. 1946. Burrowing of Plethodon jordani. Copeia 1946:102. Brucker, R.M., C.M. Baylor, R.L. Walters, A. Lauer, R.N. Harris, and K.P.C. Minbiole. 2008a. The identification of 2,4?diacetylphoroglucinol as an antifungal metabolite produced by cutaneous bacteria of the salamander Plethodon cinereus. Journal of Chemical Ecology 34:39?43. Brucker, R.M., R.N. Harris, C.R. Schwantes, T.N. Gallaher, D.C. Flaherty, B.A. Lam, and K.P.C. Minbiole. 2008b. Amphibian chemical defense: antifungal metabolites of the microsymbiont Janthinobacterium lividum on the salamander Plethodon cinereus. Journal of Chemical Ecology 34:1422?1429. Brunner J.L., D.M. Schock, J.P. Collins. 2007. Transmission dynamics of the amphibian ranavirus Ambystoma tigrinum virus. Disease of Aquatic Organisms 77:87?95. Burton, T.M., and G.E. Likens. 1975a. Salamander populations and biomass in the Hubbard Brook Experimental Forest, New Hampshire. Copeia 1975:541?546. 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. Bury, R.B. 1983. Differences in amphibian populations in logged and old-growth redwood forests. Northwest Science 57:167?178. Carey, C., N. Cohen, and L. Rollins-Smith. 1999. Amphibian declines: an immunological perspective. Developmental and Comparative Immunology 23:459?427 90 Carey, C., and M.A. Alexander. 2003. Climate change and amphibian declines: is there a link? Diversity and Distributions 9:111?121. Carey, C., W.R. Heyer, J. Wilkinson, R.A. Alford, J.W. Arntzen, T. Halliday, L. Hungerford, K.R. Lips, E.M. Middleton, S.A. Orchard, and A.S. Rand. 2001. Amphibian declines and environmental change: Use of remote? sensing data to identify environmental correlates. Conservation Biology 15:903?913. Chadwick, C.S. 1940. Some notes on the burrows of Plethodon metcalfi. Copeia 1940:50. Chatfield, M.W., B.B. Rothermel, C.S. Brooks, and J.B. Kay. 2009. Detection of Batrachochytrium dendrobatidis in amphibians from the Great Smoky Mountains of North Carolina and Tennessee, USA. Herpetological Review 40:176?179. Chinchar, V.G. 2002. Ranaviruses (family Iridoviridae): Emerging cold?blooded killers. Archives of Virology 147:447?470. Chinnadurai, S.K., D. Cooper, D.S. Dombrowski, M.F. Poore, and M.G. Levy. 2009. Experimental infection of native North Carolina salamanders with Batrachochytrium dendrobatidis. Journal of Wildlife Diseases 45:631? 636. Cogbill, C. 1976. The history and character of acid precipitation in eastern North America. Water, Air, and Soil Pollution 6:407?413. Collins, J.P., and A. Storfer. 2003. Global amphibian declines: sorting the hypotheses. Diversity and Distributions 9:89?98. Collins J.P., T.R. Jones, H.J. Berna. 1988 Conserving genetically distinctive populations: the case of the Huachuca tiger salamanders (Ambystoma tigrinum stebbinsi Lowe). In: Szaro R.C., Severson KC, Patton DR (eds) Management of Amphibians, Reptiles, and Small mammals in North America. GTR-RM-166, US Department of Agriculture Forest Service, Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO, p 45?53. Corey, S.J., and T.A. Waite. 2008. Phylogenetic autocorrelation of extinction threat in globally imperiled amphibians. Diversity and Distributions 14:614?629. Crawford, A.J., K.R. Lips, and E. Bermingham. 2010. Epidemic disease decimates amphibian abundance, species diversity, and evolutionary 91 history in the highlands of central Panama. Proceedings of the National Academy of Sciences of the United States of America. 107:13777?13782. Croissant, Y. 2010. mlogit: multinomial logit model. R package version 0.1-8. http://CRAN.R-project.org/package=mlogit Cummer, M.R., D.E. Green, E.M. O?Neill. 2005. Aquatic chytrid pathogen detected in terrestrial plethodontid salamander. Herpetological Review 36:248?249. Daszak, P., A. Strieby, A.A. Cunningham, J.E. Longcore, C.C. Brown, D. Porter. 2004. Experimental evidence that the bullfrog (Rana catesbeiana) is a potential carrier of chytridiomycosis, an emerging fungal disease of amphibians. Journal of Herpetology 14:201?207. Daszak, P., D.E. Scott, A.M. Kilpatrick, C. Faggioni, J.W. Gibbons, and D. Porter. 2005. Amphibian population declines at Savannah River Site are linked to climate, not chytridiomycosis. Ecology 86:3232?3237. Davic, R.D. 1983. An investigation of salamander guild predation in a North Carolina stream: an experimental approach. Unpublished PhD Thesis. Kent State University, Ohio. Davic, R.D., and H.H. Welsh Jr. 2004. On the ecological roles of salamanders. Annual Review of Ecology Evolution and Systematics 35:405?434. Davis, M.B. 1993. Old?growth in the east: A survey. Wild Earth, Richmond, VT Davidson, E.W., M. Parris, J.P. Collins, J.E. Longcore, A.P. Pessier, J. Brunner. 2003. Pathogenicity and transmission of chytridiomycosis in Tiger Salamanders (Ambystoma tigrinum). Copeia 3:601?607. Dupuis, L., J.N.M. Smith, and F. Bunnell. 1995. Relation of terrestrial?breeding amphibian abundance to tree?stand age. Conservation Biology 9:645?653. Elith, J., C.H. Graham, R.P. Anderson, M. Dudik, S. Ferrier, A. Guisan, R.J. Hijmans, F. Huettmann, J.R. Leathwick, A. Lehmann, J. Li, L.G. Lohmann, B.A. Loiselle, G. Manion, C. Moritz, M. Nakamura, Y. Nakazawa, J.M. Overton, A.T. Peterson, S.J. Phillips, K. Richardson, R. Scachetti-Pereira, R.E. Schapire, J. Soberon, S. Williams, M.S. Wisz, and N.E. Zimmermann. 2006. Novel methods improve prediction of species? distributions from occurrence data. Ecography 29: 129?151. Evans, R.A. 2002. An ecosystem unraveling? In B. Onken, R. Reardon, and J. Lashomb (eds.), Proceedings of the Symposium on the Hemlock Woolly 92 Adelgid in Eastern North America, pp 23?33. New Jersey Agricultural Experiment Station and Rutgers University, East Brunswick, N.J. Feder, M.E. 1983. Integrating the ecology and physiology of plethodontid salamanders. Herpetologica 39:291?310. Fielding, A.H. and J.F. Bell. 1997. A review of methods for the assessment of prediction errors in conservation presence ? absence models. Environmental Conservation 24:38?49. Forson, D.D., and A. Storfer. 2006. Atrazine increases ranavirus susceptibility in the tiger salamander, Ambystoma tigrinum. Ecological Applications 16:2325?2332. Fraser, D.F. 1976. Empirical evaluation of the hypothesis of food competition in salamanders of the genus Plethodon. Ecology 57:459?471. Frisbie, M.P., and R.L. Wyman. 1991. The effects of soil?pH on sodium?balance in the red?backed salamander, Plethodon cinereus, and 3 other terrestrial salamanders. Physiological Zoology 64:1050?1068. Frisbie, M.P., and R.L. Wyman. 1995. A field simulation of the effect of acidic rain on ion balance in a woodland salamander. Archives of Environmental Contamination and Toxicology 28:327?333. Gaffin, D.M., D.G. Hotz, and T.I. Getz. 2002. An evaluation of temperature variations around the Great Smoky Mountains National Park and their associated synoptic weather patterns. NOAA Technical Memorandum: NWS SR?221. Gahl, M.K., and A.J.K. Calhoun. 2008. Landscape setting and risk of Ranavirus mortality events. Biological Conservation. 141:2679?2689. Gesch, D., M. Oimoen, S. Greenlee, C. Nelson, M. Steuck, and D. Tyler. 2002. The national elevation dataset. Photogrammetric Engineering and Remote Sensing 68:5?11. Getis, A., and J. K. Ord. 1992. The analysis of spatial association by use of distance statistics. Geographic Analysis 24:189?206. Girvetz E.H., C. Zganjar, G.T. Raber, E.P. Maurer, P. Kareiva, and J.J. Lawler. 2009. Applied climate?change analysis: The climate wizard tool. PLoS ONE 4: e8320. doi:10.1371/journal.pone.0008320. Goka, K., J. Yokoyama, Y. Une, T. Kuroki, K. Suzuki, M. Nakahara, A. Kobayashi, S. Inaba, T. Mizutani, and A.D. Hyatt. 2009. Amphibian 93 chytridiomycosis in Japan: distribution, haplotypes and possible route of entry into Japan. Molecular Ecology 18:4757?4774. Graveland, J. 1990. Effects of acid precipitation on reproduction in birds. Experientia 46:962?970. Grant, E.H.C., L.L. Bailey, J.L. Ware, and K.L. Duncan. 2008. Prevalence of the amphibian pathogen Batrachochytrium dendrobatidis in stream and wetland amphibians in Maryland, USA. Applied Herpetology 5:233?241. Gray, M.J., D.L. Miller, A.C. Schmutzer, and C.A. Baldwin. 2007. Frog virus 3 prevalence in tadpole populations inhabiting cattle?access and non?access wetlands in Tennessee, USA. Diseases of Aquatic Organisms. 77:97?103. Gray, M.J., D.L. Miller, J.T. Hoverman. 2009a. Ecology and pathology of amphibian ranaviruses. Diseases of Aquatic Organisms 87:243?266. Gray, M.J., D.L. Miller, J.T. Hoverman. 2009b. First report of Ranavirus infecting lungless salamanders. Herpetological Review 40:316?319. Green, D.M. 2003. The ecology of extinction: population fluctuation and decline in amphibians. Biological Conservation 111:331?343. Green, D.E., K.A. Converse, and A.K. Schrader. 2002. Epizootiology and of sixty?four amphibians morbidity and mortality events in the USA, 1996? 2001. Annals of the New York Academy of Sciences 969: 323?339. Greer A.L., J.L. Brunner, and J.P. Collins. 2009. Spatial and temporal patterns of Ambystoma tigrinum virus (ATV) prevalence in tiger salamanders Ambystoma tigrinum nebulosum. Diseases of Aquatic Organisms 85:1?6. Griggs D. 2009. Polygon feature class of Frank Miller's 1938 forest types and disturbance history map at Great Smoky Mountains National Park, Tennessee and North Carolina. National Park Service, Great Smoky Mountains National Park, Resource Management and Science. Geospatial Dataset?1048440. Grobman, A.B. 1944. The distribution of the salamanders of the genus Plethodon in eastern United States and Canada. Annals of the New York Academy of Sciences 45:261?316. 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. 94 Hairston, N.G. 1949. The local distribution and ecology of the plethodontid salamanders of the southern Appalachians. Ecological Monographs 19:47? 73. Hairston, N.G. Sr. 1987. Community Ecology and Salamander Guilds. New York: Cambridge University Press. Hairston, N.G. and C.H. Pope. 1948. Geographic variation and speciation in Appalachian salamanders (Plethodon jordani group). Evolution 2:266? 278. Hairston, N. G. 1992. On the validity of the name teyahalee as applied to a member of the Plethodon glutinosus complex (Caudata: Plethodontidae): a new name. Brimleyana 18:59?64. Hairston, N.G., Sr. and R.H. Wiley. 1993. No decline in salamander (Amphibia: Caudata) populations: a twenty year study in the southern Appalachians. Brimleyana 18:59?64. Harrell Jr., F.E., and with contributions from many other users. 2010. Hmisc: Harrell Miscellaneous. R package version 3.8-3. http://CRAN.R- project.org/package=Hmisc Harris, R.N., T.Y. James, A. Lauer, M.A. Simon, and A. Patel. 2006. Amphibian pathogen Batrachochytrium dendrobatidis is inhibited by the cutaneous bacteria of amphibian species. Ecohealth 3:53?56. Harris, R.N., A. Lauer, M.A. Simon, J.L. Banning, R.A. Alford. 2009a. Addition of antifungal bacteria to salamanders ameliorates the effects of chytridiomycosis. Diseases of Aquatic Organisms 83:11?16. Harris, R.N., R.M. Brucker, J.B. Walke, M.H. Becker, C.R. Schwantes, D.C. Flaherty, B.A. Lam, D.C. Woodhams, C.J. Briggs, V.T. Vredenburg, and K.P.C. Minbiole. 2009b. Skin microbes on frogs prevent morbidity and mortality caused by a lethal skin fungus. ISME Journal 3:818?824. Heatwole, H. 1960. Environmental factors influencing local distribution and activity of the salamander Plethodon cinereus. Ecology 43:460?472. Highton, R. 1972. Distributional interactions among eastern North American salamanders of the genus Plethodon. Pp. 139?188. In Holt, P.C. (ed.), The Distributional History of the Biota of the Southern Appalachians. Part III: Vertebrates. Research Division Monograph, Virginia Polytechnic Institute and State University, Blacksburg, Virginia. 95 Highton, R. 1983. A new species of woodland salamander of the Plethodon glutinosus group from the southern Appalachian mountains. Brimleyana 9:1?20. Highton, R. 2005. Declines of eastern North American woodland salamanders (Plethodon). In M.J. Lannoo (ed), Status and Conservation of U.S. Amphibians, pp 34?46. University of California Press, Berkeley. Highton, R. and R.B. Peabody. 2000. Geographic protein variation and speciation in salamanders of the Plethodon jordani and Plethodon glutinosus complexes in the southern Appalachian Mountains with the description of four new species. Pp. 31?93. In Bruce, R.C., R.G. Jaeger and L. Houck (eds.), The Biology of Plethodontid Salamanders. Kluwer Academic/Plenum Publishers, New York. Hijmans, R.J., S.E. Cameron, J.L. Parra, P.G. Jones, and A. Jarvis. 2005. Very high resolution interpolated climate surfaces for global land areas. International Journal of Climatology 25:1965?1978. Hines, J.E., 2006. PRESENCE ? software to estimate patch occupancy and related parameters. USGS-PWRC . Hollander, M., and A.W. Douglas. 1973. Nonparametric Statistical Methods. New York: John Wiley & Sons. Pg. 185?194. Hossack, B.R., M.J. Adams, E.H.C. Grant, C.A. Pearl, J.B. Bettaso, W.J. Barichivich, W.H. Lowe, K. True, J.L. Ware, and P.S. Corn. 2010. Low prevalence of chytrid fungus (Batrachochytrium dendrobatidis) in amphibians of US headwater streams. Journal of Herpetology 44:253?260. Huheey, J.E. and A. Stupka. 1967. Amphibians and Reptiles of the Great Smoky Mountains National Park. University of Tennessee Press, Knoxville, Tennessee. Hyatt, A.D., D.G. Boyle, V. Olsen, D.B. Boyle, L. Berger, D. Obendor, A. Dalton, K. Kriger, M. Hero, H. Hines, R. Phillott, R. Campbell, G. Marantelli, F. Gleason, A. Colling. 2007. Diagnostic assays and sampling protocols for the detection of Batrachochytrium dendrobatidis. Diseases of Aquatic Organisms. 73:175?192 IPCC (Intergovernmental Panel on Climate Change). 2007. Climate change 2007: impacts, adaptation and vulnerability, contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom. 96 IUCN 2010. IUCN Red List of Threatened Species. Version 2010.4. . Downloaded on 08 March 2011. Jaeger, R.G. 1971. Moisture as a factor influencing the distributions of two species of terrestrial salamanders. Oecologia 6:191?207. Jenkins, M.A. 2007. Vegetation communities of Great Smoky Mountains National Park. Southeastern Naturalist 6:35?56. Johnson, K., Taylor, G. and Remaley, T. 2005. Managing hemlock woolly adelgid and balsam woolly adelgid at Great Smoky Mountains National Park. In: Onken, B. and Reardon, R. (Compilers). 2005. Third Symposium on Hemlock Woolly Adelgid in the Eastern United States. Asheville, North Carolina. Johnson. M.L., L. Berger, L. Philips, R. Speare. 2003. Fungicidal effects of chemical disinfectants, UV light, desiccation and heat on the amphibian chytrid Batrachochytrium dendrobatidis. Diseases of Aquatic Organisms 57:255?260. Jones, C.G. J.H. Lawton, and M. Shachak. 1994. Organisms as ecosystem engineers. Oikos 69:373?386. King, W. 1939. A survey of the herpetology of Great Smoky Mountains National Park (Tennessee). American Midland Naturalist 21:531?582. Karoly, D.J., K. Braganza, P.A. Stott, J.M. Arblaster, G.A. Meehl, A.J. Broccoli and K.W. Dixon. 2003. Detection of a human influence on North American climate. Science 302:1200?1203. Kozak, K.H., D.W. Weisrock, and A. Larson. 2006. Rapid lineage accumulation in a non?adaptive radiation: phylogenetic analysis of diversification rates in eastern North American woodland salamanders (Plethodontidae: Plethodon). Proceedings of the Royal Society B?Biological Sciences 273:539?546. Kozak, K.H., J.J. Wiens. 2010. Niche conservatism drives elevational diversity patterns in Appalachian salamanders. American Naturalist 176:40?54. Kriger, K.M., and J. -M. Hero. 2006. Large-scale seasonal variation in the prevalence and severity of chytridiomycosis. Journal of Zoology doi:10.1111/j.1469-7998.2006.00220.x 97 Kriger, K. M., and J. -M. Hero. 2007. The chytrid fungus Batrachochytrium dendrobatidis is non?randomly distributed across amphibian breeding habitats. Diversity and Distribution 13:1?8. Lawler, J.J., S.L. Shafer, B.A. Bancroft, and A.R. Blaustein. 2010. Projected climate impacts for the amphibians of the western hemisphere. Conservation Biology 24:38?50. Lips, K. R. 1998. Decline of a tropical montane amphibian fauna. Conservation Biology 12:106?117. Lips, K.R., J.D. Reeve, L.R. Witters. 2003. Ecological traits predicting amphibian population declines in Central America. Conservation Biology 17:1078? 1088. Lips, K.R., J.R. Mendelson III, A. Mu?oz-Alonso, L. Canseco-M?rquez, and D.G. Mulcahy. 2004. Amphibian population declines in montane southern Mexico: resurveys of historical localities. Biological Conservation 119:555-564. Lips, K.R., F. Brem, R. Brenes, J.D. Reeve, R.A. Alford, J. Voyles, C. Carey, L. Livo, A.P. Pessier, and J.P. Collins. 2006. Emerging infectious disease and the loss of biodiversity on a Neotropical amphibian community. Proceedings of the National Academy of Sciences of the United States of America 103:3165?3170. Link, W.A., and J.D. Nichols. 1994. On the importance of sampling variance to investigations of temporal variation in animal population size. Oikos 69:539?544. Longcore, J.E., A.P. Pessier, and D.K. Nichols. 1999. Batrachochytrium dendrobatidis gen. et sp. Nov., a chytrid pathogenic to amphibians. Mycologia 91:219?227. Longo, A.V., P.A. Burrowes, R.L. Joglar. 2009. Seasonality of Batrachochytrium dendrobatidis infection in direct?developing frogs suggests a mechanism for persistence. Diseases of Aquatic Organisms. (Online DOI: 10.3354/dao02054) MACTEC. 2010. Cloud Deposition Monitoring ? Clingmans Dome, TN ? Great Smoky Mountains National Park ? 2009. MACTEC Engineering and Consulting, Inc. Gainesville, Florida. April, 2010 Maerz, J.C., V.A. Nuzzo, and B. Blossey. 2009. Declines in woodland salamander abundance associated with non?native earthworm and plant invasions. Conservation Biology 23:975?981. 98 Marquardt, D.W. 1980. A critique of some ridge?regression methods ? comment. Journal of the American Statistical Association 75:87?91. 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. Mart?nez?Solano, I., J. Bosch, and M. Garcia?Paris. 2003. Demographic trends and community stability in a montane amphibian assemblage. Conservation Biology 17:238?244. Marvin, G.A. 1998. Interspecific aggression and spatial relationships in the salamanders Plethodon kentucki and Plethodon glutinosus: Evidence of interspecific interference competition. Canadian Journal of Zoology 76:94?103. McDonald, K.R., D. Mendez, R. Muller, A.B. Freeman, and R. Speare. 2005. Decline in the prevalence of chytridiomycosis in upland frog populations in North Queensland, Australia. Pacific Conservation Biology 11:114? 120. Merchant, H. 1972. Estimated population size and home range of the salamanders Plethodon jordani and Plethodon glutinosus. Journal of the Washington Academy of Sciences 62:248?257. Milanovich, J.R., W.E. Peterman, N.P. Nibbelink, and J.C. Maerz. 2010. Projected loss of a salamander diversity hotspot as a consequence of projected global climate change. Plos One 5:1?10. Mitchell, K.M., T.S. Churcher, T.W J. Garner, and M.C. Fisher. 2008. Persistence of the emerging pathogen Batrachochytrium dendrobatidis outside the amphibian host greatly increases the probability of host extinction. Proceedings of the Royal Society B?Biological Sciences 275:329?334. Muths, E., P.S. Corn, A.P. Pessier and D.E. Green. 2003. Evidence for disease related amphibian decline in Colorado. Biological Conservation 110: 357?365. Muggeo, V.M.R. 2003. Estimating regression models with unknown break-points. Statistics in Medicine 22: 3055?3071. Muggeo, V.M.R. 2008. segmented: an R Package to Fit Regression Models with Broken-Line Relationships. R News, 8/1, 20-25. URL http://cran.r- project.org/doc/Rnews/. 99 Murray, K.A., R.W.R. Retallick, R. Puschendorf, L.F. Skerratt, D. Rosauer, H.I. McCallum, L. Berger, R. Speare, and J. VanDerWal. 2011. Assessing spatial patterns of disease risk to biodiversity: implications for the management of the amphibian pathogen, Batrachochytrium dendrobatidis. Journal of Applied Ecology 48:163?173. NADP (National Atmospheric Deposition Program). 1981. Data report? precipitation chemistry. Vol. 3, No.2. 28 pp., Natural Resource Ecology Laboratory, Colorado State University, Fort Collins. NADP (National Atmospheric Deposition Program). 2008. National Atmospheric Deposition Program 2007 Annual Summary. NADP Data Report 2008?01. Illinois State Water Survey, University of Illinois at Urbana?Champaign. NCDC (National Climatic Data Center). 2010. Nishikawa, K.C. 1990. Intraspecific spatial relationships of two species of terrestrial salamanders. Copeia 1990:418?426. NPS (National Park Service), Great Smoky Mountains National Park. 2007. Vegetation Disturbance History at Great Smoky Mountains National Park, Tennessee and North Carolina. National Park Service, Great Smoky Mountains National Park, Resource Management and Science. Geospatial Dataset-1045868. NPS (National Park Service). 2010. Great Smoky Mountains National Park. . Accessed 18 October 2010. Nuzzo, V.A., J.C. Maerz, and B. Blossey. 2009. Earthworm invasion as the driving force behind plant invasion and community change in northeastern North American forests. Conservation Biology 23:966?974. Ouellet, M., I. Mikaelian, B. D. Pauli, J. Rodrigue, and D. M. Green. 2005. Historical evidence of widespread chytrid infection in North American amphibian populations. Conservation Biology 19:1431?1440. Padgett?Flohr, G.E. and M.E. Goble. 2007. Evaluation of tadpole mouthpart depigmentation as a diagnostic test for infection by Batrachochytrium dendrobatidis for four California anurans. Journal of Wildlife Diseases 43:690?699. Pasmans, R., P. Zwart, and A.D. Hyatt. 2004. Chytridiomycosis in the Central American bolitoglossine salamander (Bolitoglossa dolfeini). Veterinary Record 154:153. 100 Petranka, J.W. 1998. Salamanders of the United States and Canada. Smithsonian Institute Press, Washington, D.C. Petranka, J.W., M.E. Eldridge, and K.E. Haley. 1993. Effects of timber harvesting on southern Appalachian salamanders. Conservation Biology 7:363?370. Petranka, J. W., M. E. Eldridge, and K. E. Haley. 1993. Effects of timber harvesting on southern Appalachian salamanders. Conservation Biology 7:363?370. Petranka, J. W., M. P. Brannon, M. E. Hopey, and C. K. Smith. 1994. Effects of timber harvesting on low elevation populations of southern Appalachian salamanders. Forest Ecology and Management 67:135?147. Phillips, D.L., and D.J. Shure. 1990. Patch?size effects on early succession in southern Appalachian forests. Ecology 71:204?212. Phillips, S.J., R.P. Anderson, and R.E. Schapire. 2006. Maximum entropy modeling of species geographic distributions. Ecological Modelling 190:231?259. Piotrowski, J.S., S.L. Annis, J.E. Longcore. 2004. Physiology of Batrachochytrium dendrobatidis, a chytrid pathogen of amphibians. Mycologia 96:9?15. Pounds, J.A. 2001. Climate and amphibian declines. Nature 410:639?640. Pounds, JA., M.P.L Fogden, J.M. Savage, and G.C. Gorman. 1997. Tests of null models for amphibian declines on a tropical mountain. Conservation Biology 11:1307?1322. Puschendorf, R., A.C. Carnaval, J. VanDerWal, H. Zumbado?Ulate, G. Chaves, F. Bola?os, and R. A. Alford. 2009. Distribution models for the amphibian chytrid Batrachochytrium dendrobatidis in Costa Rica: proposing climatic refuges as a conservation tool. Diversity and Distributions 15:401?408. Qian, S. S., T. F. Cuffney, I. Alameddine, G. McMahon, and K. H. Reckhow. 2010. On the application of multilevel modeling in environmental and ecological studies. Ecology 91:355?361. Quinn, V.S., B.M. Graves. 1999. A technique for sexing red?backed salamanders (Plethodon cinereus). Herpetological Review 30:32. Rachowicz, L.J., R.A. Knapp, J.A.T. Morgan, M.J. Stice, V.T. Vredenburg, J.M. Parker, and C.J. Briggs. 2006. Emerging infectious disease as a proximate cause for amphibian mass mortality. Ecology 87: 1671?1683. 101 Ramsey, J.P., L.K. Reinert, L.K. Harper, D.C. Woodhams, and L.A. Rollins? Smith. 2010. Immune defenses against Batrachochytrium dendrobatidis, a fungus linked to global amphibian declines, in the South African Clawed Frog, Xenopus laevis. Infection and Immunity 78:3981?3992. R Development Core Team .2009. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org. Reaser, J.K. and A. Blaustein. 2005. Repercussions of global change. In M. Lannoo (ed), Amphibian Declines: The Conservation Status of United States Species. Pp. 60?63 University of California Press, Berkeley. Richards?Zawacki, C.L. 2010. Thermoregulatory behaviour affects prevalence of chytrid fungal infection in a wild population of Panamanian golden frogs. Proceedings of the Royal Society B?Biological Sciences 277:519?528. Richardson, D.R., B.A. Snyder, and P.F. Hendrix. 2009. Soil moisture and temperature: tolerances and optima for a non?native earthworm species, Amynthas agrestis (Oligochaeta: Opisthopora: Megascolescidae). Southeastern Naturalist 8:325?334. Robinson, R.B., T.W. Barnett, G.R. Harwell, S.E. Moore, M. Kulp, and J.S. Schwartz. 2008. pH and acid anion time trends in different elevations in the Great Smoky Mountains National Park. Journal of Environmental Engineering 134:800?808. Rodenhouse, N.L., L.M. Christenson, D. Parry, and L.E. Green. 2009. Climate change effects on native fauna of northeastern forests. Canadian Journal of Forest Research 39:249?263. Rohr, J.R., and D.M. Madison. 2003. Dryness increases predation risk in efts; support for an amphibian decline hypothesis. Oecologia 135:657?664. Rohr, J.R., and T.R. Raffel. 2010. Linking global climate and temperature variability to widespread amphibian declines putatively caused by disease. Proceedings of the National Academy of Sciences of the United States of America 107:8269?8274. Rohr, J.R., T.R. Raffel, J.M. Romansic, H.McCallum, and P.J. Hudson. 2008. Evaluating the links between climate, disease spread, and amphibian declines. Proceedings of the National Academy of Sciences of the United States of America 105:17436-17441. 102 Ropelewski, C.F. and M.S. Halpert. 1986. North American precipitation and temperature patterns associated with the El Nino/Southern Oscillation (ENSO). Monthly Weather Review 114:2352?2362. Rothermel, B.B., S.C. Walls, J.C. Mitchell, C.K. Dodd, L.K. Irwin, D.E. Green, V.M. Vazquez, J.W. Petranka, and D.J. Stevenson. 2008. Widespread occurrence of the amphibian chytrid fungus Batrachochytrium dendrobatidis in the southeastern USA. Diseases of Aquatic Organisms 82:3?18. Roulin, A., and L.F. Bersier. 2007. Nestling barn owls beg more intensely in the presence of their mother than in the presence of their father. Animal Behaviour 74:1099?1106. Rovito, S.M., G. Parra?Olea, C.R. V?squez?Almaz?n, T.J. Papenfuss, D.B. Wake. 2009. Dramatic declines in neotropical salamander populations are an important part of the global amphibian crisis. Proceedings of the National Academy of Sciences of the United States of America 106:3231? 3236. Rowe, C.L. and J. Freda. 2000. Effects of acidification on amphibians at multiple levels of biological organization. In D.W. Sparling, G. Linder and C.A. Bishop (eds.), Ecotoxicology of Amphibians and Reptiles. Pp. 545?571. Society for Environmental Toxicology and Contaminants (SETAC) Press, Pensacola, Florida. Rowley J.J.L., and R.A. Alford. 2007. Behaviour of Australian rainforest stream frogs may affect the transmission of chytridiomycosis. Diseases of Aquatic Organisms 77:1?9. Rusek, J., and V.G. Marshall. 2000. Impacts of airborne pollutants on soil fauna. Annual Review of Ecology and Systematics 31:395?423. Shaffer, H.B., R.N. Fisher, and C. Davidson. 1997. The role of natural history collections in documenting species declines. Trends in Ecology and Evolution 13:27?30. Schmidt, B.R. 2009. Count data, detection probabilities, and the demography, dynamics, distribution, and decline of amphibians. Comptes Rendus Biologies 326:119?124. Shenk, T.M., G.C. White, K.P. Burnham. 1998. Sampling-variance effects on detecting density dependence from temporal trends in natural populations, Ecological Monographs. 68:445?463. 103 Smith, K.G., K.R. Lips, and J.M. Chase. 2009. Selecting for extinction: nonrandom disease-associated extinction homogenizes amphibian biotas. Ecology Letters 12:1069?1078. Smith, C.K., and J.W. Petranka. 2000. Monitoring terrestrial salamander populations: Repeatability and validity of area?constrained cover object searches. Journal of Herpetology 34:547?557. Speare R. and L. Berger. 2000. Global distribution of chytridiomycosis in amphibians. . Accessed 11 October 2010. Stott, P.A., 2003. Attribution of regional?scale temperature changes to anthropogenic and natural causes. Geophysical Research Letters 30: doi:10.1029/2003GL017324. Stuart, S.N., J.S. Chanson, N.A. Cox, B.E. Young, A.S.L Rodrigues, D.L. Fischman, and R.W. Walker. 2004. Status and trends of amphibian declines and extinctions worldwide. Science 306: 1783?1786. Sugiura, N. 1978. Further analysis of the data by Akaike's information criterion and the finite corrections. Communications in Statistics ? Theory and Methods 7:13?26. Thurow, G.R. 1963. Taxonomic and ecological notes on the salamander, Plethodon welleri. University of Kansas Science Bulletin 44:87?108. Thurow, G.R. 1976. Aggression and competition in eastern Plethodon (Amphibia, Urodela, Plethodontidae). Journal of Herpetology 10:277?291. Tilley, S.G., and J.E. Huheey. 2001. Reptiles and Amphibians of the Smokies. Great Smoky Mountains Association. Gatlinburg, Tennessee. Todd-Thompson, M., D.L. Miller, P.E. Super, and M.J. Gray. Chytridiomycosis- associated mortality in Rana palustris collected in Great Smoky Mountains National Park, Tennessee, USA. Herpetological Review 40:321?323. Vazquez, V.M., B.B. Rothermel, and A.P. Pessier. 2009. Experimental infection of North American plethodontid salamanders with the fungus Batrachochytrium dendrobatidis. Diseases of Aquatic Organisms 84:1?7. Ver Hoef, J.M., and P.L. Boveng. 2007. Quasi?poisson vs. negative binomial regression: how should we model overdispersed count data? Ecology 88:2766?2772. 104 Voyles, J., L. Berger, S. Young, R. Speare, R. Webb, J. Warner, D. Rudd, R. Campell, and L.F. Skerratt. 2007. Electrolyte depletion and osmotic imbalance in amphibians with chytridiomycosis. Diseases of Aquatic Organisms 77:113?118. Voyles, J., S. Young, L. Berger, C. Campbell, W.F. Voyles, A. Dinudom, D. Cook, R. Webb, R.A. Alford, L.F. Skerratt, and R. Speare. 2009. Pathogenesis of chytridiomycosis, a cause of catastrophic amphibian declines. Science 326:582?585. Vredenburg, V.T., R.A. Knapp, T.S. Tunstall, and C.J. Briggs. 2010. Dynamics of an emerging disease drive large-scale amphibian population extinctions. Proceedings of the National Academy of Sciences of the United States of America 107:9689?9694. Wake, D.B., and V.T. Vredenburg. 2008. Are we in the midst of the sixth mass extinction? A view from the world of amphibians. Proceedings of the National Academy of Sciences of the United States of America 105:11466?11473. Weinstein, S.B. 2009. An aquatic disease on a terrestrial salamander: individual and population level effects of the amphibian chytrid fungus, Batrachochytrium dendrobatidis, on Batrachoseps attenuatus (Plethodontidae). Copeia 4:653?660. Wells, K.D. 2007. The ecology and behavior of amphibians. The University of Chicago Press, Chicago, IL Welsh, H.H. Jr, and S. Droege. 2000. A case for using plethodontid salamanders for monitoring biodiversity and ecosystem integrity of North American forests. Conservation Biology 15:558?569. Wetherald, R.T., V. Ramaswamy, and S. Manabe. 1991. A comparative study of the observations of high clouds and simulations by an atmospheric general circulation model. Climate Dynamics 5:135?143. Whitford, W.G. and V.H. Hutchinson. 1963. Gas exchange in salamanders. Physiological Zoology 38:228?42. Wobeser, G.A. 2006. Essentials of Disease in Wild Animals. Blackwell Publishing, Ames, Iowa. Woodhams, D.C., R.A. Alford, G. Marantelli. 2003. Emerging disease of amphibians cured by elevated body temperature. Diseases of Aquatic Organisms 55:65?67. 105 Wyman, R.L. 1998. Experimental assessment of salamanders as predators of detrital food webs: effects on invertebrates, decomposition and the carbon cycle. Biodiversity Conservation 7:641?650. Wyman, R.L. 2003. Conservation of terrestrial salamanders with direct development. In. R.D. Semlitsch (ed). Amphibian Conservation, pp. 37? 52. Washington, D.C: Smithsonian Institution. Wyman, R.L. and D.S. Hawksley?Lescault. 1987. Soil acidity affects distribution, behavior, and physiology of the salamander Plethodon cinereus. Ecology 68:1819?1827. Wyman, R.L., and J, Jancola. 1992. Degree and scale of terrestrial acidification and amphibian community structure. Journal of Herpetology 26:392?401. Young, B.E., K.R. Lips, J.K. Reaser, R. Ibanez, A.W. Salas, J.R. Cedeno, L.A. Coloma, S. Ron, E. La Marca, J.R. Meyer, A. Munoz, F. Bolanos, G. Chaves, and D. Romo. 2001. Population declines and priorities for amphibian conservation in Latin America. Conservation Biology 15:1213? 1223. Zwiers, F. and X. Zhang. 2003. Toward regional?scale climate change detection. Journal of Climate 16: 793?797.