Salamander Diversity and Distributions Data, Central Appalachian Region, West Virginia, USA 

The Central Appalachian region, USA, contains several high elevation-endemic woodland salamanders (genus Plethodon), which are thought to be particularly vulnerable to climate change due to their restricted distributions and low vagility. In West Virginia, there is a strong management focus on protection and recovery of the federally threatened Cheat Mountain salamander (Plethodon nettingi; CMS). To support this focus, there is a need for improved understanding of CMS occurrence-habitat relationships and spatially explicit projections of fine-scale contemporary and potential future habitat quality to inform management actions. In addition, there is concern among resource managers that climate change may increase habitat quality at high elevations for CMS competitors, particularly the eastern red-backed salamander (Plethodon cinereus; RBS), potentially resulting in increased competition pressure for CMS. To address these knowledge gaps, we created ecological niche models for CMS and RBS using the Random Forest classification algorithm and used the estimated occurrence-habitat relationships to assess ecological niche overlap between the species and project fine-scale contemporary and potential future habitat availability and quality. We estimated that the ecological niches of CMS and RBS were 80.5% similar, and habitat projections indicated the species would exhibit opposite responses to climate change in our region. For CMS, we estimated that amount of high-quality habitat will be reduced by mid-century and potentially lost by end-of-century, but that moderate and low-quality habitat will persist. For RBS, we estimated that amount of high-quality habitat will increase through end-of-century, and that high elevations will become more suitable for the species, indicating that competition pressure for CMS is likely to increase. This study improves understanding of important habitat characteristics for CMS and RBS, and our spatially explicit projections can assist natural resource managers with habitat protection actions, species monitoring efforts, and climate change adaptation strategies. 

We at TERN acknowledge the Traditional Owners and Custodians throughout Australia, New Zealand and all nations. We honour their profound connections to land, water, biodiversity and culture and pay our respects to their Elders past, present and emerging. 

The purpose of this study was to improve our understanding of the potential for climate change to influence habitat quality and interspecific competition pressure for the CMS. We used an ecological niche modeling (ENM) approach, which is frequently used for terrestrial species to assess species-habitat relationships and predict the relative quality of habitat within a focal area (Warren and Seifert 2011, Sillero et al. 2021). We estimated the ecological niches for CMS and RBS (within the regional distribution of CMS) based on historical occurrence data, projected their contemporary and potential future geographic distributions based on the modeled environmental relationships and climate projections, and quantified ecological niche overlap between the two species. To our knowledge, this is the first study to use an ENM approach for investigating potential impacts of climate change on interspecific competition pressure between woodland salamanders. 

We identified 9 candidate landscape predictors based on a literature review for CMS habitat associations, findings of previous ENMs for CMS and other high elevation salamanders, and expert opinion from the CMS working group. We obtained a 30 m2 resolution digital elevation model (DEM) through the U.S. Geological Survey (USGS) National Elevation Dataset (1 arc second; USGS 2017), and used the DEM to calculate elevation, heat load index (HLI), and topographical position index (TPI). We included HLI as a candidate predictor because aspect is a geophysical characteristic correlated with CMS occurrence (Green and Pauley 1987, Pauley 2008a, Kroschel et al. 2014). Because temperatures are not symmetrical around the north-south axis, slopes with afternoon sun have greater maximum temperatures than slopes with morning sun (McCune and Keon 2002). The HLI allows us to account for this variation, providing an accurate representation of the influence of aspect on habitat quality while avoiding the circular nature of aspect measurements (McCune and Keon 2002). We included TPI as a candidate predictor, which is calculated from DEMs by comparing the elevation of raster cells to the mean elevation of neighboring cells. Positive TPI values signify locations that have a higher topographic position (i.e., ridges and rises), while negative values signify areas that have a lower topographic position than surrounding cells (i.e., depressions, drainages, and valleys; Vinod 2017). Occurrence of CMS is associated with occurrence of red spruce (Picea rubens) and eastern hemlock (Tsuga canadensis; Pauley 2022). We used the LANDFIRE 2016 Remap existing vegetation type database (hereafter existing vegetation) to create four vegetation categories, including red spruce and eastern hemlock, deciduous forest, ‘other’ forest (e.g., pine forests), and non-forest (e.g., grassland, development). We also included canopy cover because woodland salamanders are typically associated with mature forests (Heatwole 1962, Petranka 1998, Means 2000, Blankers et al. 2012). We used the canopy cover layer created by Chazal et al. (2017), which was derived from the National Land Cover Database Tree Canopy Cover layer (Ruefenacht et al. 2015). We included four candidate predictors representing soil and geological characteristics. Euclidean distance to rocky habitat was derived from the West Virginia Terrestrial Habitat Map, which was created by the West Virginia Division of Natural Resources based on the Northeast Terrestrial Habitat Map (NTHM; Ferree and Anderson 2013) with some revisions specific to West Virginia, including addition of completed habitat layers, reclassification of similar ecosystems, and corrections of known errors. The NTHM was developed from NatureServe’s Ecological System Classification (Comer et al. 2003) and the National Vegetation Classification (FGDC 2008). Soil type was derived from the GSSURGO database (Soil Survey Staff 2014) and reclassified to represent the major soil orders associated with CMS occurrence (Alfisols, Inceptisols, Spodosols, and Ultisols; all other categories reclassified as ‘Other’) using the official soil description for each soil type located within the study area. Soil skeletal type was derived from the soil taxonomic class information. Rock type classification (surface geology) was created by the West Virginia Geological and Economic Survey and obtained from the West Virginia Division of Natural Resources. This layer was reclassified into 6 categories: alluvium, limestone, sandstone, shale, and siltstone (all other categories reclassified as ‘Other’). We created landscape variable layers using ArcGIS (version 10.7; Environmental Systems Research Institute, Redlands, California).