Bryce Wade
University of Tennessee and Oak Ridge National Laboratory
November 2024
Background: Habitat fragmentation is recognized as one of the most pressing threats to wildlife populations. If habitat patches become exceedingly isolated, this can lead to negative consequences such as high rates of genetic diversity loss and local extirpations. Therefore, from a conservation perspective, maintenance of sustained dispersal and gene flow between populations is essential and often the end goal of conservation action. The use of resistance surface models, or spatial layers that assign value to landscape features based on the magnitude of their impact on successful dispersal, has emerged as an important strategy for developing spatial conservation management strategies. Resistance surface models are actively being used in efforts such as the protection and restoration of important dispersal corridors, spatially informed installation of wildlife road crossings (e.g., underpasses), and ecologically-informed land use planning. Assigning values to resistance surfaces can be done using a variety of methods such as expert opinion or telemetry, but the use of population genetic data has emerged as a preferred method.
A major question in the creation of resistance surfaces using genetic data is assessing the confidence in how a model will perform at new sites. It is important that scientists and land managers are confident that the inferences regarding which landscape features influence connectivity derived from landscape genetic analyses are accurate and comprehensive. The use of a replicated landscape study design can help uncover how inferences, and ultimately conservation genetic decisions, may differ if models are developed in different places using.
I developed resistance surfaces at five different landscapes in the southeast using Four-toed Salamanders (Hemidactylium scutatum) as my focal organism. The Four-toed Salamander is a forest-dwelling, lungless salamander native to eastern North America that relies on vernal pools, boggy wetlands, and small slow-moving streams with dense moss cover for successful breeding. Four-toed Salamanders have a unique reproductive life history in which adults mate in the uplands of wetlands then females lay eggs terrestrially near the margins of wetlands within moss (or other organic matter) and subsequently guard the eggs for multiple weeks. Despite having a large geographic range, this species has a patchy distribution across its range, is often assumed to have low local abundance, and has specialized habitat requirements. Therefore, the species is considered of conservation concern in 14 U.S. states.
Methods: I sampled across 5 replicate study areas in eastern Tennessee and Kentucky where Four-toed Salamanders have been recorded historically (I refer to these as landscapes throughout). These landscapes included the Oak Ridge Reservation (ORR), the Great Smoky Mountains National Park (GSMNP), Catoosa Wildlife Management Area (CWMA), Prentice Cooper State Forest (PCSF), and the Red River Gorge Geologic Area (RRG). I sampled primarily from February to April of 2024 to coincide with the period when Four-toed Salamanders are highly detectable directly prior to nesting and during the nesting period. I flipped available cover objects (logs, rocks, etc.) directly adjacent to wetlands to locate male, juvenile, and off-nest female Four-toed Salamanders. I also searched for nests by gently parting moss along wetland banks where moss overhung the water and removed females from nests when possible. I collected non-lethal tissue samples (approximately 5 mm tail tips) from adult salamanders then immediately returned them to their point of capture.
After collecting all my samples, I processed them in the lab using “next generation sequencing” techniques. These cutting-edge genetic methods allowed me to generate immense amounts of genetic data for each of my sampling landscapes. Once I had my data, I was able to explore population genetic structure for each landscape, estimate connectivity between different ponds, and ultimately evaluate how different environmental features such as topographic roughness, wetness, forest cover, and road density may influence four-toed movement through the landscape.
Early Findings: From my initial analyses, it appears many different landscape features may influence Four-toed Salamander movement and connectivity in my sampling landscapes. Areas of high topographic ruggedness (e.g., steep cliff faces), low wetness (e.g., dry ridgetops), low canopy cover (e.g., fields or powerline cuts), and high road density (e.g., urban areas) all appear to impede movement and gene flow for Four-toed Salamanders at my sampling landscapes. However, I found that these effects are not detected at every sampling landscape and, importantly, whether or not the effect of a landscape feature is detected is associated with how diverse that feature is at a landscape. In other words, it is difficult to detect the effect of a feature on gene flow if that feature does not vary across the landscape. A good example of this in my study is the effect on canopy cover. Ecologically, Four-toed Salamanders are dependent on mature forest for reproduction and are highly vulnerable to desiccation, thus forest cover is likely essential to successful dispersal between wetlands. The positive effect of forest cover was identified as important only at sites that had forest fragmentation, but not at sites that had relatively continuous forest cover.
My study indicates that analyses relying on a single landscape may ultimately fail to identify factors that are important to dispersal and lead to incomplete resistance surfaces and thus subpar conservation management. As the use of resistance surfaces increases in applied conservation management, generalizable conclusions drawn from replicated landscape genetic analyses will be essential for accurate assessments of connectivity and, ultimately, making well-informed management decisions.
