The importance of variation and scale in landscape genetics and conservation management: a case study using four-toed salamander (Hemidactylium scutatum) metapopulations
Date Issued
May 2026
Author(s)
Wade, Bryce Steven
The University of Tennessee
Advisor(s)
Evin Carter
Additional Advisor(s)
Todd Pierson
Benjamin Fitzpatrick
William Peterman
Mona Papes
Abstract
Small, isolated populations may rapidly lose genetic diversity and suffer from a myriad of consequences including reduced adaptive capacity and a higher probability of extirpation. Therefore, ecological connectivity is a primary concern in conservation genetic management. Resistance surfaces, or spatial models with values corresponding to the difficulty of organismal movement, have emerged as a commonly used tool to predict connectivity and develop spatial conservation plans. These models are often parameterized using genetic data and landscape genetic analyses. However, challenges regarding model parameterization remain, including understanding how the variability of
landscape features contributes to the generalizability of landscape genetic inferences. Here, I focus on a threatened wetland breeding amphibian, the four-toed salamander (Hemidactylium scutatum), to demonstrate the value of resistance surfaces in conservation management, evaluate the influence of landscape variation and spatial replication for developing generalizable landscape genetic inferences, and assess the interaction between ecological connectivity and habitat loss for shaping patterns of genetic diversity. First, I provide a case study of how resistance surfaces can be integrated into infrastructure updates in which conservation was not a prior consideration
to positively influence site-wide connectivity in a cost-effective manner. Next, I use replicated landscapes to assess the impact of landscape features on salamander connectivity and assess the importance of variation and scale for detecting landscape genetic patterns. I show that four-toed salamander connectivity is maintained by areas of flat, moist forest and that landscape replication was needed to develop robust, generalizable inferences. I follow this up using demo-genetic simulations assessing how the mean and variance of landscape features affect the detectability of landscape genetic relationships. I show that landscape genetic patterns are only detectable when the landscape feature of interest is variable. Lastly, I demonstrate how genetic diversity, the
primary concern in conservation genetic management, is a product of connectivity and habitat loss in anthropogenic landscapes highlighting the role of isolation in shaping the vulnerability of populations. Altogether, these results emphasize the value of connectivity in conservation management and help provide best practices for assessing the influence of landscape features on genetic connectivity.
Degree
Doctor of Philosophy
Major
Energy Science and Engineering
File(s)![Thumbnail Image]()
Name
WADE-PRIMARY-2026.pdf
Size
37.84 MB
Format
Adobe PDF
Checksum (MD5)
c2211871f57d8db515138f118d5da10a