DISSERTATION RESEARCH: Assessing adaptative capacity in response to climate change: terrestrial salamanders in an urban heat island mesocosm

Duke University, Durham NC

Investigators

Abstract

This research will provide insights into the capacity of species to adapt to fragmented and warming habitats, and so inform actions that will be most effective in conserving biodiversity under climate change. Faced with changing climate, organisms either adapt in place or move to track conditions that they are adapted to. In the absence of these responses, they may go locally extinct. Relative to movement, adaptation in response to environmental change remains poorly understood in wild populations. Fortunately the recent development of genetic techniques for detecting adaptation in nature has made it easier to evaluate this response. Collaborations established under this project include educational outreach through the Duke Forest Office, North Carolina Central University, and the Scientific Research and Education Network. This project will study the capacity of a terrestrial salamander (Plethodon cinereus) to respond to local variability in temperature via adaptation. In the Piedmont of North Carolina, local heat islands resulting from development provide habitats that are substantially warmer than natural forests. This history of development has resulted in local warming over timescales ranging from <10 to >100 years, providing a quasi-experimental gradient in exposure time to elevated temperatures. Genetic variation within and among salamander populations will be assessed at a genome-wide scale using reduced representation, next-generation genomic sequencing methods. In combination with environmental covariates, novel genotype-environment outlier detection analyses will be used to partition genomic data into neutral and potentially adaptive variation; these data sets allow for the measurement of gene flow within and between populations and local adaptation, respectively. Integration of these data uses a Bayesian approach for modeling the covariance in neutral or adaptive allele frequencies as functions of geographic and environmental distances. Gene flow is assessed via covariance of neutral genetic markers on graph-theoretic estimates of geographic separation; local adaptation is measured in terms of covariance of genetic markers under selection on environmental dissimilarity. Relative effect sizes will identify instances of dispersal subsidy or isolation of better-adapted genotypes, as represented by locations where landscape configuration or the pace of development interferes with the ability of the resident salamanders to track local climate.