Landscape genetics of the Tasmanian devil: implications for spread of an infectious cancer
- 1.1k Downloads
Emerging infectious diseases are increasingly recognized in species’ declines and extinctions. Landscape genetics can be used as a tool to predict disease emergence and spread. The Tasmanian devil is threatened with extinction by a nearly 100% fatal transmissible cancer, which has spread across 95% of the species’ geographic range in 20 years. Here, we present a landscape genetic analysis in the last remaining uninfected parts of the Tasmanian devil’s geographic range to: describe population genetic structure, characterize genetic diversity, and test the influence of landscape variables on Tasmanian devil gene flow to assess the potential for disease spread. In contrast to previous genetic studies on Tasmanian devils showing evidence for two genetic populations island-wide, our genetic based assignment tests and spatial principal components analyses suggest at least two, and possibly three, populations in a study area that is approximately 15% of the size of the overall species’ geographic range. Positive spatial autocorrelation declined at about 40 km, in contrast to 80 km in eastern populations, highlighting the need for range-wide genetic studies. Strong genetic structure was found between devils in the northern part of the study area and those found south of Macquarie Harbor, with weaker structure found between the northeastern and northwestern portion of our study area. Consistent with previous work, we found low overall genetic diversity, likely owing to a combination of founder effects and extreme weather events thousands of years ago that likely caused large-scale population declines. We also found possible signs of recent bottlenecks, perhaps resulting from forest clearing for dairy farming in the central part of the study area. This human disturbance also may have contributed to weak genetic structuring detected between the northeastern and northwestern part of the study area. Individual-based least cost path modeling showed limited influence of landscape variables on gene flow, with weak effects of variation in elevation in the northeast. In the northwest, however, landscape genetic models did not perform better than the null isolation-by-distance model. At the larger spatial scale of the northern part of the study area, elevation and temperatures were negatively correlated with gene flow, consistent with low dispersal suitability of higher elevation habitats that have lower temperatures and dense, wet vegetation. Overall, Tasmanian devils are a highly vagile species for which dispersal and gene flow appear to be influenced little by landscape features, and spread of devil facial tumor disease to the remaining portion of the devil’s geographic range seems imminent. Nonetheless, strong genetic structure found between the northern and southern portions of our study area, combined with low densities and limited possible colonization of DFTD from the east suggest there is some time for implementation of management strategies.
KeywordsLandscape genetics Emerging infectious diseases Tasmanian devil Devil facial tumor disease
This work was supported by NSF DEB-1316549 to AS and MJ, and MJ was also supported by an ARC Future Fellowship FT100100250. Animal use was approved under IACUC protocol ASAF# 04392 from Washington State University and by the Tasmanian Department of Primary Industries, Parks, Water and Environment (DPIPWE) Animal Ethics Committee. Samples were collected by the Save the Tasmanian Devil Program (DPIPWE), particularly by Jason Wiersma, Jim Richley, Billie Lazenby, Stewart Huxtible, Clare Hawkins, Harko Werkman, and Dydee Mann. We thank Sarah Emel for help with analyses and four anonymous reviewers for comments that helped improve the manuscript.
- Bartón K (2015) MuMIn: R code package. https://cran.rproject.org/web/packages/MuMIn/MuMIn.pdf
- Bates D, Maechler M, Bolker B, Walker S (2014) R package lme4 https://cran.r-project.org/package=lme4
- Crida A, Manel S (2007) Wombsoft: an r package that implements the Wombling method to identify genetic boundary. Mol Ecol Res 7:588–591Google Scholar
- Piry S, Luikart G, Cornuet J-M (1999) BOTTLENECK: A program for detecting recent effective population size reductions from allele data frequencies. http://www1.montpellier.inra.fr/CBGP/software/Bottleneck/pub.html