Contrasting demographic and genetic estimates of dispersal in the endangered Coahuilan box turtle

Authors: Jennifer Howeth, Suzanne McGaugh, Dean A. Hendrickson

Year: 2008 (xl)

Abstract

The evolutionary trajectory of an endangered species largely depends upon the level of gene flow among subpopulations and consequently the degree of habitat patch connectivity. A conservation approach that jointly acknowledges connectivity occurring over ecological and evolutionary time scales may provide novel insight into the maintenance of genetic diversity. Here, we take this joint approach in evaluating the endangered Coahuilan box turtle (Terrapene coahuila) endemic to the isolated wetlands in the desert of Cuatrociénegas, Coahuila, México. The desert-spring ecosystem of Cuatro Ciénegas provides a patch-matrix mosaic that restricts dispersal among wetland habitats. Moreover, aquatic habitat loss and fragmentation has altered the spatial distribution of habitat patches and further inhibited interpatch connectivity. We therefore predicted that T. coahuila would exhibit metapopulation structure and limited dispersal over small spatial scales. To test this hypothesis, we 1) evaluated the spatial extent and frequency of dispersal via mark-recapture at local (< 2km) and regional (2 - 20 km) spatial scales in two subpopulations and 2) assessed gene flow among 7 subpopulations by surveying genetic diversity at 9 microsatellite loci. The mark-recapture study reveals frequent inter-wetland dispersal at the local scale and at a rate suggestive of a single subpopulation. At the regional scale, dispersal was relatively less frequent; however, one subpopulation was seasonally migratory. The population genetic analysis reveals moderate to high levels of gene flow with no indication of inbreeding. While there are low levels of spatial genetic structuring (global FST = 0.01), there is isolation by distance. This relationship is driven by one subpopulation (Laguna Grande) relatively isolated from the others by a physical barrier to dispersal. The genetic signatures provide a fingerprint of historic population structure and, in the presence of recent habitat loss (< 50 years), may not accurately reflect current conditions or predict future patterns of gene flow or the trajectory of T. coahuila. Our assessment, by jointly acknowledging ecological processes that are immediately susceptible to anthropogenic disturbance, and microevolutionary processes that provide the buffering mechanism of adaptation to such disturbance, yields a robust framework on which to build conservation recommendations for this species and other regional taxa that likely share the same evolutionary history.

Resumen

Dean will provide this soon