Estimating divergence time for two evolutionarily significant units of a protected fish species

Abstract

The evolutionarily significant unit concept provides a powerful tool for conserving biodiversity below the species level, but temporal criteria are often used explicitly or implicitly in the operational definitions of evolutionarily significant units (ESUs). Such temporal considerations have important implications for recently diverged taxa, as is the case with the White sands pupfish (Cyprinodon tularosa). This species consists of two native populations previously designated as the Malpais Spring and Salt Creek ESUs based on allele frequency differences at nuclear markers and their ecologically divergent habitats; despite a lack of reciprocal monophyly. Isolation of these two ESUs presumably occurred during the mid-Holocene, but an alternative hypothesis is that the populations were isolated due to changes in surface hydrology associated with overgrazing in the late 19th and early 20th centuries. We assayed 13 microsatellite loci and applied an Approximate Bayesian Computation analysis to estimate time of divergence between the two populations. Our reference table consisted of 1,000,000 simulated data sets, and we used three different models, each having different combinations of summary statistics. Estimates of median divergence time varied from approximately 6,500–11,000 generations (3,250–11,000 years). These findings support the hypothesis that Malpais Spring and Salt Creek having been isolated for a least a few millennia, and together with previously documented adaptive divergence, argues for continued management as separate conservation units. We consider the temporal constraints for defining evolutionary significance as it relates to recently diverged populations occupying ecologically divergent habitats.