Genetic connectivity of pupfish populations in a highly modified desert springs system

Background/Question/Methods

The magnitude of current climate change, coupled with increasing water demands from expanding human populations, poses a significant and unprecedented threat to freshwater fauna across the Southwest. Such threats could manifest in declining population sizes and diminished genetic diversity, which in turn could lead to heightened risk of extinction through demographic, stochastic, and genetic effects. Ash Meadows National Wildlife Refuge encompasses two dozen springs and their outflows near the California-Nevada border adjacent to Death Valley.  The small area (89 km²) supports at least 40 endemic species, including the Ash Meadows pupfish, Cyprinodon nevadensis mionectes. Most springs supporting C. n. mionectes are connected through anatomizing outflow channels, although some are isolated because of restrictions on water flow. We tested whether these restrictions present barriers to gene flow in C. n. mionectes as part of a larger study of the genetic diversity of pupfish in the refuge.

We exhaustively sampled pupfish from the all the springheads and most of the outflows of 12 springs during 2007-2008. We used sequenced variable 10 microsatellite loci and the mitochondrial gene ND2. We used standard population genetic and phylogenetic analyses combined with geographic information systems to describe diversity and assess connectivity of C. n. mionectes.

 Results/Conclusions

In contrast to other pupfish taxa found in the refuge, C. n. mionectes persists in large population sizes with an abundance of genetic variation within populations. Microsatellite heterozygosity ranged from 0.7 to over 0.8; mitochondrial DNA haplotype diversity ranged from 0.4 to 0.8. Interestingly, these measures of diversity were uncorrelated (Spearman’s correlation S = 28, p=0.27). Analysis of the genetic isolation of springs by comparing all populations to the most genetically diverse sample of fishes revealed two interesting patterns. First, two springs were found to be significantly isolated above expectations based on geographic distance. One of these springs is isolated from all other populations by a road. The other spring shows differentiation between the springhead and outflow, likely due to a flume at the head of the outflow. Two other springs show some indication of isolation effects by flumes. Second, the southern springs are more demographically isolated than expected, probably as a result of reduced connectivity across the entire system because of modifications to the natural hydrology. As climate change and hydrologic demands continue to impact the system, these findings have important implications for management options such as assisted migration to maintain connectivity and genetic diversity.