In the face of accelerated warming and drying, habitat specialists of riparian zones could be threatened with genetic erosion if their effective population sizes become small due to loss and fragmentation of habitat. The New Mexico meadow jumping mouse (Zapus hudsonius luteus) is Endangered due to a suite of disturbances to their riparian habitat but the degree of functional connectivity within and among the drainages they occupy is unknown. Using microsatellite and mtDNA, we examined the hypothesis of hydrologically-mediated gene flow (via riparian networks) by assessing structure, diversity, and the possibility of overland dispersal among disconnected watersheds (White Mountains, AZ). Both a priori and de novo structure supported stream-mediated gene flow. However, inferred clusters crossed watershed boundaries, and migration rates and dispersers suggest infrequent use of overland paths at nearby sub-drainages among each watershed. Analysis of mtDNA indicated that this was likely a long-term phenomenon, also suggesting that one of the watersheds (LCR) may have been more historically diverse than its contemporary snapshot (nuDNA). Our analysis led us to suspect fragmentation in its eastern fork, an area most impacted by human development. We conclude that the boundaries of drainages affect the White Mountains population structure but short, cryptic, overland paths among them likely play a role in replenishing genetic diversity. We also report an extension to the species longevity from 2 to 4 years from genetic recaptures during our six-year study period.

We generated 101 multi-locus genotypes (diploid) from 8 microsatellite DNA loci. We collected microsatellite data using multiplex PCR reactions and fragment length analysis using an ABI 3130 capillary sequencer. We called alleles using GeneMarker software. We also generated DNA sequence data (mitochondrial DNA, cytochrome B; 929 bp) for 63 of the specimens, whereas 15 of the sequences were mined from NCBI GenBank. The DNA sequences were assembled from four amplicons using custom primer sets to amplify from degraded DNA sources (e.g., feces, historical). We used bi-directional Sanger sequencing on an ABI 3130 capillary sequencer to generate the sequence data. We assembled the amplicons for each specimen to a reference sequence from NCBI GenBank using Sequencher 5.4.