Elucidating the factors that drive the genetic patterns of natural populations is key in evolutionary biology, ecology, and conservation. Hence, it is crucial to understand the role that environmental features play in species' genetic diversity and structure. Landscape genetics measures functional connectivity and evaluates the effects of landscape composition, configuration, and heterogeneity on microevolutionary processes. Deserts constitute one of the world’s most widespread biomes and exhibit a striking heterogeneity of microhabitats, yet few landscape genetics studies have been performed with rodents in deserts. We evaluated the relationship between landscape and functional connectivity, at a microgeographic scale, of the Nelson’s pocket mouse Chaetodipus nelsoni in the Mapimí Biosphere Reserve (Chihuahuan desert). We used single-nucleotide polymorphisms and characterized the landscape based on on-site environmental data and from Landsat satellite images. We identified two distinct genetic clusters shaped by elevation, vegetation, and soil. The high elevation group showed higher connectivity in the elevated zones (1250-1350 m), with scarce vegetation and predominantly rocky soils; whereas that of the Low elevation group was at <1200 m, with denser vegetation and sandy soils. These genetic patterns are likely associated with the species’ locomotion type, feeding strategy, and building of burrows. Interestingly, we also identified morphological differences, where hind foot size was significantly smaller in individuals from High elevation compared to Low elevation, suggesting the possibility of ecomorphs associated with habitat differences and potential local adaptation processes, which should be explored further. These findings improve our understanding of the genetics and ecology of C. nelsoni and other desert rodents.

Sampling individuals in the wild. Tissue samples for extracting DNA. Library preparation and sequencing were performed at the University of Wisconsin Biotechnology Center (UWBC) following a Genotyping by sequencing approach. Libraries were prepared using the ApeKI enzyme (cutting site: C[AT]G) and sequenced on a single lane of an Illumina Hi-Seq 2500 with single-end 101 bp reads. A de novo assembly was performed using raw data processed with ipyrad (Eaton and Overcast 2020; https://ipyrad.readthedocs.io/en/master/). We further processed our data using VCFtools v.0.1.13 (see publication for the bioinformatics description).