Landscape genetics is increasingly transitioning away from
microsatellites, with Single Nucleotide Polymorphisms (SNPs) providing
increased resolution for detecting patterns of spatial-genetic
structure. This is particularly pertinent for research in arid-zone
mammals due to challenges associated with unique life history traits,
boom-bust population dynamics and long-distance dispersal capacities.
Here, we provide a case study assessing the performance of SNPs versus
microsatellites in evaluating three explicit landscape genetic
hypotheses (isolation-by-distance, isolation-by-barrier, and
isolation-by-resistance) in a suite of small, arid-zone mammals in the
Pilbara region of Western Australia. Using clustering algorithms, Mantel
tests, and linear mixed effects models, we compare functional
connectivity between genetic marker types and across species, including
one marsupial, Ningaui timealeyi, and two native rodents,
Pseudomys chapmani and P. hermannsburgensis. SNPs resolved
subtle genetic structuring not detected by microsatellites, particularly
for N. timealeyi where two genetic clusters were identified.
Furthermore, stronger signatures of isolation-by-distance and
isolation-by-resistance were detected when using SNPs, and model
selection based on SNPs tended to identify more complex resistance
surfaces (i.e., composite surfaces of multiple environmental layers) in
the best-performing models. While we found limited evidence for physical
barriers to dispersal across the Pilbara for all species, we found that
topography, substrate, and soil moisture were the main environmental
drivers shaping functional connectivity. Our study demonstrates that new
analytical and genetic tools can provide novel ecological insights into
arid landscapes, with potential application to conservation management
through identifying dispersal corridors to mediate the impacts of
ongoing habitat fragmentation in the region.