INTRODUCTION:

Little is known about the phylogenetic relationships and population genetic structure of the Arizona shrew ( Sorex arizonae), a small mammal restricted to a few mountainous areas in the southwestern United States and northern Mexico. Information on genetic variability is needed, as the species is restricted to sky islands, which are small and often locally fragmented habitats.

METHODS:

Using mitochondrial cytochrome- b sequences and 12 microsatellite loci, we assessed genetic differentiation among 64 individuals from four mountain ranges in Arizona and New Mexico: Animas, Chiricahua, Huachuca, and Santa Rita ranges.

RESULTS:

We found a unique haplotype for the Animas population, while the other mountain ranges shared a common haplotype but they also each had a unique haplotype at lower frequencies. F-statistics indicate that significant population differentiation has occurred among the Chiricahua and Animas populations compared with the other two populations. The F ST statistics revealed high levels of genetic differentiation between the Chiricahua and Animas populations, compared to the other two populations that show lower but significant F ST values. Nevertheless, the STRUCTURE analysis using microsatellite markers did not show significant differentiation between Chiricahua and Animas but it did detect strong signatures of differentiation between the cluster formed by these two populations and Huachuca and Santa Rita. Phylogenetic analysis using cytochrome-b sequences confirmed that the Arizona shrew is most closely related to S. trowbridgii among North American shrews that have been sequenced for the mtDNA cytochrome-b region to date. Both species are placed in a monophyletic group sister to a clade consisting of Nearctic shrews.

DISCUSSION AND CONCLUSIONS:

Our results suggest that the S. arizonae populations have high levels of differentiation in comparison to other North American shrews. We also detected high levels of population genetic structure with both mtDNA and microsatellite although the pattern between them is different. Given the inherent vulnerability of these naturally isolated populations with their locally fragmented distributions, the genetic data support a need for conservation focused on protection of suitable habitats. As our fieldwork indicates, suitable habitats include mid- to high-elevation riparian zones and upland areas, including ridgelines and mountaintop saddles with having sufficient moisture, vegetation, and soil development to provide cover and a food base for shrews.