Abstract
Understanding the forces that drive genotypic and phenotypic change in wild populations is a central goal of evolutionary biology. We examined exome variation in populations of deer mice from two of the California Channel Islands:Peromyscus maniculatus elususfrom Santa Barbara Island andP. m. santacruzaefrom Santa Cruz Island exhibit significant differences in olfactory predator recognition, activity timing, aggressive behavior, morphology, prevalence of Sin Nombre virus, and population densities. We characterized variation in protein‐coding regions using exome capture and sequencing of 25 mice from Santa Barbara Island and 22 mice from Santa Cruz Island. We identified and examined 386,256 SNPs using three complementary methods (BayeScan, pcadapt, and LFMM). We found strong differences in molecular variation between the two populations and 710 outlier SNPs in protein‐coding genes that were detected by all three methods. We identified 35 candidate genes from this outlier set that were related to differences in phenotypes between island populations. Enrichment analyses demonstrated that patterns of molecular variation were associated with biological processes related to response to chemical stimuli and regulation of immune processes. Candidate genes associated with olfaction (Gfy,Tlr2,Vmn13r2, numerous olfactory receptor genes), circadian activity (Cry1), anxiety (Brca1), immunity (Cd28,Eif2ak4, Il12a,Syne1), aggression (Cyp19a,Lama2), and body size (Bc16,Syne1) exhibited non‐synonymous mutations predicted to have moderate to large effects. Variation in olfaction‐related genes, including a stop codon in the Santa Barbara Island population, suggests loss of predator‐recognition traits at the molecular level, consistent with a lack of behavioral aversion to fox feces. These findings also suggest that divergent pathogen prevalence and population density may have influenced adaptive immunity and behavioral phenotypes, such as reduced aggression. Overall, our study indicates that ecological differences between islands are associated with signatures of selection in protein‐coding genes underlying phenotypes that promote success in those environments.
