Genetic and ecological differentiation in the endemic avifauna of Tiburón Island

Tiburón Island is a land-bridge island in the Gulf of California, separated from mainland Sonora by 3 km. The shallow channel (13 m) separating the island and mainland is thought to have formed 10 000 years ago. Although the majority of avian resident species are not taxonomically differentiated, six species are represented by endemic subspecies (cactus wren, gila woodpecker, black-tailed gnatcatcher, Gambel's quail, canyon towhee, northern cardinal), of which all but one (black-tailed gnatcatcher) possess a pallid, ash-gray coloration compared to those on the mainland. We compared mtDNA sequences of five of the endemic subspecies (we lacked samples of northern cardinal) and one more widespread subspecies (verdin) present on the island from sequences previously published for mainland populations. For most populations, we discovered no genetic differentiation between the island and the mainland, thus questioning the taxonomic validity of the endemic subspecies. The canyon towhee and the verdin showed significant mitochondrial DNA differentiation, although neither was reciprocally monophyletic. We modeled the ecological niche for the mainland populations of the study species (plus the curve-billed thrasher, which was studied earlier) and determined if species' occurrence on the island was predicted. We found no ecological differences for the four species that showed no genetic differences and one of the species that did (verdin). In contrast, some ecological differentiation was detected for the canyon towhee and the curve-billed thrasher. We conclude that the ecological differences leading to paler plumages in Tiburón Island endemics are sufficiently subtle as to not be discovered in our ecological models, although they are likely influenced by variation in rainfall, temperature and the vegetation. In addition, the black-tailed gnatcatcher is not paler, and therefore might respond to different ecological variables. We simulated sequence data and showed that if the populations on Tiburón Island have been isolated for 10 000 years, there ought to be greater differences than we observed for black-tailed gnatcatcher, Gambel's quail, cactus wren, and gila woodpecker, suggesting that there has been gene flow connecting the mainland and island populations. If so, then the paler coloration of these Tiburón Island subspecies (excluding black-tailed gnatcatcher) has been maintained by natural selection despite gene flow. In any case, the pale coloration apparently evolved within the past 10 000 years.