干旱区生态环境知识资源中心

The methods used to detect and describe morphologically cryptic species have advanced in recent years, owing to the integrative nature of molecular and morphological techniques required to elucidate them. Here we integrate recent phylogenomic work that sequenced many genes but few individuals, with new data from mtDNA and morphology from hundreds of gecko specimens of the Gehyra variegata group from the Australian arid zone. To better understand morphological and geographical boundaries among cryptic forms, we generated new sequences from 656 Gehyra individuals, largely assigned to G. variegata group members over a wide area in Western Australia, with especially dense sampling in the Pilbara region, and combined them with 566 Gehyra sequences from GenBank, resulting in a dataset of 1,222 specimens. Results indicated the existence of several cryptic species, from new species with diagnostic morphological characters, to cases when there were no useful characters to discriminate among genetically distinctive species. In addition, the cryptic species often showed counter-intuitive distributions, including broad sympatry among some forms and short range endemism in other cases. Two new species were on long branches in the phylogram and restricted to the northern Pilbara region: most records of the moderately sized G. incognita sp. nov. are near the coast with isolated inland records, whereas the small-bodied saxicoline G. unguiculata sp. nov. is only known from a small area in the extreme north of the Pilbara. Three new species were on shorter branches in the phylogram and allied to G. montium. The moderately sized G. crypta sp. nov. occurs in the western and southern Pilbara and extends south through the Murchison region; this species was distinctive genetically, but with wide overlap of characters with its sister species, G. montium. Accordingly, we provide a table of diagnostic nucleotides for this species as well as for all other species treated here. Two small-bodied species occur in isolated coastal regions: G. capensis sp. nov. is restricted to the North West Cape and G. ocellata sp. nov. occurs on Barrow Island and other neighbouring islands. The latter species showed evidence of introgression with the mtDNA of G. crypta sp. nov., possibly due to recent connectivity with the mainland owing to fluctuating sea levels. However, G. ocellata sp. nov. was more closely related to G. capensis sp. nov. in the phylogenomic data and in morphology. Our study illustrates the benefits of combining phylogenomic data with extensive screens of mtDNA to identify large numbers of individuals to the correct cryptic species. This approach was able to provide sufficient samples with which to assess morphological variation. Furthermore, determination of geographic distributions of the new cryptic species should greatly assist with identification in the field, demonstrating the utility of sampling large numbers of specimens across wide areas.