Insight into an island radiation: the Tarentola geckos of the Cape Verde archipelago

Aim To reassess the relationships between Tarentola geckos from the Cape Verde Islands by including specimens from all islands in the range. To determine the variation within forms by sequencing over 400 specimens, thereby allowing the discovery of cryptic forms and resolving some of the issues raised previously. This extensive sampling was also used to shed light on distributions and to explain genetic diversity by comparing the ages and ecological and geological features of the islands (size, elevation and habitat diversity).

Location The Cape Verde Islands: an oceanic archipelago belonging to the Macaronesian biogeographic region, located around 500 km off Senegal.

Methods A total of 405 new specimens of Tarentola geckos were collected from nine islands with very different geological histories, topography, climate and habitats. Mitochondrial cytochrome b (cyt b) gene and 12S rRNA partial sequences were obtained and analysed using phylogenetic methods and networks to determine molecular diversity, demographic features and phylogeographic patterns.

Results The phylogenetic relationships between all known forms of Cape Verdean Tarentola specimens were estimated for the first time, the relationships between new forms were assessed and previously hypothesized relationships were re-examined. Despite the large sample size, low intraspecific diversity was found using a 303-bp cyt b fragment. Star-like haplotype networks and statistical tests suggest the past occurrence of a rapid demographic and geographical expansion over most of the islands. Genetic variability is positively correlated with size, elevation and habitat diversity of the islands, but is not linearly related to the age of the islands. Biogeographical patterns have, in general, high concordance with phylogenetic breaks and with the three eco-geographical island groups. Volcanism and habitat diversity, both tightly linked with island ontogeny, as postulated by the general dynamic model of oceanic island biogeography, as well as present and historical size of the islands appear to be the main factors explaining the genetic diversity of this group.

Main conclusions The Tarentola radiation was clarified and is clearly associated with the geological and ecological features of the islands. Two factors may account for the low intraspecific variation: (1) recent volcanic activity and high ecological stress, and (2) poor habitat diversity within some islands. More studies are needed to align taxonomy with phylogenetic relationships, whereas GIS modelling may help to predict precise species distributions.