Mainland colonization by island lizards

Aim We investigate biogeographic relationships within the lizard genus Anolis Daudin, 1802 to test the hypothesis that the mainland (Central and South American) Norops-clade species descended from a West Indian Anolis ancestor. Previous hypotheses have suggested that close island relatives of mainland Norops species (the Cuban Anolis sagrei and Jamaican A. grahami series) represent over-water dispersal from a mainland ancestor. These previous hypotheses predict that the A. sagrei and A. grahami series should be phylogenetically nested within a Norops clade whose ancestral geography traces to the mainland. If Norops is West Indian in origin, then West Indian species should span the deepest phylogenetic divergences within the Norops clade. Location Central and South America and West Indian islands. Methods The phylogenetic relationships of Anolis lizards are reconstructed from aligned DNA sequences using both parsimony and Bayesian approaches. Hypotheses are tested in two ways: (1) by reconstructing the ancestral geographic location for the Norops clade using Pagel & Lutzoni's (2002) Bayesian approach, and (2) by testing alternative topological arrangements via Wilcoxon Signed-Ranks tests (Templeton, 1983) and Shimodaira-Hasegawa tests (Shimodaira & Hasegawa, 1999). Results Our evidence supports an origin of mainland Norops anoles from a West Indian ancestor. A West Indian ancestor to the Norops clade is statistically supported, and alternatives to the biogeographic pattern [Cuban (Jamaican, Mainland)] are statistically rejected by Shimodaira-Hasegawa tests, although not by Wilcoxon Signed-Ranks tests. Main conclusions Our data support the hypothesis of a West Indian origin for mainland Norops. This result contradicts previous hypotheses and suggests that island forms may be an important source for mainland biodiversity.