Colonizing the Caribbean: New geological data and an updated land‐vertebrate colonization record challenge the GAARlandia land‐bridge hypothesis

Abstract

Over the past three decades, the hypothesized mid-Cenozoic GAARlandia walkway (34 ± 1 Ma) has featured prominently in discussions on Caribbean biogeography. However, a fundamental issue has been a lack of geological and geophysical data from the Aves Ridge with which to evaluate the proposal. Consequently, opinions have been based purely on biological data, mainly from molecular-clock studies but also from fossils and taxonomic composition. For whatever reasons, a polarization of views has occurred; some champion the land-bridge to explain the colonizations of the island group, whereas others prefer over-water dispersal. We examine the hypothesis drawing upon an updated compilation of land-vertebrate colonizations (35 clades: 6 amphibian, 6 mammal and 23 reptile) plus a newly published geological study of the Aves Ridge and the adjacent Grenada Basin (to the east). The former indicates that the colonizations were heavily filtered and assembled in a piecemeal fashion consistent with over-water dispersal; a mid-Cenozoic ecological corridor would likely have led to a richer assemblage with a wider variety of components. The latter, which draws upon drill-core and seismic data, indicates that the southern and central Aves Ridge was submerged at the aforesaid time. With the GAARlandia hypothesis geologically falsified, many biogeographical scenarios that have been proposed for the Greater Antillean biota now require re-evaluation.