Online

Chamberland, Lisa

Biology

2020

Ph. D.

The distributions of Earth's flora and fauna are shaped by a myriad of biotic and abiotic factors including dispersal capacity, and the geologic histories of landscapes. The relative importance of long-distance (overland or overwater) dispersal and vicariance, or the subsequent separations of populations via physical barriers (e.g. mountains, rivers, oceans) in shaping the disjunct distribution of taxa, are the two main biogeographic hypotheses used to explain disjunct distributions. This research explored the evolutionary histories of spiders to test how long-distance dispersal and vicariance have generated species diversity and distributions are shaped on local and global scales. In the first part of this study, we investigated the evolutionary history of Deinopis (the net-casting spider) and Gasteracantha (the spiny backed spider), in the Caribbean. The Caribbean islands are a biodiversity hotspot characterized by islands with unique geologic histories and heterogeneous landscapes rich in endemic species. Using a combination of phylogenomic data and biogeographic hypothesis testing, we tested the role of the Greater Antilles Aves Ridge (GAARlandia) in the colonization of these spiders into the Caribbean. This postulated land bridge would have connected South America to the Greater Antilles 33-35 million years ago and served as a temporary corridor for taxa to disperse overland to the islands. We found high levels of endemism within Deinopis in addition to support for a GAARlandia mediated dispersal to the Greater Antilles. In contrast, Gasteracantha, dispersed to the Caribbean long after the disappearance of GAARlandia and we find evidence of high ongoing gene flow among islands and between the islands and the mainland. For our global study, we examined the role of Gondwana, the southern portion of the supercontinent Pangaea, in evolution of the family Deinopidae. We found a complex biogeographic history of deinopids characterized by Upper Cretaceous vicariance as South America was splitting from Arica followed by multiple long-distance, transoceanic dispersals. These results suggest that, while the breakup of Gondwana was important for early deinopid diversification we nevertheless document long-distance dispersals as a more important factor in the distributions of deinopids than previously hypothesized.