Tracing the biogeographic history of the world's most isolated insular floras

Abstract

Inferring general biogeographic patterns in the sub-Antarctic region has been challenging due to the disparate geological origins of its islands and archipelagos—ranging from Gondwanan fragments to uplifted seafloor and more recently formed volcanic islands—and the remoteness of these island systems, spread around the austral continental landmasses. Here, we conduct phylogenetic reconstruction, divergence time estimation, and Bayesian Island Biogeographic analyses to reconstruct the spatio–temporal colonization histories of seven vascular plant lineages, which are either widespread across the sub-Antarctic region (Acaena magellanica, Austroblechnum penna-marina, Azorella selago, Notogrammitis crassior) or restricted to an extremely remote sub-Antarctic province (Colobanthus kerguelensis, Polystichum marionense, Pringlea antiscorbutica). Our results reveal high biological connectivity within the sub-Antarctic region, with southern landmasses (Australia, New Zealand, South America) as key sources of sub-Antarctic plant diversity since the Miocene, supporting long-distance dispersal as the primary colonization mechanism rather than tectonic vicariance. Despite the geographic isolation of the sub-Antarctic islands, eastward and westward colonization events have maintained this connectivity, likely facilitated by eastward-moving marine and wind currents, short-term weather systems, and/or dispersal by birds. Divergence time estimates indicate that most species diverged within the Plio–Pleistocene, with crown ages predating the Last Glacial Maximum, suggesting that sub-Antarctic archipelagos acted as refuges for biodiversity. Our findings highlight the role of one of the most remote sub-Antarctic archipelagos as both a refugium and a source of (re)colonization for continental regions. These results underscore the urgent need for establishing priority conservation plans in the sub-Antarctic, particularly in the face of climate change.