Search for: All records

Abstract Marine ecosystems in the wider Caribbean region (WCR) are biodiversity hotspots. They include coral reefs and provide critical societal benefits, yet climate change, pollution, and overfishing are threatening them. Marine ecosystem protection and restoration require understanding connectivity. Fish and coral larvae are actively exchanged across connected areas and larval transport promotes the replenishment of new healthy individuals after damaging events. Connectivity is dynamic and modulated by climate variability, but its evaluation with traditional tools remains elusive over spatio‐temporal scales of climate interest. Here machine learning helps exploring large‐scale connectivity in the WCR over nearly three decades. ENSO exerts the largest influence on the overall connectivity, with enhanced longitudinal connectivity in El Niño years. By combining connectivity with climate variability and thermal stress metrics, it is found that connectivity does not improve recovery potential in the WCR, in striking contrast with prior results for the tropical Pacific.