Abstract
The spinetail devil ray Mobula mobular is a data-poor species of global conservation concern with distribution and movement patterns that are linked to dynamic oceanographic processes. While some conservation measures to protect this species from diverse human threats have been implemented, the potential impacts of climate variability on this species and its main prey Nyctiphanes simplex have not yet been fully considered. We used boosted regression tree models to predict the habitat suitability and niche overlap (Schoener's D index) of predator and prey in northwestern Mexico during 2005-2015. This period included anomalously warm (El Ni & ntilde;o) and cold (La Ni & ntilde;a) years, allowing us to predict habitat preferences for both species under contrasting ENSO environmental conditions. Our model predicts that M. mobular and N. simplex distributions are associated with productive waters along the coast and with considerably less probability offshore of northwest Mexico. We found strong spatial relationships between the habitat suitability of M. mobular and its prey, with high distributional overlap inside the Gulf of California during the warm season. During the 2010 La Ni & ntilde;a event, environmental conditions resulted in particularly high habitat suitability and overlap for both species inside the Gulf during August and September, while during the 2015 El Ni & ntilde;o event, habitat suitability and overlap were higher on the west coast of the Gulf of California. We demonstrate how predicting distribution patterns of data-poor species provides vital information for marine resource managers to develop effective protection strategies for species of conservation concern, like M. mobular.