Protecting juveniles, spawners or both: a practical statistical modelling approach for the design of marine protected areas.
Fish populations undertaking ontogenetic or spawning migrations pose challenges to marine protected area (MPA) planning because of the large extent of their distribution areas. There is a need to identify the juvenile and spawner hotspots of these populations that could be set aside as MPAs. Species distribution models making comprehensive use of available monitoring data and predicting the realized juvenile and spawner hotspots of migratory fish populations will assist resource managers with MPA planning. We developed a statistical modelling approach relying on multiple, regional monitoring datasets for assisting spatial protection efforts targeting the juveniles, spawners, or both life stages, of migratory fish species and species complexes. This approach predicts juvenile and spawner hotspot indices, and critical life stage (CLS) hotspot indices, which integrate both juvenile and spawner hotspot indices. We applied the approach to 11 vulnerable species of the grouper-snapper complex of the U.S. Gulf of Mexico, which all form fish spawning aggregations (FSAs). The CLS hotspot index was predicted to be highest in the Pulley Ridge and Flower Garden Banks areas, followed by the West Florida Shelf, southwestern Florida waters and portions of the Louisiana-Mississippi-Alabama shelf. The Pulley Ridge Habitat Area of Particular Concern and Flower Garden Banks National Marine Sanctuary are two important existing MPAs of the U.S. Gulf of Mexico, whose possible expansion is being considered. The predicted CLS hotspot indices suggest that expanding these MPAs or increasing harvest regulations within them would offer substantial protection to both the juveniles and spawners of many FSA-forming species of the grouper-snapper complex. Synthesis and applications. As the number of marine protected areas (MPAs) continues to increase worldwide, statistical modelling approaches making comprehensive use of available data are urgently needed to support resource managers' abilities to establish sound and efficient spatial protection plans. The outputs of our statistical models can serve as inputs to conservation planning software packages seeking optimal marine protected areas configurations or can be directly employed by resource managers for formulating spatial protection plans.