Spatial patterns and habitat associations of targeted reef fish in and around a marine protected area in St. Croix, U.S. Virgin Islands

As marine protected areas (MPAs) become an increasingly popular method to address global declines of coral reef ecosystems, understanding how to gauge no-take reserve effectiveness through both simple metrics and spatially quantitative methods is imperative for robust design and management. In order to be an effective fisheries management tool, it must be shown that MPAs do not simply displace fishing effort, but can actually increase the biomass or density of fish in adjacent fisheries. Net export of adult fish to surrounding waters, termed the spillover effect, is one way MPAs may function to replenish fish stocks. The ability to detect the spillover effect using simple metrics of abundance and biomass for five highly targeted species was tested for an MPA in the US Virgin Islands. Moreover, it is equally important for managers to understand linkages between reef fish biomass patterns and associated habitat at multiple scales. Habitat characteristics that influence distribution patterns of targeted reef fish can be explored through spatial statistics and spatial multivariate models. In this study, the specific aim was to further evaluate seascape and local site habitat metrics to find how spatial scales of habitat composition and configuration can be used together cohesively to understand reef fish distribution patterns. In addition, global and local spatial regression models were compared to determine if there is non-stationarity of ecological processes across management zones. Results of this study reinforce the complexities of fish-habitat interactions, which make detection of spillover difficult using simple metrics. However, using a combination of both multi-scale spatial habitat metrics and local regression methods can aid in MPA management and design.