Ocean observatory data are useful for regional ­habitat modeling of species with different vertical habitat preferences

Ocean Observing Systems (OOS) now provide comprehensive descriptions of the physical forcing, circulation, primary productivity and water column properties that subsidize and structure habi- tats in the coastal ocean. We used generalized additive models (GAM) to evaluate the power of OOS remotely sensed ocean data along with in situ hydro- graphic and bottom data to explain distributions of 4 species important in the Mid-Atlantic Bight, USA, ecosystem that have different vertical habitat prefer- ences. Our GAMs explained more abundance varia- tion for pelagic species (longfin inshore squid and butterfish) than demersal species (spiny dogfish and summer flounder). Surface fronts and circulation pat- terns measured with OOS remote sensing as well as the rugosity and depth of the bottom were important for all species. In situ measurements of water column stability and structure were more useful for modeling pelagic species. Regardless of vertical habitat prefer- ence, the species were associated with vertical and horizontal current flows, and/or surface fronts, indi- cating that pelagic processes affecting movement costs, prey production and aggregation influenced distributions. Habitat-specific trends in abundance of 3 of the 4 species were well described by our OOS- informed GAMs based upon cross validation tests. Our analyses demonstrate that OOS are operationally useful for regional scale habitat modeling. Regional scale OOS-informed statistical habitat models could serve as bases for tactical ecosystem management and for the development of more sophisticated spa- tially explicit mechanistic models that couple onto- genic habitats and life history processes to simulate recruitment of organisms important to maintaining the resilience of coastal ecosystems.