Abstract
<em>Abstract.—</em> A need exists to scientifically determine optimal fish habitats to support decision making for management of essential fish habitat. Scientists have been collaborating to conduct habitat suitability index (HSI) modeling to spatially delineate fish habitats for estuarine fish and invertebrate species in Tampa Bay and Charlotte Harbor, Florida. Results from HSI modeling of juvenile spotted seatrout <em>Cynoscion nebulosus </em> in Charlotte Harbor are presented. Data obtained from 1989–1997 by fisheries-independent monitoring in the two estuaries were used along with environmental data from other sources. Standardized catch-per-unit-effort (catch rates) were calculated across gear types using fisheries-monitoring data from Charlotte Harbor and Tampa Bay. Suitability index functions were determined using three methods: (1) frequency of occurrence, (2) mean catch rates within ranges, and (3) smooth-mean catch rates determined by polynomial regression. Mean catch rates were estimated within biologically relevant ranges and, where sufficient data were available, for finer intervals across environmental gradients. Suitability index functions across environmental gradients were then derived by scaling catch rates. Gridded habitat layers for temperature, salinity, depth, and bottom type in Charlotte Harbor were also created using a geographic information system. Habitat suitability index modeling was conducted using the U.S. Fish and Wildlife Service geometric mean method linked to the ArcView Spatial Analyst module. The model integrated suitability indices associated with the habitat layers for Charlotte Harbor to create a map of the predicted distribution for juvenile spotted seatrout during the fall season. Suitability indices developed for Tampa Bay were used with Charlotte Harbor habitat layers to test transfer of the indices to another estuary. Predicted HSI maps depicted low to optimum habitat suitability zones in Charlotte Harbor. Model performance was evaluated by statistically comparing the relative ranking of mean catch rates with mean suitability indices for corresponding zones. Suitability indices obtained using polynomial regression methods yielded morereliable HSI maps for juvenile spotted seatrout than those derived using mean catch rates within biologically relevant ranges. The observed map, derived using smooth-mean suitability indices transferred from Tampa Bay, was not significantly different (Chi-square goodness-of-fit test) from the expected map derived using smooth-mean indices from Charlotte Harbor. Our modeling efforts using transferred indices indicate that it is possible to predict the geographic distributions of fish species by life stage in estuaries lacking fisheries monitoring.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have