| Coastal systems are facing many challenges including climate change, sea-level rise, storm surge, and erosion, all of which contribute to land loss. In Louisiana, this has led to the development of a coastal Master Plan supported by habitat suitability index (HSI) models to predict wildlife responses under various management scenarios. However, these models were not originally intended for this purpose and their functionality at large spatial scales is unclear. My goal was to use maximum entropy modeling, using the software MaxEnt, to predict how various bird distributions might change with coastal restoration and management. During 2015-2017, I recorded the locations of brown pelican, gadwall, green-winged teal, mottled duck, and roseate spoonbill across southern Louisiana, measuring salinity, water depth, and vegetation when the species were detected. Using environmental projections from the Coastal Protection and Restoration Authority, I predicted the probability of occurrence for each target species for current conditions and projected the distributions into the future at 25 and 50 years using sea-level rise and coastal change scenarios. Predictive models for each species under current conditions show good agreement with field observations. Future models generally show reductions in areas of potentially high habitat use, with a few notable exceptions in brown pelicans and roseate spoonbills. Both modeling approaches had advantages and disadvantages; neither were superior in predicting wildlife habitat. I recommend increasing the resolution and quality of environmental data to improve estimates of suitable habitat, habitat use, and restoration outcomes for wildlife. |
