Landscape Modeling of Coastal Habitat Change in the Mississippi Delta

Abstract

A landscape model was developed to investigate and predict the environmental factors affecting wetland habitat change within the Barataria and Terrebonne basins of coastal Louisiana, USA. The model linked an overland-flooding hydrodynamic module, using cells of 100 km2 in size and operating at a 1-h time step, and a spatially articulated ecosystem module, resolving habitat type and change for 1-km2 cells in daily time steps. Integration across different temporal and spatial scales was accomplished with interpolation routines and averaging algorithms. Forcing functions included dominant regional processes, such as subsidence, sedimentation, and sea-level rise. Hydrologic functions were calibrated against existing climate and hydrologic time series, while habitat information was compared to maps prepared by the United States Fish and Wildlife Service (USFWS) for 1978 and 1988. Spatial calibration was done by initializing the landscape pattern of the model to a 1978 USFWS habitat map. After a 10-yr simulation, the results were compared against a 1988 USFWS habitat map. Simulated maps had an accuracy of 85–90 (out of a maximum of 100), based on a multiple resolution fit algorithm. For validation, the model was initialized with a 1956 USFWS habitat map, and the results from a 32-yr simulation were compared to the 1988 USFWS habitat map. The landscape model produced reasonable regional agreement, despite the fact that small-scale processes and features were not included. The validation runs produced land-loss rates that matched historical trends with an accuracy fit above 75. The model simulated 30 years into the future, starting in 1988, testing for long-term climate variability under diverse scenarios. Results indicated that weather variability impacts land-loss rates more than replication of extreme weather years. Even when extreme dry and wet years were repeated, the model predicted lower land loss when compared to historical records. This is indicative of the ability of the simulated plant communities to adapt to repetitive climatic forcing functions.