A COMBINED REMOTE SENSING/MULTIVARIATE MODEL OF BALD CYPRESS HABITAT IN THE YAZOO BASIN, LOWER MISSISSIPPI RIVER ALLUVIAL VALLEY

Change in the timing, duration and depth of flooding of the Gulf of Mexico Coastal Plain and the Lower Mississippi River Alluvial Valley have raised concerns about the resultant influence on baldcypress (Taxodium distichum) trees, which are ubiquitous in this area. The present work investigates the influence of water level (measured as depth of standing water) and climatic factors (precipitation, temperature, relative humidity, and sunlight hours) on remote sensing data collected from the Landsat5 Multispectral Scanner (MSS). The normalized difference vegetation index (NDVI) values calculated from the MSS are used as an indicator for bald cypress productivity. A linear multivariate model using NDVI as the dependent variable helps define the relationships among the water level and climatic factors and this indicator for growth of bald cypress trees. Data for two bald cypress habitats in the Yazoo Basin portion of the Lower Mississippi River Alluvial Valley were used to construct the model to predict NDVI. The multivariate model of NDVI shows statistically significant positive relationships for temperature and sunlight hours (which comprise the seasonal components of the model) and a statistically significant inverse relationship between NDVI and water level. The regression models further indicate that extended periods of deep water negatively affect NDVI for bald cypress habitat.