Abstract
The greater Everglades ecosystem, including Florida Bay, has undergone significant anthropogenic manipulation over the past century. These actions resulted in a series of ecologically undesirable events in the Everglades ecosystem, prompting passage of the Comprehensive Everglades Restoration Plan (CERP). It is necessary to understand the variability in, and relationship between, salinity and ecology to fully evaluate the potential effects of CERP on Florida Bay. A seven-year dataset on surface salinity along with eleven-year and eight-year datasets on mesozooplankton and planktivorous fish were analyzed. Overall, mean Bay-wide salinity varied from a low of 24.2 just after the passing of Hurricane Irene in October 1999 to a high of 41.8 near the end of a drought period in July 2001. Bay-wide mean salinity exhibited dramatic decreases, up to 0.5 per day, whereas increases in bay-wide salinity were slower, with a maximum rate of 0.1 per day. Meteorological phenomena, such as tropical cyclones and ENSO, dramatically altered the salinity patterns of Florida Bay on interannual time scales. There was a large degree of spatial heterogeneity in salinity between sub-regions of Florida Bay due to differing freshwater sources and geomorphology. Mesozooplankton abundance displayed interannual variability and a positive correlation with salinity. Both of these features were also closely correlated with abundance of the dominant planktivorous fish, Anchoa mitchilli, indicating the importance of top-down control. The hypersaline periods appear to provide a refuge from predators, allowing mesozooplankton to increase in abundance during periods of increased physiological stress. The interaction between mesozooplankton and A. mitchilli, along with its correlation to salinity, was further investigated through the development of a mechanistic model of the populations in Florida Bay. The model indicated predation alone was insufficient to control mesozooplankton populations; rather, it was necessary to incorporate density-dependence utilizing a logistic prey population. With both mechanisms the model was able to replicate the observed interannual variability pattern and positive correlation between mesozooplankton and salinity. A preliminary management scenario evaluation suggests a two to six-fold difference in A. mitchilli and mesozooplankton populations between targeted and general salinity reductions. This suggests alternative freshwater management scenarios could produce drastically different ecological consequences.