Tropical Cyclone Effects on Fish Stocks and Fisheries in the Florida Keys

Severe tropical cyclones, namely, hurricanes and tropical storms, frequently affect important marine fish stocks and fisheries along the United States Atlantic and Gulf of Mexico coasts. As fish and fishery responses to these disturbances are poorly understood, tropical cyclone disturbances are not explicitly incorporated into stock assessment models. Furthermore, as it is unknown whether data used in stock assessments reflect tropical cyclone-induced changes, fishery and fish stock responses unaccounted for in assessments may be contributing error and variability to important indices used to manage fisheries. In order to determine when to appropriately account for tropical cyclone-induced changes, scientists need a better understanding of specific stock and fishery responses to tropical cyclones. However, directly observing these responses is difficult due to limited predictability of when and where tropical cyclones will strike, and further complicated by sampling limitations during and after the events. Specific stock and fishery responses to tropical cyclones were identified by analyzing data specific to the Florida Keys from seven long-term reef fish data collection programs. The Florida Keys region, from Biscayne National Park to Dry Tortugas National Park, was selected as the study site based on both the high number of hurricanes and tropical storm strikes in the past two decades, and on the high number of quality data collection programs available to the author. A total of seven datasets, six fishery dependent and one fishery independent, were obtained from state, federal, and academic studies. Hypotheses tests were developed to evaluate changes in fishery dependent variables including quantity and distribution of targeted effort; species-specific catch rates and catch size; and species composition. Fishery independent variables included abundance estimates, species composition, and size composition. Change point analysis and dissimilarity matrices proved to be promising statistical analyses to identify changes in a single variable over time and to identify trends in matrices over space and time, respectively. Following the initial hypotheses tests, further exploration included principal component analysis, and multidimensional scaling to characterize tropical cyclones and resulting environmental characteristics that are indicative of changes in fishery variables. Based on the type of stock, type of fishery data, severity of tropical cyclone and resulting environmental effects in consideration, the conclusions of this research will help fisheries scientists determine when to explicitly correct for tropical cyclone influences in their stock assessments.