Statistical and Simulation Modeling of Brown Shrimp Population Dynamics in the Northern Gulf of Mexico.

Abstract

This dissertation used statistical and simulation modeling to examine population dynamics of brown shrimp, Penaeus (Farfantepenaeus ) aztecus. Brown shrimp are ideal to use in examining recruitment because shrimp represent a common life history strategy, because they are commercially important, and because there are extensive data available. In the first section of this dissertation, I used long-term monitoring data to examine correlative relationships between annual, stage-specific abundance estimates and environmental factors. Criticisms of traditional correlative studies were addressed by using a spatially and temporally-extensive dataset, by comparing stepwise multiple regression to Bayesian model averaging, and by investigating nonlinear relationships with generalized additive models. Postlarval abundance was not well described by any statistical model. Juvenile abundance was partially described by environmental variables such as temperature, water clarity, and water level. Adult abundance was well described by early juvenile abundance, salinity, and temperature. These results suggested that juvenile abundance may be the critical component in determining the year-class strength of shrimp. In the second section of this dissertation, a spatially-explicit, individual-based simulation model was developed to further investigate relationships between estuarine habitats and juvenile shrimp survival. The model simulated the movement, mortality, and growth of individual shrimp during their residence in estuarine marshes. Relationships between shrimp survival and marsh attributes (amount of vegetation and edge habitat) were examined by overlaying the simulation on 4 habitat maps created from aerial photographs. The model was corroborated with fine-scale density patterns observed in the northern Gulf of Mexico, and sensitivity analyses were performed. Surviving shrimp grew faster, moved less, spent more time in vegetation, and experienced slightly higher local densities than shrimp that died during the simulation. Maps with more edge habitat supported higher simulated survival under baseline conditions, under high shrimp densities, under alternative rules about movement, and under realistic uncertainty in model inputs. The characteristics of surviving shrimp suggested that high-edge habitats increase survival by providing shrimp direct access to the benefits of vegetated marsh without the high-density cost that may be encountered in low-edge habitats. These results highlight the important role of marsh habitat in determining the recruitment of an estuarine-dependent species.