Modeling Early Life: Ontogenetic Growth and Behavior Affect Population Connectivity in Gulf of Mexico Marine Fish

This dissertation is an examination of growth, behavior, and dispersal during the early life stages of marine fishes in the Gulf of Mexico. Understanding movements of early life stages is a key part of managing exploited fish populations. Position in the water column can impact larval dispersal, since it determines those currents to which larvae are exposed. First, I investigated the relationship between length and age in early life stages of marine fishes. I found that demersal fish taxa tend to be represented by exponential models, while pelagic fish tend to be represented by linear models. I suggest this may be indicative of risk-taking strategies associated with a consumption/predation risk trade-off. Using some of the age models from this study to estimate the ages of sampled larvae, generalized additive models (GAM) were built using a suite of environmental variables. The GAMs were then used to predict the ontogenetic vertical migration patterns for a range of age classes and taxa of major commercial and recreational fishes in the Gulf of Mexico. These ontogenetic vertical migration patterns are used in combination with circulation data from the West Florida Coastal Ocean Model (WFCOM) to predict larval trajectories in the Gulf of Mexico. These trajectories are then used to identify spawning and settlement habitats. This creates a source/sink matrix, which can help support realistic modeling of population connectivity within ecosystem models such as Atlantis. This project lies at the intersection of predictive statistical modeling, hydrodynamic modeling, and ecosystem modeling, and represents an interdisciplinary approach to understanding larval dynamics and the impacts that they have on ecosystems and fish production