FINE-SCALE SPATIAL ECOLOGY OF AN ESTUARINE FORAGE FISH

Abstract

The spatial distribution of a species is shaped by the interactions of biotic and abiotic processes, which may cause unique patterns at any one spatial scale. This makes sampling at multiple spatial scales and across many habitat types essential to accurately determine and characterize the drivers of species abundance and distribution, though this has historically been difficult. In this dissertation, I examined the fine-scale spatial ecology of juvenile Atlantic menhaden (Brevoortia tyrannus) in a representative tidal estuarine tributary. Data were collected in the Patuxent River and its connected creeks using Adaptive Resolution Imaging Sonar (ARIS) cameras. The data were analyzed for trends in menhaden distribution at multiple spatial scales, and at aggregative extents from individual menhaden to clusters of menhaden schools. In Chapter 2, I used a generalized additive model (GAM) to characterize the density of individual menhaden across the lower Patuxent River. The highest density of menhaden was found within shallow-water tidal creeks, rather than in the Patuxent River channel. Water depth was a significant predictor of menhaden density. In Chapter 3, I developed techniques to identify the discrete schools into which individual menhaden were aggregated. Schools typically consisted of fewer than 100 individuals. GAMs were used to characterize the density of menhaden within schools and the distance of schools from their nearest neighbors. Water depth, habitat, sampling month, and tidal stage were all significant predictors of menhaden school characteristics. In Chapter 4, I used spatial point pattern analysis to examine the spatial pattern of menhaden schools. Few identifiable trends were found. However, the results indicated that the relative impact of environmental and behavioral mechanisms on menhaden school distribution likely varies with spatial scale; environmental mechanisms have a greater impact at larger spatial scales, while behavioral mechanisms are more impactful at smaller spatial scales. Chapter 5 used the results of all empirical data analysis to inform the construction of two dynamic, spatially explicit individual-based models (IBMs) of menhaden movement. One IBM represented menhaden movement as only dependent on environmental stimuli, and the other represented menhaden movement as only dependent on the movements of conspecifics. Neither model was sufficient on its own to replicate the patterns of menhaden distribution observed in the empirical data. The spatial patterns that occur at any one space-time are likely the emergent result of the interactions between the two types of forces and individual variation in the perception of and response to local stimuli. The results are an important first step towards quantifying and resolving the relative impact of schooling behavior and environmental conditions on menhaden distribution. The research in this dissertation represents a small piece of estuarine forage fish spatial ecology. However, our understanding of any species’ distribution in space and time is at its most comprehensive when it is seen as the emergent property of all its smaller components. The novel combination of fine-scale empirical data, statistical modeling, and individual-based models contribute to our understanding of the spatial ecology of forage fish in estuarine environments.