The role of piscivores in a species-rich tropical river
Much of the world's species diversity is located in tropical and sub-tropical ecosystems, and a better understanding of the ecology of these systems is necessary to stem biodiversity loss and assess community- and ecosystem-level responses to anthropogenic impacts. In this dissertation, I endeavored to broaden our understanding of complex ecosystems through research conducted on the Cinaruco River, a floodplain river in Venezuela, with specific emphasis on how a human-induced perturbation, commercial netting activity, may affect food web structure and function. I employed two approaches in this work: (1) comparative analyses based on descriptive food web characteristics, and (2) experimental manipulations within important food web modules. Methodologies included monthly sampling of fish assemblages using a variety of techniques, large-scale field experiments, extensive stomach content and stable isotope analyses. Two themes unite the information presented: (1) substantial spatial and temporal variability in food web structure, and (2) how body-size can be used to generalize species-interactions across this complexity. Spatial variability occurred at various scales, from among small fish assemblages on seemingly homogeneous sand banks, to differences among landscape scale units (e.g. between lagoons and main river channel). Seasonal variability was apparent in predation patterns, with relative prey availability and body size primarily resulting in decreasing prey sizes with falling water levels. Body size was also related to functional outcomes of species interactions, for example, a size-based response of prey fishes to large-bodied piscivore exclusion. This pattern was further substantiated at the landscape-scale, as differences in assemblage structure among netted and un-netted lagoons were largely size-based. Trophic position of fish and body size was not found to be related, likely due to the diversity of prey available to consumers, and may signify that commercial netting activity will not decrease food chain lengths. In sum, by describing human impacts within a food web context, I endeavor to provide predictive power regarding a specific human-induced environmental problem, yet still allowing for generality that will broaden the theoretical foundations and applications of food web ecology.