Applying linear programming to estimate fluxes in ecosystems or food webs: An example from the herpetological assemblage of the freshwater Everglades

We present the application of Linear Programming for estimating biomass fluxes in ecosystem and food web models. We use the herpetological assemblage of the Everglades as an example. We developed food web structures for three common Everglades freshwater habitat types: marsh, prairie, and upland. We obtained a first estimate of the fluxes using field data, literature estimates, and professional judgment. Linear programming was used to obtain a consistent and better estimate of the set of fluxes, while maintaining mass balance and minimizing deviations from point estimates. The results support the view that the Everglades is a spatially heterogeneous system, with changing patterns of energy flux, species composition, and biomasses across the habitat types. We show that a food web/ecosystem perspective, combined with Linear Programming, is a robust method for describing food webs and ecosystems that requires minimal data, produces useful post-solution analyses, and generates hypotheses regarding the structure of energy flow in the system.