Main content area

Contrast of Constant and Bioenergetics‐Based Assimilation Efficiencies in Describing Ecosystems

Jackson, Leland J.
Ecological applications 1996 v.6 no.4 pp. 1355-1364
Alosa pseudoharengus, Salvelinus namaycush, adults, animal growth, biogeochemical cycles, biomass, carbon, ecosystems, energy metabolism, food webs, game fish, lakes, mass flow, models, nutrients, polychlorinated biphenyls, predation, prediction, surveys, trace elements, Lake Ontario
Ecosystem models often follow the mass flow of matter (e.g., carbon, nutrients, trace metals, and contaminants) through food webs. Many ecosystem models (≈ 75% in a survey of recent literature) assume that carbon allocated to growth is a constant fraction of consumption. I use a simple predator‐prey model to demonstrate that the method of allocating consumed carbon to predator biomass (as a constant fraction or as a variable based on bioenergetics) can greatly affect the predicted dynamics of carbon and other matter. I then evaluate the potential disparity in the two carbon allocation methods by comparing predictions of constant and bioenergetics‐based assimilation versions of a Lake Ontario pelagic food web model under hypothetical management scenarios. Given current lake productivity and salmonid stocking, the version with constant assimilation efficiency predicted 22% more alewife (Alosa pseudoharengus) biomass than the bioenergetics‐based version at year 2005. When these models were calibrated to produce similar whole‐lake adult sport fish biomass, the constant assimilation model predicted 27% less alewife biomass at year 2005 than the bioenergetics‐based version. The two approaches predicted opposite directions of change in polychlorinated biphenyl (PCB) concentrations of stocked sport fish as their growth rates declined. Differences in the predicted PCB concentrations were as much as 450% when lake trout (Salvelinus namaycush) growth rates declined 60%. Assuming constant carbon assimilation may be acceptable for models of biomass when interest is only in one trophic level or where growth rates are static. However, constant assimilation efficiencies are inappropriate for models of trophic interactions (e.g., predictions of predation rates and prey sustainability), nutrient recycling, contaminants, trace metals, or changes in productivity.