A Field‐based Bioenergetics Model for Estimating Time‐Varying Food Consumption and Growth

Abstract

AbstractBioenergetics models are often used to describe the implications of changes in growth and consumption of specific wild populations, and yet most parameters are derived from a variety of laboratory studies on other populations or species, leading to questions regarding the validity of predictions. A novel bioenergetics approach was recently developed where many parameters are estimated from the population being modeled, but growth and consumption are assumed invariant over time, which would not hold true when manipulations to the system are known or suspected. In the present paper, a bioenergetics model with many key parameters estimated from field data are presented where temporal deviations in growth rates were directly estimated. A series of rainbow trout Oncorhynchus mykiss and northern pikeminnow Ptychocheilus oregonensis populations, which have undergone various population manipulations, were used to evaluate the model. Further, the model was fit to a series of rainbow trout size‐classes stocked into each of the study lakes to compare with their wild counterparts and evaluate intercohort differences in growth and consumption. We found the model with time‐varying consumption was more parsimonious compared with models where growth and consumption were assumed to be constant over time. Our field data demonstrated how the model can detect different patterns in growth and consumption across populations and species. The model detected highly variable growth and consumption in rainbow trout over time and between populations but did not seem to be particularly influenced by past population manipulations. By contrast, northern pikeminnow demonstrated differences between lakes, but showed little temporal variation in growth and consumption. Stocked rainbow trout demonstrated similar growth rates to their wild counterparts, helping to validate growth estimates. Our bioenergetics model moves beyond existing ones by allowing measurement and process errors to be explicitly represented, while also permitting growth and consumption to vary over time.