Bioenergetics Modeling Today: Growing Pains on the Cutting Edge

Abstract

Elaborations of the mass-balance equation that partitions energy of consumed food into its various physiological fates have flourished in the past 15 years. These bioenergetics models have been converted to powerful simulation tools and used in fisheries science, most often for predicting consumption by predators or for projecting fish growth as a function of temperature and prey availability. New uses of bioenergetics models are proliferating and range from predictions of larval survival to analysis of ecosystem function. Companion technologies such as hydroacoustics and physiological telemetry promise to further broaden the spectrum of potential applications. However, the data demands and uncertain accuracy of multiparameter models may constrain their ability to address routine fish management questions. Field measures of consumption or growth failed to corroborate bioenergetics estimates consistently (differences of at least 50%) in four of six published tests. Poor agreement could be due to inaccuracies in field measures, bioenergetics estimates, or both. Errors in bioenergetics estimates will accrue when input values for internal parameters and external variables are deficient; problems can result from unknown activity costs, extrapolation of weight-dependent power functions, unjustified borrowing of physiological values from other species, and use of nonrepresentative field data. Remediation of these deficiencies will require a major concerted research effort and extensive field corroboration. Simpler individual and production-based models show promise for predicting long-term consumption or growth but also require substantiation of input parameters and more corroboration. In the interim, I suggest that bioenergetics models will remain better suited for making relative comparisons than for making precise quantitative predictions.