Predicting food consumption and production in fish populations : allometric scaling and size-structured models

Abstract

Life-history traits in fish populations are highly correlated. A subset of these correlations are called allometric scaling, they refer to biological processes which can be described using body size as independent variable. Particularly, allometric scaling related with food consumption (Q) and biomass production (P) has gained the attention of ecologists for several decades. This thesis proposes a quantitative framework for food consumption, which allows both the identification of the mechanisms underlying the allometric scaling for Q and the development of a predictive model for consumption to biomass ratio (Q/B) in fish populations. This thesis is based on the fact that food consumption can be inferred from first principles underlying the von Bertalanffy growth model. In addition, it has been noticed in the literature that biomass production and food consumption show similar allometric scaling dependence, therefore, both can be derived from these first principles. Thus, a similar quantitative framework was used to produce models for P/B in fish populations. Once functional forms for production and food consumption were identified, a third model was developed for the ratio between production and consumption (P/Q). This ratio is usually named ecological efficiency because it determines how efficiently a population can transform ingested food into biomass. Several authors have noticed that P/Q remains invariant (independent of body size) across species. From a theoretical point of view, the results presented here allow the first quantitative explanation for the existence of the allometric scaling for Q/B and the invariance of P/Q across fish species. These results, together with the explanation for allometry in P/B reported in the literature, suggest that the regular across-species pattern for the trio {P/B,Q/B,P/Q} can be explained by basic principles that underpin life-history in fish populations. This quantitative framework for the trio {P/B,Q/B,P/Q} is based on an explicit dependence with body size, which simplifies the estimation of these quantities. Model complexity depends, in part, on which data are available. Models were applied to real data from commercially important species fished in Chile. Statistical properties of the new models were evaluated by an intensive resampling approach. The simplest possible model for the trio {P/B,Q/B,P/Q} rests on the assumption of a stable age distribution. These quantities have a key importance in ecosystem modelling because they determine population energetics in terms of food intake by predation and the transformation of this energy into population biomass of predators. Application of the new models produces results which were comparable to those given by standard methods. This thesis is a result of multidisciplinary research which attempts to make a contribution to the understanding of the mechanisms underlying the allometric scaling of food consumption and production in fish populations. It proposes models for the trio {P/B,Q/B,P/Q} and thus, has the potential to be widely applicable in fisheries science.