Abstract

The stomach content information of common hake was analyzed between 1992 and 2003, to know the trophic dynamic of this specie in the south-central Chile. The importance of each prey taxon prey is evaluated in the diet of Merluccius gayi gayi, using an artificial neural network of backpropagation that incorporates as variables: year, month, total length and weight. The analysis shows a high association between the observed values and predictions for the neural network, is this the first application carried out in the groundfish system of south-central Chile and showed the possibility to carry out predictions of the diet of this specie, demonstrating that the neural network can be used for the analysis and interpretation of trophic data in fishes.