Use of quantile regression and discriminant analysis to describe growth patterns in farmed gilthead sea bream (Sparus aurata)

Most mathematical models of fish growth on commercial farms use the evolution of average weight during the growth period, without considering weight distribution. In this paper, quantile regression techniques were used to describe the evolution of weight distribution in 20 batches of gilthead sea bream (Sparus aurata) growing on a commercial fish farm in the Spanish Mediterranean. Different Thermal-unit Growth Coefficients (TGC) were obtained for several quantiles (5, 25, 50, 75 and 95%) in a group of heterogeneous batches, whereas a single TGC was observed in another group of homogeneous batches. Thus, the growth patterns observed were either homogeneous or heterogeneous. Discriminant analysis helped to identify variables that affected the classification of batches, namely, the coefficient of variation in initial weight and daily average temperature variation during the initial growth period.