Progress in modeling quality in aquaculture: an application of the Self-Organizing Map to the study of skeletal anomalies and meristic counts in gilthead seabream (Sparus aurata, L. 1758).

One of the most common drawbacks of artificial life conditions imposed by aquaculture is the quite high presence of skeletal anomalies (SAs) in reared fish, which reduce both functional performances and marketing image ⁄ commercial value of the reared lots. Thus, skeletal malformations and their incidence are one of the most important factors affecting fish farmer s production costs, and several efforts have been due to develop appropriate tools in detecting patterns of co-variation among rearing parameters and fish quality. In this paper we explore the advantages of using Self-Organized Maps (SOMs) when dealing with the analysis of correlations between the pattern of SA presence and rearing parameters in gilthead seabream (Sparus aurata L.), that is a largely reared fish of high commercial value. SOM, which is one of the best known neural networks with unsupervised learning rules, were applied to develop a model of the occurrence of SAs, both in terms of type and quantity, in seabream lots from different rearing approaches (extensive, semi-intensive and intensive). The trained SOMs classified lots according to the variation observed in the different weights of SAs, but also allows the detection of a series of correspondence, namely between: (i) the patter of SAs occurrence and the different rearing approach currently used in seabream aquaculture; and (ii) the total SAs incidence and the variability of meristic counts, represent a completely independent dataset. Mesocosms resulted the best rearing approach to produce wild-like fish, whereas intensive rearing is characterized by the large presence of SA. Globally, results suggested that this approach is reliable to be used for estimate the distance between aquaculture products and the wild-like phenotype used as quality reference.