Consistency in the correlation of school parameters across years and stocks

Abstract

Acoustic surveys have been undertaken in different parts of the waters around Europe. In this paper we consider those taking place in NE Atlantic waters – off the NE Scottish coast, the Bay of Biscay and off the Spanish Atlantic coast – and in the Mediterranean Sea – the Catalonian Sea, the Gulf of Lion and the Aegean Sea. Retained school variables were: corrected school length (m), school height (m), school area (m2), school perimeter (m), school energy (Nautical area scattering coefficent, SA), school density (energy vs. area), school depth (m) and bottom-depth-under-school (m). STATIS (Structuration de Tableux à Trois Indices de la Statisque), a multi-table analysis based on the Principal Component Analysis was used to analyse the correlation in different schools, parameters and intra- and interstock relationships. In all stocks, the first three axes extracted explained between 86.89% (Sete) and 91.84% (Aberdeen) of the total variance, because the annual variability present in the correlation structure of each stock is not widely dispersed. In all stocks, the first factorial axis was typified by the morphological variables (school length, school height, school area and school perimeter), which showed a very clear pattern with all of them being closely related. There were major differences between all stocks in regard to the general correlation relationship of the energetic variables analysed i.e. school energy and school density. Energy and density showed less variability than the morphological variables in all stocks with the exception of one survey. There were some differences between stocks in relation to the correlation of bathymetric variables. In all cases, school density and, to a lesser extent, school energy were opposed to the bathymetric variables, so that schools with higher densities would be located near the surface and in shallower areas. A complete analysis of all the surveys of each stock showed a homogeneous pattern which was very similar to most of the stocks considered separately. All morphological variables are well grouped and show a strong positive correlation. In general, therefore, all the schools analysed increased and decreased equally in all their dimensions. Both bathymetric variables are strongly correlated.