Numeric analysis of planktonic spatial patterns revealed by echograms

Abstract

Observations of acoustic images made along a transect perpendicular to the Coast, off Villefranche-sur-Mer, allow one to recognize the hydrological zone corresponding to a surface density gradient characterizing the Ligurian Sea front (Mediterranean Sea). Combinations of statistical and geometrical parameters from the digitized echoes enable discrimination between different classes of patches in relation to the frontal structure during one transect recorded on December 20, 1990. Patch recognition is based on the extraction of the general trend of each column and each row of the digitized data matrix by Eigenvector filtering. Multivariate analysis (principal component analysis and discriminant analysis) run on a 650 patches × 19 variables matrix shows a high negative relation between statistic and geometric parameters of the patches. The two main discriminant parameters are shape and echo strength of patches. From the convergence of our results in the Ligurian front with those of previous works made in other frontal region, it is suggested that this method could be generalized for quantifying and standardizing studies of the patchiness of plankton.