The use of statistical classifiers for the discrimination of species of the genus Gyrodactylus (Monogenea) parasitizing salmonids

This study applies flexible statistical methods to morphometric measurement s obtained via light and scanning electron microscopy (SEM) to discriminate closely related species of Gyrodactylus parasitic on salmonids. For the fi rst analysis, morphometric measurements taken from the opisthaptoral hooks and bars of 5 species of gyrodactylid were derived from images obtained by SEM and used to assess the prediction performance of I statistical methods (nearest neighbours; feedforward neural network; projection pursuit regress ion and linear discriminant analysis). The performance of 2 methods, neares t neighbours and a feed-forward neural network provided perfect discriminat ion of G. salaris from 4 other species of Gyrodactylus when using measureme nts taken from only a single structure, the marginal hook. Data derived fro m images using light microscopy taken from the full complement of opisthapt oral hooks and bars were also tested and nearest neighbours and linear disc riminant analysis gave perfect discrimination of G. salaris from G. derjavi ni Mikailov, 1975 and G. truttae Glaser, 1974. The nearest neighbours metho d had the least misclassifications and was therefore assessed further for t he analysis of individual hooks. Five morphometric parameters from the marg inal hook subset (total length, shaft length, sickle length, sickle proxima l width and sickle distal width) gave near perfect discrimination of G. sal aris. For perfect discrimination therefore, larger numbers of parameters ar e required at the light level than at the SEM level.