Automatic classification of field-collected dinoflagellates by artificial neural network

Automatic taxonomic categorisation of 23 species of dinoflagellates was demonstrated using field-collected specimens. These dinoflagellates have been responsible for the majority of toxic and noxious phytoplankton blooms which have occurred in the coastal waters of the European Union in recent years and make severe impact on the aquaculture industry. The performance by human 'expert' ecologists/taxonomists in identifying these species was compared to that achieved by 2 artificial neural network classifiers (multilayer perceptron and radial basis function networks) and 2 other statistical techniques, k-Nearest Neighbour and Quadratic Discriminant Analysis. The neural network classifiers outperform the classical statistical techniques. Over extended trials, the human experts averaged 85% while the radial basis network achieved a best performance of 83%, the multilayer perceptron 66%, k-Nearest Neighbour 60%, and the Quadratic Discriminant Analysis 56%.

Feeding of anchovy Engraulis encrasicolus larvae in the northwestern Adriatic Sea in response to changing hydrobiological conditions

Results from depth integrated and vertically stratified plankton sampling in the northwestern Adriatic Sea were used for comparison of gut contents of larvae of European anchovy Engraulis encrasicolus with composition and concentration of potential prey in the plankton. Sampling was carried out over a grid of stations both before and after a period of increased wind mixing to investigate changes in food availability and larval feeding success. All larvae had empty guts soon after dusk, indicating daytime feeding and rapid gut clearance. With increasing larval length there was a greater percentage of specimens with empty guts, despite suitable food being available in the plankton for these larger larvae; this suggests differential gut evacuation during sampling-possibly related to the degree of gut development. Larval diet was principally the various developmental stages of copepods, especially calanoid and cyclopoid nauplii, which were preferentially selected by larvae, whereas selection was against harpacticoid nauplii. Lamellibranch larvae and Peridinium were generally abundant in the plankton, but were only present in the gut contents in any number when the preferred dietary organisms were present in the plankton at low concentrations. The number of food organisms in the gut contents increased with concentration of the preferred food organisms in the plankton up to a limit of similar to 50 organisms/l. Within the upper 18 m of the water column, there was a reduction in the proportion of larvae with food in their guts with increasing depth, irrespective of the vertical profile of food concentration. Following a period of wind mixing the composition of the plankton changed. This was reflected in the diet of anchovy larvae, which altered in parallel. There was also an overall 41% decrease in concentration of the preferred food particles of larvae in the plankton following the period of wind mixing, but larvae were still able to maintain their food intake. These results show that anchovy larvae can successfully adapt their diet to a changing prey field and suggest that in the conditions observed in the northern Adriatic, quite radical changes in the feeding environment were probably insufficient to affect overall larval mortality.