Studying the spatiotemporal variation of the littoral fish community in a large prealpine lake, using self-organizing mapping

Abstract

One of the most fundamental feature of freshwater systems is the spatiotemporal structure of their communities. In the present study, we used an artificial neural network model, i.e., self-organizing mapping, together with alike- lihood ratio χ2 statistic for proportions to investigate the influence of each factor of a complex sampling scheme(i.e., site, year, month, and time of day) on the littoral fish community of Lake Constance (south Germany). Based on self-organizing mapping, four clusters of samples were defined characterized by distinct fish communities. The samples gathered in clusters 1 and 2 were significantly related to the factors month and time of the day, while those in cluster 3 were related to the factors month and site and those in cluster 4 to each of the four factors. The results are discussed with regard to the temporal patterns of species succession in lakes and their similarities with the spatial patterns observable in streams, the importance of plasticity with regard to the fish nycthemeral preferences, the partitioning of habitat at a large spatial scale and its importance for the coexistence of species, and the effects of the reoligotrophication process in lakes.