Predicting asp and pikeperch recruitment in a riverine redervoir.

Abstract

Recruitment of two species, asp (Leuciscus aspius) and pikeperch (Sander lucioperca), in a riverine reservoir was studied using a novel statistical approach. Both species are piscivorous and are stocked into reservoirs for biomanipulative purposes to reduce planktivore species. Long-term data series were used, but the number of potential predictors was high. Therefore, a novel informative statistical approach based on dimension reduction methods was applied. Quality of outputs was driven by sampling methods. Main factors affecting asp recruitment were zooplankton abundance, predator density and temperature. In terms of pikeperch fry measured with seine and trawls, the number of predators was the only important factor. Gillnets underestimate small fish and the data were unsuitable for statistical modelling.