The accuracy of population density estimates of a horizontally distributed zooplankton community in Dutch fish ponds

Abstract

To elucidate the effect of the horizontal distribution of different zooplankton populations on the accuracy of population density estimates in a fish pond, 40 samples were taken according to a rectangular grid which divided the pond into 5 longitudinal rows and 10 transversal columns. The zooplankton consisted of 55% copepods (mainly Acanthocyclops viridis), 43% cladocerans (mainly Daphnia longispina and Bosmina longirostris) and 2% Rotatoria. The index of patchiness was mostly higher than 1, revealing a strongly clumped distribution. B. longirostris and D. longispina aggregated in the centre of the pond, exhibiting a shore avoiding behaviour. Chydoridae and Ceriodaphnia quadrangula were concentrated in the littoral zone. Besides this habitat related dispersion, a more passive wind induced distribution was detected for most taxonomic groups. Copepod nauplii were concentrated at the leeward side of the pond while elder copepod stages of A. viridis were found in higher densities at the windward end. Based on the calculated coefficients of variation, a table of the accuracies of the density estimates in relation to the number of samples was presented. For more detailed studies on a particular species population, about 25 samples (column samples of 10 L each) are needed to obtain an accuracy varying between 10 an 20%, depending upon the taxonomic group. A slightly higher precision can be obtained, however at a strongly increasing effort. For the practice of fish farming, two to five sample stations should suffice to estimate the total zooplankton abundance with a 30 to 10% accuracy respectively.