Effect of interreplicate variance on zooplankton sampling design and data analysis1

The variance of replicate marine and freshwater zooplankton samples is related (r2 > 0.99) to the mean population density and the volume of the sample taken. A general variance relationship calculated from 1,189 sets of replicate samples taken throughout the world is presented. This function can predict the variance of both marine and freshwater samples taken with several types of sampling gear. Only 12% of individual taxa examined showed variance relationships that departed significantly from this general function. The variance function shows how the requisite number of samples decreases with increased population density, sampler volume, and lowered precision requirements. The X0.2power transformation is recommended on empirical grounds, although caution is advised in the use of all data transformations.