The truly enormous population sizes and geographic ranges of many zooplankton species present a daunting sampling problem. Statistically sound sampling strategies must also take into account the lack of geographic and temporal stability of populations and the various spatial scales of interest. Adequate sample sizes depend in part on allele frequencies in the sample: more individuals must be assayed in order to ensure that rare alleles are sampled. As an example, 51 individuals must be assayed to attain a 95% probability of detecting an allele of frequency 0.10 [ Gregorius, 1980]. Replicate samples at some stations are a useful means of evaluating sampling error, and frequent sampling at some locations will help to establish any temporal variation in the genetic character of the samples at a given geographic location. A nested sampling approach is frequently employed, to reveal small scale structure and to avoid aliasing at particular spatial scales. Frequently, genetic heterogeneity is observed at small scales and over short time periods that does not confound the significant patterns seen at larger spatial scales (see below).