A technique for subsampling an abundant taxon while completely sorting other taxa1

Abstract

A method is described which allows an abundant taxon (Nematoda) to be subsampled while rarer taxa are completely sampled. A triply balanced F ‐square design is used to choose a quarter of the sorting field for counting the nematodes. The technique showed no evidence of bias in estimating total numbers of nematodes, and a quarter subsample was accurate to within ±0.8% when there were > 200 nematodes in a sample. Estimates of the total number of nematodes in a sample were accurate to within ± 3.2% of their actual values. The technique gave subsamples of abundant nematode species which did not differ significantly from expectation.