Laboratory Errors Associated With the Estimation of the Population Density of Heterodera Species in Soil

As unexpectedly large sampling errors were obtained in preliminary population studies on Heterodera , an examination was made of the laboratory errors associated with sampling for H. göttingiana and H. schachtii. Soil samples were taken from microplots and appeared to be satisfactory, being without the usual heterogeneity found in sampling from fields. There was little evidence of errors introduced by the technique used for estimating the number of eggs and larvae in a subsample. These errors, when they occurred, were always small, and this standard technique was deemed satisfactory. The logarithmic transformation was suitable for statistical analysis of both cyst and egg counts. Apart from the residual variation, there were two other important sources of variation when sampling for eggs. There was a large variation between sub‐samples, caused by differences in the number of cysts and in the number of eggs per cyst. There was also in some instances a difference between observers counting the same eggs of up to 20%. Where a comparative measure of density is sufficient, this bias is probably unimportant, and to increase accuracy of estimates of eelworm density, most laboratory work should be devoted to the separation of cysts from the soil sample.