SEQUENTIAL SAMPLING PROGRAMS FOR LEAFHOPPERS (HOMOPTERA, CICADELLIDAE) ON WINE GRAPES IN WASHINGTON

Wine grape vineyards in south central Washington were sampled to colle ct data to develop sampling programs for leafhoppers. Most of the vine yards had a mixture of two species: the western grape leafhopper, Eryt hroneura elegantula. Osborn, and the Virginia creeper leafhopper, E. z iczac Walsh. Both species had aggregated spatial distributions with E. ziczac the more aggregated of the two. Two sequential sampling progra ms based on Wald's sequential probability ratio test were developed fo r the two leafhopper generations for both, species combined. The sampl ing programs were validated by simulation using the Monte Carlo method . The maximum number of samples, which was limited to 50 or 100 sample s, had little effect on the operating characteristic curves. Reducing the maximum number of samples reduced the average number of samples bu t increased the percentage of terminal decisions. Terminal decisions, made after taking the maximum number of samples had unacceptable error levels, especially at means above the critical density. However, a te rminal decision indicated that leafhopper numbers were near the critic al density and sampling should be repeated in a few days. Alternativel y, sampling history or other information could be used to help make an informed pest control decision.