Robust Binomial Decision Rules for Integrated Pest Management Based on the Negative Binomial Distribution

Publisher Website
Google Scholar
Abstract
It is time consuming and expensive to count all pests on each sample unit in an integrated pest management (IPM) program to estimate mean abundance of pests. Binomial sampling using only the percentage of leaves containing any pests has been recommended on many occasions as a viable short cut, although, as shown here, this procedure can be misleading. It is possible, however, to set up a binomial procedure based on the percentage of samples with more than T pests (T > 0). If the pest species follows the negative binomial distribution, there is an optimal value of T for which the binomial scheme is most robust with respect to the value (within a prespecified range) of k, the second parameter of the negative binomial distribution. The procedure is exemplified by using IPM data for tetranychid mites in Quebec apple orchards.

This publication has 0 references indexed in Scilit: