Sequential Sampling Plan for the Armyworm1 in Maryland Small Grains2

Publisher Website
Google Scholar
Abstract
In the spring of 1979, wheat and barley fields in Maryland were monitored to determine the spatial distribution of the armyworm, Pseudaletia unipuncta (Haw). The discrete frequency data for this insect were found to statistically fit the negative binomial distribution. By using a method which compared the sample sizes in terms of time and reliability, an appropriate sample unit size was determined to be 0.9 m of row. The parameters from the mean crowding (m*) and mean density (m) regression were used to develop a sequential sampling plan for the armyworm. In 1980, fixed samples were taken from 23 wheat and barley fields in Maryland. Simulations of the sequential sampling plan were practiced on the larval counts correctly predicting a treat or not-treat decision for 88% of the cases. Overall, an average of 12 sample units were required to reach a decision. This represented a 40% saving of time when compared with the fixed-sample number currently used for small-grain pest management.

This publication has 1 reference indexed in Scilit: