Co., Alabama field sampled in 1986. There, extreme drought resulted in

retardation of plant growth, and sampling was discontinued in mid-August.

Each field was divided into 12 parts of equal area, and four random samples

taken within each.

 Sampling procedures were according to those established in previous

studies as being most appropriate for estimating nymphal and adult bigeyed bug

populations in soybean (Rudd and Jensen 1977, Shepard et al. 1974b, Turnipseed

1974). Descriptions of methods for sampling their populations in soybean are

contained in Irwin and Shepard (1980). The ground cloth method was employed.

Forty-eight, 1.8-m samples were taken in each field on each sample date when

their populations were in detectable numbers. This sample-unit size differed

from the 1.2-m size in the Waddill et al. (1974) study. For each sample, the

ground cloth was laid between two rows, the soybean plants from both sides

(0.9 m on each side) beaten onto the cloth, and the number of nymphs and

adults determined. Also, bases of the plants and adjacent soil were examined

for any bigeyed bugs.

 The variance/mean ratio was calculated for adult and nymphal sample

counts on each sample date in each field (Southwood 1978). These analyses

were performed by using Myers' (1978) FORTRAN program. Log-transformed means

and variances of adult and nymphal sample counts then were calculated for each

sample date in each field, and Taylor's (1984) power law relationship for each

field determined. A Taylor's power law relationship also was determined for

each year by combining data from all fields sampled in that year. All

Taylor's relationships were determined by using SAS programs (SAS Institute

1982a,b).