SPATIAL-DISTRIBUTION OF ADULT BEMISIA-TABACI (HOMOPTERA, ALEYRODIDAE)IN COTTON AND DEVELOPMENT AND VALIDATION OF FIXED-PRECISION SAMPLING PLANS FOR ESTIMATING POPULATION-DENSITY

We conducted studies to examine distributional patterns of adult Bemis ia tabaci (Gennadius) strain B (also referred to as Bemisia argentifol ii Bellows & Perring) ill cotton, Gussypium hirsutum L., and to develo p and validate a sequential sampling plan for estimating population de nsity. Adults were consistently more abundant on mainstem leaves from the top stratum of cotton plants than on mainstream leaves from the mi ddle and bottom strata. Counts on mainstem leaves from the top of the plant also had the lowest relative variation. Adults on the top stratu m of the plant were fairly uniformly distributed over leaves from main stem nodes 2-7 (terminal = node 1), but numbers of adults were highest and least variable on fifth-node leaves. Patterns of aggregation, as measured by Taylor's power law, did not differ among the top. middle. and bottom strata of cotton plants and were similar among the first si s mainstem leaves below the mainstem terminal. Ratios between counts o f adults on individual leaves from the top stratum of the plant and wh ole plant counts were variable and averaged (+/- SD) 0.075 +/- 0.071 B ased on fifth mainstem node leaves as the sample unit, we used Kuno's and Green's methods to develop fixed-precision sequential sampling pla ns. The underlying mean-variance models for these methods;and performa nce af the sequential stop lines were compared and evaluated using a r esampling simulation of independent data sets with means ranging from 2 to 50 adults per leaf. Compared with Iwao's mean crowding I egressio n, Taylor's power law was a less biased predictor of variance. As a re sult, Green's plan, on average, achieved the desired precision better than Kuno's plan even though neither plan consistently gave mean estim ates with tile desired precision. Further simulations provided prelimi nary adjustments in the stop lines for field implementation.