Spatial Distribution of Adult Bemisia tabaci (Homoptera: Aleyrodidae) in Cotton and Development and Validation of Fixed-Precision Sampling Plans for Estimating Population Density

Abstract

We conducted studies to examine distributional patterns of adult Bemisia tabaci (Gennadius) strain B (also referred to as Bemisia argentifolii Bellows & Perring) in cotton, Gossypium hirsutum L. , and to develop and validate a sequential sampling plan for estimating population density. Adults were consistently more abundant on mainstem leaves from the top stratum of cotton plants than on mainstem leaves from the middle and bottom strata. Counts on mainstem leaves from the top of the plant also had the lowest relative variation. Adults on the top stratum of the plant were fairly uniformly distributed over leaves from mainstem 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 six mainstem leaves below the mainstem terminal. Ratios between counts of adults on individual leaves from the top stratum of the plant and whole plant counts were variable and averaged (± SD) 0.075 ± 0.071. Based on fifth mainstem node leaves as the sample unit, we used Kuno's and Green's methods to develop fixed–precision sequential sampling plans. The underlying mean-variance models for these methods and performance of the sequential stop lines were compared and evaluated using a resampling simulation of independent data sets with means ranging from 2 to 50 adults per leaf. Compared with lwao's mean crowding regression, Taylor's power law was a less biased predictor of variance. As a result, Green's plan, on average, achieved the desired precision better than Kuno's plan even though neither plan consistently gave mean estimates with the desired precision. Further simulations provided preliminary adjustments in the stop lines for field implementation.