SPATIAL-DISPERSION AND SAMPLING OF CITRUS RUST MITE (ACARI, ERIOPHYIDAE) ON FRUIT IN HAMLIN AND VALENCIA ORANGE GROVES IN FLORIDA

Two systematic sampling designs, an area plan (n=600: five areas, five trees per area, 12 fruit per tree, two sample units per fruit) and a transect plan (n = 192: two transects, 12 trees per transect, four fru it per tree, two sample units per fruit), were used to estimate popula tion densities of citrus rust mite, Phyllocoptruta oleivora (Ashmead), at motile stages on fruit in 4-ha blocks of ''Hamlin' and 'Valencia' orange trees. The sample unit was a 1-cm2 surface area anywhere on a f ruit except where maximum or minimum sun exposure occurred. Citrus rus t mites usually were aggregated within trees, among trees and among ar eas across 4 ha but not between transects. Aggregation often was prono unced but only when mite densities were small. Assessments of the prec ision of estimates from the two sampling methods indicated each provid ed an average of 25% or less relative variation in estimates when dens ities were above about three motile mites per cm2. Density estimates u sing the two sampling methods were the same 69 and 66% of the time bas ed on untransformed and log10-transformed data, respectively. Among es timates that differed, neither sampling method consistently under- or overestimated densities; mean densities were less than one mite per cm 2 70-80% of the time that estimates from the sampling plans differed. Variance components indicated that smaller sample sizes than used woul d provide relatively good estimates depending on the density of mites being measured. Sample sizes needed to achieve a desired level of prec ision per 4 ha decreased as mite densities increased, Guidelines for s electing a sampling plan for the citrus rust mite are presented. One s ample unit per fruit from each of four fruit equally spaced around a t ree on 20 trees scattered across a 4-ha block was projected to provide an average of 25% or less relative variation in estimates when densit ies averaged 10 or more motile mites per cm2.