SAMPLING STATISTICS AND A SAMPLING SCHEME FOR THE TWOSPOTTED SPIDER MITE, TETRANYCHUS URTICAE (ACARI: TETRANYCHIDAE), ON STRAWBERRIES

Abstract

A series of samples of the twospotted spider mite, Tetranychus urticae Koch, from 16 experimental populations on “Totem” strawberry, Fragaria × ananassa Duch., were examined to determine the index of dispersion, the variance–mean relationship, the distribution, and the relationship between the mean number of T. urticae/leaflet and the proportion of leaflets without T. urticae. The slope of the variance–mean relationship, 1.64 ± 0.0355(SE), did not differ between sample dates but the intercept decreased significantly (p < 0.01) from 2.56 ± 0.0969 before harvest to 1.76 ± 0.140 during harvest. At 0.0136 T. urticae/leaflet before harvest and 0.0644 T. urticae/leaflet during harvest, the variance equaled the mean, implying that at these densities the data followed the Poisson distribution. Above these densities the data were overdispersed, most samples conforming with the negative binomial distribution but some tending towards greater dispersion than the negative binomial. There was no common k for the negative binomial nor did the data fit the expectation for 1/k that was consistent with the variance–mean relationship. A distribution-free sampling scheme based on the sample mean and the proportion of leaflets without T. urticae () was developed. Tetranychus urticae density can be quickly determined in the field using the naked eye, by iteratively observing leaflets for the presence or absence of T. urticae and referring to a small table that gives both the mean density for a given and the number of leaflets required to obtain a specified standard error.