Comparison of Sample Methods and Sample Plans for Thrips in Cotton

Thrips are a serious early-season pest of cotton throughout much of the U.S. cotton belt, and have been demonstrated to cause a 21% average yield loss to irrigated cotton on the Texas High Plains.  Currently, much of the cotton on the Texas High Plains is protected from thrips damage preventively by the use of in-furrow and seed-applied insecticides.  However, where thrips are not perennial pests, preventive insecticide use may not be necessary and foliar curative actions may be more economical.  Currently, in Texas a reliable and cost effective sampling plan for thrips management decisions is not available.  Thus, the objectives of this study were to develop then compare sample plans (enumerative vs binomial) and sampling methods (visual vs cup) to determine the most cost reliable sampling plan for thrips decision making in cotton.

Conclusions

1. Taylor's power law effectively modeled the thrips sample data from both sample methods.  Taylor's coefficients suggest that thrips  nymphs tend to be more clumped than adult thrips regardless of sample method
2. The relationship between the proportion infested cotton leaves and thrips mean density was also modeled well by using the method of Wilson and Room (1983). 
3. Development of the sample plans indicated that the binomial sample plan, regardless of sample method, required significantly fewer samples to make a management decision relative to the enumerative sample plan.
4. Sample size requirements between the sample methods for the binomial sample plan, although similar, favored the cup sample method as it required only 90% of the effort of the visual sample plan.