Site-specific agriculture uses maps to optimize within-field placement of agricultural practices. This technology introduces the potential to optimize pest management by varying pesticide or other inputs to better match within-field variation in pest density. Current sampling plans are designed to estimate mean density and may not be suitable for mapping, although useful sampling plans could be developed for map generation for integrated pest management. Using Colorado potato beetle, Leptinotarsa decemlineata (Say), adults, larvae, and egg masses as model systems, the influence of the sample unit on map validity was explored. Adapting currently used sampling plans for potato integrated pest management by spatially referencing each sampled stem failed to resolve spatial dependence and resulted in maps with poor reliability. Increasing the sample unit improved resolution of spatial dependence and map reliability for each life stage. A distance-walk sample unit for adult and late instar Colorado potato beetles which has high potential for map generation is introduced. Using this sample unit, generalizations about Colorado potato beetle spatial dependence are made to discuss issues of developing sampling programs for map generation. An iterative process of sampling, spatial analysis, and error analysis is suggested for evaluating sample units for mapping pest density in high value crops.
All Science Journal Classification (ASJC) codes
- Insect Science