|ENGLISH, PATRICK - 6066-35-00
Submitted to: National Cotton Council Beltwide Cotton Conference
Publication Type: Proceedings
Publication Acceptance Date: 1/7/2009
Publication Date: N/A
Technical Abstract: Precision-based agricultural application of insecticide relies on a non-random distribution of pests; tarnished plant bugs (Lygus lineolaris) are known to prefer vigorously growing patches of cotton. Management zones for various crops have been readily defined using NDVI (Normalized Difference Vegetation Index), apparent bulk electrical conductivity (ECa - Veris), and yield data. The objective of this study was to evaluate the interactions of field-scale heterogeneity as detected by NDVI, ECa, yield, and thermal imagery and their effectiveness in delineating management zone boundaries and transitions for the site-specific control of L. lineolaris in cotton when datasets are acquired at various spatial resolutions. Non-autocorrelated areas represent highly heterogeneous patches and define transitional areas between management zones. These non-autocorrelated areas of transition represent potential refugia and loci for reinfestation of the field by TPBs. Spatial analysis of imagery (NDVI, ECa, yield, and thermal) at 1 and 2 meter resolution showed a 15% increase in area of the non-autocorrelated regions for the lower resolution imagery; these regions have a greater risk of misclassification for scouting. Spatial autocorrelation analysis provides a very effective method to delineate these high risk areas. This analysis represents a useful tool to augment the accuracy of scouting maps and facilitate the site-specific deployment of insecticides.