Mapping Cotton Root Rot Infestations with Airborne Multispectral Imagery

Abstract

Cotton root rot, caused by the fungus Phymatotrichum omnivorum, is a serious and destructive disease that significantly reduces cotton yield and lowers lint quality. Cultural practices are commonly recommended for the control of cotton root rot, while fungicides and fumigants that may suppress the disease have been used. Because of the high costs of these chemicals, their use may be economically feasible only when the infested portions of the field are treated. The objective of this study was to evaluate airborne multispectral imagery for detecting and mapping root rot infestations in cotton fields for site-specific management of the disease. One center pivot-irrigated field and one rainfed field near Corpus Christ, Texas were selected for this study. Airborne multispectral digital imagery was taken from the two fields shortly before harvest in 2001 when the infested areas with wilted and dead plants were fully pronounced for the season. The imagery was georeferenced and then classified into healthy and infested areas using unsupervised classification. Accuracy assessment on the classification maps for the two fields indicated airborne imagery effectively and accurately identified root rot infested areas within the fields. Ground samples taken from the fields showed that cotton yield and some lint quality indices were significantly lower in infested areas than in healthy areas. Buffer zones around the infested areas were generated to account for the spread of cotton root rot on the root rot maps. The mapping procedures and maps presented in this study will be useful for site-specific management of the disease.