Submitted to: National Cotton Council Beltwide Cotton Conference
Publication Type: Proceedings
Publication Acceptance Date: 1/4/2000
Publication Date: N/A
Citation: Interpretive Summary: In cotton production, there are many factors that can reduce crop yield. One important cause is insects. Insects that cause loss to the fruit are frequently more destructive than those that damage leaves, stems and roots. This study describes an approach toward improving insect control practices directed against fruit feeding insects by integrating remote sensing imagery with on the ground sampling efforts by observers (called cotton scouts). The ecological premise at work is that just as the sensors used to acquire the image detect differences in crop status, so do many species of insects (e.g., plant bugs, boll weevils, and noctuid moths) have the sensory and behavioral capabilities to respond to differences in cotton phenological development. As a result, the first areas of a crop to be attacked by an insect pest are likely to be those areas where the crop vigor is best. Since the image can distinguish between different areas of crop vigor, the image once classified with the use of the normalized difference vegetation index (NDVI), defines the different sampling strata for the consultant to use. When both imagery and scouting indicate a pattern in the spatial distribution of a cotton pest it is possible to develop a spatially registered map prescribing which areas in a field are to be sprayed or unsprayed. Ultimately, the GPS-GIS capabilities of equipment installed on a ground sprayer determine which areas of the field are sprayed or unsprayed for a particular cotton pest, like the plant bug.
Technical Abstract: With the price of everything in farming going up except the price of cotton, there is an urgency to develop ways to reduce rising costs of production. Spectral Visions (with the support of NASA's Commercial Remote Sensing Program (CSRP)) and the USDA-ARS are exploring ways to use remote sensing in a large-scale production mode to predict where certain cotton pests (i.e., tarnished plant bugs, Heteroptera: Miridae) would gather and feed. Plant bugs prefer to feed on tender new squares found, at least early in the season, in the more healthy regions of a field and easily detected with remote sensing. Once identified, these areas can be represented in a prescription map to guide spatially variable insecticide applications. These healthy areas are then rendered into an on/off prescription map to be loaded into the controller of a GPS equipped ground sprayer. As the sprayer tracks across the field, the GPS signals its location to the controller and determines if it is in an area to be treated. During the 1999 season, we achieved insecticide reductions of nearly 40%, saving money on insecticide costs and lessening the environmental impact.