Skip to main content
ARS Home » Southeast Area » Mississippi State, Mississippi » Crop Science Research Laboratory » Genetics and Sustainable Agriculture Research » Research » Publications at this Location » Publication #96731


item Willers, Jeffrey
item Akins, Dennis

Submitted to: Journal of Cotton Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/15/1998
Publication Date: N/A
Citation: N/A

Interpretive Summary: Sampling for plant bugs at low population levels is important for developing better management tactics against this major cotton pest. The results of this study demonstrate a sampling technique that is efficient and easy to use and measures low population levels of plant bugs. The method is based on the use of the drop cloth, a readily available and familiar scouting tool. The drop cloth can be improved in its sensitivity to detect low numbers of plant bugs by arranging a series of samples into a straight line at least 8 rows long. Total numbers of plant bugs shaken onto the cloth along the line can be converted to numbers per acre using a simple formula, or a published look-up table of constants that can be applied to any row spacing of solid planted cotton, except ultra-narrow row. It was also demonstrated that improved understanding of the sample data can be achieved if remotely sensed image maps are available. Using these images on maps, differences in crop growth patterns throughout the field can be quickly distinguished. By sampling different areas of the field identified on the image map, it was demonstrated that plant bug densities differed by crop growth stage. Used together, the potential exists for reducing the sample time and effort necessary to detect and monitor plant bug abundance in cotton and yet maintain high data integrity of sample data necessary for making appropriate management decisions against this pest.

Technical Abstract: This study demonstrates a modification of line-intercept sampling (LIS) along with multi-spectral remotely sensed images for estimating the density of Lygus adults and nymphs in cotton. Better estimates of plant bug abundance result when samples are stratified according to differences in crop phenology that are determined from remote sensing image maps. Samples are collected along transect lines randomly positioned at right angles to imaginary reference baselines that lie parallel to the row direction. Ideally, the transect line and baseline lengths are selected to define a reference area that comprises one land acre, or for enhanced brevity of sample effort, quarter-fractions thereof. Sample information is collected from constant size quadrats intercepted by the transect lines. Quadrats correspond to the size of a standard drop cloth and comprise a series of adjacent sample units along the transect line. Additionally, Bayesian methods based on the hypergeometric distribution can be employed with this sampling design and technique to build a probability distribution of the numbers of plant bugs per acre. With Bayesian methods, the median, mode and selected percentiles of interest can be readily determined from the posterior distribution of population estimates. The mode obtained by Bayesian methods is equivalent to the traditional LIS estimate of the mean. Examples based on information collected from a large Mississippi Delta cotton field during 1997 are presented.