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 #218264
Title: Analysis methods for site-specific tarnished plant bug

Author
item Willers, Jeffrey
item ZHAO, YONGFENG - MISSISSIPPI STATE UNIV
item WU, JIXIANG - MISSISSIPPI STATE UNIV
item HOOD, K - PERTHSHIRE FARM
item BASSIE, J - AGRICULTURAL CONSULTANT

Submitted to: National Cotton Council Beltwide Cotton Conference
Publication Type: Proceedings
Publication Acceptance Date: 1/8/2008
Publication Date: 4/1/2008
Citation: Willers, J.L., Zhao, Y., Wu, J., Hood, K.B., Bassie, J.R. 2008. Analysis methods for site-specific tarnished plant bug sampling data. National Cotton Council Beltwide Cotton Conference. p. 1222-1234.

Interpretive Summary: This study indicates that researchers and industrial investigators would benefit from applications of count model methods to analyze insect sample counts. These count models are improvements over confidence intervals or ordinary least squares regression methods whenever large numbers of zeros occur, the data exhibit a skewed distribution, or are not otherwise normally distributed. Choice of which count model to use is dependent upon whether or not the sample counts are over-dispersed or contain excessive numbers of zeros.

Technical Abstract: Geo-referenced field samples for the tarnished plant bug (TPB) (Lygus lineolaris Miridae:Heteroptera) from the third week of July during the 2004 cotton production season were analyzed by several statistical methods. Geo-referenced imagery of a complex of cotton fields was processed into a map of three cotton habitat categories (HABITAT) and used by two different samplers (OBSERVER) to select sample locations. Analyses of these data by traditional methods such as confidence intervals or ordinary least squares regression were compared to several regression methods specifically designed to analyze counts. The latter methods better model the TPB samples as functions of the categorical explanatory variables (OBSERVER and HABITAT) than the two traditional approaches. It is concluded that count model regressions, based upon a generalized linear model approach, provide a valuable suite of tools for the research or industrial entomologist.