Nineteen different experiments were compiled to examine temporal trends in fruit retention among various insect related treatments based on end-of-season yield mapping of cotton, Gossypium hirsutum L., plants. Fourteen of these data sets were used to examine the effect of varying densities of bollworm, Helicoverpa zea (Boddie), tobacco budworm, Heliothis virescens (F.) (heliothines) and tarnished plant bug, Lygus lineolaris (Palisot de Beauvois), on survival and seed cotton weight of particular fruit cohorts at harvest. Regression equations using only sample dates where insects were present described end-of-season fruit loss better than equations using all insect sample dates. More fruit loss occurred when populations of heliothines were observed on cotton varieties not expressing an insecticidal protein from Bacillus thuringiensis Berliner (Bt) than Bt varieties. Populations of heliothines observed on non-Bt cotton at six different stages of plant development (based on total mainstem nodes) were related to decreased survival of fruit within some fruiting cohorts. Heliothine eggs and larvae on Bt cotton were related to reduced fruit survival when infestations were present on plants with four and three different total mainstem nodes, respectively. Based on these regression analyses, most damage caused by observed populations of heliothine larvae and plant bugs occurred to cotton squares ranging from 3 to 15 d old at the time of infestation. Collectively, these data indicate that important insect injury can be followed through the growing season and recorded on end-of-season yield maps. The dynamic nature of the impact and the probable role of plant compensation further support continued development of dynamic insect thresholds.
