Compatibility of Insect Growth Regulators with Eretmocerus eremicus (Hymenoptera: Aphelinidae) for Whitefly (Homoptera: Alyerodidae) Control on Poinsettias

i t a c s s w p l b g The efficacy and cost of reduced release rates of the parasitoid Eretmocerus eremicus Rose and Zolnerowich (Hymenoptera: Aphelinidae) when combined with application of the insect growth regulator buprofezin were compared to those of a higher parasitoid release rate used alone for whitefly control (Homoptera: Aleyrodidae) on poinsettia (Euphorbia pulcherrima Willd. ex Koltz.). The trial was conducted in seven greenhouses in Methuen, Massachusetts from August through December 1997 and employed commercial poinsettia production practices. Two whiteflies species, Trialeurodes vaporariorum (Westwood) and Bemisia argentifolii Bellows and Perring (5 Bemisia tabaci [Gennadius] strain B), were present. Three treatments were examined: (1) E. eremicus used alone at a release rate of three females per plant per week (two greenhouses); (2) E. eremicus at an intermediate elease rate of two females per plant per week, comined with mid-season use of buprofezin (two applicaions, spaced 1 week apart, applied in weeks 9 and 10) two greenhouses); and (3) E. eremicus at a low release ate of one female per plant per week, combined with id-season use of buprofezin, applied as in treatment (two greenhouses). In addition, observations were ade in one additional greenhouse at the site, in hich the grower used pesticides for whitefly control. rior to the start of the trial, cuttings used for all reatments experienced some pesticide use, first bamectin during rooting and later buprofezin at poting to reduce whitefly numbers, which were initially ery high. At harvest, densities of live whitefly ymphs were not statistically different among the bilogical control treatments, indicating that a low arasitoid release rate combined with buprofezin was s effective as a higher release rate of the parasitoid sed alone. Nymphal densities in separate market amples (based on smaller sample sizes) showed diferences among treatments, but all treatments, includ-