RESISTANCE RISK ASSESSMENT - REALIZED HERITABILITY OF RESISTANCE TO BACILLUS-THURINGIENSIS IN DIAMONDBACK MOTH (LEPIDOPTERA, PLUTELLIDAE), TOBACCO BUDWORM (LEPIDOPTERA, NOCTUIDAE), AND COLORADO POTATO BEETLE (COLEOPTERA, CHRYSOMELIDAE)

A new method for estimating realized heritability (h2) of resistance to insecticides from laboratory selection experiments is described. A major advantage of this method for estimating h2, the proportion of total phenotypic variation attributable to additive genetic variation, is that it requires only data that are collected routinely in selection studies (i.e., LC50 and slope before and after selection, and average mortality caused by selection each generation). Estimates of h2 of resistance calculated with this method were virtually identical to two previously reported estimates calculated by a regression method that requires estimates of LC50 for every generation. Estimates of h2 of resistance to Bacillus thuringiensis in Plutella xylostella (L.) were 0. 14, 0.17, and 0. 18 in three strains initiated from a moderately resistant field population in Hawaii compared with 0.047 in a laboratory strain from France. Estimated h2 of resistance to B. thuringiensis was 0.17 in Helicoverpa virescens (F.) and 0.09 in Leptinotarsa decemlineata (Say). Extrapolation to field populations is problematic. However, assuming that these h2 estimates can be extrapolated to the field and that populations receive prolonged and uniform exposure to B. thuringiensis causing > 90% mortality each generation, substantially increased resistance would be expected after < 10 generations for H. virescens and < 15 generations for L. decemlineata. In six cases examined here, h2 was higher in the first half of selection experiments (mean = 0.38) than in the second half (mean = 0.15). This suggests that brief selection experiments (4-6 generations) may most efficiently detect the potential for resistance development. Estimation of h2 of resistance based on the method described here provides a means for systematic, quantitative analysis of selection experiments, including many of the > 150 previously reported studies. This approach can improve evaluation of results from selection experiments and may enhance their integration into systems of resistance risk assessment.