Physiologically based demographics of Bt cotton–pest interactions: II. Temporal refuges, natural enemy interactions

Abstract

An holistic demographic physiologically based age–mass structured model of transgenic Bt cotton expressing one or two toxin of the bacterium Bacillus thuringiensis Berliner (Bt) and 10 of its pests as affected by generalist natural enemies is developed. The goal of the model is to assess factors favoring the development of resistance in pest species. Among the pest included in the model are the pink bollworm (a primary pest of cotton), and several secondary pests of the family Noctuidae (tobacco budworm, bollworm, fall armyworm, beet armyworm, cabbage looper, soybean looper) as well as a plant bug, boll weevil and whitefly. The pests have widely varying tolerance to the toxin. Resistance to the Bt toxin is assumed recessive, autosomal and controlled by a single diallelic gene. S patial refuges of non-Bt cotton have been mandated to increase heterozygosity in Bt cotton, but spatial refuges may exist for many pest species in non-Bt and/or sylvan hosts. In addition, temporal refuges may arise within Bt cotton due to innate pest tolerance to the toxin that increases with larval age, and variable toxin concentrations in the plant over time and with plant subunit age. Mortality rates and non-lethal effects that increase larval developmental time and decrease adult fecundity vary with Bt toxin concentration. All genotypes are affected but resistant genotypes are assumed less affected than susceptible ones. The effects of spatial and temporal refuges on the development of resistance and pest densities in one and two toxin cottons are examined over several consecutive seasons. Resistance to Bt toxin is more likely to develop in stenophagous highly susceptible pests such as pink bollworm and budworm than in polyphagous highly tolerant pests such as fall armyworm, beet armyworm and soybean looper that have large temporal refuges that slow and possibly reverse resistance development. Predator longevity feeding on Bt intoxicated prey is decreased ca. 28% [Ponsard, S., Gutierrez, A.P., Mills, N.J., 2002. Effects of Bt-toxin in transgenic cotton on the adult longevity of four heteropteran predators. Environ. Entomol. 31, 1197–1205.] creating tradeoffs between the mortality caused by the toxin and that due to reduced predation. Despite reduced natural enemy efficacy, levels of susceptible pests (budworm) are reduced, while levels of Bt tolerant and immune pests (e.g. fall armyworm, Lygus) increase. The use of pesticide for supplementary control of tolerant pests in Bt cotton may further disrupts natural enemies and increase pest levels.