Quantifying pesticide efficacy from multiple field trials

Simple corrected density indices (CDIs) have been used to measure reductions in pest density in fields. In the contemporary pesticide registration system, few comprehensive statistical frameworks are available that can integrate multiple datasets to evaluate how pesticidal effects are influenced by the products' properties such as mixing multiple active ingredients and possession of systemic ability. In this study, we provide a statistical framework for evaluating pesticide efficacy from multiple field trials and applying it to contemporary pesticides. In this framework, we extended the conventional CDI to a generalized linear mixed model (GLMM), which we applied to a dataset of the pesticide registration test in Japan (n = 758). The estimated mortality of a single active ingredient in reducing pest density is 88.0%, indicating the registered pesticide satisfies the “effective” criterion (roughly 70–95%) under the current pesticide registration system in Japan. Although systemic ability additionally reduced pest population to 55.5% of the post-treatment densities, the addition of active ingredients scarcely enhances efficacy (reducing population to 74.6%), suggesting that the pesticide design resulted in broadening the spectrum of target species rather than increasing toxicity.