Senthil Kumar, N., & Regupathy, A. 2005 Risk assessment of neonicotinoids applied to coffee ecosystem.. International Pest Control 47(2): 82-87.

Notes: Laboratory-derived toxicity data were used to predict the toxicity of thiamethoxam and imidacloprid to beneficial arthropods (Apis indica [A. cerana indica], A. mellifera, A. florea and Trigona iridipennis) and pests (Coccus viridis and Toxoptera aurantii) in a coffee ecosystem. The most susceptible to thiamethoxam in terms of LD SUB 50 was Toxoptera aurantii, followed by C. viridis, Trigona iridipennis, A. indica and A. florea, and A. mellifera. In terms of body weight, Toxoptera aurantii was the most susceptible, followed by A. mellifera, A. indica, A. florea, C. viridis and Trigona iridipennis. The most susceptible to imidacloprid was Toxoptera aurantii, followed by C. viridis, Trigona iridipennis, A. florea, and A. indica and A. mellifera in terms of LD SUB 50, and Toxoptera aurantii, followed by A. mellifera, A. indica, A. florea, Trigona iridipennis and C. viridis in terms of body weight. In honey bees, the body weight was negatively correlated with susceptibility. Selectivity ratios based on toxicity to Toxoptera aurantii indicated that both chemicals were highly toxic to Toxoptera aurantii compared with the toxicity to the beneficial insects. Thiamethoxam was more selective than imidacloprid based on selectivity ratios derived using data on both pests. With selectivity ratio as the sole method of evaluation, both insecticides can be recommended for use in integrated pest management (IPM). However, when probit substitution method was used to make recommendations to IPM programmes, only thiamethoxam would perhaps be recommended. Hazard ratios predicted that both chemicals were hazardous to bees at the recommended field rate. Based on the sequential testing scheme, the chemicals were considered to be more toxic, as the LD SUB 50 of both chemicals were less than 2 g per insect for all the beneficial species.