Fig. 1: Association between surrounding land-use categories and the occurrence of pesticides from the Japanese honey bee (Apis cerana japonica) colonies.
From: The impact of landscape structure on pesticide exposure to honey bees
The tested sample is honey (left: acetamiprid to glyphosate) or beeswax (right: diazinon to fipronil). Generalized linear models (GLMs) were employed with pesticide presence/absence as the response variable and land-use proportions as the explanatory variable. The models utilized a binomial distribution and a logistic link function. Statistical analysis was conducted using a two-sided Wald test without adjustments for multiple comparisons. The association was categorized as follows: positive: odds ratio > 1 and p < 0.05; negative: odds ratio <1 and p < 0.05; non-significant: others. Each GLM result is shown in Supplementary Fig. S3. Summary statistics are provided as a Source Data file.
