Because of the underlying error distribution, we used a generalized linear model with a Poisson link function to analyze the data on which this paper is based. As is clear in Supp_1.R, the analyses for Fig. 1 and Fig. 3 were performed for each year separately. The R scripts also calculate Efron's pseudo-R2. Supp_1.R also shows the analysis used in Fig. 2, where all years of pollinator data were pooled.