Dietary diversity is a significant predictor of stressor abundance. Quasi-Poisson mixed-effects models (glmm-pq) tested for association between the response (stressor abundance), and the predictor (dietary diversity, or landscape composition), controlling for experiment (location, time) as a random effect. Estimates are quasi-Poisson log-odds coefficients; P-values indicate the statistical significance of the relationship. Dietary diversity (a–c) was a significant predictor (P < 0.05) of stressor exposure, while relevant landscape parameters (d–f) were weak predictors of stressor exposure. ALC (agricultural land coverage) and ULC (urban land coverage) are landscape gradients values that indicate the relative abundance of a land-type within a 1,500 m radius of the study site. SWI (Shannon's index) are dietary gradients that indicate the relative diversity of a pollen sample.