ORIGINAL ABSTRACT

Glyphosate, the most widely used herbicide, is primarily applied to control unwanted weeds. While its impacts on pollinators such as honey bees have been extensively studied in agricultural settings, exposure from non-crop flora in urban environments remains largely overlooked. This gap is critical, as urban greenspaces function both as refuges for pollinators and as potential, underrecognized pathways of herbicide contamination. We investigated glyphosate residues in pollen collected by honey bee foragers from non-crop plants across an urbanization gradient in Suzhou, China, over an entire growing season. Using DNA metabarcoding and LC-MS/MS analysis, we identified pollen sources and quantified glyphosate residues in 71 composite samples from 11 sites along an urbanization gradient, with built-up coverage ranging from approximately 10-70 %. Glyphosate was detected in 83.1 % of samples, with concentrations up to 880 μg/kg. Residue levels showed a non-linear relationship with urbanization, with glyphosate concentrations decreasing initially and then increasing as the proportion of built-up areas increased. No significant correlation was detected between glyphosate occurrence (detection frequency or concentration) and pollen diversity. These results reveal that non-crop flora in urban greenspaces represent a season-long pathway of glyphosate exposure for honey bees, highlighting the need to integrate pollinator health considerations into urban vegetation and herbicide management strategies.