Abstract
Environmental pesticide exposure has been linked to adverse health effects, and residential proximity to agricultural land is commonly used as a proxy for exposure; however, the contribution of non-agricultural biomes remains insufficiently explored. This study examined whether the proximity and area of different biomes are associated with the detection of selected pesticides in human urine in Latvia. Urine samples were collected from 202 participants (101 adults and 101 children) within the Human Biomonitoring for Europe (HBM4EU) study during the winter and summer seasons of 2020. A suspect screening approach using liquid chromatography–high-resolution mass spectrometry (LC-HRMS) was applied and 23 pesticides were detected (8 insecticides, 12 fungicides, 2 herbicides and triclosan, an antimicrobial ingredient used in cleaning agents). Geospatial data were analysed in Quantum Geographic Information System (QGIS) to derive biome proximity and area within a 1000 m residential buffer; associations were assessed using generalized linear mixed-effects models. Agricultural land was present within 1000 m of 93.1% of residences, yet neither its distance nor area was consistently associated with pesticide detection. Boscalid was detected in 18.4% of samples and was positively associated with wetland area across seasons (p < 0.001), while fludioxonil (14.7%) showed weak and heterogeneous spatial associations and pirimiphos-methyl (10.2%) showed no significant patterns. Overall, pesticide exposure was substance-specific and influenced by landscape characteristics beyond agricultural proximity, highlighting the need to integrate non-agricultural biomes into human biomonitoring in low-intensity pesticide-use settings.