Prediction of the adsorption of ionizable pesticides in soils

Adsorption of six acidic and four basic pesticides was measured in nine temperate soils. In general, sorption of acids was weak as compared to bases and followed the order dicamba < metsulfuron-methyl < fluazifop-P < metribuzin < 2,4-D < flupyrsulfuron-methyl < fluroxypyr < terbutryn < pirimicarb < fenpropimorph. Adsorption was negatively correlated with soil pH and positively correlated with organic carbon content. Statistical analysis against a wide range of soil and pesticide descriptors was used to identify the best combination of properties that describes the variation in adsorption. A regression equation including Log D (lipophilicity corrected for pH), the soil organic carbon content, and a pesticide descriptor (related to the van der Waals volume) was selected to predict the adsorption of acids (r(2) = 72.1% on an independent dataset). The behavior of bases was more complex, and approaches specific to each compound seem to be required.