Meta-analysis on penconazole sorption to soils – Modeling variability of KOC and nonlinear isotherm parameters

Abstract

A meta-analysis was conducted to examine the variability in log KOC point-data (n = 612) and in the organic-carbon (OC) normalized Freundlich isotherm parameters (i.e., nf, Kf,OC; n = 86) of penconazole, a polar systemic triazole fungicide, in soils and biosolids-amended soils. Log KOC was not constant and varied from 2.15 to 4.24. Five existing log KOC models were tested against experimental log Kd, and they produced biased prediction with moderate to low accuracy (RMSE = 0.60 to 1.14). nf ranged from 0.38 to 1.02, with over 75% of nf < 0.85, suggesting the sorption isotherms were mostly nonlinear. Point sorption data (i.e., S, C = sorbed and aqueous concentrations) were best modeled with a nonlinear function: S=103.26fOCfOC+fclay1.21C0.81 (r2 = 0.89, RMSE = 0.26, n = 557) where fOC/(fOC + fclay) represents the relative availability of OC sorption sites. Log Kf,OC also varied considerably (i.e., 2.61–5.69) and was found to linearly correlate with nf: nf = –0.194log Kf,OC + 1.517 (r2 = 0.87). The observed sorption nonlinearity implies that assessment of subsurface mobility and ecotoxicological risk of soil penconazole is variable and site/soil-specific. The findings here challenge the conventional view of log KOC constancy currently prevailing in existing risk assessment practices and in major predictive sorption models of organic chemicals in natural geosorbents. The extent to which log Kf,OC variability is a prevailing feature in other organic contaminants is unclear, and further investigations seem warranted.