Meta-analysis on Factors Influencing Soil Sorption and Bivalve Bioaccumulation of Hydrophobic Organic Contaminants

Abstract

Sorption coefficient (K_OC) measures the tendency of a chemical to bind to soils or sediments. Biota-to-sediment accumulation factor (BSAF) measures the accumulative potential of a chemical in organisms. K_OC and BSAF are key metrics in the ecological-risk assessments of contaminants. The K_OC of a given hydrophobic organic chemical is typically assumed to be constant and unaffected by the soil properties. Estimating BSAF with equilibrium partitioning theory is extensively accepted because it is simple and easy to use. However, the variability in sorption and bioaccumulation has not been adequately explored. This thesis aimed to evaluate the validity of the constant-K_OC paradigm and the suitability of the equilibrium-partitioning theory to estimate BSAF with experimental or field data by meta-analysis. We also investigated the factors influencing the K_OC of alachlor, atrazine, chlorpyrifos, and isoproturon in soil sorption, as well as the BSAF of PCBs, PAHs, and organochlorines in bivalve bioaccumulation.

A total of 572 experimental distribution coefficients and 810 Freundlich isotherms from 132 studies on alachlor, atrazine, chlorpyrifos, and isoproturon were compiled. The experimental K_OC’s varied by a factor of 80–10 000 for each pesticide. A statistically significant negative correlation existed between Freundlich sorption coefficients and nonlinearity exponents (R² = 0.58–0.82). Sorption (K_d) prediction from soil composition (e.g., clay or silt content) (root-mean-square error (RMSE) = 0.33 and 0.30) was better than that from K_OC in a quantitative structure–activity relationship model (RMSE = 0.59 and 0.39) for alachlor and chlorpyrifos. These results indicated that the constant-K_OC paradigm may be questionable, and that the sorption nonlinearity and soil properties can introduce significant variation into soil sorption. The findings also suggested the possible convenience of estimating the sorption of pesticide by the clay sorption coefficient (K_clay) or the silt sorption coefficient (K_silt) together with K_OC.

More than 3000 field BSAFs of PCBs, PAHs, and organochlorine in bivalves were collected from 57 studies. Field BSAFs of a given contaminant in bivalves varied by 3–6.5 log units, which imposed limits on the utility of existing BSAF predictive models (e.g., BSAF approaches one based on equilibrium-partitioning theory) as a screening tool to predict the bioaccumulation potential of organic contaminants. Variations in BSAFs increased with increased complexity of hydrodynamic conditions of sampling sites. The standard deviation of log BSAF of complex hydrodynamics was about two times that of low flow or unidirectional flow. BSAFs significantly decreased with increased flow velocity (R² = 0.45). This dependence was attributed to the high-flow-induced decrease in contaminant exposure of bivalves under benthic environments. This hypothesis was based on field observations that the net diffusion of contaminants occurred from sediment to overlying water in most benthic environments. BSAF was also found to depend on the contaminant concentration in sediments. Remedial action may have to attain much lower sediment concentration to overcome the increase in bioaccumulation indicated by the higher BSAF at a lower sediment concentration.

This thesis contributes to the literature and database establishment of soil sorption and bivalve bioaccumulation of hydrophobic organic contaminants. We show that variable K_OC and isotherm nonlinearity are features of the sorption of alachlor, atrazine, chlorpyrifos, and isoproturon amongst soils. Experimental data suggest that a constant-K_OC paradigm is unreliable for ecological-risk-assessment practices of organic contaminants and model development for soil sorption. Further research is needed to enable model development for soil sorption involving soil properties and sorption nonlinearity. This thesis further highlights the importance of hydrodynamics in the bioaccumulation of organic contaminants in benthic organisms. Based on the proposed preliminary hypothesis, a simple model (BSAF = 105/(105 + v) (N=30, R² = 0.55, RMSE = 0.44 between observed and predicted log BSAFs) was developed to predict BSAF from flow velocity (v) in ecological-risk-assessment practices.

Date of Award	2 Jul 2021
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Siu Ming LO (Supervisor) & Ta Fu Dave KUO (Supervisor)