Predicting the fate and effects of tributyltin in marine systems.

ABSTRACT

The available data indicate that sediment-water partitioning, bioaccumulation, and the toxicity responses for tributyltin (TBT) are predictable when using some of the assumptions and tenets of the equilibrium partitioning method, toxicokinetic modeling (1CFOK), and critical body residue (CBR) approach. Because TBT is ionizable, its speciation is strongly affected by pH, which appears to cause large variations in the octanol-water partition coefficient. In marine systems, and in freshwater systems with high pH, TBT occurs predominantly in the hydroxide form, which may explain the hydrophobic properties and its EqP behavior. Organic carbon in sediment (> 0.2%) appears to be the major controlling factor for sediment-water partitioning. The equilibrium organic carbon-normalized sediment-water partition coefficient (Koc) in marine environments is approximately 32,000 (log10 Koc approximately 4.5), which was determined from direct measurement and confirmed by the relationship between the lipid-normalized bioconcentration factor (BCF) in porewater and the biota-sediment accumulation factor (BSAF). The conclusion that sediment-water partitioning of TBT in marine systems follows EqP is supported by the similarity between its Kow and Koc and the correlation between the sediment-water partition coefficient (Kp) and sediment TOC, which results from the influence of organic carbon on pore-water concentrations. Even though the rates of uptake and elimination control tissue residues and lipid content appears to have no bearing on the amount of TBT that is accumulated, the species specific BSAF is useful for examining bioaccumulation, sediment-water partitioning, and the toxicity response. Although TBT is hydrophobic and appears to have a propensity to accumulate in lipid, the rates of uptake and elimination, not thermodynamics, appear to control whole-body tissue concentrations. Support for a toxicokinetic approach for predicting tissue residues is found in BCF and BSAF values for several species that are far in excess of that predicted by simple thermodynamic partitioning and in the comparisons of observed and predicted bioaccumulation values based on toxicokinetic coefficients. This observation is counter to the assumption of EqP that the route of uptake is of no consequence under equilibrium conditions. For TBT, it appears that kinetics determine tissue residues and that body lipid is important only for regulating the toxic response, not the amount bioaccumulated. Unlike those for neutral hydrophobic organic compounds, the toxicokinetics for this one toxicant are highly variable in diverse species but relatively accurate in predicting the amount bioaccumulated and the resulting toxicity response. For the CBR approach to be useful, a relatively constant tissue residue for a given biological response is necessary. Several studies support the CBR approach because certain biological effects, such as mortality and growth inhibition, occur at a relatively constant TBT tissue concentration. For TBT, the lethal whole-body tissue concentration affecting 50% of individuals (LR50) exhibits little variation, occurring at approximately 48 micrograms/g (166 nmol/g) dry weight in a range of species. Direct evidence and correlation of the LC50 and the bioconcentration factor (BCF) support this observation. Impaired growth, a sublethal response, also appears to be associated with a relatively constant tissue concentration, which has also been demonstrated by direct measurement and indirectly by regression of the BCF and LOEC. The lowest-observed-effect tissue residue (LOER) associated with impaired growth for several species was approximately 3 micrograms/g (10.4 nmol/g) dry wt. Because of the small number of studies linking growth impairment and tissue concentrations, additional studies are needed to confirm these values. (ABSTRACT TRUNCATED)