Evaluation of some physico-chemical techniques for the determination of the fraction of dissolved copper toxic to the marine diatom nitzschia closterium

Abstract

The toxicity of copper to the marine diatom Nitzschia closterium was determined by growth rate measurements in the presence of a range of copper complexing agents, both natural and synthetic. The measurements were made in raw, unenriched sea water to avoid the reaction of copper with silicate or colloidal hydrated iron(III) oxide which occurs in standard culture media. The algae remained in exponential growth for at least 72 h in unenriched sea water and, in the presence of copper, produced an exudate which decreased the concentration of labile copper. Labile copper was measured in the algal assay solutions by using anodic stripping voltammetry (a.s.v.) at different deposition potentials, by separation on iminodiacetate (Chelex-100) and thiol resins, and by extraction with hexane/n-butanol (9:1) to simulate lipid solubility. No consistent correlation was observed between the toxic fraction of copper measured by algal assay, and the labile copper determined by the physico-chemical techniques. Although some of the naturally-occurring ligands (e.g., fulvic acid and iron—humic acid colloid) gave reasonable agreement between the toxic fraction and a.s.v.-labile copper, the chelating resins usually grossly overestimated toxicity. Lipid-soluble complexes of copper with synthetic ligands (e.g., 8-quinolinol and diethyldithiocarbamate) were highly toxic; as little as 2 μg Cu l −1 in the presence of 5 × 10 −5 M 8-quinolinol caused complete depression of algal growth. It is proposed that the extreme toxicity of lipid-soluble copper complexes results from their ability to catalyse the intercellular formation of highly destructive hydroxyl free radicals (OH.) from molecular oxygen, in a Fenton-type reaction.