Diffusive gradients in thin films sampler predicts stress in brown trout (Salmo trutta L.) exposed to aluminum in acid fresh waters.

Increased levels of aluminum ions released from nutrient-poor soils affected by acid rain have been the primary cause of fish deaths in the acidified watersheds of southern Norway. The complex aluminum chemistry in water requires speciation methods to measure the gill-reactive species imposing toxic effects toward fish. Previously, aluminum speciation has mainly followed the fractionation principles outlined by Barnes/Driscoll, and several analogues of these fractionation principles have been used both in situ and in the laboratory. Due to rapid transformation processes, aluminum speciation in water samples may change even during short storage times. Thus, results obtained by laboratory fractionation methods might be misleading for the assessment of potentially toxic aluminum species in the water. Until now, all in situ field fractionation methods have been time and labor consuming. The DGT technique (diffusive gradients in thin films) is a new in situ sampler collecting a fraction of dissolved metal weighted according to the rate of diffusion and dissociation kinetics. In a field experiment with acid surface water we studied the DGT sampler as a new prediction tool for the gill accumulation of aluminum in trout (Salmo trutta L.) and the induced physiological stress responses measured as changes in blood glucose and plasma chloride. Aluminum determined with DGT (DGT-AI) was higher than labile monomeric aluminum (Ali) determined with a laboratory aluminum fractionation procedure (PCV--a pyrocatechol violet analogue of Barnes/Driscoll), a difference due to collection of a fraction of organically complexed aluminum by DGT and a reduction of the Ali fraction during sample storage. DGT-AI predicted the gill uptake and the aluminum-induced physiological stress responses (increased blood glucose and decreased plasma chloride, r2 from 0.6 to 0.9). The results indicate that DGT-AI is a better predictor for the stress response than laboratory-determined Ali, because the DGT sampler collects a more correct fraction of the gill-reactive aluminum species that induces the stress.