Summary
Prior to environmental legislation in the 1980s, anthropogenic waste in Antarctica was often deposited into landfill sites or into the sea. This resulted in metal contamination in terrestrial and near-shore marine environments. In this study, we assess the feasibility of using both past and present diatom assemblages to reconstruct and monitor past and future metal contamination. Our dataset included the analyses of both surface sediment samples and sediment cores from a contaminated site near Casey Station, Antarctica. Redundancy analyses indicated a strong relationship between metal concentrations and the composition of diatom communities. Within the surface sediment samples, tin and lead individually explained 43% of the variation observed in the diatom data; copper and iron explained 42% of this variation. In the sediment cores, tin and lead individually explained 53% of the variation in diatom community composition. In the same samples copper explained 47% of this variation, with iron explaining 46% of the observed variation. Once one metal had been selected, incorporating further metal data into the analyses added little extra information. Modern analog technique (MAT) analyses showed a strong correlation between actual and predicted values within one dataset (R2: Cu 0.75; Pb 0.86; Sn 0.89; p<0.05 for each). MAT reconstructions of metal concentrations closely followed measured concentrations, with both high and low concentrations recorded. MAT analyses performed favorably when compared to predictive techniques based on multivariate linear regression and multilayer perceptron neural networks. This study demonstrates that the composition of benthic diatom communities is a good indicator of anthropogenic metal contamination, and may be useful in monitoring the success of environmental remediation strategies in Antarctica and elsewhere.