Ninety-nine lakes were sampled at varying distances up to 75 km from the Horne smelter at Rouyn-Noranda, Quebec, to study the influence of the smelter versus other factors on metal concentrations in lake sediments. Most of these lakes lie within the Abitibi Greenstone Belt, a zone of extensive base metal and gold mineralization and the focus of a mining and smelting economy for almost a century. Lake sediment cores, c. 25 cm long, were collected and sampled at the top (0–2 cm) and the bottom (18–20 cm) to capture sediment that was deposited after the smelter was in operation (‘post-industrial’) and well before the mining and smelting activity was started (‘pre-industrial’). Additionally, nine cores were sampled in 1 cm increments to depths of up to 50 cm to study temporal patterns and potential element remobilization in detail. The cores were analysed for an extensive suite of elements. This paper focuses on those elements that are emitted by the smelter for which there are records of emissions through time, namely As, Cd, Cu, Pb and Zn.
A spatial statistical approach – a logistic model of metal content versus distance from the smelter – was used investigate the relationship of sediment chemistry with smelter emissions and other possible influences. Using Cu as a representative proxy for the other emitted metals, this analysis demonstrates that elements are enriched in lake sediments by a factor of about three times around the smelter, that the impact of the smelter is detectable in lakes to a distance of at least 50 km, and that there is no obvious association between sediment Cu concentration and bedrock geology, land-use, lake pH, or lake morphometry (lake area/lake catchment area).
The nine lakes studied in detail show enrichment towards the sediment–water interface (SWI) and relatively steady concentrations below depths of c. 10 cm. However, depth profiles do not match changes in the magnitude of smelter emissions through time, nor do they match changes in emission chemistry (element ratios) through time. Element ratios do generally move towards the chemistry of the emissions, suggesting smelter influence, but do not do so predictably. For example, (i) trends in the Cu/Pb ratio continue to the very bottom of cores into material deposited hundreds of years before industrialization, and (ii) proximity to the smelter does not lead to greater similarity between sediment and emission chemistry. These results suggest that significant element remobilization is occurring and that it differs from lake to lake and from element to element.
We conclude that lakes within 50 km of the smelter have elevated metal concentrations in their near-surface sediments due to stack emissions but, due to element cycling and mobility, it is difficult to quantitatively determine the magnitude of metal increase attributable to the smelter. We also suggest that due to upward remobilization, the duration of industrial metal enrichments in surface sediments (the residence time) may be increased, thereby making surface enrichments more persistent than would be predicted by the sedimentation rate.