Sources of atmospheric metal(loid) pollution recorded in Thompson Manitoba lake sediment cores within the Canadian boreal biome.

Global atmospheric emissions and subsequent deposition of numerous metal(loid)s has increased markedly since the industrial revolution. Due to a paucity of long-term metal(loid) flux measurements, the magnitude and timing of change are largely unknown, resulting in limited ability to predict time-scales of ecosystem recovery in response to emission decreases. In the absence of long-term data, palaeo-reconstructions provide continuous records of atmospheric metal(loid) deposition on an ecosystem, and landscape, scale. Here, we use high-resolution dated lake sediment cores to reconstruct the last c. 100 years of atmospheric anthropogenic deposition of a full suite (40) of metal(loid)s near a large nickel (Ni) and copper (Cu) smelter in an other-wise largely "pristine" region of northern Canada (Thompson, Manitoba). Anthropogenic depositional fluxes were compared to other regions of Canada including Kejimkujik National Park in Nova Scotia, Experimental Lakes Area in Ontario, as well as the Flin Flon, Manitoba Cu and zinc (Zn) smelter, located ~200 km southwest of Thompson. Deposition of 12 metal(loid)s were enriched above baseline (pre-1915) levels: antimony (Sb) > palladium (Pd) > bismuth (Bi) > mercury (Hg) > cadmium (Cd) > Ni > lead (Pb) > arsenic (As) > strontium (Sr) > Cu > platinum (Pt) > Zn. Spatio-temporal patterns in depositional fluxes and inventories demonstrate that 6 of these metal(loid)s were sourced primarily from the smelter, while As, Hg, Pb, Pt, Sb and Zn were sourced primarily from global and/or regional sources. Comparison of anthropogenic fluxes and inventories to available emissions data showed that Cu and Ni deposition has plateaued since the late 1970s despite dramatic smelter emission decreases between 2005 and 2014. We hypothesize that this discrepancy is due to releases of terrestrial metal(loid)s by climate-driven permafrost degradation, which is widespread across the region and will likely continue to drive increased metal(loid) fluxes to northern Canadian lakes for unknown time-scales.