Giant Mine, located in the city of Yellowknife (Northwest Territories, Canada), is a dramatic example of subarctic legacy contamination from mining activities, with remediation costs projected to exceed $1 billion. Operational between 1948 and 2004, gold extraction at Giant Mine released large quantities of arsenic and metals from the roasting of arsenopyrite ore. We examined the long-term ecological effects of roaster emissions on Pocket Lake, a small lake at the edge of the Giant Mine lease boundary, using a spectrum of palaeoenvironmental approaches. A dated sedimentary profile tracked striking increases (approx. 1700%) in arsenic concentrations coeval with the initiation of Giant Mine operations. Large increases in mercury, antimony and lead also occurred. Synchronous changes in biological indicator assemblages from multiple aquatic trophic levels, in both benthic and pelagic habitats, indicate dramatic ecological responses to extreme metal(loid) contamination. At the peak of contamination, all Cladocera, a keystone group of primary consumers, as well as all planktonic diatoms, were functionally lost from the sediment record. No biological recovery has been inferred, despite the fact that the bulk of metal(loid) emissions occurred more than 50 years ago, and the cessation of all ore-roasting activities in Yellowknife in 1999.
The impacts of permafrost thaw do not manifest uniformly across the Arctic, and this presents challenges for predicting how permafrost‐affected lakes will respond to climate change. Here, we leveraged long‐term field data collection and studies of a particular permafrost lake type (predisposed to thaw slump disturbance) in the Tuktoyaktuk Coastlands region to develop a conceptual model that can shed light on, and facilitate further testing of, lake ecosystem impacts associated with polycyclic shoreline retrogressive thaw slumps. To inform the development of the model, we compared regional changes in slump activity and lake water quality variables known to be impacted by slumping between 2005 and 2017, and showed that the timeline to recovery in lake dissolved organic carbon (DOC) following slump stabilization was longer than the time between cycles of slump growth, stabilization, and re‐initiation in most lakes. We also analyzed geochemical changes in sediment cores that incorporated the last several hundred years of history in 18 lakes that spanned a gradient in thaw slump disturbance, as well as a core dating back to ca. 1365 CE in a lake with an ancient slump. Results indicate that sediment geochemical changes were most pronounced at the initial onset of slumping, with only muted changes in sediment characteristics upon slump growth or re‐initiation. Our conceptual model can guide future investigations into the extent to which lakes with polycyclic slumps are predisposed to, or buffered against, significant limnological changes under anthropogenic climate warming. While our model is specific to lake systems in ice‐cored morainal terrain impacted by retrogressive thaw slumping, similar approaches would be highly useful for understanding ecosystem response under a range of permafrost disturbance categories.