Assessing mercury contamination patterns of fish communities in the Laurentian Great Lakes: A Bayesian perspective.

We examine the spatio-temporal trends of mercury, a well-known global legacy contaminant, in eleven fish species across all of the Canadian Great Lakes. These particular fish species are selected based on their ecological, commercial, and recreational importance to the biodiversity and fishing industry of the Great Lakes. We present a two-pronged Bayesian methodological framework to rigorously assess mercury temporal trends across multiple fish species and locations. In the first part of our analysis, we develop dynamic linear models to delineate the total mercury levels and rates of change, while explicitly accounting for the covariance between fish length and mercury levels in fish tissues. We then use hierarchical modelling to evaluate the spatial variability of mercury contamination between nearshore and offshore locations, as well as to examine the hypothesis that invasive species have induced distinct shifts on fish mercury contamination trends. Our analysis suggests that the general pattern across the Great Lakes was that the elevated mercury concentrations during the 1970s had been subjected to a declining trend throughout the late 1980s/early 1990s, followed by a gradual stabilization after the late 1990s/early 2000s. The declining trend was more pronounced with top fish predators, whereas benthivorous fish species mainly underwent wax-and-wane cycles with a weaker evidence of a long-term declining trend. Historically contaminated regions, designated as Areas of Concern, and bays receiving riverine inputs are still characterized by mercury concentrations that can lead to consumption restrictions. Lake Erie displayed the lowest mercury levels across all the fish species examined. However, several species of commercial importance showed a reversing (increasing) trend in the 2000s, although their current levels do not pose any major concerns for consumption advisories. These recent trend reversals can be linked with systematic shifts in energy trophodynamics along with the food web alterations induced from the introduction of non-native species, and the potentially significant fluxes from the atmosphere.