Can remotely sensed catchment to lake area ratios predict mercury levels in subarctic fishes?

Mercury concentrations ([Hg]) in fish reflect a complex array of interacting biogeochemical and ecological variables. In northern regions where fish are a critical subsistence food, understanding and predicting fish [Hg] can be particularly difficult, largely due to a paucity of comprehensive data associated with the logistical challenges of field sampling. Building on previous work where we elucidated causal relationships between fish [Hg] and a variety of catchment, water quality, and ecological variables in subarctic lakes, we investigated whether using only ratios of catchment area to lake area (CA:LA) can predict [Hg] in northern freshwater fish species. As CA:LA can be sensed remotely, they may be more feasible and practical to obtain than field data in far northern regions. Our study included thirteen remote lakes that represent a CA:LA gradient of 6.2-423.5 within an ∼66,000 km2 subarctic region of Northwest Territories, Canada. We found that size-standardized [Hg] in three widespread fish species, including Lake Whitefish (Coregonus clupeaformis), Walleye (Sander vitreus), and Northern Pike (Esox lucius), were significantly and positively related to CA:LA (p < 0.007, r2 = 67-80%), indicating higher fish [Hg] in smaller lakes surrounded by relatively larger catchments. Our findings provide compelling evidence that remotely sensed CA:LA can be used to predict [Hg] in northern fishes and aid in prioritizing understudied and subsistence fishing lakes of the Canadian subarctic for [Hg] monitoring programs.

Estimates, spatial variability, and environmental drivers of mercury biomagnification rates through lake food webs in the Canadian subarctic.

Biomagnification of mercury (Hg) through lake food webs is understudied in rapidly changing northern regions, where wild-caught subsistence fish are critical to food security. We investigated estimates and among-lake variability of Hg biomagnification rates (BMR), relationships between Hg BMR and Hg levels in subsistence fish, and environmental drivers of Hg BMR in ten remote subarctic lakes in Northwest Territories, Canada. Lake-specific linear regressions between Hg concentrations (total Hg ([THg]) in fish and methyl Hg ([MeHg]) in primary consumers) and baseline-adjusted δ15N ratios were significant (p < 0.001, r2 = 0.58-0.88), indicating biomagnification of Hg through food webs of all studied lakes. Quantified using the slope of Hg-δ15N regressions, Hg BMR ranged from 0.16 to 0.25, with mean ± standard deviation of 0.20 ± 0.03). Using fish [MeHg] rather than [THg] lowered estimates of Hg BMR by ∼10%, suggesting that the use of [THg] as a proxy for [MeHg] in fish can influence estimates of Hg BMR. Among-lake variability of size-standardized [THg] in resident fish species from different trophic guilds, namely Lake Whitefish (Coregonus clupeaformis) and Northern Pike (Esox lucius), was not significantly explained by among-lake variability in Hg BMR. Stepwise multiple regressions indicated that among-lake variability of Hg BMR was best explained by a positive relationship with catchment forest cover (p = 0.009, r2 = 0.59), likely reflecting effects of forest cover on water chemistry of downstream lakes and ultimately, concentrations of biomagnifying MeHg (and percent MeHg of total Hg) in resident biota. These findings improve our understanding of Hg biomagnification in remote subarctic lakes.

Mercury bioaccessiblity in freshwater fish species from northern Canada.

Indigenous communities in northern Canada rely on locally harvested traditional foods, including fish, which provides them with nutritional, cultural, and social benefits. However, mercury exposure from fish consumption can pose a health risk for populations that consume large amounts of fish with elevated mercury concentrations. The bioaccessiblity of mercury in the tissue of northern Canadian freshwater fish is not yet known. To address this, samples from five commonly consumed freshwater fish species (Lake Trout, Northern Pike, Walleye, Lake Whitefish, and Burbot) caught from lakes in the Northwest Territories and Yukon, Canada were examined. Total mercury concentrations, fatty acid composition, and total mercury bioaccessibility differed significantly among fish species and lakes. Mean total mercury bioaccessibility using an in vitro gastrointestinal model ranged between 56 % and 96 % in muscle tissue across fish species and waterbodies examined and was 39 % in liver tissue from Burbot. Mean total mercury bioaccessibility was much lower (range: 38 % to 42 %) for a subset of samples run through only the gastric phase of the digestion model. Total mercury bioaccessibility was significantly lower (on average 40 % lower) in a subset of samples that were pan-fried in water. Thus, although cooking increased total mercury concentrations in pan-fried fish samples (likely due to moisture loss), bioaccessible concentrations of total mercury were lower (on average 32 % lower). Results from this study contribute to addressing a large knowledge gap in the literature regarding bioaccessibility of total mercury in northern freshwater fish species. To the best of our knowledge, this is the first study to examine mercury bioaccessibility in raw and cooked liver samples from freshwater fish. It also adds to the growing literature indicating that mercury bioaccessibility varies among fish species, locations, and cooking/preparation methods.

Understanding among-lake variability of mercury concentrations in Northern Pike (Esox lucius): A whole-ecosystem study in subarctic lakes.

Mercury concentrations ([Hg]) in fish reflect complex biogeochemical and ecological interactions that occur at a range of spatial and biological scales. Elucidating these interactions is crucial to understanding and predicting fish [Hg], particularly at northern latitudes, where environmental perturbations are having profound effects on land-water-animal interactions, and where fish are a critical subsistence food source. Using data from eleven subarctic lakes that span an area of ~60,000 km2 in the Dehcho Region of Northwest Territories (Canada), we investigated how trophic ecology and growth rates of fish, lake water chemistry, and catchment characteristics interact to affect [Hg] in Northern Pike (Esox lucius), a predatory fish of widespread subsistence and commercial importance. Results from linear regression and piecewise structural equation models showed that 83% of among-lake variability in Northern Pike [Hg] was explained by fish growth rates (negative) and concentrations of methyl Hg ([MeHg]) in benthic invertebrates (positive). These variables were in turn influenced by concentrations of dissolved organic carbon, MeHg (water), and total Hg (sediment) in lakes, which were ultimately driven by catchment characteristics. Lakes in relatively larger catchments and with more temperate/subpolar needleleaf and mixed forests had higher [Hg] in Northern Pike. Our results provide a plausible mechanistic understanding of how interacting processes at scales ranging from whole catchments to individual organisms influence fish [Hg], and give insight into factors that could be considered for prioritizing lakes for monitoring in subarctic regions.