S2 Canadian Clean Air Regulatory Agenda (CARA) Mercury Science Program

The effects of atmospheric, edaphic, topographic, and geologic specifications of watersheds and wet-area coverage associated with Geological Survey of Canada (GSC) sampling locations for stream and lake sediments were investigated for total mercury (THg) variations in Nova Scotia. Geospatial analyses were carried out by using high resolution digital elevation models (10 m resolution) to determine mercury hotspot watersheds. The result showed that THg concentrations were higher in lakes (average: 360 ppb) than streams (average: 90 ppb) for both upland and lowland sampling locations and were related to lake area, presence/absence of wetlands, wetland types, and atmospheric mercury deposition. The analyses were focused on case studies including Cap Breton, KejimKujick National Park, Mercy River, and Pock-Bowater Water watersheds in Nova Scotia.

The Clean Air Regulatory Agenda of Environment Canada funded three years of mercury-related research and monitoring across Canada. One aspect of this scientific program was the Ecological Risk Mapping Project. There were six component studies under this project. The FISHg project monitored fish mercury trends in forage and predatory fish in 15 study lakes across Canada. The Wildlife Mercury study monitored mercury levels in forage fish and common loons in four study areas across the country. The Fish and Wildlife Risk Assessment study compiled a national fish mercury database and then used it to spatially assess mercury risks to predatory fish and fish-eating wildlife. The Quebec Fish Mercury Trends study looked at trends in fish mercury levels in 4 study lakes and the environmental factors that may be driving those trends. The Fish Health study investigated the impacts of elevated mercury levels on the health of yellow perch in southwest Nova Scotia. Lastly, the Mercury Methylation study evaluated a new biotechnology for measuring mercury methylation in situ in lakes to identify hotspots. Several studies were co-located in areas downwind of major mercury point sources or in mercury sensitive ecosystems. We will present a summary of the six studies and the teams of scientists involved. The results of this scientific program will feed into the Canadian Mercury Science Assessment Report, which will be published by Environment Canada in 2012.

The HERMES model uses a reduced set of inputs to predict concentrations of total mercury and methyl mercury in the water column and sediments of lakes. The model was tested on a number of lakes in Canada and has proved to be a good predictive screening-level tool.

As part of CARA ecosystem modelling effort, model applicability was improved through the development of better estimation methods and linked with other air, land and water mercury models. This model provides an alternative for comparison with D-MCM predictions of THg and MeHg in the water column and sediments. The long term goal is to add a food web module to predict MeHg in fish.

Mercury (Hg) dynamics in boreal lakes is of special concern as human populations may be exposed to the contaminant through fish consumption. In Northern Quebec, forestry and mining represent major economic activities and watersheds which never underwent perturbations over the past decades are rare. It has been argued that logging and mining activities induce an increase of mercury fluxes in the environment and consequently may result in enhanced mercury concentrations in fish commonly consumed by Northern Quebec populations. Here we report variations on Hg concentrations in fish flesh of both pike and walleye over a 30 years period for undisturbed lakes and lakes with watersheds that have been highly impacted by logging and/or mining activities. Watersheds history over this period was estimated by satellite imaging analysis using GIS whereas the estimation of mercury fluxes reaching the lakes over the time was based on sediment core analyses. Lignin biomarkers (indicators of both quantity and quality of terrigenous organic matter) were also measured and allowed to differentiate forestry from mining inputs of both organic matter and associated mercury from the watershed to the lake.

Previous studies have shown that the Hg concentrations in fish from different lakes can be related to atmospheric Hg inputs, watershed variables (e.g. wetland area) and in-lake variables (e.g. water chemistry and trophic dynamics). The Freshwater Inventory and Surveillance of Mercury (FISHg) is a project under Environment Canada’s Clean Air Regulatory Agenda (CARA) Mercury Science Program. The FISHg network has been sampling fish and water Hg concentrations in over 20 lakes across Canada during the last 3 years. The different lakes comprise a wide range of atmospheric Hg inputs (several lakes were located in proximity to point-sources of Hg emissions; while others were in remote regions), watershed characteristics, and fish species. The objective of this presentation is to try to identify which variables are likely having a dominant influence on the spatial variability of fish Hg concentrations between lakes across Canada. In addition, short-term inter-annual variability of fish and water Hg concentrations was assessed to identify the relative importance of spatial and temporal variability of fish Hg concentrations. Initial results have shown a large range of Hg concentrations in the predatory fish (0.05 to 3.65 ug/g) and forage fish (0.01 to 1.22 ug/g) as well as water total-Hg concentrations between lakes (0.2 to 7.8 ng/L). Within a given lake, there was a wide range in Hg concentrations between different fish species related to differences in foraging behaviour (i.e. benthic versus pelagic) and trophic position (identified through stable isotope analysis). Within-lake differences in fish Hg concentrations for a given species were related to fish length. GIS analysis of watershed variables and atmospheric Hg inputs is ongoing and will help identify the relative importance of within-lake variables and external-lake variables on the concentrations of Hg in fish at a national level.

In this study, the goal is to assess the temporal and spatial variation of Hg isotope ratios by analyzing depths profiles in sediment cores at different distances from a point source, a metal smelting operation. The study site is the typical mining town-- Flin Flon at Manitoba. Three lake cores were selected along the main wind direction (NW-SE) from the smelter: Phantom Lake, 5km; Cleaver Lake, 23km and Mclurg Lake, 73km. All three lake cores had distinct Hg isotopic compositions, which were significantly different among lakes (P<0.0000001). The d²⁰²Hg of the Phantom core ranged from -0.67 to -1.48‰, while d²⁰²Hg of the furthest Mclurg core varied from -1.44 to -2.71‰. The Cleaver core showed a wider range of d²⁰²Hg from -0.79 to 2.48‰. Phantom Lake was dominated by heavier isotopes with higher Hg content. Mclurg Lake served as the Hg background site and was enriched in lighter isotopes. Cleaver Lake had the mixture of two different Hg sources: smelter and background. Mass independent fractionation (∆¹⁹⁹Hg and ∆²⁰¹Hg) was not detected in those sediment cores. Moreover, by applying a binary mixing model, we demonstrated the relative contribution of anthropogenic sources in different samples and also provided evidence of the smelter impact on those sediment cores.

This study provided another example to successfully use Hg isotopic composition to track and quantify Hg pollution sources and ultimately to better understand Hg biogeochemical cycling in the environment.

This paper presents a Geological Survey of Canada (GSC) data survey of mercury (Hg) concentrations in stream sediments (n » 142,000) for selected areas across Canada based on the general geological and topographic elevation differences between upland and lowland streams. Hg concentrations vary widely from 5 to 9,400 ppb per sample, with regional averages from 29 (central Nunavut) to 165 ppb (Vancouver Island, BC). An extensive cluster (100x400 km, oriented SE-NW) of Hg concentrations > 4000 ppb is located in the eastern central portion of the Yukon Territory. Regionally and locally, Hg concentrations tend to be consistently higher in stream sediments in higher uplands elevations than lowlands. The upland to lowland delineations were derived nationally and selected regionally, from coarse- (national, 300 m) and fine-gridded (local, 30 m) digital elevation models (DEMs).

Although there are several models that seek to describe the behaviour of mercury (Hg) in natural and semi-natural environments, there are no existing catchment-scale models that can be used to link the effects of changing Hg deposition to soil and water Hg processes. Such a model is needed to provide estimates of Hg fluxes from the terrestrial to the aquatic environment and to parameterize models of Hg contamination in aquatic food webs. INCA-Hg is based on the Integrated Catchment biogeochemical modelling framework, and provides a model of mercury cycling and transport in boreal and temperate forested catchments. INCA-Hg simulates three species of Hg: elemental (Hg⁰), ionic (Hg²⁺) and methyl mercury (MeHg). Inputs to the model include land use within catchments and sub-catchments, climate and hydrologic data, and atmospheric deposition of mercury. The model output includes estimates of the flux from the catchment of the simulated Hg species to downstream lakes. We have used INCA-Hg to model mercury transport in several sub-catchments of two headwater lakes in the southern boreal region of Ontario (Harp and Dickie), and have been successful in calibrating the model for both total and methyl mercury at these sites. Our goal is to link INCA-Hg with atmospheric deposition and lake Hg models to provide an integrated whole-catchment capability, and to link INCA-Hg with DOC and sediment transport models already developed in the INCA framework so that DOC- Hg complexes and particulate Hg can be included as chemical species.