Using mercury stable isotope fractionation to identify the contribution of historical mercury mining sources present in downstream water, sediment and fish

Abstract

Ecosystems downstream of mercury (Hg) contaminated sites can be impacted by both localized releases as well as Hg deposited to the watershed from atmospheric transport. Identifying the source of Hg in water, sediment, and fish downstream of contaminated sites is important for determining the effectiveness of source-control remediation actions. This study uses measurements of Hg stable isotopes in soil, sediment, water, and fish to differentiate between Hg from an abandoned Hg mine from non-mine-related sources. The study site is located within the Willamette River watershed (Oregon, United States), which includes free-flowing river segments and a reservoir downstream of the mine. The concentrations of total-Hg (THg) in the reservoir fish were 4-fold higher than those further downstream (>90 km) from the mine site in free-flowing sections of the river. Mercury stable isotope fractionation analysis showed that the mine tailings (δ²⁰²Hg: -0.36‰ ± 0.03‰) had a distinctive isotopic composition compared to background soils (δ²⁰²Hg: -2.30‰ ± 0.25‰). Similar differences in isotopic composition were observed between stream water that flowed through the tailings (particulate bound δ²⁰²Hg: -0.58‰; dissolved: -0.91‰) versus a background stream (particle-bound δ²⁰²Hg: -2.36‰; dissolved: -2.09‰). Within the reservoir sediment, the Hg isotopic composition indicated that the proportion of the Hg related to mine-release increased with THg concentrations. However, in the fish samples the opposite trend was observed-the degree of mine-related Hg was lower in fish with the higher THg concentrations. While sediment concentrations clearly show the influence of the mine, the relationship in fish is more complicated due to differences in methylmercury (MeHg) formation and the foraging behavior of different fish species. The fish tissue δ¹³C and Δ¹⁹⁹Hg values indicate that there is a higher influence of mine-sourced Hg in fish feeding in a more sediment-based food web and less so in planktonic and littoral-based food webs. Identifying the relative proportion of Hg from local contaminated site can help inform remediation decisions, especially when the relationship between total Hg concentrations and sources do not show similar covariation between abiotic and biotic media.

Keywords: abandoned mines; contaminated sites; mercury; reservoirs; stable isotopes.

FIGURE 1

Map showing locations where samples were collected. Fish were collected over a larger geographic area than the abiotic parameters (soil, sediment, and water). Colors denote collection locations of different matrices. Circles represent locations downstream of the mine and diamonds represent background samples.

FIGURE 2

(A): Comparison of least square mean (LSM ± standard error—SE) of YOY Largemouth Bass fish tissue THg concentrations between Cottage Grove Resevoir (n = 14) and downstream sample locations in Willamette River (n = 38; fish length was a signficant covariate in the GLM analysis). (B): Comparison of LSM ±SE of YOY fish tissue δ²⁰²Hg compositions between Cottage Grove Resevoir and downstream sample locations in Willamette River (fish species was a significant covariate in the GLM analysis). (C): Comparison of LSM ±SE of YOY fish tissue Δ¹⁹⁹Hg compositions between Cottage Grove Reservoir and downstream sample locations in Willamette River (fish species was a significant covariate in the GLM analysis).

FIGURE 3

(A) shows the mean with standard error δ¹⁵N versus δ¹³C isotopic compositions in fish sampled in Cottage Grove Reservoir. (B) shows the relationship between tropic position (represented as δ¹⁵N) versus THg concentrations (R² = 0.64, p < 0.0001; error bands represent 95% confidence intervals).

FIGURE 4

(A) showing the δ²⁰²Hg values plotted against the inverse THg concentration (μg/g); and (B) showing the Δ¹⁹⁹Hg versus δ²⁰²Hg values. Note: all soil samples are plotted as diamonds and all sediment samples are plotted as circles. The tailings samples from the old furnace are distinguished from the newer furnace with a heavier black outline.

FIGURE 5

(A) shows the THg concentrations and δ²⁰²Hg values in water samples for particulate and filtered phases and (B) shows this relationship for the Δ¹⁹⁹Hg values in water. The background samples were collected approximately 3 km upstream of the mine in Garoutte Creek (−3 km on the graph). The 0 km location marks the location of the Black Butte Mine and Furnace Creek. Furnace Creek flows into Little River, which was sampled 1 km downstream of the mine. Little River flows into the Coast Fork Willamette (CFW) river and was sampled 9 km downstream of the mine. The Cottage Grove (CG) reservoir is 15 km downstream of the mine and at 19 km downstream a sample was collected below the reservoir on the CFW River. Note that the secondary y-axis in (A) is plotted on a log-scale. Uncertainty of measurements was assessed by the 2SD of processing standard NIST 3133, which was 0.09‰ for both δ²⁰²Hg and Δ¹⁹⁹Hg.

FIGURE 6

(A) shows the δ202HgCOR values for different adult fish species as a function of length normalized THg concentrations in Cottage Grove Reservoir. (B) shows the δ202HgCOR values for different fish species as a function of the δ 15N value. (C) shows the δ202HgCOR values for different fish species as a function of the δ13C value.