Mercury in the pelagic food web of Lake Champlain

Abstract

Lake Champlain continues to experience mercury contamination resulting in public advisories to limit human consumption of top trophic level fish such as walleye. Prior research suggested that mercury levels in biota could be modified by differences in ecosystem productivity as well as mercury loadings. We investigated relationships between mercury in different trophic levels in Lake Champlain. We measured inorganic and methyl mercury in water, seston, and two size fractions of zooplankton from 13 sites representing a range of nutrient loading conditions and productivity. Biomass varied significantly across lake segments in all measured ecosystem compartments in response to significant differences in nutrient levels. Local environmental factors such as alkalinity influenced the partitioning of mercury between water and seston. Mercury incorporation into biota was influenced by the biomass and mercury content of different ecosystem strata. Pelagic fish tissue mercury was a function of fish length and the size of the mercury pool associated with large zooplankton. We used these observations to parameterize a model of mercury transfers in the Lake Champlain food web that accounts for ecosystem productivity effects. Simulations using the mercury trophic transfer model suggest that reductions of 25-75% in summertime dissolved eplimnetic total mercury will likely allow fish tissue mercury concentrations to drop to the target level of 0.3 μg g(-1) in a 40-cm fish in all lake segments. Changes in nutrient loading and ecosystem productivity in eutrophic segments may delay any response to reduced dissolved mercury and may result in increases in fish tissue mercury.

Fig. 1

Location of lake-segments sampled on Lake Champlain. Lake Champlain is situated primarily within the United States on the New York and Vermont border, with the northern end of the lake extending into Québec Province, Canada. Missisquoi Bay (north east end) and the South Lake (southern end) segments are the most eutrophic waters. Mallets Bay (east central) contains the most oligotrophic waters in the lake system.

Fig. 2

General linear model for (a) fish tissue mercury in µg g⁻¹ wet weight as a function of (b) fish length in mm×0.0022093 ± 0.000244 (p = 0.0003) and (c) the pool of total mercury associated with large (> 202 µm) zooplankton in pg L⁻¹×0.3857262 ± 0.091475 (p = 0.0084), intercept = −0.830067 ± 0.153292 (p = 0.0029), overall p = 0.0008, n = 8, adjusted r² = 0.94). Mallets Bay (open circle in panel a is an outlier. The fish tissue mercury concentration in Mallets Bay is estimated from the ratio of the observed to predicted value from the general model (see Table 3).

Fig. 3

Schematic diagram of the empirical mercury trophic-transfer model developed for the pelagic ecosystem of Lake Champlain detailed in table 3. As described in the text and table 3, Secchi depth serves as proxy for primary productivity. The specific absorbance at 254 nm is related to the balance between algal vs. terrestrial sources of DOC and is thus modified by primary productivity. The dissolved total mercury (DissTHg) in the system may vary with changes in atmospheric deposition and tributary loading. Piscivore tissue mercury (PiscivoreHg ppm) is a function of fish length, primary productivity and dissolved mercury levels as modulated by the biomass dynamics of the zooplankton community.