S6 Mercury rising in Kejimkujik National Park, Nova Scotia

Tuesday, 26 July, 2011

TS6-O1 — 8:30-8:45
Author: KIDD, Karen1
(1)Canadian Rivers Institute, kiddk@unbsj.ca

A study in the mid 1990s showed that loons (Gavia immer) in Kejimkujik National Park, Nova Scotia had the highest mercury (Hg) concentrations of any population in North America, due to the high Hg concentrations in their main prey yellow perch (Perca flavescens). In 2006 and 2007, we revisited 16 of these acidic lakes (pH ranges from 4.3 to 6.1) to assess whether there have been any changes in perch Hg concentrations over the past decade. In both the 1996/97 and 2006/07 studies, ~ 9 yellow perch from each of 3 size classes (5-10 cm, 10-15 cm, and 15-20 cm) were captured from the lakes in late summer, measured for length and weight, and scales were removed for ageing. Total Hg concentrations were measured in whole body homogenates of the perch, and their trophic position was assessed using nitrogen isotopes (d15N) of perch (both studies) and food web organisms (2006/07 only). Between 1996/97 and 2006/07, Hg concentrations (wet weight) in the perch increased an average of 29 % (range of 10 – 58%) in 10 lakes, decreased by 21% (range of 11 – 36%) in 3 lakes, and did not change in the other three systems (polynomial regressions, Hg versus length). Some of the increases within lakes were related to lower condition (6 lakes) or higher d15N (3 lakes; assuming no changes in baseline over time) of the fish. The percentage increases in fish Hg (mean 2006/mean 1996 x 100%) were greatest in lakes with higher pH (r2=0.39, p=0.01) or lower total organic carbon (ranges of 2.6 to 15.4 mg/L; r2=0.39, p=0.01), aqueous THg (0.8 to 7 ng/L; r2=0.40, p=0.01), conductivity (20 – 40 µS•cm-1; r2=0.79, p=0.02) and aluminum (35 to 234 µg/L; r2=0.45, p=0.005). Although water concentrations of total N (TN) increased and sulfate decreased significantly in most lakes over time, no physical (e.g. lake or watershed area) characteristics or changes in water chemistry explained the temporal trends found in perch. As a result of these as yet unexplained increases, Hg concentrations in perch now pose a greater risk to loon health than a decade ago. In recent years in 12 of 16 lakes, mean Hg concentrations in perch exceeded the 0.21 µg/g ww threshold known to reduce loon productivity by 50%. Results of this study indicate that Kejimkujik National Park remains a Hg “hotspot” in North America.

TS6-O2 — 8:45-9:00
Authors: BATCHELAR, Katharina L.1, KIDD, Karen A.1, MUNKITTRICK, Kelly R.1, DREVNICK, Paul E.2, BURGESS, Neil M.3
(1) Canadian Rivers Institute, University of New Brunswick , k.fischer@unb.ca; (2) INRS-ETE, Université du Québec; (3) Environment Canada.

Kejimkujik National Park and National Historic Site (KNPNHS), located in southwestern Nova Scotia, Canada, has been designated a Hg “hotspot” due to high concentrations of Hg found in common loons (Gavia immer) and their main prey, yellow perch (Perca flavescens) within the park. Concentrations of total Hg in yellow perch in KNPNHS have increased over the last decade, and many currently exceed the tissue threshold effect level for juvenile and adult fish of 0.2 µg Hg g-1 (ww). To determine whether the elevated Hg is negatively affecting the general and reproductive health of this species, sexually mature male and female yellow perch were collected in the fall of 2009 and 2010 from 12 lakes known to contain perch with differing Hg concentrations. Several health variables (condition, liver somatic index (LSI), gonadosomatic index (GSI), macrophage aggregates, and oocyte development) were examined in these fish, and compared between lakes. Mean total mercury concentrations in dorsal muscle of these perch ranged from 0.08 – 2.13 µg Hg g-1 (ww) (n = 17 to 51/lake) across lakes. Significant among-lake, within sex differences in condition and GSI do not appear to be related to muscle Hg concentrations, and no effects of Hg on oocyte development were found. However, LSI is positively related across lakes to total Hg in muscle, and the proportion of macrophage aggregates (an indicator of oxidative stress and tissue damage) found in the liver, kidney and spleen were positively related to total Hg in muscle. Analysis of plasma sex steroid concentrations, and the expression of stress related genes in the liver, is ongoing. The effects of Hg seen in these wild perch may be also occurring in other wild fish populations; a concern given that Hg is a global pollutant.

TS6-O3 — 9:00-9:15
Authors: CLAYDEN, Meredith G.1, KIDD, Karen A.1, KIRK, Jane L.2, MUIR, Derek C.G.2, O’DRISCOLL, Nelson J.3
(1) University of New Brunswick, m.clayden@unb.ca; (2) Environment Canada; (3) Acadia University.

A recent study in Kejimkujik National Park, Nova Scotia, Canada, found that levels of total Hg (THg) in yellow perch (Perca flavescens) increased by an average of 29% in ten lakes between 1996 and 2006 and that methyl Hg (MeHg) concentrations in predatory invertebrates were the strongest predictors of THg in perch across lakes. It is unclear why Hg in yellow perch is increasing in Kejimkujik, since concentrations of this contaminant have declined in biota elsewhere in response to stricter industrial emission controls. In 2009 and 2010, we conducted food web studies on seven lakes in the park representing a range of physical and chemical characteristics. Our aims are: 1) to understand the structure of the food webs; and 2) to determine which biotic or abiotic factors account for the anomalously high levels of Hg in these perch. To do this, we measured MeHg and THg in zooplankton, macroinvertebrates, fishes, sediments and water, as well as sources of energy (using sulphur and carbon isotopes) to and trophic positions (using nitrogen isotopes) of the biota, and chemical characteristics of the lakes. Results of the carbon isotope analyses indicate that perch in these lakes feed mainly on littoral invertebrates. As in other studies, the regression slope of log Hg versus nitrogen isotopes is being used to model Hg biomagnification through the food web of each lake. Although our study lakes are acidic and oligotrophic (pH 4.8-6.0, chlorophyll-a 1.0-2.0 µg/L), regression slopes ranged from 0.18 ± 0.01 to 0.20 ± 0.01 (r2 = 0.64-0.80; p < 0.001), similar to those from more neutral and eutrophic systems. Ongoing analyses are examining whether the y-intercepts of these regressions reflect among-lake differences in MeHg concentrations at the base of the food web. We are also investigating the effects of lake chemistry (eg. pH, DOC, dissolved iron) on biotic Hg concentrations, and conducting sulphur isotope analyses on biota, sediments and water to improve our understanding of Hg pathways to fish. Results of this research will expand our understanding of the factors that affect Hg concentrations at the base of freshwater food webs and will help to explain what is driving the increasing levels of Hg in yellow perch in Kejimkujik National Park.

TS6-O4 — 9:15-9:30
Authors: NASR, Mina1, CLAIR, Tom C. 2, ARP, Paul A. 1
(1) Faculty of Forestry and Environmental Management, University of New Brunswick, mnasr@unb.ca; (2) Environment Canada;

This study presents a spatial and temporal framework for modelling upland-lowland mercury (Hg) transfer in forested watersheds of Kejimkujik National Park, Nova Scotia. With this framework, the two geo-spatial sampling studies were used to identify differences in total organic carbon (TOC), total and methyl Hg concentrations by type of freshwater environment (surface water and groundwater-fed systems) and seasonal high to low flow conditions. It is demonstrated that scavenged atmospheric Hg by the forest canopy is allocated to forest biomass production and partitioning process. Hydrological calculations were used to determine the rate of dissolved organic carbon (DOC) and Hg leaching from the forest soil based on the soil-internal DOC production. The subsequent leaching rate was then partitioned into surface and sub-surface water flow, with the former carrying the soil produced Hg-DOC combination, while the latter loosing this combination through continued DOC and Hg retention within the soil. The model-estimated Hg concentrations were calibrated with the corresponding field determinations for (i) the forest biomass components, (ii) the upland soil layers, and (iii) the concentrations in soil leachates and stream water.

TS6-O5 — 9:30-9:45
Author: VOST, Emma1
(1)Acadia University, emmavost@hotmail.com

Mercury is a globally distributed, toxic environmental contaminant. Quantifying the release of volatile mercury from aquatic ecosystems is integral to predicting mercury-sensitive freshwaters. Inorganic mercury (Hg+2) in freshwater lakes can be reduced to volatile elemental mercury (Hg0) through reactions with dissolved organic matter (DOC) and energy from solar radiation. Current predictive models lack the gross oxidation and reduction rate kinetics of these reactions and how they change with water characteristics. This research will attempt to address this current knowledge gap. Freshwater samples were collected from ten lakes in Kejimkujik National Park, Nova Scotia (spanning a range of DOC concentrations from 3.78- 13.28 mg L-1) in May, 2008, May 2009, and August 2010 (DOC ranging from 2.6-15.4 mg L-1). Samples were frozen until analysis of DOC, total mercury, and mercury photo-reduction and oxidation rates. Both filter-sterilized and unfiltered waters were analyzed using a quartz photoreaction cell in combination with a Luzchem photo-reaction chamber (combined UVA UVB irradiation) and a TEKRAN 2537 gaseous elemental mercury analyser to quantify gross and net reduction rates. Rate constants were derived using a kinetic fitting model for a pseudo first-order reaction. Results from batch experiments and gross reduction experiments on the 2008 lakes showed that unfiltered samples of lake water released more mercury (mean of 279 ± 48 pg) when exposed to UV radiation than 0.2 µm filtered samples (185 ± 37 pg); net photoreduction rate constants were much more variable and had much higher associated error (range from 1.28 x 10-5 to 4.50 x 102 for filtered and 3.20 x 10-1 to 2.97 x 103 for unfiltered) calculated gross oxidation rates (1.42 x 10-3 to 3.42 x 10-3 for filtered, and 1.90 x 10-3 to 2.78 x 10-3 for unfiltered) were not significantly different than the gross reduction rates, suggesting that the system is in equilibrium. DOC was related to rate constants for gross reduction, net reduction, and gross oxidation, and an exponential decay relationship was found to exist (r2 = 0.49, p = 0.0065; r2=0.32, p = 0.065; and r2= 0.39, p = 0.0233; respectively).

TS6-O6 — 9:45-10:00
Authors: O’DRISCOLL, Nelson1, VOST, Emma1, QURESHI, Asif 2, MACLEOD, Matthew2, LUKEMAN, Matthew1, NEUHOLD, York-Michael 2, HUNGERBÜHLER, Konrad2
(1) Acadia University, nelson.odriscoll@acadiau.ca; (2) ETH Zurich;

Photo-reduction and photo-oxidation are fundamental mechanisms controlling mercury volatilization and accumulation in both freshwaters and ocean water. However differences between these processes have not been examined in these ecosystems. Oceans are often regarded as a source of atmospheric mercury in global transport and some lakes are known areas of retention but the reasons for these contrasts were previously not clear. Oceans account for about one-third of total mercury flux to the atmosphere, in the form of dissolved gaseous mercury (DGM) or elemental mercury, which evades from the water surface. In all surface waters DGM is produced as a net result of the reduction of reducible mercury which believed to be primarily (Hg2+) bound to specific carbon-based ligands and the oxidation of elemental mercury ( Hg0). These two processes control the amount of DGM available for evasion across the water-air interface. However, the fundamental rate constants and mechanisms of these reactions in ocean water and freshwaters are poorly studied.

Here, we review recent progress our group has made and present rate constants derived from two recent projects that quantified both the gross photo-reduction and gross photo-oxidation of mercury in water samples from the open Atlantic Ocean and a series of freshwater lakes in Kejimkujik National Park, Nova Scotia, Canada. A Luzchem temperature-controlled photo-reactor was used to irradiate 200 mL filtration sterilized samples in quartz beakers exposed to two intensities of either ultraviolet (UV-A or UV-B) radiation or total UV radiation. Pseudo-first order rate constants were derived as a function of time and cumulative radiation energy. Both abiotic and biotic samples were analyzed to quantify the relative importance of biotic and abiotic redox reactions.

We conclude that reduction and oxidation of mercury in ocean water does not follow a simple reversible reaction pathway; instead, Hg(0) can be oxidized to another (presently unidentified) mercury species that is not available for further reduction. This reaction step plays a major role in the net formation of DGM in surface oceans and leads to the depletion of photo-reducible mercury and DGM over relatively short time frames. In contrast, the net-photo-oxidation rates for freshwaters are so low that it appears that reduction and oxidation are very close to being in balance and there is no evidence of a depletion of photo-reducible mercury over the 24 time frame of the experiments.

TS6-O7 — 10:00-10:15
Authors: PANNU, Ravinder1, O’DRISCOLL, Nelson2, SICILIANO , Steve 1, RENCZ, Andy3
(1)University of Saskatchewan, r.pannu@usask.ca; (2) Acadia University; (3) Natural Resources Canada.

Elemental mercury is a volatile metal at standard temperatures which can be transformed into several species in ecosystems some of which are persistent, bio-accumulative and highly toxic. Natural emissions and re-emissions of mercury from soils have been identified as a major contributor to the global mercury budget and conservative estimates of global mercury fluxes suggest a total of 700 to 1000 tonnes anumm-1 volatilized from soils (Lindqvist, 1991). The growth of research in soil Hg emissions has made us recognize a large number of uncertainties associated with our estimation of the contribution of soil Hg emissions to the atmospheric Hg pool and the effects on global Hg cycling. Quantification of mercury flux from natural soils as influenced by environmental variables presents a large gap in current predictive models for a better understanding the role of natural surfaces in the mercury biogeochemical cycling and developing emission budget.

This research quantifies the effect of soil moisture (measured as percent water filled pore space (WFPS) - 15, 30, 45 and 60%) on the kinetics of elemental mercury Hg(0) released from 14 soils collected from Kejimkujik National Park and Antigonish County, NS, Canada. The samples represent a wide range of texture, pH, organic matter and total mercury contents to cover different types of soils present in the region. Quartz beaker chamber was used to measure mercury flux by manipulating variables of interests under controlled conditions Mercury flux increased from 15% to 30% to 45% WFPS and decreases at 60%. The soils maintained at 45% WFPS were found to facilitate greatest release of elemental mercury and negligible mercury was emitted as the soils tend to attain near saturation levels. First-order exponential reaction for values (the reaction rates constant) depicting mercury release kinetics from soils derived by fitting the first order exponential curve to cumulative Hg released over 24 hour period were found to decreases as the WFPS increases from 15% to 45% indicating limited amount of water addition tend to liberate more elemental mercury but over a longer period of time. We will also present the preliminary data on the effect of soil surface temperature manipulations (5, 10, 20 and 30 oC) on mercury fluxes from the same soil samples.

TS6-O8 — 10:15-10:30
Authors: EDMONDS, Samuel1, O’DRISCOLL, Nelson2, EVERS, David1, HILLIER, Kirk2
(1) BioDiversity Research Institute, sam.edmonds@briloon.org; (2) Acadia University;

Mercury burdens to the Rusty Blackbird were examined across the range of this rapidly disappearing species to identify the regions and seasons with the highest exposure and to determine if the contaminant represents a realistic threat to the species. The far eastern breeding region including the Canadian Maritimes and northern New England States had the greatest exposure to the species (geometric mean 0.94 µg g-1 Hg in blood, 8.26 µg g-1 Hg in feathers), over an order of magnitude greater than the winter regions sampled in the southern U.S. and > 3x that of the other breeding region sampled, Alaska. The greatest blood-Hg concentration (3.4 µg g-1) was for a male breeding in the Kejimkujik/Tobeatic region of southern Nova Scotia—an area that has previously been identified as having high Hg exposure to wildlife. Overall, a large proportion of the breeding population in the Northeast exceeds estimated adverse effects levels based on both blood and feather Hg concentrations, supporting the case for Hg as a contributing factor to the ongoing population decline. Invertebrates and water samples from Northeastern wetlands with breeding adults were collected for analysis of MeHg and THg, along with six additional water column parameters, to identify those water characteristics that promote MeHg production and bioavailability at these sites. DOC (mean part correlate = –0.20), pH (–0.20), and DO2 (–0.38) were frequently related to bird and invertebrate MeHg concentrations, suggesting they most influenced bioavailability of MeHg at these sites. Redox potential (part correlate = 0.76) and DO2 (–0.80) were strongly correlated with MeHg production (based on water %MeHg). Of four assumed prey invertebrate taxa, MeHg concentrations of three were significantly related to bird blood-Hg concentrations (Araneae, r2 = 0.49; Ephemeroptera, r2 = 0.84; Trichoptera, r2 = 0.34), with Odonata or two assumed non-prey taxa not significantly correlated (Corixidae and Gerridae). A combination of a habitat having a high bioavailability and the species’ high trophic position likely explains the observed elevated Hg concentrations in the Northeast.

Tuesday, 26 July, 2011