Mark J. Peterson

Long-term increased bioaccumulation of mercury in largemouth bass follows reduction of waterborne selenium.

Average mercury concentrations in largemouth bass from Rogers Quarry in east Tennessee were found to increase steadily following the elimination of selenium-rich discharges of fly ash to the quarry in 1989. From 1990 to 1998, mean mercury concentrations (adjusted to compensate for the covariance between individual fish weight and mercury concentration) in bass rose from 0.02 to 0.61 mg/kg. There was no indication that the rate increase was slowing or that mercury concentrations in fish were approaching a plateau or steady state. Mean selenium concentrations in bass declined from 3 to 1 mg/kg over the first five years of the study, but remained at 1-1.5 mg/kg (about twice typical concentrations in bass from local reference sites) for the last three years of the study. Gross physical abnormalities were common in fish from the site in the first three years after elimination of fly ash discharges but disappeared after two more years. Although it remains possible that other chemical or physical changes related to fly ash disposal in the system were associated with increased mercury bioaccumulation, the most likely explanation is that selenium played a critical role. It appears as though aqueous selenium enrichment was capable of having a profound effect on mercury bioaccumulation in this system but at the cost of causing a high incidence of gross abnormalities in fish. However, it is possible that selenium concentrations between the national ambient water quality criterion for the protection of aquatic life, 5 microg/l, and that now found in Rogers Quarry (<2 microg/l) could reduce mercury bioaccumulation without causing adverse effects on aquatic biota and fish-eating wildlife.

Form of mercury in stream fish exposed to high concentrations of dissolved inorganic mercury

The form of mercury predominating in mercury-contaminated fish from both pristine and industrialized waters in North America and Europe has almost universally been methylmercury. Sunfish (Lepomis auritus) living in a stream contaminated with 0.5-1 micrograms/L dissolved inorganic mercury accumulated greater concentrations of total mercury at headwater sites, where the dissolved mercury concentrations were greatest, than they did at downstream sites. However, despite evidence from laboratory studies that dissolved inorganic mercury is rapidly accumulated by fish without transformation to methylmercury, methylmercury constituted 85% or more of the total mercury concentration in fish at all sites.

ASSESSING RECOVERY IN A STREAM ECOSYSTEM: APPLYING MULTIPLE CHEMICAL AND BIOLOGICAL ENDPOINTS

Recovery dynamics in a previously disturbed stream undergoing remediation was assessed by measuring several chemical and biological endpoints. Three major trophic levels (periphyton, benthic macroinvertebrates, and fish communities) were analyzed along with various biological indicators of fish health including sub-organismal (biochemical and physiological) and individual-level responses. Ambient toxicity testing, water quality measurements, and fish bioaccumulation studies were also conducted to provide additional indicators of stream recovery. Information on chemical loading, biological exposure (bioaccumulation), and bioassays help identify the underlying or mechanistic basis of recovery, while endpoints measured on instream biota, and particularly those at higher levels of biological organization, are key responses for understanding the dynamics and ecological significance of recovery. When assessing recovery in aquatic ecosystems, it is important to conduct long-term field studies incorporating a variety of response variables that represent a wide range of sensitivities and response time scales. The value of different endpoints for assessing recovery depends on the inherent characteristics that are related to response variability and sensitivity, specificity and causal relationships to stressors, and ecological relevance. Long-term data sets incorporating a variety and range of these endpoints are needed to improve our understanding of natural variability in streams and provide a baseline against which effects of disturbance and subsequent recovery processes can be evaluated. The complexity of aquatic systems and their variable recovery dynamics suggest that no single measure is adequate for assessing aquatic ecosystem recovery and that a suite of chemical and biological endpoints is required for a more complete understanding of ecosystem dynamics and status during both the recovery and the post-disturbance periods. Such information should be valuable to environmental managers and regulators in helping to make more informed decisions regarding effective management and mitigation practices in disturbed systems.

Response of mercury contamination in fish to decreased aqueous concentrations and loading of inorganic mercury in a small stream.

Approximately 250 000 kg of mercury was lost towater and soils at the U.S. Dept. of Energy Y-12 Plantin Oak Ridge, Tennessee in the 1950s and early 1960s. A creek originating within the plant receivedcontinuous inputs of waterborne mercury, predominantlyas dissolved inorganic mercury, from groundwater,streambed contamination, and sump and process waterdischarges to the contaminated storm sewer network.These produce aqueous total mercury concentrations of1–2 μg L-1 in the upper reaches of the stream,decreasing to about 0.1–0.2 μg L-1 in its lowerreaches. A program to reduce mercury concentrationsin the creek identified specific sources (buildingsumps, contaminated springwater seeps, foundationdrains, and contaminated piping) and rerouted wateraround contaminated portions of the drain system orcollected and treated mercury-contaminated waterbefore discharging it. As a result, waterbornemercury concentrations in the creek and total mercuryloading were reduced from 1.8 μg L-1 to0.6 μg L-1 and 100 to 20 g d-1, respectively, in the last 5 yr.Mean mercury concentrations in fish nearest sourceareas in the creek headwaters decreased at roughly thesame rate as waterborne total mercury concentrationsover the past five years, but at the facility boundarydownstream the decline in mercury bioaccumulation wasmuch less. At sites 5–15 km farther downstream, nodecrease was evident. Dissolved methylmercury tendedto increase with distance downstream in a patterninverse to that noted for its dissolved inorganicmercury precursor.Improvements in water quality and modification ofweirs to allow the passage of fish have resulted inthe establishment of large populations of fish inmercury-contaminated headwater areas previously devoidof fish. It may be that the accumulation, retention,and eventual downstream transport of this reservoir ofbiologically incorporated methylmercury has acted tobuffer against expected reductions in mercury in fishat downstream sites.

Changes in concentrations of selenium and mercury in largemouth bass following elimination of fly ash discharge to a quarry

Elimination of slurried fly ash discharges to a water-filled quarry was followed by a steady increase in concentrations of mercury in the axial muscle of resident largemouth bass (Micropterus salmoides). Average mercury concentrations in bass (adjusted for covariance with fish weight) increased from 0.02 micrograms/g to 0.17 micrograms/g in three years. Aqueous selenium concentrations in the quarry decreased from 25 micrograms/L to < 2 micrograms/L after elimination of fly ash discharges, but selenium concentrations in bass remained about three times background levels. Previous studies have shown selenium addition to be a viable means of ameliorating mercury contamination in fish in low alkalinity, low pH waters of northern Europe and Canada. These results suggest that selenium may also be effective at blocking the accumulation of methylmercury in harder, more alkaline waters.

Bioaccumulation Factors for Mercury in Stream Fish

The bioaccumulation of methylmercury in fish is a complex process affected by many site-specific environmental factors. The US Environmental Protection Agency (USEPA) recently recommended changing the basis for expressing the ambient water quality criterion for mercury from an aqueous concentration to a measure of the methylmercury concentration in fish. This change would make the regulation of mercury in surface waters a site-specific exercise in which fish-based bioaccumulation factors (BAF; the ratio of mercury concentration in fish to the concentration of mercury in water) are used to calculate aqueous concentration limits for total mercury. These limits would then be used to allocate mercury loading among various point and nonpoint sources and guide regulatory actions. In order for this approach to succeed, it is critical that the site-specific BAFs and methylmercury:total mercury conversion factors be independent of aqueous total mercury concentration (HgT). We investigated this relationship by measuring aqueous methylmercury and HgTs and mercury in fish in ecologically similar warm-water streams in the southeastern United States. Bioaccumulation factors based on HgT in water were found to decrease with increasing HgT, primarily as a consequence of the reduction in the ratio of aqueous methylmercury to total mercury with increasing HgT. Methylmercury-based BAFs did not vary as a function of HgT. The implication of this relationship is that site-specific determination of aqueous HgT limits at contaminated sites may use BAFs that would be underestimates of the appropriate BAFs to describe mercury bioaccumulation in the system after mercury inputs have been reduced. In such cases, regulatory limits set using site-specific BAFs might not achieve their intended purpose of reducing mercury contamination in fish to acceptable concentrations.

Evasion of added isotopic mercury from a northern temperate lake

Isotopically enriched Hg (90% 202Hg) was added to a small lake in Ontario, Canada, at a rate equivalent to approximately threefold the annual direct atmospheric deposition rate that is typical of the northeastern United States. The Hg spike was thoroughly mixed into the epilimnion in nine separate events at two-week intervals throughout the summer growing season for three consecutive years. We measured concentrations of spike and ambient dissolved gaseous Hg (DGM) concentrations in surface water and the rate of volatilization of Hg from the lake on four separate, week-long sampling periods using floating dynamic flux chambers. The relationship between empirically measured rates of spike-Hg evasion were evaluated as functions of DGM concentration, wind velocity, and solar illumination. No individual environmental variable proved to be a strong predictor of the evasion flux. The DGM-normalized flux (expressed as the mass transfer coefficient, k) varied with wind velocity in a manner consistent with existing models of evasion of volatile solutes from natural waters but was higher than model estimates at low wind velocity. The empirical data were used to construct a description of evasion flux as a function of total dissolved Hg, wind, and solar illumination. That model was then applied to data for three summers for the experiment to generate estimates of Hg re-emission from the lake surface to the atmosphere. Based on ratios of spike Hg to ambient Hg in DGM and dissolved total Hg pools, ratios of DGM to total Hg in spike and ambient Hg pools, and flux estimates of spike and ambient Hg, we concluded that the added Hg spike was chemically indistinguishable from the ambient Hg in its behavior. Approximately 45% of Hg added to the lake over the summer was lost via volatilization.

Trace Element Contamination in Benthic Macroinvertebrates from a Small Stream near a Uranium Mill Tailings Site

Direct measurement of the accumulation of non-radioactive traceelements in aquatic biota near uranium mining or processing sites has been relatively rare, with greater focus on the radiological activity in the adjacent soils and groundwater. To evaluate the potential ecological concern associated with trace elements at a former uranium mill site in southeasternUtah, benthic macroinvertebrates were collected and analyzed for 17 trace elements from multiple locations within a small on-site stream, Montezuma Creek, and a nearby reference stream. Key questions of this study relate to the spatial and temporalextent of contamination in aquatic biota, the potential ecological risks associated with that contamination, and the usefulness of benthic macroinvertebrates as a monitoring tool at this site. Composite samples of similar macroinvertebrate taxa and functional feeding groups were collected from each site over atwo year period that was representative of normal and dry-yearconditions. In both years, mean concentrations of arsenic,molybdenum, selenium, and vanadium were significantly higher (afactor of 2–4 times; P < 0.05) in macroinvertebrates collectedfrom one or both of the two Montezuma Creek sites immediatelydownstream of the mill tailing site in comparison toconcentrations from reference locations. Mean uraniumconcentrations in invertebrates immediately downstream of themill site were more than 10 times higher than at reference sites.The site-to-site pattern of contamination in Montezuma Creekinvertebrates was similar in 1995 and 1996, with mill-relatedtrace elements showing a downstream decreasing trend. However,nine of seventeen contaminant concentrations were higher in thesecond year of the study, possibly due to a higher influx of deepgroundwater during the drier second year of the study. Apreliminary assessment of ecological risks, based on the benthicmacroinvertebrate bioaccumulation data, suggests that aquatic andterrestrial population risks are low. Benthic macroinvertebratesappeared to be sensitive integrators of trace element inputs tothe aquatic environment from a former uranium mill tailing site,and provided useful spatial and temporal patterns ofcontamination not easily obtained using conventional surfacewater or groundwater measures.

Monitoring mercury in fish in a stream system receiving multiple industrial inputs.

Sunfish and a minnow species were used as indicators of anthropogenic mercury contamination in an east Tennessee stream system receiving multiple point and non-point discharges. The monitoring of bluegill (Lepomis macrochirus) and redbreast sunfish (Lepomis auritus) identified bioavailable mercury near three geographically separate industrial facilities, and was able to detect decreases in contamination with distance away from these facilities. In general, total mercury concentrations in the tissue of sunfish in this study were low in comparison to the most commonly cited human health threshold limits, although concentrations at some sites exceeded 1 μg/g. Caged blacknose dace (Rhinichthys atratulus) were monitored in conjunction with resident fish as an indicator of more discrete sources in selected headwater streams where fish movement was deemed a potential factor affecting mercury body burdens. Mercury concentrations in muscle tissue of caged dace after 12 weeks exposure were generally low (<0.2 μg/g) at all sites but higher than in fish from reference streams. mercury accumulation varied between species (sunfish vs. dace) and monitoring method (caged vs. resident) at the same site, with sunfish tending to accumulate higher concentrations of mercury than resident dace which, in turn, contained about twofold higher concentrations than caged dace. However, the site-to-site pattern of mercury accumulation was similar. This study demonstrates the utility of using small stream dwelling fish with restricted home ranges as a tool for identifying and evaluating the bioavailability of mercury sources in large industrial or urbanized settings.

Monitoring Fish Contaminant Responses to Abatement Actions: Factors that Affect Recovery

Monitoring of contaminant accumulation in fish has been conducted in East Fork Poplar Creek (EFPC) in Oak Ridge, Tennessee since 1985. Bioaccumulation trends are examined over a twenty year period coinciding with major pollution abatement actions by a Department of Energy facility at the stream’s headwaters. Although EFPC is enriched in many contaminants relative to other local streams, only polychlorinated biphenyls (PCBs) and mercury (Hg) were found to accumulate in the edible portions of fish to levels of human health concern. Mercury concentrations in redbreast sunfish were found to vary with season of collection, sex and size of individual fish. Over the course of the monitoring, waterborne Hg concentrations were reduced >80%; however, this did not translate into a comparable decrease in Hg bioaccumulation at most sites. Mercury bioaccumulation in fish did respond to decreased inputs in the industrialized headwater reach, but paradoxically increased in the lowermost reach of EFPC. As a result, the downstream pattern of Hg concentration in fish changed from one resembling dilution of a headwater point source in the 1980s to a uniform distribution in the 2000s. The reason for this remains unknown, but is hypothesized to involve changes in the chemical form and reactivity of waterborne Hg associated with the removal of residual chlorine and the addition of suspended particulates to the streamflow. PCB concentrations in fish varied greatly from year-to-year, but always exhibited a pronounced downstream decrease, and appeared to respond to management practices that limited episodic inputs from legacy sources within the facility.