Milltown Reservoir Research Articles

Relative effectiveness of CaCO3 and Ca(OH)2 in minimizing metals solubility in contaminated sediment

Reclamation and re-vegetation of excavated sediments can represent major environmental challenges. This study was conducted to compare CaCO3 and Ca(OH)2 as amendments for pH control in the environmental management of a circum-neutral Mn-, Zn-, and Cu-contaminated sediment in which re-vegetation was unsuccessful. The specific objectives were to evaluate the relative impacts of these treatments on localization and solubility of Mn, Zn, and Cu, as well as potential bioavailability. Sediment samples were collected from Milltown Reservoir, Montana, USA and analyzed before, during, and after oxidative incubation. The mineralogical analyses included composition (X-ray diffraction), control of metals solubility (by chemical equilibrium), and localization of Cu, Mn, and Zn (SEM-EDS; selective extraction procedures). The concentrations of Mn, Zn, and Cu in solution were determined by ICP-OES and atomic absorption spectroscopy. The sediment was composed principally of quartz, feldspars, phyllosilicates, gypsum, trace sulfide minerals, and 15.5 g kg−1 organic carbon. Upon CaCO3 treatment, pH increased from 7.0 to 7.5 from day 1 to day 48, compared to the Ca(OH)2 treatment which resulted in an initial pH > 11, subsequent decrease to pH 8–8.5 within the same period, and the concomitant precipitation of secondary carbonate and silicate phases. Due to the different chemistries of the CaCO3 and Ca(OH)2 reaction systems, Ca(OH)2 was significantly (p < 0.05) more effective than CaCO3 in reducing soluble, ion-exchangeable, and adsorbed Mn and Zn concentrations. The solubilities of Mn and Zn were controlled by rhodochrosite and willemite, respectively. Slight increases in Cu solubility with the Ca(OH)2 treatment were attributable to partially non-reversible increased dissolved organic matter and Cu-organic complex concentrations. The results of the study reflected the different chemistries of the CaCO3 and Ca(OH)2 systems. With the Cu-, Mn-, and Zn-polluted Milltown sediment, the Ca(OH)2 treatment offered significant advantages compared to the CaCO3 treatment in metals immobilization; however, each reclamation site can have unique sediment properties and should be considered independently. The solubilities of Cu-, Mn-, and Zn- minerals controlling equilibria are highly pH dependent; hence, pH control may be an essential component of effective site management for circum-neutral sediments.

English

The of Montana will restore the confluence of Clark Fork and Blackfoot rivers following the removal of the Milltown Dam and some of the contaminated sediments located there. The Restoration Plan for the Clark Fork River and Blackfoot River Near Milltown Dam - October 2005, describes the restoration actions that are currently being planned (WestWater Consulting et al., 2005). In order for the restoration activities to occur, a portion of the project needs to be integrated with the EPA Superfund remedy for the Milltown Reservoir Sediments Operable Unit (USEPA, 2004). EPA, the State of Montana, ARCO, Northwestern, and Envirocon all participated in this integration process through the negotiation of the consent decree and it's attachments, principally the Statement of Work, which outlines who will implement various aspects of the remedial and restoration actions (Envirocon, 2005). The integration limits full restoration of all the natural river processes, but this is a concession that is outweighed by the benefits and cost savings of integration. All river and stream restoration projects have site constraints, e.g., landownership, roads, property lines, etc. The Milltown restoration project's integration with remediation results in additional constraints beyond the typical ones: legal constraints required by a consent decree, contaminated material remaining within the restoration work area, geographic limits to the project area, and negotiated actions. In order, to be successful the State and the other natural resource trustees have established goals and objectives for this project that considers the integration.

Quantifying Metal Contributions from Multiple Sources to the Clark Fork River, Montana, U.S.A.

Identifying and quantifying the contributions of multiple sources of trace elements to stream sediments in a basin containing several possible inputs presents a unique problem related to the investigation of rivers impacted by industrial activity. A multi-source dilution–mixing model was developed and applied to determine the relative contributions to As, Cu and Pb burdens in the Clark Fork River, Montana, a recipient of historical mine wastes as a result of over a century of mining and milling operations. The results identified the Flint Creek drainage as a major source of anthropogenic As (47%) and Pb (35%) to sediments of the Clark Fork River and the Milltown Reservoir, in addition to the major sources associated with mining operations in Butte, MT. The Little Blackfoot River also contributes anthropogenic As (3%) and Pb (4%) to the Clark Fork River, while minor inputs of Cu (1%) and Pb (2%) emanate from the Blackfoot River. The model allows source quantification, and an understanding of the fate and transport of mine wastes in a basin, allowing identification and eventual prioritization of sites destined for remediation.

Food chain analysis of exposures and risks to wildlife at a metals-contaminated wetland.

A food chain analysis of risks to wetland receptors was performed in support of a baseline ecological risk assessment at the Milltown Reservoir Sediments Superfund site in Montana. The study area consisted of over 450 acres of primarily palustrine wetland contaminated with metals from mining wastes transported from upstream sources (average of 465 mg/kg for Cu in sediments, and 585 mg/kg in soils). The food chain analysis focused on several species of terrestrial and semiaquatic animals indigenous to montane wetlands of the northern Rocky Mountains. Receptors consisted of mice, voles, muskrats, beaver, various waterfowl species, osprey, bald eagles, and deer. Samples of aquatic and terrestrial invertebrates, small mammal tissues, fish tissue, aquatic and terrestrial vegetation, soils, sediment, and surface water were collected and analyzed for As, Cd, Cu, and Zn. A linear multimedia food-chain model was constructed to estimate daily intakes of the metals for each receptor, with assumed values for ingestion of aquatic and terrestrial food items, ingestion of local surface water, and incidental ingestion of soils and/or sediments. Evaluation of health risks to the receptors was performed by comparison of exposures expressed as daily intakes to a suite of toxicity values. The range of values consisted of the lower end of chronic toxicity data found in toxicology databases or the literature for the same or similar species, modified to account for extrapolation uncertainties. Daily intakes of chemicals of concern were below or within the range of toxicity values for all receptors. The weight of evidence from the food chain analysis and earlier bioassessment and ecological studies suggest that the health of the wetland receptors is at minimal risk due to the presence of elevated metals in sediments, upland soils, water, or food items at the site.

Sampling Benthic Communities for Sediment Toxicity Assessments Using Grab Samplers and Artificial Substrates

In this study we compared macrobenthic communities collected from the Saginaw River (Michigan), Buffalo River (New York), Indiana Harbor (Indiana), Little Scioto River (Ohio), and Mill-town Reservoir (Montana) at the same locations using artificial substrates and grab samplers. Oligochaete worms ( Limnodrilus hoffmeisteri) were the predominant benthic organisms in sediment samples collected with grab samplers. Artificial substrates also collected nothing but oligochaetes, but only at the stations with the most contaminated sediment in Indiana Harbor and the Little Scioto River. Artificial substrate samples from the Saginaw River, Buffalo River, and Milltown Reservoir generally contained a more diverse assemblage of benthic taxa. Routine assessments of sediment toxicity based on analyses of benthic macroinvertebrate communities should sample all available benthic habitats. Sediment samples collected with a grab sampler and artificial substrates should both be used for evaluating benthic communities in sediment toxicity assessments.

Precipitation of arsenic during bacterial sulfate reduction

Contaminated sediments from the Milltown Reservoir in western Montana release arsenic and various heavy metals (e.g., Cu, Cd, Pb, Zn, Mn) into an underlying alluvial aquifer as redox conditions in the sediments change with seasonally fluctuating water levels. Porewater analyses indicate that sulfate is depleted with depth. In this study, the feasibility of inducing As(III) precipitation through bacterial reduction of sulfate was evaluated in laboratory microcosms established under strictly anaerobic conditions. As(lII), Fe(II), and sulfate concentrations were routinely monitored in the aqueous phase as sulfate was reduced to sulfide. Both As(III) and Fe(II) concentrations in the sediment microcosms decreased as sulfide was made available. Energy‐dispersive x‐ray (EDS) analysis indicated that some of the arsenic was precipitated as an iron‐arsenic‐sulfide solid phase. The precipitation of arsenic observed in this laboratory study suggests that bacterial sulfate reduction may be a process by which heavy met...

Chemical characterization of sediments and pore water from the upper Clark Fork river and milltown reservoir, Montana

AbstractThe upper Clark Fork River basin in western Montana is widely contaminated by metals from past mining, milling, and smelting activities As part of a comprehensive ecological risk assessment for the upper Clark Fork River, we measured physical and chemical characteristics of surficial sediment samples that were collected from depositional zones for subsequent toxicity evaluations Sampling stations included five locations along the upper 200 km of the river, six locations in or near Milltown Reservoir (about 205 km from the river origin), and two tributary reference sites Concentrations of As, Cd, Cu, Mn, Pb, and Zn decreased from the upper stations to the downstream stations in the Clark Fork River but then increased in all Milltown Reservoir stations to levels similar to uppermost river stations Large percentages (50 to 90%) of the total Cd, Cu, Pb, and Zn were extractable by dilute (3 n) HCl for all samples Copper and zinc accounted for greater than 95% of extractable metals on a molar basis Acid‐volatile sulfide (AVS) concentrations were typically moderate (0 6 to 23 μmol/g) in grab sediment samples and appeared to regulate dissolved (filterable) concentrations of Cd, Cu, and Zn in sediment pore waters Acid volatile sulfide is important in controlling metal solubility in the depositional areas of the Clark Fork River and should be monitored in any future studies Spatial variability within a sampling station was high for Cu, Zn, and AVS, therefore, the potential for toxicity to sediment dwelling organisms may be highly localized

Characterization of ecological risks at the milltown reservoir‐clark fork river sediments superfund site, montana

AbstractA comprehensive field and laboratory approach to the ecological risk assessment for the Milltown Reservoir‐Clark Fork River Sediments Site, a Superfund site in the Rocky Mountains of Montana, has been described in the preceding reports of this series. The risk assessment addresses concerns over the ecological impacts of upstream releases of mining wastes to fisheries of the upper Clark Fork River (CFR) and the benthic and terrestrial habitats further downstream in Milltown Reservoir. The risk characterization component of the process integrated results from a triad of information sources: (a) chemistry studies of environmental media to identify and quantify exposures of terrestrial and aquatic organisms to site‐related contaminants; (b) ecological or population studies of terrestrial vegetation, birds, benthic communities, and fish; and (c) in situ and laboratory toxicity studies with terrestrial and aquatic invertebrates and plants, small mammals, amphibians, and fish exposed to contaminated surface water, sediments, wetland soils, and food sources. Trophic transfer studies were performed on waterfowl, mammals, and predatory birds using field measurement data on metals concentrations in environmental media and lower trophic food sources. Studies with sediment exposures were incorporated into the Sediment Quality Triad approach to evaluate risks to benthic ecology. Overall results of the wetland and terrestrial studies suggested that acute adverse biological effects were largely absent from the wetland; however, adverse effects to reproductive, growth, and physiological end points of various terrestrial and aquatic species were related to metals exposures in more highly contaminated depositional areas. Feeding studies with contaminated diet collected from the upper CFR indicated that trout are at high risk from elevated metals concentrations in surface water, sediment, and aquatic invertebrates. Integration of chemical analyses with toxicological and ecological evaluations of metal effects on the wetland and fishery has provided an important foundation for environmental decisions at this site.

Ecological assessment for the wetlands at milltown reservoir, missoula, montana: Characterization of emergent and upland habitats

AbstractWetlands in mining districts in the western United States are frequently impacted by heavy metal contaminated sediments The present study summarizes a soil contamination evaluation and an ecological assessment completed for a Superfund site located at Milltown Reservoir wetlands (MRW) in western Montana For wetlands, as well as upland habitats adjacent to wetlands, biological evaluations in the field and laboratory should be considered critical components in the ecological risk assessment process Depending upon habitat type, field and laboratory methods have been developed for hazard and risk assessment that lend themselves directly to the Superfund ecological risk assessment process, and that consider contaminant bioavailability and subtle expressions of adverse biological effects associated with chronic exposures As part of an ecological risk assessment for MRW, field surveys and a variety of biological test methods (e g., terrestrial and aquatic tests) were critical to the wetland evaluation For evaluating heavy metal effects at MRW, field and laboratory methods within the ecological assessment included wetlands delineation and preliminary plant and wildlife survey, vegetation tests in emergent and upland habitats, soil macroinvertebrate (earthworm) tests, preliminary studies using amphibian and bacterial test systems, soil characterizations, and chemical analysis of soils, sediments, and biological materials In conjunction with chemical analyses, these biological and ecological evaluations yielded an integrated evaluation of ecological effects and exposure at MRW The data gathered from laboratory and field work at MRW suggested that biological and ecological effects were subtle in their expression in the wetland In conjunction with sediment contamination evaluations, these studies should reduce the uncertainty associated with the baseline ecological risk assessment for MRW

Toxicity of metal‐contaminated sediments from the upper clark fork river, montana, to aquatic invertebrates and fish in laboratory exposures

AbstractSediments of the upper Clark Fork River, from the Butte and Anaconda area to Milltown Reservoir (230 km downstream), are contaminated with As, Cd, Cu, Pb, Mn, and Zn primarily from mining activities. The toxicity of pore water from these sediments was determined usingDaphnia magna(48‐h exposure), rainbow trout (96‐h exposure), and Microtox®. However, pore‐water data from these exposures were questionable because of changes in the toxicity of pore‐water samples after 5 to 7 d of storage. Whole‐sediment tests were conducted withHyalella azteca(28‐d exposure),Chironomus riparius(14‐d exposure), rainbow trout (Oncorhynchus mykiss) 21‐ to 28‐d exposure andDaphnia magna(2‐ to 22‐d exposure). Sediment samples from Milltown Reservoir and the Clark Fork River were not generally lethal to test organisms. However, both reduced growth and delayed sexual maturation of amphipods were associated with exposure to elevated concentrations of metals in sediments from the reservoir and river. Relative sensitivity (most sensitive to least sensitive) of organisms in whole‐sediment toxicity tests was:Hyalella azteca &gt; Chironomus riparius&gt; rainbow trout &gt;Daphnia magna. Relative sensitivity (most sensitive to least sensitive) of the three end points evaluated withHyalella aztecawas: length &gt; sexual maturation &gt; survival. The lack of lethal effects on organisms may be related to temporal differences in sediment, acid‐volatile sulfide, or organic carbon.

Use of benthic invertebrate community structure and the sediment quality triad to evaluate metal-contaminated sediment in the upper clark fork river, montana

The upper Clark Fork River, above Flathead River, is contaminated with large amounts of As, Cd, Cu, Pb, Mn, and Zn ores from past mining activities. The contaminated area extends from the Butte and Anaconda area to at least 230 km downstream to Milltown Reservoir. Both the upper Clark Fork River and Milltown Reservoir have been designated as U.S. Environmental Protection Agency Superfund sites because of metal-contaminated bottom sediments. We evaluated the impacts of past mining activities on the Clark Fork River ecosystem using benthic invertebrate community assessment, residue chemistry, and toxicity testing. Oligochaeta and Chironomidae generally accounted for over 90% of the benthic invertebrate community in the soft sediment depositional areas. Taxa of Oligochaeta and Chironomidae were predominantly pollution tolerant. Higher numbers of Chironomidae genera were present at stations with higher concentrations of metals in sediment identified as toxic by the amphipod Hyalella azteca in 28-d exposures. Frequency of mouthpart deformities in genera of Chironomidae was low and did not correspond to concentrations of metals in sediment. Total abundance of organisms/m2 did not correspond to concentrations of metals in the sediment samples. Chemical analyses, laboratory toxicity tests, and benthic community evaluations all provide evidence of metal-induced degradation to aquatic communities in both the reservoir and the river. Using a weight-of-evidence approach-the Sediment Quality Triad - provided good concurrence among measures of benthic community structure, sediment chemistry, and laboratory toxicity.

Planning and implementation of a comprehensive ecological risk assessment at the milltown reservoir‐clark fork river superfund site, montana

AbstractThe U.S. Environmental Protection Agency (EPA) initiated an innovative ecological risk assessment program at the Milltown Reservoir‐Clark Fork River Sediments Superfund Site in 1989. The site is located in the Clark Fork River (CFR) basin of western Montana, and consists of 160 km (100 miles) of the CFR and an 820‐acre wetland at Milltown Reservoir. The focus of the assessment was to identify chronic risks to ecological receptors from metals‐contaminated sediments that have deposited in aquatic and terrestrial habitats of the site. Lack of established sediment or soil quality criteria, the extent of habitats within the watershed, and anticipated complex sediment and soil chemistry led to a unique laboratory and field‐based approach for the baseline risk assessment. Study tasks evaluated aquatic, wetland, and terrestrial habitats across multiple trophic levels. The risk assessment integrated the results of these studies as a triad of chemical, toxicological, and ecological information. The comprehensive nature of the risk assessment was designed to help determine a remedial action for the wetland operable unit and set the basis for additional studies in the upper Clark Fork River watershed. As the problem formulation step in the EPA's risk assessment methodology, this article describes the background of the site, the approach and conceptual model for the risk assessment, and the planning process for implementing the various field and laboratory studies.

Bioavailability of metals and arsenic to small mammals at a mining waste-contaminated wetland.

In support of a baseline ecological risk assessment evaluating the impacts of mining wastes at the Milltown Reservoir Sediments Superfund site in Montana, a food chain transfer analysis was performed for resident small mammals. Deer mice (Peromyscus maniculatus) and meadow voles (Microtus pennsylvanicus) were trapped from a 200 A portion of a mixed upland and palustrine wetland, and concentrations of As, Cd, Cu, Pb, and Zn in carcass, liver, kidney, and testes were quantified. Concurrent to small mammal trapping, samples of grasses, forbs, and soils were collected and analyzed for metal and As residues. Using a linear multimedia food-chain model, assuming a forage base of wetland vegetation from the site, ingestion of local surface water, and incidental ingestion of soils with vegetation, body burdens of the metals and As in the herbivores were estimated. As a means of estimating potential element bioavailability at the site, the modeled body burdens were compared with measured element concentrations in herbivore tissues. The analysis indicates that the bioavailabilities of As, Cd, Cu, Pb, and Zn were equal to or less than 0.2% for internal organs and 0.1% for carcasses, on a microgram/g tissue wet weight basis. Available site data on soil pH, cation exchange capacity, and extractable elements fraction supported the limited release of soil elements. These results suggest that the bioavailable fraction of mining waste metals in riparian wetland soils may be quite small, and, for the Milltown Reservoir site, lower than originally anticipated.

PLANNING AND IMPLEMENTATION OF A COMPREHENSIVE ECOLOGICAL RISK ASSESSMENT AT THE MILLTOWN RESERVOIR-CLARK FORK RIVER SUPERFUND SITE, MONTANA

The U.S. Environmental Protection Agency (EPA) initiated an innovative ecological risk assessment program at the Milltown Reservoir-Clark Fork River Sediments Superfund Site in 1989. The site is located in the Clark Fork River (CFR) basin of western Montana, and consists of 160 km (100 miles) of the CFR and an 820-acre wetland at Milltown Reservoir. The focus of the assessment was to identify chronic risks to ecological receptors from metals-contaminated sediments that have deposited in aquatic and terrestrial habitats of the site. Lack of established sediment or soil quality criteria, the extent of habitats within the watershed, and anticipated complex sediment and soil chemistry led to a unique laboratory and field-based approach for the baseline risk assessment. Study tasks evaluated aquatic, wetland, and terrestrial habitats across multiple trophic levels. The risk assessment integrated the results of these studies as a triad of chemical, toxicological, and ecological information. The comprehensive nature of the risk assessment was designed to help determine a remedial action for the wetland operable unit and set the basis for additional studies in the upper Clark Fork River watershed. As the problem formulation step in the EPA's risk assessment methodology, this article describes the background of the site, the approach and conceptual model for the risk assessment, and the planning process for implementing the various field and laboratory studies.

USE OF BENTHIC INVERTEBRATE COMMUNITY STRUCTURE AND THE SEDIMENT QUALITY TRIAD TO EVALUATE METAL-CONTAMINATED SEDIMENT IN THE UPPER CLARK FORK RIVER, MONTANA

The upper Clark Fork River, above Flathead River, is contaminated with large amounts of As, Cd, Cu, Pb, Mn, and Zn ores from past mining activities. The contaminated area extends from the Butte and Anaconda area to at least 230 km downstream to Milltown Reservoir. Both the upper Clark Fork River and Milltown Reservoir have been designated as U.S. Environmental Protection Agency Superfund sites because of metal-contaminated bottom sediments. We evaluated the impacts of past mining activities on the Clark Fork River ecosystem using benthic invertebrate community assessment, residue chemistry, and toxicity testing. Oligochaeta and Chironomidae generally accounted for over 90% of the benthic invertebrate community in the soft sediment depositional areas. Taxa of Oligochaeta and Chironomidae were predominantly pollution tolerant. Higher numbers of Chironomidae genera were present at stations with higher concentrations of metals in sediment identified as toxic by the amphipod Hyalella azteca in 28-d exposures. Frequency of mouthpart deformities in genera of Chironomidae was low and did not correspond to concentrations of metals in sediment. Total abundance of organisms/m2 did not correspond to concentrations of metals in the sediment samples. Chemical analyses, laboratory toxicity tests, and benthic community evaluations all provide evidence of metal-induced degradation to aquatic communities in both the reservoir and the river. Using a weight-of-evidence approach-the Sediment Quality Triad - provided good concurrence among measures of benthic community structure, sediment chemistry, and laboratory toxicity.

Wetland risk assessment

Wetlands represent unique environments for assessing ecological risks. Habitats may vary from riverine to basin type and include such diverse media as surface waters, sediments, soils, and ground water, with both terrestrial and aquatic biota. Given the diversity of wetland habitats, a number of species may be expected to be fairly unique to a particular site. Wetland ecosystems may be impacted by chemical contamination or by nonchemical stressors such as temperature or suspended solids. A key to assessing ecological risks to chemically contaminated wetlands is determining the degree of contaminant bioavailability from multiple environmental media. Chemical and physical factors of the various wetland habitats must be evaluated for their role in chemical release, transformation, and availability. Approaches to assessing ecological risks may extend from simple benchmark or literature comparisons to direct measurement of exposure and toxicity through laboratory and/or field tests. To increase the utility of wetland risk assessments, the uncertainty inherent in the complex habitats and in the chemistry that governs contaminant bioavailability should be minimized. This can be most readily accomplished by applying an assessment methodology triad of ecology, chemistry, and toxicology to characterize ecological risks. Literature toxicity information and laboratory and field data are used to evaluate potential threats to individuals or species in each trophic level of the wetland food web. The ecological data are integrated with this information to assess whether the concentrations of contaminants and the observed or predicted toxicity relate to actual ecological effects. The ecological relevance of the expected or measured biological responses is of prime importance in predicting risks to the wetland ecosystem. Applications of this approach to risk assessments are presented as case studies of metals-contaminated wetlands at Milltown Reservoir, Montana, and Kesterson Reservoir, California.

Ecological risk assessment of a metals-contaminated wetland: reducing uncertainty

For over 125 years, mining wastes containing arsenic, cadmium, copper, and zinc have contaminated the Clark Fork River in Montana, USA. Mining wastes have partially filled Milltown Reservoir, situated approximately 100 miles (240 km) downstream of major sources of the wastes. Previous studies suggested that benthic invertebrate communities have been impacted by the sediment metals, and trout populations are not at the expected carrying capacity in most sections of the river. An ecological risk assessment has been initiated under the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA) to evaluate impacts of the wastes on terrestrial and aquatic plant and animal species. U.S. EPA has organized the risk assessment through public work groups consisting of representatives of a number of government, university, citizen, and private organizations. The ecological risk assessment focuses on lower trophic levels species and populations in the Milltown Reservoir wetland and riverine environment. Although wildlife surveys have indicated good species diversity and abundance, possible subtle effects on plant and invertebrate species, as well as food chain transfer of metals to trout, have been the primary test objectives. To minimize uncertainty in the assessment and to address anticipated subtle biological effects, field and laboratory studies were instituted to derive site-specific concentration-response information. Studies have included soil and wetland bioassays of earthworms and terrestrial and aquatic vegetation endpoints, in addition to measurements of metals uptake in plants consumed by small mammals and waterfowl. Additional studies on the benthos and geochemistry are nearing completion. Preliminary results of the wetland assessment indicate subtle biological effects at high metals concentrations at major depositional areas. Final results will provide managers with substantial ecotoxicological information to assist in rational decisions for the site.

Extraction and speciation of arsenic in lacustrine sediments

Arsenic was partially extracted with 4.O M hydrochloric acid, from samples collected at 25-cm intervals in a 350-cm column of sediment at Milltown Reservoir, Montana and from a 60-cm core of sediment collected at the Cheyenne River Embayment of Lake Oahe, South Dakota. The sediment in both reservoirs is highly contaminated with arsenic. The extracted arsenic was separated into As(III) and As(V) on acetate form Dowex 1-X8 ion-exchange resin with 0.12 M HCl eluent. Residual arsenic was sequentially extracted with KClO 3 and HCl. Arsenic was determined by graphite-furnace atomic-absorption spectrometry. The analytical results define oxidized and reduced zones in the sediment columns.