ScholarWorks at University of Montana

The upper Clark Fork River in western Montana is an ideal site in which to test long term effects of remediation on the sediment quality of a large basin. Numerous EPA Superfund sites are located along the mining-impacted river from the headwaters at Silver Bow Creek to Milltown, 240 river kilometers downstream. Between 1991 and 1998, remediation efforts chemically treated and stabilized floodplain sources of metal-contaminated sediments. Settling ponds, first constructed in the 1950’s, were upgraded at the head of the Clark Fork River, immediately downstream of Silver Bow Creek, a major source of As and trace metals to the Clark Fork River. To determine the effect these actions had on bed sediment composition, paired basin surveys in August 1991 and August 1998 were conducted, collecting bed sediments at 44 sites on the upper Clark Fork River. Samples were dissolved and analyzed for major and trace elements using ICP-ES. This study found that major element concentrations appear to have remained constant or increased slightly between 1991 and 1998. The more mobile trace elements, such as Cd and Zn, slightly decreased in concentration in the middle part of the basin. Less mobile trace elements, such as Pb and Cu, showed no apparent change between 1991 and 1998. These data suggest that response to remediation may be a slow process or that stabilization of floodplain sources may not have a measurable effect on metal contamination in fine-grained bed sediments.

Metal toxicity is affected by how organisms internally compartmentalize metals. This study compared the concentrations and intracellular distributions of Cd and Cu among hydropsychid caddisflies, Hydropsyche spp. and Arctopsyche grandis, and mayflies, Baetis spp. and Serratella tibialis, from contaminated sites in the upper Clark Fork river, Montana, and from an uncontaminated tributary (Blackfoot River). Bioaccumulation patterns exhibited species and metal specificity. Relative to uncontaminated samples, intracellular metal accumulation was greater in mayflies than in caddisflies. For example, Cd concentrations in the cytosol (soluble cytoplasm) were 7 - 13 (g/g in Serratella and Baetis and 0.4 - 2 (g/g in Hydropsyche and Arctopsyche. Cytosolic Cu concentrations were highest in Serratella (111 (g/g), and similar in Baetis, Hydropsyche and Arctopsyche (27 - 44 (g/g). The cytosol was a major accumulation site for Cd. Cadmium and Cu accumulation in the cytosol of Arctopsyche and Baetis was accompanied by a shift in the distribution of metal from “heat stable” (metal binding) to “heat denatured” (essential) proteins. The relatively high intracellular metal concentrations in Serratella and Baetis are supportive evidence of the reported metal sensitivity of mayflies. Species - specific toxicity may be further modified by how excess metal is partitioned among cytosolic ligands.

Lateral channel movement on the upper Clark Fork River of western Montana has resulted in loss of valuable agricultural land and delivery of sediment and mine tailings into the river. In spring 1996, we initiated a study to evaluate the effectiveness of streambank stabilization treatments to reduce the potential for bank erosion. The treatments implemented in this study focus on the use of native riparian vegetation to stabilize banks instead of traditional "hard" treatments such as rip rap. This study examines the effectiveness of 21 different bank stabilization treatment combinations for reducing bank erosion on a large river system. The treatments incorporate coir (coconut husk) fabric, conifer revetments, log barbs, rock barbs, rock toe stabilization, coir fascines, willow (Salix spp.) and red-osier dogwood (Cornus stolonifera) fascines, willow cuttings, containerized seedlings, mature shrub transplants and rock rip rap. Treatments were installed in fall 1996, spring and summer 1997, and fall 1998 on 24 reaches totaling 1,740 m (5,708 ft) in length. Typical sites are on the actively-eroding, concave side of channel meanders and consist of 5-ft. tall, nearly-vertical banks. 140 permanently-monumented cross sections have been monitored before construction and after the construction for all treatments, after ice events in 1997 and 1998 for treatments established at that time, and after peak flow events in 1997 and 1998. A total of 100 bank surface profiles have been monitored by total station and changes in streambank surface volume have been calculated. Survival rates of various vegetative treatments were also monitored, and costs of construction for each individual treatment were calculated from detailed monitoring of construction activities.The 1996-97 and 1997-1998 bankfull discharge ice event caused little erosion. However, the 1997 flood event, rare in its volume and duration, caused substantial erosion of treatment and control banks. 1997 and 1998 flood-caused erosion rates varied between treatments. First year survival was high for mature transplants (100%), containerized seedlings (90%), and vertically-planted willow cuttings (88%). Second year survival was also high for mature transplants and constainerized seedlings, but survival rates dropped drammatically in the second year for vertical willow stakes. Cost of various treatment implementation ranged widely from a low of $5.58/ft to a high of $82.29/ft.

A meander bend in the Upper Clark Fork River, just upstream of the confluence with Dempsey Creek, actively eroded in recent years. The outer bank had no deep-rooted vegetation. Bank soils consist of an upper cohesive layer of silts and clays and a lower cohesionless layer of sand and gravel. The bank was undermined as the cohesionless layer was eroded, and the overlying cohesive layer would eventually collapse into the river.

The purpose of the demonstration project was to stabilize the streambank by increasing bankform roughness using natural materials (large willows, small shrubs, and willow cuttings). The bankform was intermittently cutback at a 25-foot interval to create a “scalloped” shape. Additional shade was provided with large willows, reducing water temperature and improving aquatic habitat. The design draws from research of Dr. J. Duncan Smith of the U.S. Geological Survey (USGS). This voluntary demonstration project was undertaken by ARCO Environmental Remediation, LLC with the cooperation and assistance of the local landowner (Allen Johns), the U.S. Environmental Protection Agency, USGS, the Montana Department of Fish Wildlife and Parks, and other agencies.

Construction began on December 15, 1999 and was completed on January 24, 2000. Approximately 664-lineal feet of streambank were stabilized. A total of 20 “scallops”, on approximately 25-foot centers, were constructed. One hundred twenty-five large willows and 100 containerized shrubs were installed. Sod was cut prior to excavation and re-installed afterwards, and disturbed areas were seeded with a native riparian grasses seed mix. Willow stakes will be cut and planted in spring, 2000. Coconut fabric and geo-coir fabric were installed to provide short-term erosion protection until vegetation is established. Riprap was installed at the toe of the bank to anchor the geo-coir fabric and provide additional erosion protection. Riprap was installed below the water surface to preserve a natural look. Less riprap was used than typically associated with a “hardened toe” approach. Fencing was installed to prevent disturbance from livestock until vegetation is established. Work was completed at a unit cost of $127 per lineal foot, including design, oversight, and construction. To evaluate effectiveness of the technique, monitoring will quantify erosion rates, based on surveys after ice-flow and spring runoff 2000.

Since 1990, a number of demonstration projects have been completed in the Upper Clark Fork River Basin. Projects have addressed the riparian zone, the floodplain, and upland areas. The projects include the , the RIT Project, the Governor’s Demonstration Project , the University of Montana Riparian and Wetland Research Program (RWRP) Streambank Stabilization Pilot Study, the South Deer Lodge Entryway Improvement Project, the Upland Till Demonstration Project, and, most recently, the Allan Johns Bank Stabilization Demonstration Project. ARCO has been the source of funding for most of these projects. Since the original construction, these demonstration projects have been monitored to assess their effectiveness at mitigating pathways of metals exposure to human and ecological receptors.

These projects demonstrate technologies to remediate areas that have been impacted by historical mine waste. Some projects have focused on eroding streambanks; others have addressed stabilization and revegetation of tailings deposits in the floodplain Similar techniques have been applied to treat upland areas impacted by river sediments carried by historical irrigation practices. This poster provides a brief overview of the projects, their effectiveness, and how information gathered from these projects will be used in preparing the Feasibility Study for the Clark Fork River Operable Unit Superfund Site.

Lost Creek, a tributary to the Upper Clark Fork of the Columbia, is listed on Montana’s 303(d) list as impaired for a number of beneficial uses, including aquatic life support, drinking water supply, and cold water fishery. Lost Creek is undergoing major riparian restoration and grazing management changes which will be the basis of a Total Maximum Daily Load (TMDL) for nutrients and sediment for the lower 17 stream miles. Therefore the objectives of this project include the following:

1) assess current conditions in Lost Creek including kinds and degrees of impairment;

2) provide baseline data to evaluate benefits of restoration work;

3) evaluate Lost Creek as a nutrient source to the nutrient-impaired Clark Fork River;

4) evaluate nutrient sources along Lost Creek;

5) make recommendations for TMDL development for Lost Creek, and how it should relate to the Clark Fork VNRP (which calls for a 20% reduction in nonpoint sources of nutrients).

Water samples were collected from May through August 1999 at sites along the creek which bracketed suspected sources. Samples were analyzed for nutrients (nitrate/nitrite, total Kjeldahl nitrogen, soluble reactive phosphorus, and total phosphorus) using an EPA-approved protocol. Riparian health assessments were performed on the lower 20 miles of Lost Creek using the University of Montana’s Riparian and Wetland Research Program’s Lotic Inventory Form. Riparian inventories are used to identify and prioritize problem areas and provide detailed baseline information for gauging the success of restoration projects on Lost Creek.

Lost Creek does not provide good habitat for attached algae growth, but in some areas aquatic plants may be a problem. Hence, the main reason for reducing nutrients in Lost Creek is to reduce the load to the Clark Fork. Phosphorus levels in Lost Creek were below those considered to be a problem for streams according to the Clark Fork VNRP. Total nitrogen (particularly nitrate/nitrite) levels are high enough to be a concern. Nitrate/nitrite levels increase in the area near Dutchman reservoir. Although wetland disturbance by cattle grazing is a likely source of nutrients in this area, it appears likely that irrigation water from the land application of Anaconda’s municipal wastewater is leaching into groundwater from nearby hay fields and from storage ponds in the Dutchman Creek drainage. Riparian inventories found 30% of riparian areas were not performing their functions while the other 70% were at risk to become nonfunctional.

In terms of TMDL development for Lost Creek, the conservation practices being undertaken by landowners with state and federal funding will likely improve habitat and reduce nutrient loads. Success should be judged by periodic reevaluation of riparian condition and nutrient loads. Lost Creek does provide a significant TN load to the Clark Fork, and this is probably best addressed by riparian wetland restoration and land application of Anaconda wastewater over a larger area at an appropriate agronomic rate. Additional recommendations for monitoring and TMDL development are detailed in the full report.

This work was supported by the Montana University System Water Center with funds from the USGS Section 104 Program. Our grateful thanks to Montana Dept of Fish, Wildlife and Parks, the US NRCS and landowners in the Lost Creek Basin for their efforts to restore Lost Creek.

Samples of bed sediment (<63um) and aquatic insect larvae have been collected annually since 1986 at stations encompassing 380 km of the Clark Fork river, Montana. With these data, we have assessed yearly changes in the relationships between sediment contamination levels and metal bioaccumulation relative to streamflow and remediation activities in the upper Clark Fork. In the earliest years of the study (1986-94), the gradient of metal bioaccumulation showed a clear trend in concentrations from upstream to downstream. Concentrations were highest in the upper 70 Km (200-220 ppm Cu; 3-7 ppm Cd) and moderate in the middle reach (70-180 Km: 30-40 ppm Cu; 0.5-1.0 ppm Cd). Since 1995, concentrations of metals in the upper 70 Km have decreased to 80-100 ppm Cu and 1.0-1.5 ppm Cd; Cd in sediments has also declined. Bioaccumulation concentrations in the middle reach have increased almost twofold since 1995 (80-120 ppm Cu; 1.5-2.0 ppm Cd), but metal concentrations in sediments remain unchanged. The upstream decline in bioaccumulated metal may be partly due to remediation. Intensive remediation activities have occurred in the upstream segment of the river since 1992. Hydrologic factors influence metal bioaccumulation in the middle reaches and may contribute to some of the interannual differences, however, the source of the uncoupling of bioaccumulation from sediment metals remains unclear. Copper concentrations in two bioindicators, “Hydropsyche spp.” and “Arctopsyche grandis” (order Trichoptera), correlate significantly with Cu in sediment in every year, but in some years of high streamflow, there is enhanced uptake at some stations. Correlations between Cd bioaccumulation and Cd in sediment were only apparent in low-moderate flow years (1990-92; 1994).

Milltown Reservoir is located at the confluence of the Clark Fork and Blackfoot Rivers, approximately seven miles upstream of the city of Missoula. It was created by the construction of Milltown Dam in 1907 and has collected contaminated sediments over the years from mining, milling, and smelting activities upstream in Butte and Anaconda.

Approximately 6.6 million cubic yards of sediment has accumulated in the reservoir, containing thousands of tons of arsenic, copper, zinc, iron, and manganese. Periodic scouring events mobilize metal-contaminated sediments, causing water quality standards to be exceeded in the Clark Fork River below the dam. A plume of contaminated groundwater covering approximately 110 acres has developed below and downgradient of the reservoir, with arsenic concentrations exceeding 20 times the drinking water standard. The dam is a barrier to migrating fish, including endangered Bull Trout.

Milltown Reservoir was declared a Superfund site in 1982, and the Environmental Protection Agency is currently evaluating alternatives for remedial action at the site. The Missoula Water Quality Advisory Council has developed a list of remediation outcomes essential to the success of any remedial action program for Milltown Reservoir:

1. Surface and groundwater standards should be met downgradient of the site. Surface water should meet Montana WQB-7 standards for total recoverable metals.

2. Scouring incidents that mobilize metal-contaminated sediments should be prevented.

3. There should be passage for all fish species during their migration season.

EPA’s alternatives for remedial action fall into four categories:

1. No action beyond institutional controls and monitoring.

2. Dam modification.

3. Dam modification and partial sediment removal.

4. Dam and sediment removal.

The Council has concluded that, of these alternatives, only the last has potential to achieve their desired outcomes. The poster explains this view.

About 200,000 abandoned hardrock mining sites are distributed throughout the Mountain West of which about 10% pose hazardous conditions in need of remediation. Risks include both environmental degradation and physical threats to human safety. Water quality is compromised locally and throughout watersheds by heavy metals contamination and acid discharges. Open shafts and adits invite potentially fatal entry and exploration into decaying underground systems. Some preliminary statewide surveys of abandoned mine sites are completed in Montana. Some remediation is complete while other projects are in various stages of completion. However the vast majority of sites remain incompletely surveyed. Where the need for remediation is identified, most projects are under-funded or completely unfunded.

The Linton site contains many hazards typical of abandoned hardrock mine sites. It serves as an effective example of the challenges facing remediation teams. The Linton Mine is located along Cramer Creek, waters from which drain directly into the Upper Clark Fork River. The site is under the jurisdiction of the Bureau of Land Management, whose existing site maps are cursory and inadequate for planning use. This poster presents a comprehensive site map, incorporating GPS mapping and air photo interpretation, to document on-site hazard factors and features. Local and regional maps show the scope and magnitude of the problem and the distribution of abandoned mines in different jurisdictions.

The Clark Fork valley of western Montana is adversely affected by metals derived from past mining and smelting activities. To aid remediation planning by the U.S. Environmental Protection Agency, the sources, transport, and deposition of metals were assessed. This required examining the geomorphic history and deposition of flood- plain tailings, determining river migration rates across the flood plain, and tracking the current transport and accumulation of sediment and metals through the 120-mile reach from the Warm Springs Ponds to Milltown Reservoir.

Flood-plain tailings were mapped and the range of metal concentrations determined. Average annual meander-migration rates of the Clark Fork in the Deer Lodge valley were quantified from aerial photographs taken in 1960 and 1989. Estimated sediment and copper loads from individual sources were integrated in mass-balance calculations to estimate input, transport, and deposition. Transport rates were calibrated to water-quality monitoring data and hydrologic conditions of 1985-95.

Under post-1990 source conditions and 1985-95 hydrology, mass-balance calculations indicate that streambank erosion is the largest source of copper input to the river, comprising about 56 percent of the total copper input along the 120-mile reach. Upstream input (Silver Bow and Warm Springs Creeks) accounts for about 5 percent, whereas tributaries and streambed exchange each account for about 8 percent. Flood-plain runoff and ground water together account for about 24 percent of the copper input. Not all of the copper input to the river is transported downstream; instead, about 47 percent is deposited on point bars. When 1985-95 estimates of copper input are adjusted to long-term (1930-95) flow conditions, the percent contributions from individual sources are similar, with bank erosion still providing the largest input of copper (60 percent).

High flows in 1996 and 1997 resulted in larger sediment and copper loads than those measured in 1991-95. Average annual copper yields (tons per river mile) for the expanded 1991-97 period, which had flows similar to long-term hydrology, were 2-3 times larger than yields for 1991-95. In particular, reaches between Deer Lodge and Turah contributed substantially more copper per river mile in 1996-97 than during the low-flow years of 1991-95. These reaches actually contributed as much copper per mile as the reach above Galen. The largest copper yield per mile is contributed from the reach between Galen and Deer Lodge, regardless of flows.

This material is published in U.S. Geological Survey Water-Resources Investigations Report 98-4170.

Cottonwood Creek is located in Powell County, Montana and flows through the town of Deer Lodge. The Natural Resource Conservation Service (NRCS), will be working with ranch owners along Cottonwood Creek on restoration and conservation projects in the spring of 2001. In September and October of 2000, six graduate students from the University of Montana completed a baseline assessment of the lower Cottonwood Creek watershed, including the tributary Reese Anderson Creek. The assessment is the start of a more comprehensive assessment to be completed in the summer of 2001. The purpose of this baseline assessment is to characterize the current condition of Cottonwood Creek.

The objectives of this study include:

1. To assess the current condition (“health”) of Cottonwood Creek’s riparian areas.

2. To provide baseline data needed to evaluate the benefits of conservation and restoration projects.

3. To gather information about the current and historical land-uses in the Cottonwood Creek Watershed.

4. To make recommendations on a landowner monitoring system.

Study Approach

The Fall 200 study synthesized existing data and conducted new field observations. Existing data includes maps, soil information, geology, climate, and historically and current land uses. The major component of the field research was the evaluation of the riparian corridor using University of Montana’s Riparian and Wetland Research Program’s (RWRP) Lotic Assessment Form. This method breaks up the riparian corridor into about ¼ mile to ¾ mile sections, called polygons, and evaluates each polygon’s vegetation, stream bank stability, and invasive species. Each polygon is then rated as “Properly Functioning”, “Functioning but at Risk”, or “Nonfunctioning”. Other field data collected included measurement of the stream cross section, photo documentation of the riparian area, and stream discharge.

Summary of results

The Fall 2000 assessment broke Cottonwood Creek into 11 polygons and Reese Anderson Creek into 4 polygons. The RWRP Lotic Health Assessment scored five polygons as Functioning but at Risk, and the remaining ten polygons, including all 4 on Reese Anderson Creek, were found to be Non-Functioning systems. A big factor in reduced functioning along Cottonwood Creek is over grazing which resulted in a streambank instability and lack of woody vegetation. This is most evident on Reese Anderson Creek and the downstream portion of Cottonwood Creek. Other concerns in the riparian corridor included dewatering, use of rip-rap and invasive species.

Comparing the cross sections measurements made in each polygon showed a widening trend downstream. A likely cause for this trend is the lack of woody vegetation along the stream bank downstream. Also, the limited discharged data collected suggested that the stream loses water as it flows downstream. This is likely because of the amount of diversions along the stream. However, further measurements need to be taken in the spring of 2001 to gain a better understanding of how much of a loss there is.

The Voluntary Nutrient Reduction Program (VNRP) for the Clark Fork River has set a target of reducing non-point source nutrient inputs by 20%. In-stream target levels are 20 ppb for total phosphorus and 300 ppb for total nitrogen. In order to assess whether this was being worked towards and to identify areas of concern, long-term monitoring was of nutrient input to the Clark Fork River and tributaries of concern was proposed in the spring of 1998. Monitoring began in the fall of 1998, and has continued to the present. In 2000, some sites on the Bitterroot River were added.

Current reclamation efforts to revegetate hardrock minelands in western Montana have met with limited success. In the Upper Clark Fork River Basin, for example, there remains vast areas of barren and unproductive land. The majority of native species currently being seeded on hardrock mine reclamation projects were developed for revegetaion of coal strip-mines in the dry, high pH soils of eastern Montana. This plant material is not well adapted to the acid/metalliferous soils and local climatic conditions found at hardrock mine sites. The Development of Acid/Heavy Metal-Tolerant Cultivars project seeks to address this problem by selecting plant ecotypes indigenous to western Montana that demonstrate superior tolerance to acid/heavy metal soil conditions. In 1995, two initial evaluation plantings were constructed on the Anaconda Smelter Superfund Site. Plant materials in this study were assembled from both wildland collections and commercial seed sources. The plots collectively tested 95 species consisting of 51 grass, 29 forb, 14 shrub, and 1 tree species. After three growing seasons, the superior performing entries were identified. These better performing collections are presently being tested and compared to other accessions and cultivars of the same species in a comparative evaluation planting (CEP) near Anaconda. Concurrently, 13 grass, 6 forb, and 7 shrub species are being grown at the Bridger Plant Material Center (BPMC) to determine cultural management techniques and to increase the supply of seed. The results from the CEP and the success of seed production will provide valuable information for the selection of locally indigenous plant materials. Formal plant release will be via the pre-varietal process as Selected Class material. Foundation seed for the releases will be maintained at the BPMC for distribution through the Montana and Wyoming Crop Improvement Associations to commercial seed producers.

The Clark Fork River basin in western Montana is an excellent area to study solute metals in contaminated systems. Past flooding events have deposited mine tailings in the river's floodplain. Surface runoff flowing over, and groundwater flowing through, these floodplain deposits provide non-point sources to the dissolved-phase (

The headwaters region produced elevated concentrations of As, Cu, Fe, and Mn in surface waters with relatively high pH values (8.5 to 9.2). Treatment ponds downstream from the river's headwaters precipitate trace metals due to the addition of liming agents to the surface water. Most trace element concentrations decrease through this treatment system, with the exception of As. On a basin-wide scale, the concentration trends for trace elements in the system vary from reactivity to domination by dilution with the surface water discharge of the Clark Fork and its tributaries. Major ions appear to have downstream sources associated with tributaries or groundwater. This contrast in concentration trends illustrates the difficulty in studying geochemical trends on the large scale of a basin.

Tables and Figures referred to in McKinnon’s full paper are linked at the bottom of McKinnon’s abstract page. (Note: Figure 1, map of upper Clark Fork, is not available).

Previously published studies conducted on the Upper Clark Fork River (UCFR) suggest adverse effects due to metals-enriched sediments found in the depositional areas that comprise approximately 4% of the riverbed. While these studies measure exposure concentrations from depositional areas, effects measurements (e.g., benthic invertebrate abundance and diversity measures and tissue residues) have been predominantly obtained from coarse substrate riffle areas. Comparing exposure data from one habitat to effects data from another is problematic. An integrated approach was conducted to assess effects data by sampling benthic macroinvertebrates for community composition measurements, tissue residues, and sediment contaminant concentrations. Thirteen co-located sampling sites along the UCFR were sampled in both depositional and riffle habitats in 1996. Metals concentrations in bulk sediments and benthic invertebrate tissues decreased with distance from metals sources in the headwaters, but sediment porewater concentrations did not. Although there were some significant relationships between bulk sediment metals concentrations and tissue metals residues, BMI community metrics did not appear to vary along this gradient, or with exposure concentrations. This is consistent with an evaluation of sediment toxicity tests performed using sediments from Warm Springs Ponds, as well as the UCFR, which do not suggest that metals are of potential concern to benthic macroinvertebrates.

A sediment quality assessment was developed to evaluate the potential bioavailability of metals from riverbed sediments. The proposed approach consisted of multiple assessment methods using bulk sediment metals concentrations, equilibrium partitioning measures, and bulk sediment and pore water toxicity tests. Metals bioavailability was evaluated using: 1) EPA’s theoretical guidelines which compare acid-volatile sulfide concentrations to simultaneously extractable metals concentrations, and compare sediment pore water metals concentrations to ambient water quality criteria; 2) correlative guidelines, which compare bulk sediment metals concentrations against sediment no-effect concentrations (NECs), and 3) sediment toxicity tests. Sediment toxicity studies were also used to derive site-specific NEC values. The study site chosen for this assessment was the metals-enriched upper Clark Fork River (UCFR) located in southwestern Montana, USA. Results of this assessment indicate only nominal risk to most aquatic organisms posed by sediment metals concentrations in depositional areas in the UCFR. The uncertainty associated with using only one of the approaches for evaluating sediment contamination should be reduced by using the combined approach outlined here.

Risk-based decision making associated with chemically contaminated waste sites generally emphasizes the reduction or elimination of actual or potential toxic exposures to receptor populations of concern. Wildlife inhabiting contaminated sites can be front line indicators of chemical exposure and effects due to their intimate association with site-related contaminated media. This association provides a sensitive means of detecting exposure and effects-associated adverse responses. Measures of exposure and adverse responses in wildlife can provide a means for testing predictive models of exposure and effects and determining the effectiveness of proposed remedial actions. In this preliminary study, we examined small mammals inhabiting test remediation plots on the Anaconda Smelter Hill and the Anaconda Tailings Pond.

The Anaconda Regional Water, Waste, and Soils Operable Unit (ARWW&S OU) covers about 300 square miles in the southern Deer Lodge Valley and the surrounding foothills in southwestern Montana, and is part of the Clark Fork River watershed. This OU is included in the Anaconda Smelter Superfund Site and contains large volumes of wastes, debris, and contaminated soils from the processing of ores to recover metals, chiefly copper. These milling, smelting, and refining activities continued for a period of 96 years, from 1884 until 1980.

In 1998, The U.S. Environmental Protection Agency and the Montana Department of Environmental Quality identified land reclamation as the general remedial action for major portions of the Operable Unit (OU). Recognizing that the intensity of land reclamation is a technology continuum and parallels the continuum of ecological function within the OU, a distinct scientific approach was required to evaluate field conditions and then select the appropriate level of reclamation intensity. The Land Reclamation Evaluation System (LRES) was developed to fill this need. This decision-making tool contains several components: 1) a description of potential human and ecological risk, followed by an assessment of the nine National Contingency Plan criteria; 2) a quantitative scoring system for the existing vegetation communities and the potential for COC transport; 3) an identification of modifying factors that may play significant roles in determining the level and extent of land remediation; and 4) decision diagrams to help guide the decision makers in identifying remedial actions and levels of reclamation intensity. The LRES has been applied to significant areas within the OU and preliminary remedial units have been delineated and assigned remedial actions. Additional data and LRES quantitative scores for extensive acreage were generated in 1999 to support remedial designs.

This poster describes the LRES, remedial action triggers, selection of land reclamation remedial actions and their definitions, and post-reclamation criteria for growth media and vegetation.

In the Upper Clark Fork River, Montana, as well as other mining areas in the western United States, concentrations of arsenic, cadmium, copper, lead, and zinc in aquatic biota may be elevated compared to levels in background tributaries. For fish, the ingestion of aquatic invertebrates containing elevated metals concentrations may be a significant route of exposure. Although some testing previously has been conducted to evaluate dietary metals toxicity, this information has not been summarized and synthesized to establish toxicity reference values for use in ecological risk assessment. Using standardized assessment criteria, test results from the existing published literature for rainbow trout were systematically evaluated and, when possible, dietary exposure NOAEL and LOAEL values (expressed as mg/kg dry weight) were established. Specifically, NOAEL concentrations were 40 mg/kg for arsenic, 150 for cadmium, 603 mg/kg for copper, 210 mg/kg for lead, and 1900 mg/kg for zinc. LOAEL concentrations were 44 mg/kg for arsenic, 10,000 for cadmium and 664 mg/kg for copper; LOAELs could not be determined for lead and zinc since adverse effects were not demonstrated in any study not confounded by elevated exposure to multiple metals. Dietary toxicity reference values were compared to benthic invertebrate metals concentration data from the UCFR to evaluate the likelihood of adverse effects to trout via dietary exposure. During the period 1991 to 1998 no benthic tissue concentrations reported for selected species by the USGS or reported from community tissue monitoring samples collected by the USFWS and ARCO were found to exceed NOAEL or LOAEL values.

Some irrigated lands along the Clark Fork River, near Deer Lodge, Montana contain elevated levels of arsenic, cadmium, copper, lead and zinc, and low soil pH. This contamination originated from river-transported mine wastes associated with metal mining, milling and smelting operations in the Butte and Anaconda, Montana areas. In May 1999, the U.S. Environmental Protection Agency ordered the responsible party (ARCO) to reduce contamination on historically irrigated lands along the Eastside Ditch near Deer Lodge, Montana. These lands consist of residential properties and adjacent pastures. The preferred action for the residential yards is excavation and replacement of soils (up to 18 inch depth) whose average arsenic values exceed the trigger value of 120 mg/kg which is the 95% lower confidence interval value of the human health risk-based residential concentration of 150 mg/kg. The preferred remedial action for adjacent pastures is tilling and liming to varying depths (up to 24 inches) to evenly distribute arsenic and metals into the soil profile, and neutralize potential and active acidity.

The Reclamation Research Unit, in conjunction with W.B.E, Inc. and CH2MHill performed oversight for sampling and remedial actions along the Eastside Ditch for the U.S. EPA. Soil samples were collected from 17 residences and adjacent pastures, and 2 pastures in 1998 and 1999. The soil samples were analyzed for arsenic, copper, lead, zinc and pH. Soils were excavated and replaced in 8 residential yards as of November 1999 (remaining soil arsenic levels are typically below 70 mg/kg). In addition, 10 adjacent pastures were tilled and limed as of November 1999. Further sampling and remedial activities are expected to continue in Spring 2000.

Many risk assessments rely on ambient water quality criteria (AWQC) to evaluate potential risks to aquatic biota from contaminant exposure. The AWQC are not intended to consider risk from sediment or dietary exposure pathways and few assessments account for site-specific bioavailability of metals. Risks to a variety of aquatic biota from metals in the Upper Clark Fork River (UCFR), Montana were evaluated using assessment tools specific to each exposure medium- water, sediment, and diet. Each medium was evaluated using extensive site-specific exposure data and, when possible, effects criteria that reflect site-specific metals bioavailability. Risk from exposures to water column metals (As, Cd, Cu, Pb and Zn) was evaluated by comparing observed dissolved metals concentrations (an average of long term monitoring) to chronic AWQC and a site-specific rainbow trout chronic toxicity reference value (TRV) developed by EPA based on testing UCFR water. (When environmental concentrations are below TRV’s, risks are usually assumed nominal.) Sediment metals exposure was evaluated using EPA draft Equilibrium Sediment Guidelines (ESGs--i.e., simultaneously extractable metals - acid-volatile sulfide and sediment porewater summed as toxic units) and using bulk Sediment Effects Concentrations (SEC) derived from toxicity tests with UCFR bed sediments. Trout dietary exposures were addressed by comparing extensive benthic macroinvertebrate tissue data with dietary TRVs derived from selected literature. Dissolved metals concentrations were generally lower than the water column TRVs for each metal. Metals in pore water and bulk sediments were below ESGs and SECs. Furthermore, benthic macroinvertebrate tissue concentrations did not exceed dietary TRVs. These results predict nominal risk to most UCFR aquatic biota under observed conditions for each of the exposure pathways evaluated.

The benthic macroinvertebrate (BMI) communities of the Upper Clark Fork River (UCFR) Watershed in Montana have been the subject of an extensive long-term monitoring program by the State of Montana. However these investigations have not included a characterization of habitat at sampling stations in the Clark Fork River or reference streams. Additionally, the temporal patterns of annual average dissolved metals concentrations and how they co-vary with community data has not been presented for sites with co-located average water chemistry measurements. Biomonitoring reports have implied effects from historic mining and smelting activities based on a biotic index that includes multiple community metrics with impairment defined relative to scores observed at a reference stream (Blackfoot River). In this paper the differences in habitat between the reference stream and biomonitoring sites in the Clark Fork River Watershed are documented. When compared with UCFR sites, reference sites were found to differ significantly and ecologically in habitat features such as embeddedness and riffle and pool-run pebble count substrate particle size distribution. Macroinvertebrate community measures (taxa richness, EPT richness and abundance) were not correlated with annual average dissolved or total metals levels or with available USGS invertebrate tissue monitoring data for sites on the UCFR during the period of record 1991-1998. These data suggest that macroinvertebrate communities of the UCFR are not significantly affected by metals exposures.

Editor’s note: I suggested that the final sentence read:

This correlation analysis provides no evidence that UCFR macroinvertebrate communities are affected by metals exposure.

but the author and I have agreed to disagree on this.

Potential sublethal effects of heavy metals in stream macroinvertebrates were examined with two cellular and biochemical biomarkers in larvae of two caddisflies indigenous to the Clark Fork River, Montana, - Hydropsyche spp. and Arctopsyche grandis. Stress proteins, in particular members of the HSP70 family, are involved in cellular protein homeostasis and repair, and are induced by a variety of stressors, which either damage cellular proteins directly or cause cells to synthesize aberrant proteins. Lysosomes are intracellular organelles that play key roles in the detoxification of both organic and inorganic xenobiotic compounds. Larvae of Hydropsyche spp. were collected from four sites on the Clark Fork (Galen Gage--4.7 km, Goldcreek--85.6 km, Turah--189.7 km, above Flathead--381 km) and a reference site (the Blackfoot River). Larvae of A. grandis were collected from the same sites minus the Galen site. Samples were immediately frozen in liquid nitrogen for HSP70 analysis, or preserved with Tissue Tek, then frozen in liquid nitrogen for the lysosomal stability assay. HSP70 was analyzed by western blotting using monoclonal antibodies. Lysosomal integrity was measured in cryosections by acid labilization with acid phosphatase as a marker enzyme. Results to date show elevated tissue concentrations of Cd, Cu, Pb and Zn and significantly increased levels of HSP70 in Arctopsyche from Goldcreek compared to reference samples. Lysosomal integrity also was compromised in samples from Goldcreek. In Hydropsyche, tissue concentrations of Cd, Cu and Pb from Galen Gage were elevated (4-7 times) relative to the Blackfoot River, but levels of HSP70 did not differ between the two sites. These preliminary results indicate that sublethal effects of metal exposure may differ between species.