9

Risks of Metals Contamination in Coeur d’Alene Lake

As part of its statement of task, the committee was asked to discuss the relevance of metals release in the Coeur d’Alene (CDA) Lake to human and ecological health. This was premised on the possibility that nutrient inputs to the Lake may be increasing, which might promote growth of algae and consequently lead to low oxygen conditions in the Lake that could promote release of sediment-bound metals.

The remedy for the Bunker Hill Superfund site was based on removing sources of metals contamination and/or breaking exposure pathways linking metals to human and ecological receptors in the CDA Lake watershed. The remedy is supported by human health (Terra Graphics/URS Greiner/CH2M Hill, 2001) and ecological (CH2M Hill and URS Corp., 2001) risk assessment reports that were reviewed by a committee of the National Academies in 2005 (NRC, 2005). The U.S. Environmental Protection Agency (EPA) has not selected a remedy for CDA Lake (see Chapter 1); hence, evaluations of the Lake are not included in the published risk assessments.

This chapter specifically focuses on exposure pathways and attendant risks of metals in CDA Lake. For human health, this includes occupational and recreational exposure and exposure from fish consumption and drinking water (for those who derive drinking water from waterbodies hydrologically connected to the Lake, such as the Rathdrum Prairie Aquifer). In addition, the chapter also delves into some of the risks related to nutrient enrichment that go beyond exposure to metals, in particular the possibility of harmful algal blooms (HABs). For ecological health, the discussion is much more expansive given the Lake’s complex ecology and includes risks from exposure to metals in pelagic (phytoplankton, zooplankton, fish) and benthic communities of animals and plants.

HUMAN HEALTH RISKS

As discussed in Chapter 1, the Bunker Hill Mining and Metallurgical Complex was designated as a Superfund site based on the high blood lead levels in children and contamination of the local environment by lead, arsenic, cadmium, and zinc. As part of the Superfund process, a human health risk assessment (HHRA) was conducted for the CDA basin to determine the extent of contamination, understand the potential risk to humans who come into contact with contaminated media, and provide information to support remedial activity and cleanup benchmarks (Terra Graphics/URS Greiner/CH2M Hill, 2001). The HHRA also considered the practice of subsistence lifestyles by members of the CDA Tribe. NRC (2005) provided an independent review of the progress on the remediation and the HHRA, and five-year updates of the Superfund remedy have evaluated remediation progress (the most recent being EPA Region 10, 2021).

The areas considered in the 2001 HHRA were the populated areas of the Bunker Hill Box and the upper and lower CDA basins, where nearly all historic primary mineral extraction activities and releases of primary wastes were located. A screening-level risk assessment was also completed in 1999 on 24 common-use areas along the shoreline of CDA Lake and the Spokane River (Appendix B, Terra Graphics/URS Greiner/CH2M Hill, 2001) including eight locations south of the city of Coeur d’Alene. The goal was to assess potential health risks to recreational users of beaches and picnic areas. Although the Lake’s shoreline and access points were considered in the HHRA, there is no formal assessment for CDA Lake itself, even though highly contaminated secondary and tertiary particulate and dissolved contaminants have been deposited for the past 120 years in the sediments of CDA Lake.

The set of contaminants selected for the HHRA included eight metals: antimony, arsenic, cadmium, iron, lead, manganese, mercury, and zinc. The primary human health concerns are associated with arsenic, lead, and mercury. Lead is a known neurotoxin posing particular risk to sensitive subpopulations. Arsenic is a potential carcinogen in skin, bladder, kidney, lung and liver; it also poses a risk via the ingestion pathway for pre-cancer and non-cancer effects. Mercury is also considered here given its potential for human health effects and the advisory in place warning of risks from eating fish potentially contaminated with mercury. The following sections briefly summarize conclusions from the 2001 HHRA, the 2005 NRC study and the latest five-year review (EPA Region 10, 2021) as they might apply to human health exposure pathways relevant to CDA Lake.

Lead

Lead is of human health concern in the CDA basin because widely distributed lead-enriched¹ mine wastes have contaminated soils, dust, sediments, and waters that people contact, incidentally ingest, and inhale. Understanding the effects of lead on humans is essential to appreciating the risks of accidental exposures to these wastes. Because a full review of the rich literature that exists on the effects of lead toxicity in humans is beyond the scope of this report, this section briefly summarizes the state of understanding and its application to CDA Lake.

Effects of Lead in Humans

Naranjo et al. (2020) recently summarized the molecular mechanisms by which lead manifests its effects within cells:

The most notable effects of lead lie in its ability to harm the sensitive and complex processes of central nervous system development. These effects result from exposure to lead during pregnancy (on the growing fetus) and in early childhood. The degree of damage to the fetus or the growing child depends upon the length of exposure and the cumulative amount of lead in the body. Naranjo et al. (2020) note that a “cascade of neurological dysfunction” occurs with the slightest disruption of developmental processes in the central nervous systems of a growing child. At that age there are limited opportunities to repair or compensate for such changes. In particular, exposure of the developing fetus or growing child to lead can target dopamine and the hippocampus in the developing brain, both

___________________

¹ Pb-enriched is defined by Bookstrom et al. (2013) as concentrations in excess of 1,000 μg/g dry weight.

² Lead and calcium have similar physical and chemical characteristics and follow similar intracellular metabolic pathways (Potula and Kaye, 2005).

of which are involved in learning and memory. The developing child is particularly sensitive to lead exposure because the blood-brain barrier, which helps block accumulation of lead in adult brains, is immature and less effective in children. There is also an increased absorption of lead from food in the gut of children compared to adults. Normal childhood behaviors like putting dust-contaminated objects, lead-enriched dirt, or lead-contaminated toys in their mouths increases the likelihood of incidental exposure. The effects that manifest are tragic and well documented over years of study (Needleman and Bellinger, 1991; Finkelstein et al., 1998; Lidsky and Schneider, 2003; Needleman, 2004).

The effects of lead on neurological development manifest as poorer scores on behavioral and developmental tests, including IQ tests (Bellinger et al., 1987; Canfield et al., 2003; Lanphear et al., 2000; Schwartz, 1994; Wasserman et al., 1994), psychosocial morbidity (Bellinger et al., 1994; Wasserman et al., 1998) and juvenile delinquency (Needleman et al., 1996). Naranjo et al. (2020) summarized the current state of understanding. “Longstanding exposure to any amount of lead has been associated with intellectual disability in a dose-dependent manner, ranging from delay or loss of developmental milestones to reduced cognitive function and academic achievement. Other symptoms are shortened attention span, impaired executive function, delayed processing speed, and impairments in visual and verbal memory and visuospatial skills.” Lead exposure, especially before age three, appears to affect learning and behavior even with only low levels of lead detected in the blood, and these effects then persist through childhood and adolescence. Effects from early exposure that appear to be manifested later in life include anxiety and depression. Aggressive, criminal, and antisocial behaviors have been described with chronic lead exposure. Renal disease, hypertension (high blood pressure), and degenerative diseases are found in adults who suffered childhood exposure to lead. Exposure to lead during development is also associated with increases in future susceptibility to nerve degeneration and the likelihood of developing Alzheimer disease later in life (Naranjo et al., 2020).

Once in the blood stream, the way lead is passed to tissues is independent of exposure route. In pregnant women, lead in the blood can cross the placenta and then the blood brain-barrier in the developing fetus. From the blood, lead can also distribute to other soft tissues and eventually to bone. It progressively accumulates in bone during exposure, where it can persist for decades (Potula and Kaye, 2005). Bone stores of lead can become dynamic and remobilize back into the blood under some conditions, especially when changes in bone turnover occur during stages of life (e.g., during pregnancy, lactation, or menopause; Potula and Kaye, 2005).

The critical concentration of lead in blood at which effects begin to manifest in children was the subject of debate when the early studies were published (Needleman, 2004) because of the difficulties with differentiating between the effects of multiple stressors. In 1991, the Centers for Disease Control and Prevention (CDC) felt the evidence was sufficient to suggest a blood lead level (BLL) of 10 μg/dL in children as the lowest level of concern. Exceeding the level of concern in a local population usually resulted in actions to reduce exposures. There is now both toxicological and epidemiological evidence that adverse health effects are associated with a BLL < 10 μg/dL (100 μg/L) (EPA Region 10, 2021). In 2012, the CDC lowered the blood lead reference value (level of concern) to < 5 μg/dL (Ruckert et al., 2021) and in 2021, the CDC further lowered it to 3.5 μg/dL (CDC, 2021). According to some literature, there is no concentration below which there are no adverse effects of lead exposure, leading some to suggest lowering the BLL of concern to about 2 μg/dL (Gilbert and Weiss, 2006). The U.S. government pledged in Healthy People 2020 to reduce mean BLLs to 1.6 μg/dL among children ages 1–5 (Digman et al., 2019). EPA has not yet changed its national lead health risk policy; thus, the remediation action objective for the CDA basin remains at a mean BLL of 10 μg/dL (EPA Region 10, 2021).

Routes of Exposure

People can be exposed to lead through ingestion, inhalation, or dermal contact. Direct lead uptake through the skin is not expected to be a major route because the most abundant forms of lead in the CDA basin are inorganic and would not efficiently penetrate the skin. Lead-enriched particles on the skin or clothing can lead to incidental ingestion, contribute to dust coatings on tools and household items, or contribute to dust that can be inhaled. In regions with a mineral extraction history, inhalation of lead-impacted ambient particulate matter in the air (e.g., blowing dust) is a significant exposure pathway. Resuspension of lead-enriched household dust may contribute to indoor exposure, with children being at higher risk compared to adults because of their higher

respiratory rate, their closeness to the ground relative to adults, and their higher fractional deposition of particles in the lungs (Kastury et al., 2019). Inhalation studies with fine particulate material (≤ 0.25 μm) suggest that about one-fourth of the inhaled lead particles deposit in the lungs, and up to 95 percent of the lead on those particles is then absorbed (Kastury et al., 2019).

Compared to inhalation, incidental ingestion of soil and surface dust is a less common exposure pathway, except for children (Naranjo et al., 2020). Incidental ingestion of lead can result from hand–mouth activity typical in 1- to 4-year old children. This can be exacerbated by tracking lead-contaminated soils into the home or working in contaminated clothes and shoes, all of which can generate dust as well.

Diet and a complicated array of other factors affect intestinal lead absorption once lead is ingested. Early studies showed that, on average, about 8–18 percent of the ingested lead was absorbed into the blood and tissues of humans; 10 percent was the most frequently quoted average (NRC, 2005). But recent studies show that absorption rates may be higher (Naranjo et al., 2020). For example, data from the CDA basin show that 2.5–39 percent of the lead ingested was absorbed in the body; average absorption efficiency was 26 percent (EPA Region 10, 2021).

A major goal of the Superfund remedy was to reduce exposure of humans to lead. Box 9-1 discusses the most recent progress as related in the 2021 Five-Year Review (EPA Region 10, 2021).

Measures of Lead Exposure

Blood lead levels are the earliest reflection of exposure to lead and are the metric used to determine human exposures within the preceding 30 days. Blood lead levels have been directly related to adverse outcomes in adults and children (NRC, 2005).

Declines in average lead concentrations in domestic dust and soils in the CDA basin have been accompanied by a downward trajectory in blood lead levels of children ages 0–9. Figure 9-1 shows trends by age since the mid-1990s in OU-3 (the area of the Superfund complex outside the Box). The greatest progress in reducing blood lead levels occurred early in the remediation, but in the past ten years the decline appears to have stalled. The average blood lead levels for the 0–9 age group remain in the 2–4 μg/dL range, fluctuating around or near the latest CDC reference value (3.5 μg/dL; CDC, 2021). For OU-3, 6–22 percent of households exceeded the 5 μg/dL threshold (EPA Region 10, 2021), although the accuracy of that estimate was affected by a small sample size (Table 9-1). For comparison, the average incidence of blood lead levels in excess of 5 μg/dL among seven western states with data³ was 2.02 ± 0.43 percent (CDC, 2021). Thus, percentage-wise, average OU-3 blood lead levels are typically two to four times higher than found on average in other western states for which data are available.

Although the blood lead levels are much improved since the 1990s, several challenges with understanding lead exposures of children in the basin remain, which could affect any future analysis specific to the Lake (as discussed in EPA Region 10, 2021).

___________________

³ States with data in CDC (2021) are Colorado, Nebraska, New Mexico, Oklahoma, Oregon, Texas, and Washington.

and Native American communities disproportionately suffer from elevated human health risks, including elevated blood lead levels (Needleman, 2004). The collapse of the mineral extraction industry by 1990 exacerbated economic problems for some people in the basin, increasing vulnerability to adverse effects from lead. NRC (2005) concluded that “American Indians who practice traditional lifestyles likely would have higher risks than other residents of the Coeur d’Alene River basin” and reviewed the importance of better understanding those risks. Concerns remain about participation being biased by a fixed-site approach to sampling and possible overrepresentation of children thought to be at lower exposure and risk (EPA Region 10, 2021).

Possible Exposure Pathways Involving CDA Lake

From what is known about lead enrichment in the Lake and the human health risks in the basin, some potential pathways of exposure relevant to the Lake itself can be identified beyond the lakeshore exposures that were the subject of the 1999 screening-level assessment. These include human exposure to contaminated lake water and lake sediments—in particular, those whose work or recreation involves incidental exposure to suspended or bed sediments in the Lake, eating fish from the Lake, and other dietary exposure pathways related to subsistence lifestyles.

TABLE 9-1 Number of Children Tested for Blood Lead Levels and Percent Exceedances of the 2012 CDC Reference Level of 5 μg/dL

SOURCE: EPA Region 10 (2021).

Studies of lake sediments (Horowitz et al., 1995; Harrington et al., 1998a; Toevs et al., 2006; Morra et al., 2015; EcoAnalysts, Inc., 2017) show that the highest lead concentrations are associated with finer-grained sediments in the Lake—those sediments most likely to be mobilized by disturbances like wind or recreation or to be redeposited on land during floods. As discussed in previous chapters, lead concentrations in lake sediment vary widely and range from 100 to 10,000 μg Pb/g (Bookstrom et al., 2013; Horowitz et al., 1993; EcoAnalysts, Inc., 2017). Patches of extreme lead concentrations (that exceed the action levels for soils in the upper basin Superfund site) occur near Harrison, including Carlin Bay, Rockford Bay, and other bays surrounding the confluence of the CDA River; and Cougar Bay near the head of the Spokane River (Horowitz et al., 1993; Spears et al., 2007; EcoAnalysts, Inc., 2017; Scofield et al., 2021). In contrast, annual lead sampling at the beach at Harrison showed no exceedances of EPA’s lead soil target level (EPA Region 10, 2021). A systematic investigation of soils and sediments in the areas where human contact is most likely seems warranted, in order to update the 1999 screening level shoreline investigation (which showed that lead exposures from soils and sediments on the lakeshores were not universally high) and focus on where current populations are most likely to recreate.

The EPA and CDC action level for lead in drinking water is 15 μg/L (the maximum contaminant level goal is zero). CDA Lake is not a source of drinking water, and dissolved lead concentrations rarely, if ever, exceed 1 μg/L during the summer months, when incidental exposure to Lake water during recreation is most likely. Thus, incidental water ingestion during swimming in the Lake is less likely to be a significant exposure pathway than exposure to Lake sediments (EPA, 2001). Ingestion of fish fillets is also less likely to be a primary pathway of lead exposure because fish do not transport most divalent cations to their muscle tissue; rather, they trap these metals in internal organs (especially liver) and perhaps in bones. The Idaho Department of Health and Welfare (Idaho Department of Health and Welfare [IDHW], 2019) analyzed ~ 150 samples of fish fillets from a variety of edible species from the Lake and found that none of the samples exceeded 0.3 mg Pb/kg dw,⁴ with all concentrations lower than consumption advisories. However, risks could be higher from consumption of whole fish from the Lake or unpeeled water potatoes from the shoreline of the Lake north of the mouth of the St. Joe River (especially to those practicing a subsistence lifestyle). No data are available on lead concentrations in whole fish, but Scofield et al. (2021) reported that the tubers of water potatoes could contain up to 30 mg Pb/kg tissue (dw).

***

Although it is possible that human health risks associated with the Lake itself are lower on average than those in the areas already covered by the HHRA to date, there is the issue of subpopulations that could be experiencing elevated exposure (as discussed by Moody and Evans, 2011). With that in mind, it seems that exposure pathways relevant to CDA Lake deserve further study, including perhaps a formal HHRA. Several suggestions from the most recent five-year review (EPA Region 10, 2021) are related to the potential lead risks inherent to CDA Lake. These include development of a shoreline-specific soil/sediment lead exposure level, updated evaluations of recreational exposures in common-use areas, better understanding of effects of flooding and flood control in lakeside recreational areas, and improved participation in blood lead testing. A thorough Lake-wide survey of lead concentrations in fine-grained fractions of shoreline soils and sediments could be a first step in determining whether and where further risk assessment would be of value, especially near areas of greatest recreational use. One study showed that “proactive mitigation approaches to cope with potential environmental degradation in lake ecosystems can have significant economic benefits to owners of lakefront properties and local communities” (Liao et al., 2016).

Arsenic

Arsenic is a naturally occurring trace element that poses a threat to human and ecosystem health, particularly when incorporated into food or water supplies. The greatest risk imposed by arsenic to human health results from contamination of drinking water. Ingestion of drinking water with hazardous levels of arsenic can lead to arsenicosis

___________________

⁴ The most reliable metal concentrations in tissues of plants and animals are reported on a dry weight (dw) basis because water content differs among organisms and affects results reported on a wet weight (ww) basis.

and cancers of the bladder, skin, lungs and kidneys. The level of concern for arsenic in drinking water defined by both the World Health Organization and EPA is > 10 μg/L, although adverse effects have been observed at lower concentrations (Ahmad and Bhattacharya, 2019).

Within waters and sediments of CDA Lake, arsenic (As) is present primarily in one of two oxidation states: arsenate [As(V)] or arsenite [As(III)]. In solution, arsenic exists primarily as oxyanionic acids; arsenate exists as H₂AsO₄⁻ and HAsO₄²⁻, while arsenite exists as H₃AsO₃ (Goldberg and Johnston, 2001). Within soils, microbial activity has been shown to methylate arsenic, leading to chemical species such as dimethyl arsenate (Zhao et al., 2013). Another group of arsenic species, the thioarsenates, have recently been detected in flooded soils and include inorganic thioarsenates and methylated thioarsenate (Wang et al., 2020). Methylated species are usually not abundant in aqueous solutions compared to inorganic forms of arsenic (Smedley and Kinniburgh, 2002; Chen et al., 2021).

Effects of Arsenic in Humans

Numerous publications describe substantial epidemiological evidence that high concentrations of arsenic in drinking water are associated with several detrimental effects on human health, including skin lesions and cancer of the lung, bladder, kidney, and liver, among other effects (Ahmad and Bhattacharya, 2019; Fendorf et al., 2010; NRC, 1999, 2005). Upon ingestion of arsenic-bearing water or food, arsenate is taken up through phosphate transporters, while arsenite is absorbed through aquaglyceroporins (Mukhopadhyay et al., 2002). When arsenate is taken up by a cell, it is reduced to arsenite before secretion or sequestration (Mukhopadhyay et al., 2002). Arsenite, whether absorbed directly or from conversion of arsenate, may go through several methylation steps within the liver, forming monomethylarsenicals and dimethylarsenicals.

Chemical species of both arsenic oxidation states, inclusive of methylated and thiolated species, are toxic to organisms (prokaryotic and eukaryotic). Arsenate is a chemical analog to phosphate and thus may substitute for phosphate in biomolecules. In particular, it interferes with phosphate binding sites in adenosine triphosphate (ATP), resulting in the formation of adenosine diphosphate (ADP)-arsenate, which limits cellular energetics (Brazy et al., 1980; Liebl et al., 1995). Arsenite has a high affinity for thiol (-SH) groups, forming a dihydrolipoylarsenite chelate, which disrupts the structure of proteins and enzymes (Muehe and Kappler, 2014) and thus interferes with critical cellular functions that include chromosomal abnormalities, oxidative stress, altered DNA repair, altered DNA methylation, altered growth factors, cell proliferation, promotion/progression, gene amplification, and p53 gene suppression (Kitchin, 2001). Furthermore, DNA damage may be induced by methylated trivalent arsenicals that are mediated by reactive oxygen species (ROS) (Nesnow et al., 2002; Mass et al., 2001). Early onset of arsenicosis shows an erythematous flush that leads to melanosis, hyperkeratosis, and desquamation. Long-term cutaneous complications include the development of multicentric basal cell and squamous cell carcinomas (Pershagen et al., 1981), and continual exposure leads to even more serious health impacts.

Routes of Exposure

The Superfund remedy considers human exposure to arsenic in the CDA basin via a number of pathways that are the same as with lead, including ingestion of arsenic from contaminated soil and sediment in residential, commercial, and undeveloped areas; and inhalation of contaminated airborne dust generated at these locations. Ingestion of arsenic from drinking water wells is an exposure pathway less relevant to lead that has been explored in the CDA basin, as evidenced by the evaluation of private wells in OU-3 for arsenic contamination (EPA Region 10, 2021, Table 6-2). Any drinking water exposure pathway for arsenic requires that arsenic partitions from the solid phase into the aqueous phase. Processes favoring such partitioning can be grouped into four categories: (1) ion displacement, (2) desorption at pH values > 8.5, (3) reduction of arsenate to arsenite, and (4) mineral dissolution, particularly reductive dissolution of iron and manganese oxides. Processes 3 and 4 generally occur in concert (Tufano et al., 2008). Although various processes may liberate arsenic from solids, for lake systems with pH values < 8.5, a transition from aerobic to anaerobic conditions and commensurate arsenic and iron reduction appears to be a dominant means by which high concentrations of dissolved arsenic are generated (see Chapter 7 for details; Smedley and Kinniburgh, 2002; Fendorf et al., 2010).

Metrics of Arsenic Exposure

In the CDA region, measurements of arsenic exposure have not been gathered from urine or hair samples, although this is often done to characterize chronic arsenic exposure. Instead, the risks from arsenic were mainly assessed by modeling human exposures based on arsenic concentrations in environmental samples (NRC, 2005). Cancer risks and non-carcinogenic hazards were calculated on the basis of total arsenic concentrations in each area. Arsenic is naturally occurring in this region, but there is little question that soils and sediments influenced by mineral extraction activities are enriched with arsenic relative to background concentrations (Harrington et al., 1998b). It should be noted that both NRC (2005) and the most recent five-year review (EPA Region 10, 2021) emphasized that developing a human health metric for evaluation of arsenic exposures in the CDA basin was an important need.

The two main arsenic exposure pathways noted in the 2001 HHRA that present the greatest cancer risks are those associated with subsistence lifestyles among CDA Tribal members and exposure to arsenic in drinking water from private wells (EPA, 2001). For the subsistence scenarios (both traditional and modern), arsenic and iron in soil and sediment were the most notable contributors to non-cancer hazard. Non-cancer risk for non-tribal members is at an elevated level of concern solely in the Burke/Nine Mile area and only associated with well water supplying the home. The Burke/Nine Mile area had the highest neighborhood risks and hazards because of the waste pile exposures evaluated for this area. Waste piles had the highest concentrations of non-lead metals. The lower basin had the highest concentrations of arsenic and iron in soil and sediment (except for waste piles).

Possible Exposure Pathways Involving CDA Lake

Exposure pathways that involve CDA Lake directly would be if arsenic is ingested from drinking Lake waters, migration of arsenic from the Lake water into groundwater wells (see, e.g., Moore and Woessner, 2003), or migration of mobilized arsenic into major aquifers. There is some evidence that private wells in OU-3 have been evaluated for arsenic contamination (EPA Region 10, 2021, Table 6-2), but exposure pathways directly involving the Lake are not known to currently exist. However, as discussed in Chapter 1, CDA Lake is hydrologically connected to the Spokane Valley-Rathdrum Prairie Aquifer, which is a regional drinking water supply. As described in the 2011 Rathdrum Prairie Comprehensive Aquifer Management Plan (Idaho Water Resources Board[IWRB], 2011), between the outlet of CDA Lake and the confluence with the Little Spokane River, the Spokane River loses flow to the aquifer in some reaches while gaining in others. Water withdrawals from the Spokane Valley-Rathdrum Prairie Aquifer are anticipated to increase in the future as development in the region continues. As such, to the extent that in the future areas in and/or near the northern bays of CDA Lake could experience anoxic conditions and release of metals of significance to human health, such as arsenic, there exists a potential pathway for water quality in CDA Lake to impact water supplies outside the Lake area.

It is unclear what the current risks from arsenic release into bottom waters of CDA Lake are to humans. Fendorf et al. (2010) described “three environmental requirements for groundwater arsenic concentrations to increase: water saturation (which limits diffusion of atmospheric oxygen), a limited supply of sulfur, and a source of organic carbon to drive microbial dissolution of Fe oxides.” Where the above conditions are met (i.e., sediments become anoxic and sulfur is consumed by FeS), there is the potential for arsenic in sediments in CDA Lake to be mobilized into porewaters, bottom waters of the Lake, or eventually groundwater (see Chapter 7). Analyzing 206 cores from CDA Lake, Harrington et al. (1998b) found that Lake sediments averaged 211 ± 11 μg As/g dw (with a range of 2–568 μg/g)—17 times higher than concentrations found in the St. Joe River sediment cores. Release of dissolved arsenic into porewaters and bottom waters from anoxic sediments has been shown to occur in CDA Lake at C6 (Figure 7-10), in the deep bend waters studied by the CDA Tribe (Chess, 2021) and in lateral lakes (Sprenke et al., 2000). Concentrations of arsenic in bottom waters at C6 can reach the human health level of concern (10 μg/L) during short periods (one or two months) of anoxia that occur annually at those locations. If anoxia were to occur in bottom waters of the deeper portions of CDA Lake (C1–C5), where arsenic concentrations in sediments are about 17 times higher than at C6 (Harrington et al., 1998b), greater release of arsenic than that seen at C6 might be expected. Sprenke et al. (2000) observed arsenic concentrations in the interstitial water of lateral lakes in the CDA basin of up to 217 mg/L, and as high as 422 mg/L in the water within the uppermost meter of lake sediment. These were sediments with 30–300 μg As/g dw, similar to sediment in CDA Lake. Thus, the large reservoir of arsenic in the sediments of CDA Lake has the potential to release high concentrations of arsenic into

bottom waters and porewater if conditions change in the future. Understanding the connections between the Lake and groundwater that might provide drinking water for lakeshore populations, and beginning to monitor those sources of drinking water, could facilitate effective responses should such changes occur.

Mercury

Mercury (Hg) in aquatic systems arises from both local (e.g., mining) and global (e.g., combustion) sources, but once present in a watershed, it presents a risk to aquatic food webs and humans who consume animals from those food webs. Mercury is usually released into the environment as inorganic mercury, but the risks arise in aquatic environments because of the formation of monomethylmercury (CH₃Hg⁺ or MeHg) in anoxic sediments. MeHg is primarily formed by sulfate and iron-reducing bacteria in suboxic conditions (Fitzgerald and Lamborg, 2007). Hence, MeHg production is enhanced in ecosystems with dynamic oxygen gradients, such as the water column and/or sediments of lakes, reservoirs, floodplains, and wetlands (see Figure 9-2) (Alpers et al., 2014; Beutel et al., 2020;

Conaway et al., 2008; Fuhrmann et al., 2021; Selin, 2009). Higher levels of mercury and longer periods of anoxia can lead to greater MeHg formation. Mercury is also highly volatile when present as elemental Hg or vaporized by combustion. Widespread mercury contamination of the atmosphere has resulted from a combination of the cumulative effect of discharges from mercury sources and local sources of combustion (e.g., coal-fired power plants).

Monomethylmercury, which accounts for < 5 percent of mercury in aqueous environments, poses the greatest threat to human and ecosystem health because it is a potent neurotoxin that biomagnifies in the food web, reaching potentially toxic concentrations in predator fish, such as bass (Benoit et al., 2003; Cossaboon et al., 2015). MeHg concentrations in high-trophic-level fish can be > 10⁶ times higher than the water in which they live, and their body burden increases as they grow and age (Cossaboon et al., 2015; Lavoie et al., 2013). Thus, the primary exposure route of MeHg to humans is the consumption of fish (Sunderland, 2007; Wang, 2012), and mercury-driven fish consumption advisories exist for waterbodies throughout the 50 states (Wentz et al., 2014), including CDA Lake.

Mercury in the Coeur d’Alene Lake Region

Not surprisingly, there is mercury present in the CDA River and in CDA Lake. Mercury was used in mining (and processing) of gold and silver in the CDA watershed during the late 19th and early 20th centuries and was, undoubtedly, released into the local environment (as was typical of mining technology of the times). Bookstrom et al. (2013) listed mercury as an element that sometimes co-occurred in elevated concentrations with lead in the CDA basin. Sediment cores from the lateral lakes along the CDA River, discussed in Sprenke et al. (2000), illustrate the legacy of mercury contamination, often with the highest peaks in the subsurface (presumably deposited historically) (see Figure 9-3). Elevated total mercury concentrations in sediments, but not in food webs, were reported even earlier by Gebhardt et al. (1971): “The highest mercury concentration in water and sediment samples (among 93 locations in Idaho) were collected from the lower Coeur d’Alene River. . . . (but) fish and water samples collected above the confluence of the South Fork Coeur d’Alene River were low in mercury.” The concentrations in sediments are for total mercury, which includes forms with varying levels of toxicity, not necessarily indicative of food web contamination.

In a 2009 national study conducted by EPA, 48.8 percent of lakes in the United States exceeded the EPA’s recommended tissue-based mercury water quality standard of 0.3 mg/kg wet weight.⁵ In all of Idaho there are 341 river miles and more than 143,000 acres of lake that are listed as being impaired by mercury. The detection of greater than 0.3 mg/kg of mercury in fish (four of 30 samples in the northern Lake, seven of 40 in the central Lake) in 2016 (IDHW, 2020) has led to CDA Lake being listed as impaired for cold water aquatic life, salmonid spawning, and primary contact recreation in 2020 (IDEQ, 2020). The highest concentrations were found in bass (0.056–0.798 mg/kg ww) and pike (0.056–0.479 mg/kg ww), both top predators. Furthermore, the CDA Lake fish consumption advisory was updated in 2020 because “[h]igh mercury levels were found in some fish species, including bass, bullhead, northern pike, panfish, and kokanee,”⁶ but the data upon which such conclusions were drawn are minimal.

The relative contribution of mercury to the CDA food web coming from local contamination compared to atmospheric inputs is unknown. Bechard et al. (2009) compared mercury concentrations among six locations in Idaho in the feathers of bald eagles, another top predator prone to elevated concentrations of MeHg. Concentrations in the CDA basin were not significantly different than concentrations in the Salmon, Boise, Payette, and Snake River basins. Thus, the listing of CDA Lake for mercury contamination is not necessarily an indicator of a local source of contamination, and the degree of contamination in the food web does not indicate unusually high local concentrations of methylated mercury.

Compiled data from the U.S. Geological Survey (USGS) from the past 40 years (Figure 9-4) suggest that total mercury concentrations in the watershed may be declining (similar to the trends observed for other metals in the watershed—see Chapter 3), but the data are sparse and no statistical techniques were used to analyze trends in the data. If such declines were ever confirmed, they would coincide with decreases in atmospheric deposition in North America and Europe (Zhang et al., 2016). Locally, reduced mining activity, better source control at combustion sources, and burial of higher concentration materials in the sediment by deposition of cleaner materials could be factors contributing to declines in mercury concentration in various media.

___________________

⁵ https://www.epa.gov/fish-tech/national-study-chemical-residues-lake-fish-tissue-results

⁶ https://healthandwelfare.idaho.gov/news/fish-advisory-coeur-dalene-basin-revised-after-high-levels-mercury-found-some-species

Efforts to reduce mercury inputs to watersheds are a focus nationwide. Although more data on mercury cycling and mercury in the food web would be informative in the CDA basin, findings to date support the suggestion that the food web contamination with mercury in CDA Lake is more influenced by atmospheric sources. MeHg in some fish exceeds warning targets but is lower than some other regional exposures, which is enough to suggest that the major sources are regional (atmospheric) rather than local. However, a likely ongoing issue with regard to mercury would be increased duration or extent of anoxia in CDA Lake (or lateral lakes) leading to a greater extent of formation of MeHg. This would counteract efforts to reduce mercury inputs by having conditions that lead to greater formation of the most problematic, bioaccumulative form of mercury.

Localized Risk Associated with Algal Blooms in an Altered Chemical Environment

A concern resulting from changes in dissolved metals concentration from natural processes or remediation activities is the increase in bioproductivity of CDA Lake, leading to a eutrophic state. While nutrient and bioproductivity classification of CDA Lake is considered oligotrophic (see Chapter 5), an overabundance of bioproductivity in a localized area could negatively affect water quality and ecosystem services (Liao et al., 2016). Two potential consequences of an algae-induced eutrophic event resulting from degradation of algal detritus in CDA Lake are considered below: (1) suspended algal blooms of toxic cyanobacteria and (2) growth of filamentous algal species.

Harmful Algal Blooms

HABs involve any type of phytoplankton or periphyton blooms that cause water quality or environmental degradation (e.g., toxins, anoxia, taste and odor compounds). Toxin-producing cyanobacteria are the most common taxa that form HABs in lakes. Cyanotoxins may be toxic for aquatic and terrestrial biota and can also impair the recreational use of lakes and especially their use as drinking water sources. HABs are most common in aquatic systems with high nutrient concentrations (phosphorus and nitrogen). When cyanobacteria blooms die, they often accumulate at the surface and can appear as foam, scum, paint, or mats on a waterbody’s surface. Harmful blooms are most commonly caused by cyanobacteria (of genera such as Microcystis, Anabaena, Aphanizomenon, Planktothrix, and Cylindrospermopsis) that produce toxins like microcystin, anatoxin-a, saxitoxin, and cylindrospermopsin. Multiple toxin exposure routes are possible, including skin contact with water containing the toxins, drinking water containing toxins, breathing in contaminated water droplets in the air, and eating fish or shellfish that contain toxins. Many of the toxins produced by HABs have no antidote, and medical intervention consists of managing symptoms. HABs present a difficult challenge to managers of aquatic ecosystems (both for source water and recreation), and their frequency is expected to expand with increases in global temperature (Paerl and Huisman, 2008, 2009; Carey et al., 2012; Paerl and Scott, 2010).

While many of the toxins that are produced by toxic cyanobacteria have been characterized, the underlying mechanisms of bloom formation (e.g., the cascade of genetic factors that trigger cyanobacteria to produce toxins) are poorly understood. In fact, while there are known species of cyanobacteria that commonly are present during a HAB event, the genes that encode toxin production may not be present or may be incomplete in different populations of the same species of cyanobacteria. Currently, the only means to determine if an algal bloom is harmful involves monitoring of toxins (commonly microcystin) in water samples or direct experimental assessments of toxicity.

Understanding the potential processes that may lead to HABs is key to predicting whether they might play a role in CDA Lake water quality in the future. One process involves production of reactive oxygen species (ROS). According to Paerl and Otten (2013), algal “blooms increase reactive oxygen species . . . proportional to the dissolved organic carbon concentration and light intensity”. Reactive oxygen species (ROS) are free radicals of hydroxide (^•OH) and molecular oxygen (O₂^•−, superoxide) generated by natural biological and photochemical processes in high-light-intensity aquatic environments, such as fresh surface waters, that damage the cells of resident microorganisms. ROS can damage DNA, proteins, or other cellular materials or rapidly form strongly oxidizing peroxides (e.g., H₂O₂). Some cyanobacteria in fresh waters produce microcystin, which has been identified in conferring protection from ROS (Paerl and Otten, 2013). That is, microcystin binds (via thioester bonds) to key proteins to confer protection from the oxidative damage and proteolysis brought about by light (Zilliges et al., 2011), although the precise mechanism is unknown.

One might assume that the oxidative protection conferred by microcystin may allow organisms that produce this toxin to gain a competitive advantage over non-toxin-producing phytoplankton species in freshwaters. Indeed, this appears to be the case during bloom initiation, as these toxin-producing organisms can take advantage of highlight intensities at the top of the water column (Kardinall et al., 2007; Davis et al., 2009). However, in light-limited constructed and natural systems, there can be a gradual decrease in microcystin and microcystin producers over time during a bloom event (shown in Figure 9-5A and C) (Kardinall et al., 2007; Davis et al., 2009). That is, phytoplankton growth will eventually lead to light attenuation, favoring those species with more efficient light-harvesting capabilities, especially in wind-mixed lake systems.

In CDA Lake, high dissolved metals concentrations could both enhance ROS accumulation and confer a competitive advantage to some cyanobacterial taxa. As shown in Figure 9-5B and D, elevated concentrations of metals that occur in CDA Lake, such as cadmium and lead, can induce oxidative stress in resident organisms (Stohs and Bagchi, 1995; Nies, 1999), although the situation is more complex for zinc (Marriero et al., 2017; Xia et al., 2008). In addition, the toxins produced by some cyanobacteria can bind to metals, rendering the metals less toxic and thereby conferring another competitive advantage over non-toxin-producing, metal-sensitive algal species.

Neither dissolved organic carbon nor nutrients are especially high in the main body of CDA Lake. Thus, the likelihood of dense blooms of harmful algae under present conditions is low in the Lake overall. But occurrence of the conditions conducive to such blooms in protected localities, some littoral habitats of the Lake, or in the lateral lakes, could be enhanced by the presence of metal contamination. Monitoring such locations could provide an early warning about the beginnings of more-widespread conditions conducive to HAB formation.

Filamentous Algal Blooms and Other Invasive Submerged Aquatic Vegetation

Filamentous algal blooms (FABs) and other invasive aquatic plants have been increasing within littoral, or nearshore, regions in oligotrophic lakes, like CDA Lake, despite measured low concentrations of phosphorus (Vadeboncoeur et al., 2021). FABs are typically composed of green algae (Chlorophyta) that are not usually toxic, but may provide a reservoir of carbon and nutrients to microbial communities in the sediment and increase potential for localized eutrophication. Eurasian water milfoil is a taxon of submerged aquatic vegetation with an ability to tolerate and grow in a wide range of water temperatures, depths, and turbidities—all of which contribute to its success as an invader. Liao et al. (2016) cited surveys showing that milfoil infestations prevailed in at least ten out of 28 bays of CDA Lake during the summers of 2011–2014.

Occurrence of invasive plants like milfoil and FABs in littoral areas could be an indication of a lake system trending toward a eutrophic state. A combination of factors including nutrient availability, decreased grazing by benthic communities, changes in hydrology, and warming of water temperatures due to climate change often facilitate such invasions (Vadeboncoeur et al., 2021). CDA Lake may be uniquely vulnerable to such outbreaks if the elevated concentrations of toxic metals in lake sediments are adversely affecting benthic communities that graze on early life stages of these plants.

***

In summary, the combination of ROS protection and metal sequestration could bestow a competitive advantage to toxin-producing cyanobacteria that allows them to extend successional dominance in metal-contaminated aquatic environments. However, HAB risk is likely low in the majority of CDA Lake due to a combination of factors including primarily low-nutrient, oligotrophic conditions. Occurrence of the conditions conducive to such blooms, such as an increase in temperature and/or nutrients, is more likely in the littoral habitats of the Lake or in the lateral lakes. Similarly, an event of higher probability is an increase in FAB resulting in localized degradation of water quality in nearshore Lake regions most often used by humans.

ECOLOGICAL HEALTH

Although the greatest concern for the ecological health of lakes has traditionally been eutrophication driven by excess nutrients, in CDA Lake, the ecology is threatened by profound contamination with heavy metals, including arsenic, cadmium, lead, and zinc. Much of the ecological disturbance and loss of ecosystem services in the rivers of the CDA basin has been attributed to the extreme concentrations of these metals (CH2M Hill and URS Corp., 2001). Although primary waste deposition did not occur on the lakeshore, CDA Lake is the ultimate repository for the metals mobilized from the watershed. This section explains how lake ecosystems generally respond to metal exposures, discusses benchmarks for different levels of ecological disturbance, evaluates metal stress in the CDA Lake ecosystem, and discusses the limited data available on CDA Lake ecology that reflects the risks of metal exposure.

Hierarchy of Effects of Metals in Lake Environments

Ecological responses to metal contamination are complex and difficult to separate from responses to other drivers in an ecosystem. Any adverse effects of metal(loid)s enrichment are manifested at multiple levels of biological organization (Luoma and Rainbow, 2008) as shown in the hierarchy in Figure 9-6. The first toxic effects of trace metals occur at the level of molecules within cells. With increasing availability of the toxin, effects at the molecular level become more serious and are manifested at higher levels of biological organization (McCarthy and Shugart, 1990). Thus, an early sign of metal-driven malfunction is disruption of biochemistry and the structure and function of cells.

The effects of metals on ecological receptors range from harmless to sublethal (chronic malfunctions) to lethal (toxicity). At extreme exposures, individuals can experience mortality within a few days (acute toxicity) or over

longer periods of time (chronic toxicity). More common are adverse effects expressed as sublethal signs of malfunction, such as reduced size of individuals, poor physiological condition, changes in behavior, or changes in morphology. Metrics for determining functional disturbances include impaired reproductive success or changes in growth rates. The health of the population of a species is at risk when enough individuals are adversely affected at some life stage, such that recruitment of new individuals declines or mortality rates increase. When recruitment failures or mortality are large enough, the vulnerable population will disappear from the community (termed extirpation).

Loss of species changes the composition of the community. Starting with the most significant, metrics for community disturbance include absence of species or higher taxonomic groups most sensitive to the metals (e.g., mayflies can be the most sensitive taxa in freshwater streams and lakes—Milani et al., 2003); reduced numbers of taxa; a lower community diversity index; or exceptionally low abundance of individuals from all taxa (e.g., Clements et al., 2021). Loss of taxa from a community also can affect interactions among taxa and food webs, cascading into broader effects on the ecosystem. The ultimate effect of disturbances from contaminants is ecosystem simplification (Woodwell, 1970), degradation of ecosystem functions, and loss of ecosystem services. The greater the metal exposure, the further the cascade of responses will proceed.

In reality, ecological responses to metals exposure are not necessarily detected (or even expressed) in a linear fashion, nor do all responses necessarily lead to ecosystem simplification. Compensatory reactions can occur at every level of organization, analogous to detoxification at the molecular level. At all levels of organization, adverse effects occur when the compensatory mechanism is overwhelmed and/or when compensation imposes secondary costs.

Detecting ecological responses to metals (and distinguishing them from effects of other processes) is more difficult as one goes from simpler to more complex levels of organization (Figure 9-6). Thus, it is a challenge to detect metal-specific malfunctions with sensitivity or pinpoint cause and effect when considering population sizes, community structures, and ecological function. Lower-level responses are more readily detectable and easier to study in controlled experiments. Indeed, controlled studies to define thresholds of toxicity at different levels of organization and different life stages for different species are critical to writing the criteria embedded in governmental environmental regulations. However, the thresholds defined by single-species toxicity tests and the thresholds of metal-driven change in populations, communities, and ecosystems may not be the same (Cairns, 1986; Buchwalter et al., 2017; Clements et al., 2021).

To ecologically evaluate metals in CDA Lake, a weight of evidence approach is necessary, supported by a body of work across multiple levels of organization and using different tools and approaches. Figure 9-7 shows how single species toxicity tests (laboratory bioassays) are used to generate hypotheses about the thresholds of metal toxicity, controlled experiments under field conditions (mesocosms) with multiple species can be used to test these hypotheses, and systematic field surveys or observations can be employed to validate hypotheses or generate new, testable questions. Scale and relevance to nature are lowest in toxicity tests and greatest in field studies, while control and replication are greatest in toxicity tests but lowest in field surveys.

Data on Ecological Risks from Metals in Coeur d’Alene Lake

Understanding how the CDA Lake ecosystem will change in response to changes in metal and nutrient inputs and factors like climate change requires knowing the current status of the lake at each level of biological organization, described below using the available information.

Exposure: Metal Concentrations

The first step in an assessment of ecological health and future risks in CDA Lake is determination of exposure. Zinc, cadmium, lead, and arsenic enrichment in the water column and sediments of CDA Lake is extreme by the standards of other large, oligotrophic lakes (see Chapters 1 and 6). Because water-column metal concentrations change seasonally, spatially, and year to year, mean annual metal concentrations in the water column at a few locations are not adequate indicators of ecological risk. Rather, the intensity and duration of peak metal concentrations, the period of recovery between peaks, and the timing of peaks in the life cycle of key taxa define the level of toxicity at different levels of biological organization (Groenendijk et al., 1999; CH2M Hill and URS Corp., 2001). Physiological or biochemical disturbances and even overt toxicity can result from short exposures to high concentrations of metals.

In CDA Lake, metal concentrations in the biologically active euphotic zone are highest in spring at the northern Lake monitoring locations (C1 and C4), but peak later in the year at the southern Lake location (C5). Dissolved metal concentrations during the periods of highest exposure might be the most effective metric to determine if any ecological effects might be expected, especially if such exposures last 30–60 days. Heat maps of water-column metal concentrations, such as Table 6-2, could provide a simple means of illustrating monthly changes in exposure at different locations.

It is also important to consider metal concentrations other than dissolved concentrations. Suspended sediments in the Lake could be incidentally ingested by zooplankton, adsorbed onto macroflora and biofilms, or enmeshed in periphyton or microflora (e.g., aufwuchs) and thereby be a pathway for exposure for some taxa in the Lake. Balistrieri et al. (2002) found that in the watershed, metal concentrations per unit mass (the measure of relevance ecologically) of suspended material were of the same magnitude as bed sediments and soils (thousands of μg/g for lead and zinc—see Figure 9-8; tens of μg/g for cadmium and arsenic). Kuwabara et al. (2006) analyzed suspended material assumed to be primarily biogenic (defined as “phytoplankton”) in June 2005 and found concentrations of zinc of about 1,300 μg/g dw at C5 and concentrations of 2,700–4,490 μg/g dw at a location near the mouth of the CDA River. Concentrations of lead at these two locations were 128–193 μg/g dw and 744–1,244 μg/g dw, respectively. Thus, the biogenic material from the surface waters of the Lake was highly enriched with zinc and lead (Kuwabara et al., 2006) but to a lesser extent than the dominantly inorganic suspended material from the watershed (Balistrieri et al., 2002). Greater understanding of seasonality, spatial distributions, and trends in metal concentrations on suspended material in the Lake is important to evaluating exposures because this material represents the base of the water-column food web.

Metal concentrations in surface sediments affect the food webs at the bottom of the Lake. Table 9-2 shows metal concentrations in surface sediments of the Lake. The spatial variability of metal concentrations in lake-bottom sediments is substantial; Horowitz et al. (1993) found lower metal concentrations at the backs of some bays and at the confluence of the St. Joe River and the highest concentrations in profundal locations. In general, however, highly enriched sediments occur Lake-wide. Metal concentrations in bottom sediments are also strongly influenced by particle size. Horowitz et al. (1993) found that the four most metal-enriched sediments were characterized by extremely fine-grained sediments (< 20 μm). More recent studies of sediment samples at five locations in the Lake showed that most of the substantial variation among locations and depths was driven by differences in particle size (EcoAnalysts, Inc., 2017), as evidenced by strong correlations of lead, cadmium, and zinc with aluminum.

TABLE 9-2 Cadmium, Lead, and Zinc Concentrations (mg/g dw) in Surface Bed Sediments from Five Different Studies at Different Times

NOTES: *Median lake-wide, **one location, ***transects from 5-m water column depth to 30-m depth, respectively. **** Carlin Bay, north of the mouth of the CDA River, was separated from the rest because it represents the highest concentrations observed in the more recent studies.

Bioavailability: Waterborne Metals

Exposure is a function of the concentration of bioavailable⁷ metal, not total metal. Hence, understanding both total concentrations in the environment and the bioavailability of each metal in that environment is required to evaluate ecological risks (e.g., CH2M Hill and URS Corp., 2001; Balistrieri and Blank, 2008; EcoAnalysts, Inc., 2017). The bioavailability of dissolved metals is determined by speciation (i.e., the distribution of a metal among different ligands or forms). Speciation is “a function of the water composition, presence of inorganic and organic ligands that bind with the metals, the presence of competing ions, and the properties of the metals themselves” (Smith et al., 2015). For example, lead forms strong inorganic and organic complexes, leaving little uncomplexed lead (low free ion concentrations), and is thus of low bioavailability. Dissolved zinc and cadmium form relatively weak inorganic and organic complexes (their free ion concentrations are a higher proportion of total concentrations), such that they have greater bioavailability than dissolved lead (Smith et al., 2015).

Speciation of a dissolved metal is difficult to measure directly but can be approximated by a variety of computation and analytical techniques. The most advanced method of applying speciation concepts to metal toxicity is the Biotic Ligand Model (BLM). The BLM determines uptake at a generic biological receptor (termed the biotic ligand) using equilibrium speciation/competition calculations that consider effects of interactions between the metal and major ions, inorganic ligands, and organic ligands (dissolved organic matter). It then statistically relates this uptake to outcomes of acute or chronic toxicity tests. Alternatively, diffusive gradient in thin films (DGT) is an example of a physical measurement tool designed to estimate bioavailability. All techniques available to approximate bioavailability have important limitations (e.g., Slaveykova and Wilkinson, 2005; Luoma and Rainbow, 2008) but are valuable constructs if employed within the proper context.

Dissolved metal bioavailability has not been systematically studied in CDA Lake, but studies in the watershed are informative. Balistrieri and Blank (2008) compared BLM calculations and DGT measurements to dissolved concentrations of cadmium, lead and zinc at seven mining sites in the South Fork of the CDA River and two locations with little contamination. DGT-labile lead was as low as 16 percent of dissolved lead concentrations. Up to 30–40 percent of the total lead (in two models) appeared to be adsorbed to colloidal iron oxides or strongly complexed with organic substances of low bioavailability. Similar speciation/partitioning is likely in the lake during spring runoff when lead in the water column is at its highest concentrations. For example, the colloidal and

___________________

⁷ Bioavailability describes the fraction of total ambient metal that an organism actually takes up from both dissolved forms and forms ingested in its diet. Metals from each pathway have a separate bioavailability. Bioaccumulation is the concentration of metal that accumulates in the tissues of an organism as a result of that combined uptake. Bioaccumulation depends upon the bioavailable metal in each pathway (diet or dissolved) summed across all pathways. When metal is bioavailable to aquatic life, it has the potential to generate adverse ecological effects (toxicity). Thus, for the purposes of determining toxicity thresholds, exposure is a function of the concentration of bioavailable metal, not total metal.

nanoparticulate forms of metals common in spring runoff from the CDA River pass through 0.45-μm filters and thus are classified as dissolved metals in CDA Lake, but their bioavailability is low compared to truly dissolved metal forms. Thus, a small but variable fraction of the lead in solution in the lake is likely bioavailable. This means that benchmarks based upon total lead in water would overprotect from lead toxicity, all other things being equal.

In contrast, nearly all the zinc and cadmium in the South Fork of the CDA River were present as potentially bioavailable free ion and inorganic complexes in most situations (Balistrieri and Blank, 2008; Smith et al., 2015) with little or no colloidal metal found. There is competition for biological uptake between major ions (represented by hardness, calculated as mg/L calcium carbonate) and zinc, so high hardness will reduce bioavailability/uptake of zinc. However, waters in the northern CDA Lake have low hardness (≤ 25 mg Ca/L), so such competition is minimal. Balistrieri and Blank (2008) concluded that the bioavailability of water-column cadmium and zinc was minimally restricted by organic complexation or association with colloids, a conclusion similar to that of Kuwabara et al. (2006). Given their high concentrations and high bioavailability, cadmium and zinc are likely to be the dissolved metals of greatest toxicological concern.

Bioavailability: Diet and Sediments

The proportion of particulate or foodborne metals taken up by animals after ingestion is defined by the assimilation efficiency of the metal from that food (what proportion of the total is absorbed by the animal; Wang et al., 1996). Ingestion rate, the type of food ingested, and metal concentration in the food also affect how much metal is taken up from food. Experimental studies show that metal bioavailability to copepods feeding on marine algae is largely driven by the concentration of metal within the soluble fraction of the phytoplankton cell (Reinfelder and Fisher, 1991), with the two metals most bioavailable from ingestion of algae (other than selenium) being zinc and cadmium. Less than 10 percent assimilation efficiency was seen for metals that were bound strongly to the surface of the phytoplankton (as would be expected for lead). Zinc and cadmium concentrations in the water column of CDA Lake are high enough to suggest that zooplankton are exposed to these metals via feeding upon phytoplankton. The zooplankton would then be a pathway of exposure for pelagic-feeding fish.

Birds, fish, and other predators also feed on the organisms that live in the lake sediments. This benthic food web plays a central role in supporting higher trophic-level production in most lakes (Vander Zanden and Vadeboncoeur, 2002). Thus, disturbance of the benthic food web by the extremely high concentrations of lead and zinc in the sediments of CDA Lake could have ramifications across the entire ecosystem of the Lake.

Like all deep oligotrophic lakes, sediment-feeding fauna dominate the benthos in CDA Lake, including the oligochaetes, chironomids, and sphaerid bivalves (EcoAnalysts, Inc., 2017). As shown in Figure 9-9, these organisms can differ in the strategies by which they achieve their need for both oxygen and nutrition (Luoma and Ho, 1993; Lee et al., 2000). For example, Lee et al. (2000) showed that bioavailable metal exposures of taxa feeding at the oxic surface of the sediment column were greater than the exposure of animals feeding from deeper, anoxic sediments. In CDA Lake, this could ultimately result in lower risks of toxicity to head-down feeders like tubificid oligochaetes (the most abundant taxa in the benthos of CDA Lake) compared to sphaeriid bivalves that feed at the sediment surface (and are present in some locations but not others). EcoAnalysts, Inc. (2017) identified the benthic taxa present at five different locations in CDA Lake. Evaluation of differences in functional ecology among those taxa could be a sensitive measure of how the metal contamination of the sediments is affecting the benthos (e.g., Janssen et al., 2011).

Metals in Plant and Animal Tissue

How much metal accumulates into the tissues of plants and animals is a direct determinant of metal bioavailability. Such “biomonitoring” is a well-used alternative or a complement to geochemical measures and models for estimating metal bioavailability (Bryan and Hummerstone, 1977; Phillips and Rainbow, 1994).⁸ Determining the

___________________

⁸ A biomonitor is any organism that accumulates metal(loid) in its tissue in proportion to metal concentrations and bioavailability in its environment, thereby providing a relative measure of the bioavailability of that metal from all routes of exposure (Luoma and Rainbow, 2008).

amount of metal that has accumulated in taxa from nature has long been used to evaluate (1) if ecotoxicologically relevant exposures might be occurring and at what level, (2) geographic distributions of such exposures, (3) temporal variability, and more recently (4) ecosystem exposure in dose–response estimates from nature (Luoma et al., 2010; Clements et al., 2021).

There have been few attempts to determine metal bioaccumulation in the food webs of the CDA basin and CDA Lake. Scofield et al. (2021) measured arsenic, cadmium, lead, and zinc concentrations in emergent and submergent aquatic macrophytes from three lateral lakes on the CDA River, a reference location in the St. Joe River watershed, and two locations in CDA Lake (Harrison Slough and Cottonwood Bay). They found that aquatic macrophytes adsorb metals externally into their tissues during growth. The plants may be directly consumed by some species, but more importantly, they release metals back into the environment as organic detritus upon senescence and decomposition of the shed tissue (Jackson and Kalff, 1993; Weis and Weis, 2004). These metals then enter the lake food web when ingested by herbivores, omnivores, and detritivores. Scofield et al. (2021) found that arsenic and lead concentrations in all three submergent macrophytes were highest at Harrison Slough compared to the lateral lakes, reaching as high as 150 μg As/g dw and 1,250 μg Pb/g dw. Cadmium and zinc concentrations were higher in macroflora from the littoral zone of Cottonwood Bay (across the Lake from Harrison) than in the lateral lakes or Harrison Slough. Scofield et al. (2021) also determined metal concentrations in the tubers of the water potato (Sagittaria latifolia) from the locations above.

Water potato is a macrofloral species of cultural significance to the CDA Tribe as well as an important food item for native birds, including swans. All metal(loid) concentrations were higher in the leaves of the macrophytes than in the tubers of the water potatoes. Lead concentrations in the tubers in this study were similar to concentrations in S. latifolia observed in a 1999 study (up to 30 μg/g lead associated with dirt/outer skin; Audet et al., 1999); this is an indication that any decline over time in metal sediment concentrations is not yet detectable in the food web.

Birds that eat from the lake sediments (e.g., swans) or lower on the food chain have the highest dietary exposures to lead, which is not passed efficiently from prey to predators. Lead and zinc were long suspected to be a cause of widely publicized waterfowl poisonings, as indicated by analysis of soil, plant, and waterfowl specimens in the CDA region dating back to the 1950s (Chupp and Dalke, 1964). Given the high concentrations of metal(loid)s at the base of the food web, it is not surprising that elevated lead and cadmium concentrations are found in the tissues of Canada geese, American robins, and mallard (Blus et al., 1995; Henny et al., 2000; Spears et al., 2007). Ground-feeding terrestrial birds, including the American robin, song sparrow, and Swainson’s thrush, are examples of riparian animals that show elevated lead concentrations in their tissues (e.g., Sample et al., 2011). Tundra swans, Canada geese, and ducks are aquatic omnivores that are exposed to lead in macrophyte tissues, detritus, and sediment incidentally ingested with their food. Beyer et al. (2000) showed that lead concentrations in the blood of ducks, swans, and goslings increased linearly with lead concentrations in food containing lead-contaminated sediments⁹ from the CDA basin, illustrating that sediment lead is bioavailable via diet. Sediments were first sieved to remove artifactual lead (lead shot, sinkers etc); hence, the bioavailability resulted from mining-derived lead in sediments.

Arsenic is also bioaccumulated into at least riparian zone food webs near mine sites. Similar to lead, arsenic concentrations are higher in lower trophic level taxa. For example, arsenic concentrations in submerged plants, mosses, algae, and biofilm reach tens to hundreds of mg/kg dry weight near the Stibnite mine site in Central Idaho (Dovick et al., 2015). Tadpoles accumulated hundreds of mg/kg arsenic at the site, primarily from ingestion of algae and biofilms, while levels of arsenic in macroinvertebrates and trout were lower (maximum 30 mg/kg) than in the tadpoles. Arsenic bioaccumulation in the riparian food webs of CDA Lake have not been studied, but these could also be vulnerable to arsenic exposure.

Farag et al. (1998) measured cadmium, lead, and zinc concentrations in sediments and biofilms,¹⁰ benthic invertebrates (primarily insect larvae), whole fish, and trout kidney at the confluence of the North and South Forks of the CDA River and in the mainstem CDA River near Harrison. They compared those concentrations to a reference location in the St. Joe River watershed. Biofilms and sediments are a food source for invertebrates and the invertebrates are a primary route of exposure for predaceous fish (Farag et al., 1998). Concentrations of metal(loid)s by all measures were patchy and site-specific, without a clear gradient between Cataldo and Harrison. Concentrations of arsenic, cadmium, and lead in invertebrates were 10–40 times higher at Harrison than in invertebrates in the St. Joe River watershed, while concentrations of zinc were 2–3 times higher; the lesser magnitude is a result of the high natural concentrations of zinc (an essential metal) in invertebrates.

Metal concentrations declined according to the ranking: sediments = biofilm > invertebrates > whole fish (Farag et al., 1998). Thus, tissue concentrations of arsenic, cadmium, zinc and lead do not increase from prey to predator (biomagnify), but all four are bioavailable to invertebrates and passed on, to a degree, to fish predators (Table 9-3). The enhanced exposure derived from the contamination in this food web is clear evidence that metal(loid)s originating from the CDA River and depositing in the Lake are not in an innocuous form unavailable to aquatic life and therefore have the potential to generate adverse ecological effects. Direct study of metal bioaccumulation in fish from the Lake itself would be an important next step in understanding the implications of this bioavailability.

___________________

⁹ They determined that incidentally ingested sediments composed ~ 22 percent of the food eaten by swans in the field. The food in the experiments contained 22 percent sediments from the CDA basin with different lead concentrations.

¹⁰ Biofilms consist of algae, bacteria, and associated fine detrital material attached to rocks and other substrates in water bodies. They are an important food source for an especially metal-vulnerable functional group of benthic invertebrates that scrape hard surfaces for food.

TABLE 9-3 Metal Concentrations in Two Food Sources (Sediments and Biofilm) to Three Different Taxa: Benthic Invertebrates, Perch (Whole Body) and the Kidneys of Cutthroat Trout

NOTE: Ref = St. Joe River basin; HSR = Harrison in the lower CDA basin. *sampled upstream of Harrison.

SOURCE: Farag et al. (1998).

Evidence from Biomarkers

Few studies have looked for biochemical signs of metal stress (biomarkers) in organisms from either the CDA watershed or CDA Lake but those that have indicate that bioaccumulated exposures are sufficient to elicit stress responses at the molecular level. Beyer et al. (2000) showed that depressed levels of δ-aminolevulinic acid dehydratase (an enzyme inhibited by lead) accompanied lead bioaccumulation in ducks fed a diet contaminated with CDA sediments. Ducks and tundra swans that bioaccumulated lead from littoral sediments in the CDA basin had blood lead levels high enough to indicate lead poisoning and also showed depressed levels of δ-aminolevulinic acid dehydratase (Blus et al., 1995, 1999; Henny et al., 2000; Spears et al., 2007). Several lines of evidence showed that the lead exposure came from the mine-contaminated soils, not lead shot (Blus et al., 1999). Osprey are also exposed to lead through their diet of fish and show depressed levels of δ-aminolevulinic acid dehydratase, although effects on reproduction rates were not detectable (Henny et al., 2000).

Biomarker studies in invertebrates and fish can yield valuable metrics in assessing both bioavailability and potential for stress in aquatic food webs. A diet of metal-contaminated invertebrates had physiological effects and led to metal accumulation in tissues of rainbow trout (Farag et al., 1994). Farag et al. (1999) fed a diet of metal-enriched invertebrates that had been collected at Pinehurst and Cataldo on the South Fork of the CDA River to cutthroat trout. They observed reduced feeding activity, “increased number of macrophage aggregates and hyperplasia of cells in the kidney, degeneration of mucosal epithelium in the pyloric caecae, and metallothionein induction” in the trout; they concluded that these effects would likely reduce growth and survival of fish in the wild. While both diet and water chemistry affected metals levels in the fish, diet was the primary route of exposure in such experiments (see, e.g., Woodward et al., 1994), and it disrupted digestive physiology, thus affecting growth and survival (e.g., Hansen et al., 2009). Even if benthic communities are not themselves disturbed, contaminated benthos can be an important pathway of potentially harmful metal exposures to their predators. No determinations of either physiological or biomarker responses in fish from different food webs (e.g., pelagic vs. benthic) in CDA Lake have been conducted. The studies with birds illustrate the potential value of adding such metrics to any comprehensive evaluation of metal effects on the ecological health of the Lake.

Whole Organism: Overt Toxicity

Wildlife provide one of the most obvious ecosystem services delivered by CDA Lake. The most overt evidence that the legacy of mineral extraction is damaging wildlife is the observations of hundreds of water bird mortalities over the years in the watershed. In the 1990s, over 600 animals (29 species of birds, including waterfowl, songbirds, and birds of prey; six species of mammals, including vole, muskrat, mink, and beaver; and amphibians and reptiles) were found sick or dead in the CDA River basin, compared to only 40 in the St. Joe River basin (Audet et al., 1999). Lead poisoning was documented in the vast majority of these animals in the Coeur d’Alene area,

and over 90 percent of these poisoned animals had not ingested lead artifacts, such as shot or fishing sinkers. In the St. Joe River, lead poisoning is also observed, but it is related to ingestion of lead artifacts.

Fewer data on bird toxicity were available from CDA Lake itself until Spears et al. (2007) assessed lead concentrations in sediments and the blood of mallards and wood ducks from 22 shallow bay locations and 11 wetlands within and around the lake. Blood lead from ducks from the Lake was correlated with concentrations of lead in sediments. Blood lead concentrations exceeding thresholds for “clinical poisoning” (> 5 μg/dL) or severe clinical poisoning (> 10 μg/dL) occurred in ≥ 50 percent of the mallards from four of the eight locations sampled. Fecal samples from 19 Canada geese and three mallards indicated that these animals were “exposed to lead by ingesting contaminated lake sediment.” Spears et al. (2007) calculated that adverse effects (physiological to population parameters) could occur over the range 147–944 μg/g dw lead in sediment. They also concluded that overt mortality of mallards utilizing CDA Lake could occur at lead concentrations of 1,652 mg/kg dw in sediments. These thresholds for toxicity estimated by Spears et al. (2007) are similar to those determined by Beyer et al. (2000) from experimental data with swans, where it was shown that physiological stress occurs above 530 μg Pb/g dw in sediment and that 1,800 μg Pb/g dw caused overt toxicity. Spears et al. (2007) concluded that “locations of Harrison Slough, Powderhorn Bay, Cottonwood Bay, Blackwell Island and Cougar Bay near the Spokane River outflow of CDA Lake were the areas of greatest concern for waterfowl exposure to lead-contaminated sediment.” Because there is no evidence that lead concentrations in the Lake have changed detectably since the date of this study, it seems likely that similar bird exposures apply in 2022.

In summary, it appears that lead exposures of wildlife in some locations in CDA Lake can be as significant as exposures that occur in the watershed, but exposures in the Lake are patchy. In some areas, sediments are sufficiently contaminated to cause overt toxicity, and long-term adverse effects are possible. There are also areas for which contamination is insufficient to result in detectable effects. It is important to recognize that only a few locations have been studied in this large lake; no systematic picture is available. Careful management of wildlife depends upon a clearer understanding of the frequency and location of metal-driven adverse effects and studies of exposures and biomarkers in more types of birds and other wildlife. More experiments with or data on blood lead in wildlife in areas of median level exposure, or across multiple gradients of exposure, would also help clarify thresholds where sublethal effects are damaging wildlife. Clearly, remediation of the areas of extreme contamination would benefit the ecosystem services provided by wildlife for the Lake community and Idaho as a whole.

Experimental Studies of Toxicity

The toxicity of the metals of interest in CDA Lake is generally known, but toxicity testing that could be most relevant to the Lake is limited. Woods and Beckwith (1997) stated that several early direct studies of the toxicity of CDA Lake waters to fish, invertebrates and phytoplankton were cited by Savage (1986). In 1993 and 1994, two phytoplankton taxa (the diatoms Achnanthes minutissima and Cyclotella stelligera) were isolated from limnetic locations in the Lake with different zinc concentrations (Woods and Beckwith, 1997). Nominal zinc concentrations of 19.6 and 39.2 μg/L, which are typical of the Lake, both strongly inhibited growth of both species of diatoms. Later, Kuwabara et al. (2006, 2007) isolated two algal species common to the Lake, Chlorella minutissima and Asterionella formosa. The goal of the experiment was to test the hypothesis that, at high enough bioavailable concentrations, zinc suppressed cell division (growth) and inhibited utilization of the limiting nutrient phosphorus by the algae. This would allow phosphorus to accumulate intracellularly without stimulating growth. Functionally, the hypothesis meant that above some bioavailable concentration, zinc would limit the algal growth below the potential for growth defined by phosphorus. They tested only three concentrations of zinc and three concentrations of phosphorus, so the study was not designed to define a threshold for the effect. Growth in both isolates was inhibited at zinc concentrations typical of the Lake, although the chlorophyte C. minutissima was more affected by zinc than was the diatom isolate A. formosa. The adverse effect of zinc on growth rate was statistically significant for both species, but changes in cell concentration (the number of cells per volume of culturing suspension) were not statistically significant at the 95 percent confidence level for the diatom.

No experiments have been published that were designed to test potential toxicity of zinc, cadmium, lead, or arsenic to zooplankton under conditions directly relevant to the Lake. Studies determining relative sensitivities of zooplankton taxa that occur or might be expected to occur in the Lake could be of particular interest given

the sensitivity of zooplankton community structure observed in other metal-enriched lakes (Yan and Strus, 1980; Marshall et al., 1983; Keller et al., 2007; Valois et al., 2010).

Extrapolations of the results of single-species toxicity tests to conclusions about the effects of metals on communities and populations have limitations. Mesocosm studies could be more relevant to CDA Lake. Such studies employ a reasonable representation of the natural community or exposure to one taxon over an entire life cycle. The most realistic studies equilibrate dissolved and dietary exposure, and capture some population or community responses (examples include Marshall et al., 1983; Cairns, 1989; Iwasaki et al., 2018; Mebane et al., 2020).

Controlled multi-species mesocosm studies with Florida flora and fauna provide one example of such an experimental approach (Hoang et al., 2021; see Box 9-2). Box 9-2 shows that complex changes occurred in phytoplankton community structure (e.g., with chlorophytes, cyanobacteria and cryptophytes) during an exposure to zinc lasting months. The most sensitive responses occurred at 8 μg Zn/L, including reduced diversity and reduced abundance of Chlorophyta (both phytoplankton and periphyton) and Chrysophyta. Cyanobacteria took over the community after 30 days of zinc exposure at extreme concentrations (80–100 μg/L). The authors concluded that the No Effects Threshold for zinc in this system was 14 μg/L (similar to the recommendation of Wong and Chau, 1990, for lakes, in general). Although results from one particular study cannot be directly extrapolated to CDA Lake, they illustrate the complexity of the phytoplankton community responses that might be expected as zinc concentrations change in the Lake. CDA Lake also seems amenable to in situ experimentation with simplified water column communities (see Marshall et al., 1983) given the north-to-south gradients in zinc exposure and the relatively high concentrations of zinc in the northern Lake.

Two mesocosm studies of zinc (or zinc and cadmium in combination) toxicity to the aquatic insect communities typical of Rocky Mountain streams are useful examples of approaches to developing concentration thresholds for water-column concentrations where rocky bottom (and perhaps related) communities began to simplify (see, e.g., Schmidt et al., 2011; Mebane et al., 2020). Such studies consistently show that mayfly populations declined then disappeared at lower-concentration metal treatments than did other taxa. Lake mayfly taxa, such as Hexagenia spp. (which is found in the southernmost CDA Lake locations) have been suggested (Milani et al., 2003; Reynoldson et al., 1989) as a sensitive indicator of adverse anthropogenic effects on benthic lake communities and as potential test organisms for sediment toxicity testing.

TABLE 9-4 Comparison of Sediment Quality Criteria to Measured Cadmium, Lead, and Zinc Concentrations in CDA Lake Sediments

NOTE: First row gives the ranges of sediment quality criteria across 15 different approaches to developing guidelines. Second row gives the range of median cadmium, lead, and zinc concentrations in CDA Lake sediments (see Table 9-2). Third row gives the magnitude by which the low and high medians estimated for CDA Lake sediments exceed the low and high criteria.

In lieu of site-specific studies, general toxicity thresholds from the literature can be used to estimate levels of concern for zinc and lead in CDA sediments. These are termed sediment quality criteria (SQG). These thresholds were mostly developed using single-species or multiple-species tests directly on sediments from different environments.

Many factors can affect the toxicity of sediments, including metal concentration, particle size, the type of organism, how that organism interacts with the sediment, and the geochemistry of the sediment (MacDonald et al., 1996). Thus, the threshold of toxicity may differ among lakes. To take this variability into account, a “weight-of evidence approach” was developed based upon a database of toxicity tests from different environments, with different levels of contamination, different geochemistry, and a variety of organisms. Studies showing adverse effects are ranked by concentration of the chemical, and then a percentile is chosen (e.g., the median concentration at which some adverse effect occurs) as the criterion (e.g., Effects Range Median, National Oceanic and Atmospheric Administration). These rankings have many limitations when used for precise predictions for any specific environment but are useful in establishing the general range within which chronic or sublethal toxicity occurs. Hubner et al. (2009) listed criteria established by 15 different approaches (first row of Table 9-4). The ranges shown in Table 9-4 include criteria used in local studies to assess the meaning of sediment metal concentrations in the CDA basin (Maret et al., 2003¹¹); or in the Lake (Morra et al., 2015¹²; EcoAnalysts, Inc., 2017).

The magnitude by which the sediments of CDA Lake exceed observations from toxicity tests on sediments suggests a high likelihood of toxicity to at least some taxa and of effects on the benthic food web of the Lake. Median concentrations of cadmium, lead, and zinc in different studies exceed the median concentration expected to cause toxicity by 4- to 16-fold. In a multi-metal exposures like CDA Lake, the sum of the magnitudes by which each metal exceeds the criteria provides an estimate of the toxicity of the mixture, termed the cumulative toxic unit (assuming toxicity of different metals is additive—Maret et al., 2003). The cumulative toxic units for median cadmium, lead, and zinc in the Lake suggest that the mixture of contaminants in the sediments of the Lake exceed by 25-fold the median (not the lowest level) concentration that typically causes toxicity in a controlled test. Hubner et al. (2009) concluded that “although such criteria are not definitive measures of toxicity, they can have a high predictive ability and are a vital tool for identifying areas with potentially adverse biological effects.” Clearly, CDA Lake is an environment where sediment toxicity would be expected, based upon controlled testing with sediments in general. Conducting specific studies with relevant taxa seems an important next step (EcoAnalysts, Inc., 2017).

Community Responses

Community structure, function, and productivity are driven by a wide variety of factors in lakes (see, e.g., Caires et al., 2013). Unambiguous demonstrations of community impacts for any individual stressor, including metals, are inherently challenging. Changes in food webs in the basin upstream from CDA Lake provide evidence that metals can affect food webs dramatically. The 2001 ecological risk assessment concluded: “Toxic effects of

___________________

¹¹ Maret et al. (2003) cited probably effects level (sediment criteria) of Cd 3.5, Pb 91, and Zn 315, all in μg/g dw.

¹² Morra et al. (2015) compared EPA criteria to surface sediment concentrations in a core collected from near Harrison: Cd = 2.49 vs. core 25.3; Zn = 384 vs. core 3,326; Pb 161 vs. core 3,000-4,000, all in μg/g. The cumulative toxic unit would be about 30.

contaminated sediment are believed to contribute to adverse effects on aquatic life in. . . the entire South Fork, the Coeur d’Alene River, the Spokane River, and, possibly, some parts of Coeur d’Alene Lake.” Hoiland et al. (1994) reported that taxonomic richness (number of species) in the South Fork of the CDA River at Smelterville was zero (no macrofaunal life) in 1968. Life began to reappear 20 years after tailings ponds were completed in 1968. Taxa richness increased to six species in 1987 and 15 in 1991. The number of taxa in 1991 was about half the number found at reference sites on the North Fork.

Maret et al. (2003) found seven to nine EPT¹³ taxa at two locations on the CDA River near Cataldo, and zero to one metal sensitive Ephemeroptera (mayflies) in 2000 (compared to 10 to 19 EPT taxa and three to seven mayfly species in reference areas). In 2010 the presence of zinc and cadmium was linked to lower populations of plecopterans (stoneflies) and reduced aquatic insect diversity in the CDA Lake region (Lefcort et al., 2010). Fish communities were depauperate in numerous places and some taxa typical of watersheds (sculpins) were completely missing from rivers and streams in the CDA basin (Maret and MacCoy, 2002). Thus, effects of metal contamination on the benthic and fish communities were unambiguous in the basin upstream from the Lake. The degree to which such conclusions apply to the Lake at present depends upon understanding the characteristics of the pelagic and benthic communities of the system.

Ecological Communities in Lake Water Columns

The community of organisms adapted for life in suspension in the water column of lakes, rivers, and oceans is referred to as the plankton, with the photosynthetic members of the plankton being the phytoplankton. Collectively, phytoplankton communities are formed by various species of eukaryotic photosynthetic protists (“algae”) that include, at the taxonomic level of division, the Chlorophyta (“green algae,” such as Chlorella or Scenedesmus), the Chryosophyta (the “golden brown algae”, which include diatoms), the Cryptophyta (e.g., Cryptomonas, Rhodomonas), and the Dinophyta (e.g., Ceratium, Peridinium). Importantly, included among the phytoplankton are the photosynthetic prokaryotes, the cyanobacteria. These include very small and ubiquitous unicellular taxa (e.g., Synechococcus) as well as colonial taxa that can be toxic and/or fix atmospheric nitrogen (e.g., Microcystis, Anabaena).

Phytoplankton are important in forming the basis of pelagic food webs. They are consumed by filter-feeding zooplankton species that themselves form the food base of planktivorous fishes, which support piscivorous fish often prized by anglers. Phytoplankton are also key determinants of lake water quality because phytoplankton growth is often limited by nutrients such as nitrogen and phosphorus. Their growth can lead to algal blooms and decreased water quality when nutrient inputs increase, frequently from anthropogenic sources (e.g., septic leachate, fertilizer runoff). Under some conditions of increased nutrient loading, some phytoplankton species, especially cyanobacteria, can produce toxic substances harmful to fish and humans. The species composition of the phytoplankton reflects environmental conditions, with diatoms, green algae, and small unicellular cyanobacteria dominating low-nutrient, “oligotrophic” conditions and colonial cyanobacteria becoming important in eutrophic systems.

Phytoplankton Data in Coeur d’Alene Lake and Lateral Lakes

A phytoplankton community assessment of CDA Lake was performed at stations C1 (Tubbs Hill) and C4 (University Point) across several seasons from 2007 to 2017 (EcoAnalysts, Inc., 2020). The diatom Asterionella formosa was found to be the most abundant taxon at both stations during runoff and warm stratified periods. During the cold clear period, the potentially toxic cyanobacterium Microcystis was important at C1 while the diatom Tabellaria flocculosa and the cryptomonad Cryptomonas spp. were dominant at C4. The study showed that different seasons of the year harbored different phytoplankton communities. Analysis of annual data indicated that 2007–2008 was distinct from subsequent years in terms of community structure at both sites, but no consistent trends are obvious during the last ten-year interval. The relatively high abundance of the potentially toxic cyanobacterium Microcystis at C1 (Tubbs Hill) is of potential concern, but this is largely associated only with the cold, clear period (winter). Otherwise, the dominant taxa are diatoms, which are generally indicative of good water quality.

___________________

¹³ EPT refers to the three orders of aquatic insects most common in cobble bottom streams and rivers (Ephemeropteral, Plecoptera, and Trichoptera). As stress increases, these three taxa groups typically decline in abundance and are replaced by other, more tolerant, groups.

Phytoplankton population and community data are also available for sites in the southern part of the Lake (C5 and C6), the lower St. Joe River, and two lateral lakes (Benewah Lake, Round Lake) for 2011–2020 (CDA Tribe and Avista Corporation, 2017, and courtesy of D. Chess, CDA Tribe). According to the report for 2011–2015 (CDA Tribe and Avista Corporation, 2017), phytoplankton cell numbers were dominated by cyanobacteria (largely small unicells). However, based on biovolume (biomass), taxonomic composition was distributed relatively evenly among major groups (cyanobacteria, diatoms, greens, and cryptophytes/chrysophytes). Site C5 had the lowest average biomass. Community differences (assessed by non-metric multidimensional scaling, a method of reducing complex multivariate data into fewer dimensions to facilitate interpretation) identifed years with higher runoff as distinct due to increases in abundances of benthic diatoms. At the genus and species level, taxa typical of eutrophic conditions (Anabaena circinalis and Aphanizomenon spp.) were abundant (based on relative biovolume) at C6 and the lateral lakes but not at C5, where diatoms (Fragilaria crotonensis, Tabellaria fenestrata), Chlamydocapsa spp. (green), and Euglena (euglenoida) were more prevalent. Ordination analysis (compression of multivariate species composition data into reduced dimensions) of phytoplankton community composition found that community composition was associated with a number of environmental variables (orthophosphorus, specific conductance, temperature, chlorophyll, nitrate, pH, fluorescence, and dissolved oxygen). Notably, little discussion of possible temporal trends within the five-year synthesis period was presented, nor were comparisons with previous years of data discussed.

Longer-term data for phytoplankton community composition were provided to the committee for 2007–2019 (courtesy of D. Chess, CDA Tribe), including summertime (May–September) total phytoplankton biovolume, cyanobacteria biovolume, and percentage of biovolume contributed by cyanobacteria.¹⁴ As shown in Figure 9-10A, at C5 phytoplankton biovolume was high during the initial 2 years but rapidly declined after 2008, increasing to moderate levels in 2015–2016 and then declining again. Cyanobacteria biovolume was low throughout the

___________________

¹⁴ Text here and throughout the chapter was edited after report release to include the full phytoplankton dataset.

observation period except in 2008 when cyanobacteria dominated the community throughout the year, contributing ~ 65 percent on average. Dominant among the cyanobacteria at C5 during 2008 was Microcystis, a potentially toxic taxon. Excluding 2008, cyanobacteria contributed ~ 12 percent (on average) to phytoplankton biomass across the years reported (16 percent if 2008 is included). Data for C6 indicate phytoplankton biomass levels somewhat higher than those at C5 during the same period (Figure 9-10B). Also, cyanobacteria played a more important role at C6 than at C5, contributing 31 percent to phytoplankton biomass. Important taxa at C6 included the potentially toxic forms Anabaena circinalis, Microcystis spp., Aphanizomenon, and Planktothrix. This is consistent with other metrics at C6 that indicate a more eutrophic state relative to other areas of the Lake.

Data have also been provided for epilimnetic phytoplankton for two lateral lakes (Swan Lake, Thompson Lake) in 1 year (2015) with six sampling dates between mid-July and mid-September. The data consist of biovolume concentrations (mm³/L) obtained by microscopic identification and counting of cells combined with estimates of per-cell volumes. Identification was to the level of genus. Overall, phytoplankton biovolume was dominated by Chlorophyta (green algae such as Euglena, Oocystis, Gloeococcus, and Tetraedron; ~ 48 percent) and Cryptophyta/non-diatom Chrysophyta (Chroomonas, Dinobryon, Komma, and Cryptomonas; ~ 24 percent). Cyanobacteria (Chroococcus, Synechococcus, with some Microcystis) made up 13 percent of biovolume, on average, in these samples.

Overall, C6 and some of the lateral lakes are relatively productive and display some eutrophic characteristics in terms of higher cyanobacteria abundance (and nutrients and oxygen depletion), but these conditions diminish moving into the main Lake (e.g., at C5), with a transition to communities more characteristic of oligotrophic conditions in northern parts of the Lake. From 2007 to 2019 at C5, phytoplankton biovolume was highest in the early part of the observation period but has been lower in recent years. Although a single year (2008) that involved high biomass of cyanobacteria-dominated (~ 65 percent) phytoplankton was observed, C5 maintains relatively low levels of cyanobacteria overall. Based on these data, it is problematic to infer patterns in the temporal dynamics of phytoplankton communities in CDA Lake except to note that since 2007 water quality conditions in terms of phytoplankton biomass and cyanobacteria dominance seem to have improved. Nevertheless, observation of high phytoplankton biomass with cyanobacteria dominance in 2008 indicates the possibility for the Lake to support blooms of potentially harmful cyanobacteria given suitable conditions.

In their toxicity studies, Kuwabara et al. (2006) noted that the results were consistent with the observations that zinc-sensitive C. minutissima decreased in cell concentration away from the mouth of the CDA River, and the more tolerant A. formosa increased in abundance as metal concentrations increased in 2005. They concluded that “significant differences in response by the phytoplankton isolates in this study suggest that observed longitudinal shifts in phytoplankton community composition may represent a response to longitudinal gradients in solute (Zn) concentrations.” They suggested implications for management of the Lake: “If dissolved Zn can be reduced in the water column from > 500 nM (i.e., current concentrations near and down stream of the Coeur d’Alene River plume) to < 3 nM (i.e., concentrations near the southern St. Joe River inlet) such that the Lake is truly phosphorus limited, management of phosphorus inputs by surrounding communities will ultimately determine the limnologic state of the lake.” This implies that zinc inhibits phytoplankton growth in the Lake at present, and has become the basis for a current concern that if the presumed zinc inhibition were reduced, the Lake might be more prone to eutrophication than at present. The conclusions of Kuwabara et al. (2006, 2007) were based upon studies conducted at one point in time in a complex lake in which zinc exposures themselves are complex in both time and space. Although the experimental observations were consistent with the field collection, they did not have the benefit of the 14 years of consistent seasonal monitoring now available for the Lake (as described above) and the limnological understanding that has developed since 2005. Statistical tests of the influence of zinc on phytoplankton biomass (chlorophyll a [chl a]) performed by the committee did not support zinc inhibition (see Chapter 5). Although the committee cannot exclude the possibility of zinc inhibition of the growth of some phytoplankton taxa in the Lake, the community dynamics described above are not consistent with zinc being an important driver of phytoplankton community structure. For example, euphotic-zone zinc concentrations at C5 are below typical thresholds of toxicity in spring due to seasonal inputs from the St. Joe River, then increase dramatically to above experimentally derived thresholds in the late summer and fall. But changes in the relative abundance of Chlorella spp. and A. formosa are not coincident with those changes in zinc concentrations. This is an illustration of both the challenges in extrapolating results from limited single-species toxicity testing to the limnological processes that drive

phytoplankton community dynamics in a large, complex lake and of the value of multiple lines of evidence when evaluating metal influences.

Macrobenthic Invertebrate Community Structure, Distributions, and Dynamics

Macrobenthos refers to the community of larger organisms that lives on the bottom (the benthic habitat) of lakes and rivers. The illuminated region of the benthic zone is referred to as the littoral zone, and the dark, poorly illuminated region is called the profundal zone. This community includes an important set of larger invertebrate animals (macroinvertebrates) such as the larval or adult forms of various insects (mayflies, dipterans, dragonflies, damselflies), mollusks (snails, mussels), and segmented worms (oligochaetes). Macroinvertebrates are ecologically important in the littoral zone for transferring primary production to higher trophic levels and in the profundal zone for transferring detrital materials to higher trophic levels. Macroinvertebrates can also transfer contaminants from sediments to higher trophic levels (Lavoie et al., 2013) and participate in bioturbation, which can affect the exchange of oxygen, nutrients, and metals between sediments and the water column. Macrobenthic taxa differ in their sensitivity to various environmental factors, such as oxygen or metal concentrations, and thus are often used as bioindicators of chemical conditions in rivers and lakes (Hauer and Lamberti, 2011).

The benthos of large oligotrophic and mesotrophic lakes is typically characterized by low abundances (total number of individuals) and high variability in both time and space compared to more productive lakes. Physical differences in substrate (such as grain size), instability of the substrate, differences in depth, differences in the availability of food, and other factors drive seasonal, year-to-year, and spatial variability in abundance and other community measures (White and Miller, 2008; Caires et al., 2013; Hayford et al., 2015). Organisms with broad environmental tolerances or life cycles suited to these changing conditions do best in large, deep oligotrophic or mesotrophic lakes.

In CDA Lake, highly enriched concentrations of arsenic, cadmium, lead, and zinc are an additional potentially influential variable. Sediments from Lake locations north of C5 have metal concentrations that fall within the window or exceed the window predicting toxicity according to sediment quality criteria (EcoAnalysts, Inc., 2017; Table 9-4). Such heavily contaminated conditions can select for species that avoid bioavailable metal forms, confounding broad measures of community change like abundance and even taxa richness. Thus, changes in community structure attributable to metal enrichment alone are typically difficult to distinguish from changes caused by other perturbations unless the impact on the benthos is dramatic (Nalepa and Landrum, 1988).

These challenges are evident in the studies to date of the benthos of CDA Lake. Horowitz et al. (1993, 1995) first noted that sediment cores from profundal locations in CDA Lake often included varved (layered) sediments that clearly reflect undisturbed material. Varved sediments reflect minimal sediment reworking by the benthic community, typical of a depauperate community, suggesting that was the case in at least deep lake locations in the past. The varving was sufficiently distinct that the authors used it to estimate when effects from mineral extraction activities began. Deeper layers (deposited before 1980) in the core studied by Morra et al. (2015) showed similar varving.

Benthic production supports higher trophic levels in most lakes but can often go underappreciated. In a survey across a northern temperate lake, Vander Zanden and Vadeboncoeur (2002) found that lake fish relied on the benthos for 65 percent of their food consumption and that the benthic food web played “a central role in supporting higher trophic level production and ecosystem processes in the pelagic zone.” Given the anecdotal suggestion from Horowitz et al. (1993) of a depauperate benthos, it is important to have a better understanding of benthic production, benthic community structure and function, and how they are affected by the “profound enrichment” of arsenic, cadmium, lead, and zinc in the sediment of CDA Lake.

Data from four surveys (1972, 1996, 1999–2005 and 2011–2015) of macrobenthos communities in various localities within CDA Lake are available. While at coarse time resolution and involving somewhat different localities, the studies permit some insight into the status and dynamics of macrobenthos in the Lake, as described in Box 9-3.

Although the emphasis of the two most thorough benthic studies (EcoAnalysts, Inc., 2017; Kuwabara et al., 2006) differed somewhat, many results were comparable between the two. Figure 9-11 shows that the frequency distributions of density and taxa richness among samples were similar between the two studies when similar

depths were compared.¹⁵ Figure 9-11A also compares the two CDA Lake studies to studies from oligotrophic lakes with little metal contamination—the 17 lakes in the Convict Creek basin (Reimers et al., 1955) and a study using comparable methods from Lake Tahoe (Caires et al., 2013). The main difference was the high frequency of locations with densities less than 2,000 per m² in CDA Lake. Caires et al. (2013) found no locations in Lake Tahoe

___________________

¹⁵ Water depth influences both invertebrate density and richness in all lakes. To control for that, in Figure 9-11A only samples > 10 m depth from EcoAnalysts, Inc. (2017) were used to compare to Kuwabara et al. (2006) because the latter study included no shallow-water samples. Richness was compared between shallow-water samples in EcoAnalysts, Inc. (2017) and the location from the St. Joe River, because water depths were shallow. In the comparison with Caires et al. (2013), only samples from similar depths were compared.

with macroinvertebrate densities as low as several of the CDA Lake sites. While many factors affect invertebrate densities in oligotrophic lakes, locations as depauperate as many of those in CDA Lake are rare.

Distributions of taxa richness were reduced in the contaminated portion of CDA Lake compared to a reference location (Figure 9-11B). As noted earlier, taxa richness in benthic communities in the basin of the CDA River has recovered from zero before 1968 to about half the richness of reference sites in the most recent samplings (Hoiland et al., 1994; Maret et al., 2003). It would not be surprising if similar observations applied to the Lake.

At a qualitative level, the macrobenthos communities of CDA Lake have been consistently dominated by chironomids and oligochaetes throughout the period 1972–2015. In general, this is a community typical of large, deep oligotrophic/mesotrophic lakes. This ecological stability over the window of observation limits our ability to draw inferences about response of macrobenthos to changing metal concentrations or nutrient levels during this interval.

Locations with unusually depauperate benthic communities are characteristic of at least some deeper waters in CDA Lake. Whether the frequency of such locations is sufficient to affect benthic production overall cannot be determined from the limited data available to date. As to the role of metal enrichment, the conclusion made by EcoAnalysts, Inc. (2017) probably applies to all studies to date:

Understanding benthic responses to either nutrients or metals could be improved by careful analysis of differences in sensitive/tolerant species and functional attributes, especially within the family Chironomidae (e.g., Caires et al., 2013) or the subclass Oligochaeta (Vivien et al., 2020). Changes in those groups can signal the causes of differences (Vivien et al., 2020) or have implications for processes such as reduced productivity (Caires et al., 2013).

Benchmarks for Different Metals

NRC (2005) suggested that an important outcome of evaluating ecological health was to establish potential remediation goals. In a simple world, one should be able to identify the threshold concentration beyond which each trace metal begins to elicit adverse effects in natural waters. The relationship between the concentration of trace metal in a waterbody and the influence of the resulting contaminant concentration on life in that environment depends upon detecting metal “toxicity” in nature. As is clear from the analysis above, quantifying the response of ecological health to metal enrichment, or the threshold for harmful exposures, is fraught with uncertainty and, in some cases, controversy (Luoma and Rainbow, 2008). The uncertainties stem, at least partly, from the complexity of the problem; from different perspectives and methods for determining toxicity; and the choices of data used for the benchmarks, criteria, or goals (Buchwalter et al., 2017).

Defining when remediating the legacy of mineral extraction is complete for CDA Lake may be beyond EPA’s remit at present. But it is nevertheless important to define expectations of ecological health or benefits from ecosystem services that would derive from reducing metal impacts. Is remediation complete when the EPA’s National Ambient Water Quality Criteria (AWQC) are met; when the Lake Management Plan (LMP) targets are met; when risks to biodiversity, native fisheries, and other ecosystem services are reduced or removed; when traditional foods are safe to eat; or when the geochemical aspects of the Lake’s ecology become similar to other lakes of its size and character? Each of these goals would require a different (progressively lower) benchmark for each metal(loid). Defining transparent, explicit goals (perhaps in ecological terms) and linking those goals to a benchmark (i.e., target) is an essential step in a sustainable plan for the future of CDA Lake, even if those goals differ among interest groups. Examples of benchmarks (considering zinc only) for different expectations from lake ecology are proposed below as guides to evaluating if changes in the lake are sufficient to improve ecosystem services.

TABLE 9-5 Chronic EPA Ambient Water Quality Criteria for Dissolved Cadmium, Lead, and Zinc Corrected for Hardness Ranging from 10 to 100 mg/L CaCO₃

SOURCES: Balistrieri et al. (2002) and EPA (2016) for cadmium.

The appropriate benchmarks to protect ecological health in a specific environment, like CDA Lake, depend upon goals designed around expectations of the ecosystem after remediation. For example, Table 1-6 shows that most large lakes in the United States have open-water dissolved zinc concentrations of < 1.0 μg/L. Van Genderen et al. (2008) found median concentrations of zinc of 2–3 μg/L in a variety of lakes from different countries in Europe. Zinc concentrations in the St. Joe River and reference areas upstream of most historic mineral extraction activities fall in a similar range, 0.2–2 μg/L (Kuwabara et al., 2007; Chapter 6). Thus, a goal of 1–3 μg/L might represent pre-mining zinc concentrations that are typical for large lakes subject to some human development on their shores. These values are close to concentrations observed in the St. Joe River near the confluence with CDA Lake. It can be assumed most such lakes provide a variety of ecosystem services and minimally disturbed ecological functions.

The simplest goal for CDA Lake is compliance with regulatory criteria. Existing criteria are designed to protect 95 percent of species and limit the most overt ecological damage. The EPA’s National AWQC, from which local criteria are derived, correct the national values for local water hardness (which varies in CDA Lake from 15 to 65 mg/L CaCO₃; Balistrieri et al., 2002). Chronic AWQC for dissolved cadmium, lead, and zinc in the CDA basin, across a range of hardness, are presented in Table 9-5.

Most criteria, guidelines, targets, or benchmarks are defined by a single number, for simplicity, but are corrected for water quality characteristics that are themselves variable. This can result in variability in criteria among jurisdictions within the same region. For example, EPA Region 10’s hardness-corrected target for zinc in the CDA basin, as described to the committee, is 58 μg Zn/L, while the LMP hardness-corrected target is 36 μg Zn/L. Presumably, the different targets reflect differences in the hardness correction between jurisdictions. DeForest and Van Genderen (2012) used a more inclusive geochemical modeling approach (the Biotic Ligand Model) to correct for local water quality conditions and calculated the appropriate range of benchmarks to be 18–27 μg Zn/L for CDA Lake. Thus, a single target concentration for cleanup is useful pragmatically, but it likely reflects jurisdictional interpretations of water quality and is not necessarily indicative of a threshold for ecological effects in a region.

In developing criteria, EPA does not allow data other than single-species toxicity testing for most metals (Buchwalter et al., 2017). Yet, compared to single-species toxicity testing, mesocosm studies can include a more realistic assemblage of species and can consider both dissolved and dietary exposures (e.g., Mebane et al., 2016), thereby providing more realistic estimates of the potential for ecological effects. In most cases, the thresholds of toxicity derived in tests that more carefully mimic nature are lower than those derived from single-species tests (Mebane et al., 2020; Clements et al., 2021)—that is, ecosystems are usually more sensitive to metals than suggested by single species toxicity testing. It has long been argued (e.g., Cairns, 1986) that combining single-species toxicity testing with mesocosm and field data might result in more realistic criteria or benchmarks than the present approach. Such criteria might better meet a goal of maximizing biodiversity and ecosystem services by minimizing simplification of the Lake community.

Hoang et al. (2021) (see Box 9-2) showed that the first effects of zinc on a mixed assemblage of phytoplankton taxa began at 8 μg Zn/L. They calculated a lowest detectable effect concentration of 14 μg Zn/L for changes in group abundance and 21 μg Zn/L when chlorophyll a or zooplankton toxicity was the end point in mesocosm studies of zinc toxicity. Two additional mesocosm studies of zinc (or zinc and cadmium in combination) toxicity to the aquatic insect communities typical of Rocky Mountain streams provide another example of a benchmark

TABLE 9-6 Thresholds of Water-Column Zinc Toxicity by Different Approaches

relevant to CDA Lake. These studies evaluated dietary and dissolved metal bioavailability in concert with evaluation of effects on the community. Both were conducted under water quality conditions typical of streams in the CDA basin. Mebane et al. (2020) observed that mayfly populations (the most sensitive taxa) declined then disappeared from the community at zinc concentrations of about 10 μg/L. Schmidt et al. (2011) showed that the bioaccumulated zinc began to result in the disappearance of sensitive mayfly species at 5.4 μg/L equilibrated in the stream, and this was associated with adverse effects in the aquatic community as a whole.

Based upon what is known to date, a variety of choices for a zinc threshold or benchmark are possible for CDA Lake, depending upon ecological and ecosystem goals (Table 9-6). Greater protection from adverse effects of zinc toxicity would result from using targets built around mesocosm studies from the literature compared to compliance with existing regulations alone.

CONCLUSIONS AND RECOMMENDATIONS

The widespread distribution of mining wastes has affected human and ecological health in the CDA region. Although metal concentrations in the Lake waters and sediments are well described and some are monitored regularly, human exposures and ecological risks associated with the Lake itself have not been the subject of comprehensive study compared to the systematic ecological and mechanistic evaluation of risks in the basin upstream of the lake. Trends in the basin indicate that human exposures to hazardous materials like lead have declined and severely damaged ecosystems in the rivers and streams have improved. But there is limited information about exposure pathways that specifically apply to the Lake.

Ultimately, a body of knowledge developed across multiple levels of biological organization will be necessary to comprehensively understand ecological implications of metal contamination in CDA Lake, as well as potential benefits from remediation into the future. Identifying present-day ecological implications of the legacy of metal contamination in CDA Lake is a first step toward addressing the future viability of ecosystem services, such as biodiversity, ecological functions, fisheries, wildlife, and support for activities ranging from recreation to a subsistence life style. The following conclusions and recommendations build from studies of the basin, considering the data that exist for the Lake and drawing analogies from other environments.

1. Blood lead levels in children have declined in the Coeur d’Alene region but have stalled in recent years at a level about equal to the 2021 CDC level of concern (3.5 μg/dL). Remediation of lead-contaminated lands in OU-1 and OU-2 reduced mean blood lead levels in children (aged 0–9) in the basin to below the EPA level of concern of 10 μg/dL. Yet blood lead levels within the Superfund jurisdiction are about two to four times higher than in other western states for which data are available. These estimates suffer from questions about underrepresentation in economically disadvantaged and indigenous communities, low participation in the testing programs (especially in the lower basin), and age groups studied.

2. Assessments of human health risks specifically associated with CDA Lake would help address remaining potential pathways of lead and arsenic exposure. Occupations that expose people to Lake water or Lake sediments, recreational exposures to water or Lake sediments, and indirect exposure to metals in the Lake via groundwater used as drinking water have not been the focus of studies to date because the Lake is outside the Superfund jurisdiction.
3. A more comprehensive characterization of the sources of mercury in the food web of CDA Lake could benefit resource allocation decisions with regard to mercury. Few investigations have assessed mercury biogeochemical cycling in this system or addressed the source of MeHg (local vs. regional). By evaluating mercury dynamics in CDA Lake, environmental managers will be in a better position to consider options for reducing, over time, MeHg concentrations in high-trophic-level fish.
4. Expansion of existing monitoring to include a few sensitive nearshore environments could provide an early warning system for the onset of harmful algal blooms and expansion of nuisance-attached algae and of invasive plants. While on a lake-wide basis, the Lake remains oligotrophic with some mesotrophy in the south, experience elsewhere suggests the first signs of changes in trophic status can occur in nearshore, local waters in the form of blooms of attached algae. Expanded lakeshore monitoring could aid in detecting those changes before they become widespread.
5. Systematically developing a body of knowledge on how CDA Lake food webs are influenced by the legacy of mineral extraction will inform decisions about remediation and efforts to maximize ecosystem services. High priorities for better understanding ecological processes in the Lake include (1) expanded characterization of benthic and pelagic food webs; (2) evaluation of metal exposures in key components of the food web, and (3) experiments with benthic and water-column mesocosms to identify thresholds below which the Lake ecosystem will improve. CDA Lake may be the single greatest asset in the region.
6. Zinc concentrations in lake waters in many locations, as well as zinc and lead concentrations in many of the sediments of CDA Lake, exceed thresholds that suggest they could be toxic to some aquatic species. However, evidence of disturbance in phytoplankton and benthic communities in CDA Lake is ambiguous. There is no evidence among existing field data of phytoplankton that supports the concept that reducing zinc concentrations in Lake waters will increase the risk of eutrophication in the Lake. Existing data from monitoring program and limited ad hoc studies are insufficient to clarify the contradictions in the line of evidence above, and existing data do not allow an assessment of the influence of metal contamination on the CDA Lake ecosystem.
7. Multiple benchmarks could aid in characterizing remediation successes relevant to CDA Lake. These would include goals for the Lake ecosystem and ecosystem services and targets for metal concentrations below which such goals could be achieved. Different goals (e.g., compliance with state law vs. avoiding effects on water column communities) will require different benchmarks. The potential benchmark for zinc in the water column that would allow the Lake to return to pre-mining reference conditions and regain lost ecological functions and ecosystem services could be as low as 2 μg/L.

REFERENCES

This page intentionally left blank.