Lecture No. 14. Mining-Derived Metal Contamination (Continued)

Nimick and Moore proposed an index based on the known effects of acids and metals on aquatic life. Others have proposed similar "contaminant indices."

The hazard level for a given metal is the highest tolerated concentration or the lowest toxic concentration for a metal, i.e., the point at which the metal becomes a hazard. For iron the authors used a chronic hazard level, because there was no known acute hazard level.

See the Nimick and Moore paper, page 641. The PI correlates well with pollution effects in the upper Clark Fork basin.

The correlation with pH is strictly a local relationship in the Clark Fork. In another basin the metal solubilities and metal availabilities could be different, and the pollution index might have a very different relationship to pH. While the acidity itself is toxic, the specific dissolved metals also are toxic. Consider, for example, a drainage basin with very low available copper and low availability of other transition metals. Here the source of acidity might be only weakly coupled to metal concentrations, if the acidity came from something like hot springs or geysers. Lowering the pH would not put much metal into solution until it got low enough to dissolve aluminum out of the clay minerals. On the other hand, a basin with low copper but with high lead and mercury might exhibit a much faster increase of PI with decreasing pH.

Nevertheless, pH makes a very handy surrogate for pollution index in the upper Clark Fork.

Harold Bergman (University of Wyoming) did a set of experiments with a mixture of metals to mimic a real-world pulse event such as the ones that caused the massive fish kills on the Clark Fork. His starting mixture (called 1P) was used at various dilutions. Multiples of 1P were also used.

Metal	1P (ppb)	Red Water (ppb)
Zn	230	14,000
Cu	20	13,000
Pb	3.2	30
Cd	2.0	85

Experiments were done in a divided aquarium with the sections not well mixed. Fish did in fact avoid higher metal concentrations, beginning at extremely low levels. At higher levels, sensory overload set in.

Fish are exposed to contaminant metals in ways other than by direct immersion in highly contaminated water. The following data from Moore and Luoma show that aquatic insects (a major part of the trout diet) contain elevated levels of copper and cadmium in the upper Clark Fork and even as far downriver as Missoula. These data are for caddis fly larvae. The reference tributary is an uncontaminated stream.

The fish are stressed for a third reason. The contaminated stretch of river has a much smaller than normal number of insect taxa; only the metal-resistant ones live there. These insects do not make up a balanced diet for trout, and this probably explains the fact that the trout’s growth seems to be stunted from the second year of life on.

Mortality from serious disease has been uncommonly high in the upper Clark Fork Basin. Silver Bow County has extremely high death rates from several diseases and conditions, including

There is always the problem of separating environmental causes of disease from causes related to occupational hazards, lifestyle (e.g., smoking, diet, etc.), and heredity. However, it does appear that a large portion of the health problems in Butte comes from the environment. Several elements known or suspected to impair human health are present in the soil and elsewhere as a result of mining activity. These elements include mercury, lead, cadmium, and arsenic, among others. We have previously discussed the acute effects of cadmium (itai-itai disease). Chronic low-level cadmium poisoning is thought to cause hypertension. Chronic mercury poisoning causes central nervous system, bone, teeth, and kidney damage. Chronic arsenic exposure can lead to cancer. Chronic lead poisoning effects include bone and nerve damage and infertility in adults and more severe effects in children.

The problem is probably worse in Deer Lodge County, especially the area around Anaconda. The reason is bioavailability. Metals around mining areas tend to be tied up in sulfides, which weather and release the metals relatively slowly. In the vicinity of a smelter, the metals often occur in oxides, which release the metals in soluble form much more quickly. A person breathing in or ingesting sulfide dust will not absorb much of the metal, unless the dust is very extensively altered; someone breathing or ingesting flue dust from a smelter will absorb a much larger percentage of the metal.

There was concern over the high lead levels in soil and house dust in Butte and Walkerville. Young children have a habit of putting their hands in their mouths, and this causes them to swallow whatever has been on their hands, including dirt. It was therefore believed that local children had elevated blood lead levels. The Butte Silver Bow Health Department did a study in conjunction with the University of Cincinnati. Local children’s blood lead levels were, surprisingly, well below the natural average. Why? One possible cause might be the very high zinc levels in the local soil. Zinc may suppress lead uptake. Another contributing factor is probably the fact that most local lead is in the form of sulfide minerals. Each particle will have a weathering rind, normally anglesite (PbSO₄), but most of the lead is in the form of very insoluble sulfides. Even anglesite is not terribly soluble.

Another study was done by Lewis and Clark County, the State of Montana, and the Federal Government on children living in East Helena near the lead smelter. Since a number of cattle and horses grazing downwind of the smelter had died of lead poisoning, there was reason to suspect the children might be at risk. The East Helena children had blood lead contents that ranged from somewhat high to extremely high. One child had to go on chelating treatments immediately. (He had a habit of eating snow.) The source of their blood lead was flue dust containing litharge (PbO) and other readily soluble lead compounds.

Sometimes the contaminants from mining enter the food chain in odd ways. American Antimony has a plant on Prospect Creek west of Thompson Falls. Their tailings disposal pond recharges the local aquifer, and the groundwater is contaminated with antimony for a few hundred meters down the valley. However, that’s not the problem. Deer were using the sediments in the pond as a salt lick! They were attracted by the high sodium levels, but they were taking up significant arsenic and antimony. A fence had to be built around the pond.

We will get back to this subject later. Contaminants get distributed in all sorts of ways.

So you have a river contaminated with metal-rich tailings and a continuing supply of the tailings coming downstream. Moreover, soil in areas near the mines and smelters is contaminated. And a reservoir almost 200 km downstream is filling up with contaminated sediments. And the ground water near that reservoir is contaminated with arsenic from the tailings. This is a brief description of the country’s largest Superfund site, the Clark Fork drainage from Butte to Milltown.

Moore and Luoma lay out a strategy for designing a remediation project. They point out that we just plain don’t know (from practical experience) how to deal with some of the problems. There are a number of important considerations, such as

After taking a good look at a site to identify and quantitate the problems, one may find there are some things that just have to be taken care of immediately, on an emergency basis. The authors recommend using these emergency actions as learning experiences. Every step should be documented, and every conceivably useful datum should be measured.

Another point they make is that the contamination should be removed, not just fenced off or isolated.

They also point out that a site like the upper Clark Fork can never be put back the way it was. Silver Bow Creek was once a stream that an early explorer compared to Diana’s silver bow. Now the creek is completely gone where the Berkeley Pit was excavated, and the rest of its upper reach is called the Metro Storm Drain. It then passes through a "black canyon" made of 19^th-Century slag heaps. The original landscape around Butte has been replaced by bare rock cliffs, several big holes in the ground, and a number of waste rock dumps. If the health hazards and the fish and wildlife hazards are corrected, we will have done all we can.

One other point about the cleanup: the EPA and the State are going about it one step at a time, starting upstream. This is a wise way to go about it. Some critics want the river from Warm Springs to Galen cleaned up NOW. The problem is that the pollution will keep coming from upstream until it is cleaned up. If they clean up the downstream reaches now, they will have to do it all over again in a few years.

Moore and Luoma talk about making the area a "National Environmental Disaster Monument." That will never happen, of course; Travel Montana and every Chamber of Commerce in the State would scream bloody murder. But it’s an interesting idea. People should be aware of the kind of damage that 130 years of bad industrial practice has caused.

This is a river that is only slightly to moderately impacted by mining. The mining was on a much smaller scale that that around Butte, and there are some accidents of geography and geology that limited the damage done by mine wastes. This case study is detailed in a paper by Moore, Luoma, and Peters.

You are all familiar with the recent controversy over the Macdonald gold project near Lincoln. It is not as widely known that there has been mining in the Blackfoot drainage since the mid-1860’s. Most of the mines closed down before 1960. There was not a large amount of milling done, but there was some, and tailings were dumped here and there. The metal deposits were mostly hosted by the Middle Belt Carbonate group of low-grade metamorphic rocks. There were a lot of small gold mines near Rogers Pass and some placer gold operations. The tunnels of some of the underground gold mines drained through adits. There is plentiful groundwater in the mining districts, and in many places the rock is pervasively fractured. The tunnels generally intercepted multiple fractures and gave the groundwater a path to drain toward the adits. At the same time, the tunnels introduced air into the anoxic rock, and the usual pyrite oxidation happened, with accompanying acid production. Thus the water draining out of the gold workings was normally acidic and high in iron. The gold mining operations also left waste rock dumps here and there, which provided another source of acid.

There were also mines where sulfide ore deposits were worked for silver and for copper, zinc, and other base metals. Two of these were the Mike Horse and Carbonate mines. The water draining from these mines carried dissolved contaminant metals, especially copper, zinc, cadmium, and lead, with lesser amounts of arsenic and nickel. Iron, manganese, and aluminum were also in solution as a result of the low pH.

The mainstem of the Blackfoot River officially begins at the confluence of Bear Trap Creek and Anaconda Creek, just a few miles below Rogers Pass and about a mile from the stream of acid water that used to flow out of the Mike Horse Mine. Immediately below this confluence the river flows into the first of three wetlands. There are a total of three swamps and marshes in the first 20 km of the river. Below these wetlands, the river is generally recognized as clean, and trout fishing is practiced all along the river by locals and tourists. The Blackfoot is not as superb a stream as Rock Creek is supposed to be, but it is no slouch.

It has generally been believed that a wetland acts to prevent downstream contamination. It does this two ways. First, it is a sediment trap just as a lake is (although perhaps not as efficient as a lake). Second, the conditions in its sediments are anoxic, so that sulfate tends to be reduced to sulfide, and heavy metals precipitate as sulfides. Authigenic pyrite forms in the sediments, consuming acid. Moore and co-workers set about investigating this hypothesis and measuring the levels of metal contamination in various species at various parts of the river.

A number of different insect taxa were collected at sampling stations along the full length of the river in 1988 and 1989. Whole insects and fish livers were digested and analyzed for the contaminant metals and for iron, manganese, and aluminum. Sediment samples were collected, totally digested by a method similar to the USGS total-in-sediment method, and the digests were analyzed. Dissolved metals were analyzed in river water collected at the various stations. Samples were also collected on contaminated tributaries, uncontaminated tribs (such as the Clearwater River) and in an uncontaminated headwater stream immediately adjacent to the contaminated tributaries but outside the Blackfoot drainage. This last sampling location afforded the authors a direct comparison between clean and dirty streams draining the same Belt rocks. The map in the paper (Figure 1) shows the sampling locations.

The first overhead shows some water chemistry along the mainstem and along contaminated and uncontaminated tributaries. Notice that pH is as low as 3 in the worst cases upstream, but it settles in between 7 and 8 by about 200 km above Milltown. Sulfate is initially very high and then drops, but it doesn’t drop as fast as pH. It is still somewhat elevated farther downstream; on the other hand, sulfate is not toxic. Alkalinity is low in the acid-contaminated area, then it rises to about 200 mg/L. Finally it drops to 150 mg/L or so. The drop in downstream regions is probably the result of water draining limestone and dolomite being diluted by water draining off the low-alkalinity rocks farther west, such as the Bonner and McNamara Formations.

The next overhead shows dissolved metal concentrations. Iron, manganese, copper, cadmium, and lead all drop below the limits of detection by about River Kilometer 190. Zinc takes longer to drop, and it still is present at elevated concentrations until Kilometer 100. The next overhead shows concentrations in sediments. All metal concentrations settle in to background levels (as indicated by samples from clean tributaries) by about Kilometer 175. This is more than 10 km below the last marsh, so contaminated sediments are getting through the marshes.

The contaminated reaches of the mainstem and tributaries were easy to spot because of the orange oxyhydroxide coatings on rocks. As the acidic, metal-laden water from the mines is aerated, Fe²⁺ is oxidized to the +3 state and precipitates as ferrihydrite, goethite, etc. The manganese oxyhydroxides such as MnOOH coprecipitate with them. Another clue to contamination was the absence of mayflies, which are metal-sensitive.

Bioavailability varied from metal to metal. Cadmium accumulated most consistently. Everything that lived in the upper reach contained elevated cadmium. In the middle reach, only the long-lived predators such as brown trout and stoneflies had elevated Cd. Also in the middle reach, zinc was elevated in some insects but not in fish. Upstream of the marshes, brook trout were the only targeted taxon found alive. In that area they contained high Cd, Zn, and Cu.

Arsenic and nickel showed no trends in concentration and were not present at elevated levels in any animals.

Surveys of fish populations had been made in 1973 and 1975 (before and after an abandoned tailings dam collapsed). Another survey was made in 1988 after there were widespread complaints of dropping fish populations. Trout populations did in fact drop severely between 1973 and 1975, and between 1975 and 1988. The 1973-75 drop was obviously due to the tailings dam failure and the metal contamination it added to the river. In the upstream area, the "young of the year" brook trout were present in 1973 and completely absent in 1975; in other words, one whole spawning season was lost. In 1988 the age 0 brook trout had returned, but the populations were only 33% as large as in 1973. At 200 km (between the 2^nd and 3^rd wetlands), numbers of cutthroat trout aged 1 and older fell between 1973 and 1975 and were even lower in 1988. Brown trout populations in downstream areas were lower in 1988 than in the 1970’s, but the young-of-year trout were present. It appears that in the middle and lower portions of the river, something other than metal contamination must be blamed for the loss of trout. Overfishing, invasion by northern pike, the drought conditions in the late 1980’s, and disease are all possible causes.

The data suggest that marshes and swamps may slow the movement of metal contaminants but may not stop their transport altogether. One possible reason for this is that rivers tend to cut channels through marshes in flood years.

We are revisiting this subject yet again, this time from the standpoint of mechanisms.

This is a very complicated set of processes; some people build careers on studying metals with just one substrate. However, there are some simple generalities that can be useful.

Recall the zero point of charge of a substrate — that pH where the surface charge is zero.

See the figures from Elder’s paper. Figure 8 shows surface charge as a function of pH. The more basic materials stay positive at a higher pH than do the more acidic ones. Figure 5 shows sorption of metals to humic acids in the range pH 1 to pH 7. From this figure it is obvious that different metals can have widely varying affinities for humic acid, but that any metal tends to be sorbed more strongly as pH increases. Also note in Figure 4 that cadmium sorption to ferric hydroxide in the range pH 5 to pH 9 is sensitive to dissolved iron concentrations. If the solution concentration is higher, cadmium sorbs at a lower pH. Iron is apparently not acting as a competing ligand here. It is not significantly affecting ionic strength either, since the sodium nitrate concentration (0.1 M) swamps any changes in other metals.

Possibly the dissolved iron is affecting the properties of the Fe(OH)₃ surface.

The next overhead is a figure from a paper by Ficklin and others on mine drainages and natural drainages in mineralized areas in Colorado. The authors did some modeling of sediment vs. solution distributions of three metals (Cu, Zn, and Cd) in sediments collected from Saint Kevin’s Gulch, a heavily contaminated stream near Leadville, CO. The sediments were exposed to various pH conditions, and the percentage of each metal sorbed to the sediments was calculated and measured. They then compared the model predictions with actual data at three sites and at two times (summer and fall of 1989). Not surprisingly, the metals all exhibited pH-dependent sorption curves.

River Kilometer	Copper (m g/g dry weight)	Cadmium (m g/g dry weight)
3–59	173 ± 5	2.8 ± 0.7
61–106	66 ± 19	4.5 ± 2.7
107–163	59 ± 17	1.6 ± 1.0
163 – 200	29 ± 8	0.7 ± 0.2
Reference Tributary	15	0.2