Bioremediation by Sulfate Reducing Bacteria of Acid Mine Drainage

Mining activity produces metal sulfide wastes particularly pyrite which remain in the mine long after operations cease. When water percolates through a mine where oxygen is present, a series of chemical reactions occur which produce extremely low pH and high concentrations of toxic sulfate and metal ions. This toxic leachate can cause severe aquatic habitat degradation downstream of the mine. This study addresses this environmental problem by harnessing the metabolism of sulfate reducing bacteria, whose ability to reduce sulfate produces carbonate which neutralizes acids and sulfide, which chemically stabilizes toxic metal ions as solid metal sulfides. Batch reactors were set up with mine leachate, bacterial culture, a growth medium, and various sources of organic carbon. Results have shown that bacterial reactions caused copper and zinc reductions of 100%, pH increases of up to 2, and decreases in toxicity of 100%. Introduction Acid mine drainage (AMD), a major environmental hazard that affects aquatic ecosystems around the world, results from the oxidation of metal sulfides, particularly pyrite (FeS2). Discharges from abandoned sulfide ore mines (the most commonly mined sulfides being sulfur, copper, zinc, lead, gold, silver, and uranium) often contain high concentrations of metal sulfides. In lotic environments such as streams, these metal sulfides become rapidly oxidized though chemical and microbial processes (Gray 1998, Bonnissel-Gissinger et al. 1998, Evangelou and Zhang 1995). The resulting sulfate and hydrogen ions lower pH significantly as well as add toxic metal ions to the environment (Gray 1998). Acid mine drainage causes the degradation of aquatic systems through acidification, high concentrations of iron and sulfate, and elevated levels of soluble toxic metals. Under the acidic conditions resulting from AMD, the oxidation of pyrite proceeds by the following reaction (BonnisselGissinger et al. 1998): FeS2 + 14Fe + 8H2O → 15Fe + 2SO4 + 16H This reaction demonstrates the polluting capability of the oxidation of pyrite – every mole of pyrite becomes converted to 16 moles of hydrogen and 2 moles of sulfate. This reaction serves as a template for the similar oxidation reactions of most metal sulfides, which also contribute acidity, sulfate, and toxic metal ions to the aquatic environment. Once oxidation has occurred, its products damage the ecosystem in a number of ways. First, the ferric precipitate common to AMD destroys vegetation by blanketing the soil layer and clogging the substrate interstices (Gray 1998). Evidence has also shown that AMD has been responsible for marked declines in aquatic species and ecosystem diversity, as well as productivity. And the toxicity of extreme concentrations of heavy metals and acidity has been linked to total elimination of certain fish species in some aquatic ecosystems (Gray 1998). This study investigated bioremediation as a solution to AMD. In past experimentation, the most common method of combating AMD has been the construction of a treatment wetland downstream from the mine (Machemer et al. 1993; Mitsch and Wise 1997). Such wetlands take up the polluted water, and through microbial processes similar to those used in this experiment purify the water. The wetlands further remove metals through uptake into plant tissues. The drawback of this method is that the problem is not dealt with at the source, but rather after the polluted water has already had an environmental effect. The goal of this project was to determine if sulfate reducing bacteria can be used inside the mine itself such that the leachate is not toxic. Prior research has shown that bacterial metabolism can be of significant use in the removal of metals from wastewater (Dvorak et al. 1992). Sulfate reducing bacteria oxidize simple organic molecules using the sulfate ion as an electron acceptor. This process produces hydrogen sulfide(H2S) and the bicarbonate ion(HCO3). Hydrogen sulfide readily reacts with heavy metal ions to immobilize the metals as insoluble metal sulfides, while the bicarbonate ions buffer the pH to significantly higher levels (Dvorak et al. 1992). Thus, sulfate is removed as hydrogen sulfide gas and immobile metal sulfides, metals are removed as metal sulfides, and pH is raised, improving water quality. In order to maintain bacterial metabolism, the bacteria must be given both an organic carbon source (as food) and some growth substrate for attachment (the bacteria cannot survive in open water). AMD from an abandoned mine site was treated in batch reactors with an organic carbon source, a growth substrate, and a culture of sulfate reducing bacteria. The mine site lies in the hills east of Oakland, California. Streams originating in the watershed currently show strongly visible signs of habitat degradation such as vegetation loss, low pH, and coating of riparian substrate with oxidized ferric precipitate. It is hypothesized that the bacterial reactions described above will improve this severely toxic water to quality levels similar to unpolluted streams. Methods Study Site All AMD water samples were taken from the Leona Heights Mine, an abandoned sulfur mine in the hills east of Oakland, California. The small creek running out of the mine showed strong visible signs of metal pollution including the characteristic brownorange coloration of iron precipitate. In addition, toxic levels of several metals and sulfate were found in the creek. Both this creek and the terminal conduit of the watershed, located at the inlet to Lake Aliso several miles downstream in Oakland have been found to be toxic to Ceriodaphnia dubia, a zooplankton species commonly used in aquatic metal toxicity testing. Bacterial Remediation Experiment This study cultured sulfate reducing bacteria in the presence of AMD, an organic carbon source, a substrate on which to grow, and conditions otherwise similar to those inside a mine shaft. The goal was a determination of whether SRBs can have a significant effect of reducing concentrations and toxicity of metal and sulfate ions while raising pH in water sampled from abandoned mine sites. The bacterial remediation experiment consisted of filling three one-liter sample bottles for each treatment with AMD water from a polluted mine site, a carbon source, a bacterial culture, and dried cattails as a growth substrate. The three experimental treatments contained varied organic carbon sources for the bacterial reaction as listed in Table 1. The control reactors contained only AMD water and cattails. Name of Treatment Bacteria Carbon Source Growth Medium Control NO NONE Dead Cattail PCI YES Pine Needles Dead Cattail OCI YES Oak Leaves Dead Cattail ECI YES Eucalyptus Leaves Dead Cattail Table 1. Bacterial Remediation Experiment, Site 1, Reactor Treatments. Only the three treatments PCI, MCI, and SCI include all ingredients necessary for bacterial metabolism – growth substrate and carbon source. The dead cattail clippings were added as an intended substrate for the bacteria to grow on, while the pine needles, oak leaves, and eucalyptus leaves were intended as organic carbon sources for the bacteria. Before use in the reactors, the dead cattails were soaked in water for two weeks in order to leech out as much carbon as possible, thereby making their contribution as a carbon source negligible. The reactors were stored in covered, insulated boxes to keep temperature fluctuation to a minimum and to keep out light. This helped to approximate actual mine shaft conditions. Every week, 5mLs were removed from the sample bottles for analysis. The sample bottles were analyzed for sulfate, iron, zinc, and copper concentrations on a weekly basis. In addition, pH was measured on a biweekly basis. Sulfate measurements were taken with an Ion Chromatograph (Dionex DX 100) with a pulse electrochemical detector using a Dionex Ionpac AS4A-SC column (detection limit~0.1ppm). Metal analysis was performed by Flame Atomic Absorption Spectroscopy (detection limit~1ppm), and pH measurements were taken with a digital pH meter. After every sampling, reactors were purged of any oxygen with nitrogen gas in order that anoxic conditions typical of an actual mine shaft be maintained. Two months into the experiment, approximately four grams of carbon source were added to two of the three batch reactor replicates for each treatment. Toxicity Testing Toxicity tests were performed on Ceriodaphnia dubia, a species of zooplankton classically used for its intolerance to many aqueous heavy metals (Nimmo et al. 1990, Ribeiro, et al. 2000). Tests were performed according to standard EPA protocol (USEPA 1993). Ten replicates were used for each water sample. Table 2 summarizes each treatment. Sample 1 C. dubia culture water Sample 2 Control Batch Reactor water (diluted 1:10 w/Horseshoe Creek) Sample 3 Eucalyptus-remediated water (diluted 1:10 w/Horseshoe Creek) Sample 4 Pine-remediated water (diluted 1:10 w/Horseshoe Creek) Sample 5 Oak-remediated water (diluted 1:10 w/Horseshoe Creek) Sample 6 Horseshoe Creek water Table 2. Toxicity Testing Experiment Water Samples Each sample replicate contained one individual C. dubia and was investigated after fortyeight hours for mortality and after every 24 hours thereafter for five days for number of new live births of C. dubia young. In treatments including water from the batch reactors, a 1:10 dilution of sample water to Horseshoe Creek water was used for the toxicity test. Horseshoe Creek is an unpolluted stream in the same watershed as the Leona Heights mine. From flowrate measurements throughout the watershed, it was determined that the impacted creek directly below the mine is diluted 1:10 when it reaches the watershed terminus at Lake Aliso. This realistic dilution level was used for the toxicity tests. Acute toxicity was determined by mortality rates of C. dubia after the first 48 hours of incubation. Chronic toxicity was determined through the number of births of C. dubia juveniles per day over the five day period following the acute test. Results Bacterial Remediation Experiment The data from the Oakland mine site shows a general downward trend in copper concentrations for averages over the three replicates for the three experimental reactors (pine needles, eucalyptus, and oak). In all three cases, copper concentrations were brought to below detectable limits within a four week period (Figure 1). The control reactors showed a slight rise over the course of the experiment. Figure 1. Copper concentration with time in batch reactors. Detection limit for analysis is ~0.1ppm. Values below equipment detection limit are plotted as zero. Zinc measurements over the course of the experiment did not fall in all cases as occurred with copper concentrations. The experimental reactors containing oak leaves as the carbon source show a sharp decline in zinc concentration in the fourth week after the start of the experiment (Figure 2). From that point on, zinc levels were below detection limits. The reactors with pine needles as a carbon source showed no sign of zinc sulfide precipitation until the addition of carbon to the reactors 73 days into the experiment. After that carbon addition, the pine needle reactors showed a decrease in zinc concentration all the way to below detectable limits. The eucalyptus reactors showed no decline in zinc concentration throughout the experiment, even after the second carbon addition. 0 0.5 1 1.5 2 2.5 3 3.5 4