HEAVY METAL CONTAMINATION OF THE PILCOMAYO RIVER, BOLIVIA

Jerry R. Miller*, Department of Geosciences & Natural Resources Management, Western Carolina University, Cullowhee, NC 28723, USA (jmiller@wcu.edu); Karen A. Hudson-Edwards, Department of Earth Sciences, Birkbeck College, University of London, London WC1E 7HX, UK; Mark G. Macklin, Institute of Geography and Earth Sciences, University College of Wales, Aberystwyth, Ceredigion, SY23 3DB, UK; and Paul J. Lechler, Nevada Bureau of Mines and Geology, University of Nevada, Reno, Reno, NV 89557, USA.

ABSTRACT

The Rio Pilcomayo heads on the Cerro Rico de Potosí precious metal-polymetallic tin deposits of Bolivia. Mining of the Potosí deposits has continuously occurred since 1545, releasing large quantities of waste materials to headwater areas. In addition, a tailings dam breach at the Porco mine in 1996 released an estimated 235,000 m³ of tailings and fluid into the Rio Pilaya, the largest tributary to the Pilcomayo. Concentrations of Cu, Zn, As, Ag, Hg, Cd, Sb, Tl, and Pb in contemporary channel sediments upstream of the Pilaya confluence are significantly elevated above background values. Downstream of the Pilaya confluence, concentrations of Cu, Zn, and Pb are slight elevated, whereas Ag, Cd, Sb, and Tl cannot be distinguished from background levels suggesting that the long-term impacts of the Porco spill are limited. Significant downstream declines in concentration occur within 15 km, and again between 150-200 km, from the mines. The latter decrease is related to changes in sediment transport regime and the influx of “clean” sediment from several large tributaries. With the exception of Hg, concentrations within historic deposits are generally lower than observed within contemporary sediments. Moreover, metal concentrations within channel deposits generally increased between 1996 and 1998. These data suggest that while the erosion of historic deposits is an important source of metals to the river, the release of waste materials from modern mining operations is significant.

INTRODUCTION

The Rio Pilcomayo heads in the central Andes and drains the Cerro Rico de Potosí precious metal-polymetallic tin deposits. From its headwaters, the Pilcomayo flows in a southeasterly direction until reaching a low-relief, fan-shaped body of sediment referred to as the Chaco Plains near the town of Villa Montes (Fig. 1). Within Bolivia, the Pilcomayo basin encompasses an area of 98,000 km², and exhibits a low- and high-water discharge of 80 m³/s and 3,600 m³/s, respectively.

The mineral deposits within the large Cerro Rico de Postosí mining district consist of mineralized hydrothermal veins developed in Ordovician slate, dacitic tuff and tuff breccia, and other dacitic rocks (Cunningham et al., 1991). Mining of the deposits began in 1545 with the exploitation of the very rich (4.25-25.5%) Ag ores in the upper part of Cerro Rico. By 1573, the rich, upper Ag deposits were exhausted, but Ag mining continued using the mercury patio process. After 1852, most of the silver was extracted from low-grade ores by large mining companies, but low Ag prices resulted in a change from predominantly Ag to Sn production. In 1952, the national government privatized all major mines under the Corporaci\n Minera de Bolivia (COMIBOL). With the weakening of COMIBOL in 1985, small- and medium-sized mines have become the dominant producers and production has reverted to primarily Ag.

R. Poco

Poco

Figure 1. Location map showing sampling locations in 1998 and 1996.

The mining activities within the vicinity of Cerro Rico have resulted in the release of huge quantities of waste materials to the upper most reaches of the Rio Pilcomayo, including Hg enriched sediments during the region’s early mining history. In addition, approximately 235,000 m³ of waste materials were released to the Rio Pilaya, when a tailings dam was breached on the 19^th of August and the 1^st of September 1996. The spill resulted in massive fish kills hundreds of km downriver, and Macklin et al. (1996) suggested that a ‘spike’ of sediment from the Porco tailings dam reached as far as Villa Montes. Approximately 8 weeks after the spill, the mining company COMSUR initiated a cleanup operation, removing much of the tailings trapped along the river.

METHODS

Sediment samples were collected from the channel bed and historic floodplain and terrace deposits at 8 locations between Potosí and the Argentine border to (1) assess the spatial and temporal variations in metal concentrations along the Rio Pilcomayo, and (2) determine the impacts of the Porco spill on the system from a long-term (century scale) perspective. Samples were also obtained from pre-historic alluvial deposits in order to establish background concentrations. The samples were homogenized and subdivided for various analyses. One of the subsamples was air-dried, disaggregated and screened using a 2 mm sieve. It was then ground, digested in HNO₃ and analyzed by an ICP-MS at the Institute of Geography and Earth Sciences, University of Wales, Aberystwyth. A separate subsample was analyzed at the Nevada Bureau of Mines and Geology for Hg, also by ICP-MS methods. Both precision and accuracy were generally within 10 %. In addition, selected subsamples were analyzed for grain mineralogy by X-ray diffraction methods using CuKa₁ radiation from a primary beam and tube operating at conditions of 45 kV and 45 mA. Relative proportions of each mineral were calculated according to the method of Batchelder and Cressey (1998).

RESULTS AND DISCUSSION

The collected geochemical data show that upstream of the Pilaya confluence, concentrations of Ag, As, Cd, Cu, Hg, Pb, Sb, Tl, and Zn in contemporary channel bed sediments are significantly elevated above background values. Downstream of the Pilaya confluence, concentrations of Cu, Hg, Zn, and Pb are only slight elevated, and Ag, Cd, Sb, and Tl cannot be distinguished from background concentrations (see fig. 2 for representative examples). The data imply that while the Porco tailings spill may have had a significant short-term affect on water and sediment quality along downstream reaches of the Rio Pilcomayo, the long-term effects have been minimal. This is supported by a general increase in metal concentrations between 1996 and 1998 both up- and downstream of the Rio Pilaya confluence. Most of the sediment-borne heavy metals and arsenic appear to be derived from the upper most reaches of the Rio Pilcomayo. The highest concentrations of the metals were observed in sediments at R0, 1.2 km from Potosí where tailings effluent is currently entering the river. Moreover, concentrations decline downstream of the Potosí mining district (fig. 2). These downstream patterns in metal values correspond remarkably well with the amount of pyrite and other sulphide minerals within the samples (fig. 3). Given that the only major source of sulphides in the upper Pilcomayo basin is the Potosí mining district, the relations between sulphides and metal content suggest that mine and mill tailings are the major source of contaminants to the Rio Pilcomayo. Microprobe analyses demonstrate that sulphides are indeed major carriers of these elements.

The observed downstream declines in metal concentrations are not uniform, but (with the exception of Zn) are concentrated between sampling points R0 and R1 (~1.2-10 km downstream), and again between R3 and R4 (~150-200 km downstream). The former drop is undoubtedly the result of the dilution of nearly pure tailings effluent released to the river from processing mills near Potosí by “clean” sediment. These unpolluted sediments are derived from tributaries and the erosion of pre-mining bank deposits. The latter decline appears to be related to a combination of factors, including dilution associated with the influx of uncontaminated debris from two relatively large tributaries (the Rio Mataca and the Rio Poco Poco). Perhaps of more importance, however, is a change in the local sediment transport regime. In particular, the influx of sediment from tributary rivers between

Figure 3. Downstream variations in valley width.

reaches R3 and R4 has led to a significant episode of channel aggradation as indicated by (1) the deposition of sediment that extends from one side of the valley floor to the other, and (2) a paucity of Holocene terraces present farther up- and downvalley (and that are apparently buried along the aggraded reach). Presumably, the downvalley movement of metals from Potosí is inhibited at 150-200 km from the mines as sediment-borne metals are deposited and stored in the aggrading reach. Data from other studies where similar geomorphic processes have occurred (e.g., Gilbert 1917; Knighton, 1991) suggest that the metals may eventually be moved downvalley as the aggradational processes are reversed and channel stabilization occurs. Thus, the most significant impacts of metal contamination may not be realized in downstream areas of decades, centuries, or even millenia.

Figure 3. Percent pyrite in samples along the Rio Pilcomayo.

REFERENCES

Batchelder M and Cressey G (1998) Clays and Clay Minerals, 46: 183-194.

Cunningham CC, McNamee J, V<squez JP, and Ericksen GE (1991) Econ. Geol. 86: 415-421.

Gilbert, G K, 1917. U.S. Geological Survey Professional Paper 105.

Macklin MG, Payne I, Preston D, and Sedgwick C (1996) Rept. to Overseas Development Agency, UK, 33p.

Knighton, A D (1991) Geomorphology, 4: 205-219.