Environmental assessment of mercury contamination from the Rwamagasa artisanal gold mining centre, Geita District, Tanzania

https://doi.org/10.1016/j.scitotenv.2004.09.042Get rights and content

Abstract

This study presents the results of an environmental assessment of mercury (Hg) contamination in the Rwamagasa artisanal gold mining area, northwest Tanzania, and the potential downstream dispersion along the River Malagarasi to Lake Tanganyika. At the time of sampling, generally low concentrations of Hg (<0.05 mg/kg) occurred in most cultivated soils although higher Hg (0.05–9.2 mg/kg) was recorded in urban soils and vegetable plot soils where these are impacted by Hg-contaminated water and sediment derived from mineral processing activities. Hg in vegetable and grain samples is mostly below the detection limit of 0.004 mg/kg Hg, apart from 0.007 and 0.092 mg/kg Hg in two yam samples and 0.011 to 0.013 mg/kg Hg in three rice samples. The standardized (i.e., standardized to 10 cm length) Hg concentrations in Clarias spp. increase from about 0.01 mg Hg/kg for the River Malagarasi delta to 0.07, 0.2, and 1.6 mg/kg, respectively, for the Rwamagasa ‘background’, moderately and most contaminated sites. For piscivorous (Lates, Brycinus, and Hydrocynus spp.), insectivorous (Barbus spp.), and planktivorous (Haplochromis spp.) fish species, the 10-cm standardized Hg concentrations increase from about 0.006 mg/kg for the River Malagarasi-Lake Tanganyika area to 0.5 and 3.5 mg/kg, respectively, for the Rwamagasa moderately and most contaminated sites. The low concentrations of Hg in fish from the Malagarasi River delta and Lake Tanganyika indicate that Hg contamination from the Rwamagasa area does not have a readily discernible impact on the biota of Lake Tanganyika. Many of the fish samples from Rwamagasa exceed guidelines for human consumption (0.5 mg/kg) as well as the WHO recommended limit for vulnerable groups (0.2 mg/kg). Tissue total Hg (THg) of all fish collected from the River Malagarasi-Lake Tanganyika subarea is well below these guidelines. Potential human exposure through consumption of 300 g/day of rice grown on Hg-contaminated soils is 5.5 μg/week. Consumption of 250 g Nile perch (Lates spp.), 500 g tilapia (Oreochromis spp.), and 250 g of catfish (Clarias spp.) each week would result in an intake of 65 μg Hg/week for people consuming only fish from the Mara and Mwanza regions of Lake Victoria and 116 μg Hg/week for people in the Rwamagasa area consuming tilapia and Nile perch from Lake Victoria and catfish from mining-impacted streams. This is lower than the Provisional Tolerable Weekly Intake (PTWI) of 300 μg for Hg in the diet set by the WHO and the FAO. Inadvertent ingestion of soil containing 9 mg Hg/kg at a rate of 80 mg/day would give an additional estimated weekly intake of 5 μg THg, whereas the persistent and purposeful consumption of soil (geophagia) at a rate of 26 g soil/day would produce an additional chemical exposure of 230 μg Hg/day.

Introduction

Artisanal gold mining provides income to some of the world's poorest people, many of whom are women and children, but it is also one of the major sources of mercury (Hg) contamination, especially in developing countries. Whereas the gold extraction process (known as amalgamation) is a simple technology, it is potentially very dangerous and contaminates the air, soil, rivers, and lakes with Hg. The health of the miners and other people living within the area affected by Hg contamination may be negatively affected through inhalation of Hg vapour, direct contact with Hg and through eating fish and other food affected by Hg contamination. The Global Environmental Facility (GEF) of the UN funds a project Removal of Barriers to the Introduction of Cleaner Artisanal Gold Mining and Extraction Technologies (also referred to as the Global Mercury Project (GMP; http://www.unites.uqam.ca/gmf/intranet/gmp/front_page.htm) that is currently being executed by UNIDO in six developing countries (Brazil, Indonesia, Laos, Sudan, Tanzania, and Zimbabwe). This paper presents the results of an environmental assessment in the Rwamagasa (or Rwamagaza) artisanal gold mining area, which was selected by UNIDO as the GMP demonstration site for Tanzania. Primary artisanal workings in the Rwamagasa area are centred on quartz veins in sheared, ferruginous, chlorite mica schists. Grab samples of vein and wall rock grade 6–62 g/t Au (Spinifex 2002 Annual Report). One of the objectives of the current study was to assess the impact of Hg contamination on international waters as well as in the immediate vicinity of the Rwamagasa ‘mining hotspot’. Consequently, the field programme was carried out in two subareas: (a) the Rwamagasa ‘mining hotspot’ subarea and (b) the River Malagarasi–Lake Tanganyika subarea (Fig. 1). Dispersion of Hg from Rwamagasa to Lake Tanganyika is probably relatively unlikely because contaminant Hg will be adsorbed by organic material in the extensive Moyowozi and Njingwe Swamps and flooded grassland area, located from 120 to 350 km downstream of Rwamagasa (Fig. 1). Whereas the swamps will act as a potential biomethylation zone, they will also act as an environmental sink for Hg contamination which is likely to inhibit migration of Hg into the lower reaches of the Malagarasi River and Lake Tanganyika, some 430 km downstream from Rwamagasa. The swamp area was inaccessible within the logistical and budgetary constraints of the current project.

The only legal mining in the Rwamagasa area is carried out within the boundaries of the Primary Mining Licence held by Blue Reef Mines (Fig. 2) where approximately 150 people are involved in mining and mineral processing activities. This is the only site in the Rwamagasa area where primary ore is being mined underground. All other mineral processing activity of any significance is concentrated at the northern margin of Rwamagasa, especially on the land sloping down to the Isingile River (Fig. 2, Fig. 3). In this area, there are about 30 groups of historic and active tailings dumps and about 10 localities where small (200 l) ball mills are operating. The number of people actively involved, at one particular time, in ball milling, sluicing, and amalgamation is probably no more than 300.

Amalgam is burned in a small charcoal fire, which releases Hg to the atmosphere. Retorts are not used. Amalgamation mainly takes place adjacent to amalgamation ponds, which are usually formed of concrete, but sometimes have only wood walls even though environmental legislation dictates that the Hg-contaminated mineral concentrates and tailings should be stored in concrete lined structures. The field programme was carried out during the dry season at which time there was little evidence that large quantities of contaminated tailings were being washed into the Isingile River. However, waste water and tailings from amalgamation ‘ponds’ were observed at one site to be overflowing onto an area where vegetables were being grown. If large quantities of Hg-contaminated tailings are dispersed onto the seasonal swamp (mbuga) area adjacent to the Isingile River during the wet season, then this may lead to the significant dispersion of Hg both into the aquatic system and onto agricultural sites being used for rice, maize, and vegetable cultivation (Fig. 2, Fig. 3).

Previous studies in the Lake Victoria Goldfields area indicate that dispersion of Hg from tailings is relatively restricted, not least because Fe-rich laterites and seasonal swamps (mbugas) act as natural barriers or sinks attenuating the widespread dispersion of Hg in sediments and soils (van Straaten, 2000b). The Blue Reef Mine is reported to produce about 1 kg Au per month whereas artisanal miners re-working tailings produce about 0.5 kg per month (personal communication, John Nayopa, October 2003). On this basis, approximately 27 kg of Hg will be released to the environment from the Rwamagasa area each year. Of this, atmospheric emissions from amalgam burning will be about 14 kg from the Blue Reef mine site and 7 kg from the other amalgamation sites. About 2 to 3 kg Hg will remain in heavy mineral tailings in the amalgamation ponds, which are frequently reprocessed.

The geology and gold mineralization of the region are described by van Straaten (1983), Barth (1990), and Borg (1994). Background information on Hg contamination associated with artisanal gold mining in Tanzania is available in a number of published reports and scientific papers (Gladher et al., 2000, Asano et al., 2000, Campbell et al., 2003a, Campbell et al., 2003b, Harada et al., 1999, Ikingura and Akagi, 1996, Ikingura and Akagi, 2002, Kinabo, 1996, Kinabo, 2002a, Kinabo, 2002b, Kinabo and Lyimo, 2002, Kishe and Machiwa, 2003, Machiwa et al., 2003, Mutakyahwa, 2002, Semu et al., 1989, Sindayigaya et al., 1994, University of Dar es Salaam, 1994, van Straaten, 2000a, van Straaten, 2000b).

Section snippets

Drainage sediment and water, tailings and soil

Drainage sediment and water were collected from the lower reaches of the Malagarasi River near where it enters Lake Tanganyika (Ilagala) and also from near Uvinza, approximately 60 km upstream (Fig. 1). At Ilagala, sediment samples were collected using a Van Veen grab sampler whilst at Uvinza sediment was obtained from appropriate sites along the bank of the river. Streams in the Rwamagasa area were either completely dry or ponded. Bottom sediment samples, each of 150–200 grams, were collected

Water, sediment, tailings, and soil

Drainage sediment, soil and tailings were prepared by BGS and analysed by ACME Analytical Laboratories, Vancouver, Canada. Waters were analysed by BGS.

Hg in filtered waters was determined by cold vapour atomic fluorescence spectroscopy (CV-AFS) to a practical detection limit of 20–30 ng/L. Water samples were subjected to a bromination stage, prior to analysis, to break down any organo-Hg compounds. Arsenic was determined by hydride generation AFS to a practical detection limit of 0.25 μg/L.

Water

In the Rwamagasa area, Hg in filtered water samples from the immediate vicinity of Rwamagasa do not exceed concentrations detected at the two background sites (0.04–0.05 μg/L) apart from one sample with 0.07 μg/L, located in the Isingile River about 1 km downstream of the nearest Hg amalgamation ponds (Table 1). Filtered water from a well located within Rwamagasa village, which was not used for the abstraction of drinking water, contained 0.01 μg/L Hg whilst filtered water from a Hg

Acknowledgements

This paper is published with the permission of the Executive Director, British Geological Survey. The assistance of the UNIDO (Vienna) Coordinating Unit staff, Dr Marcello Veiga and Ludovic Bernaudat, the UNIDO Country Representative in Tanzania, Felix Ugbor, and the Geita Mines Officer, Mr John Nayopa is gratefully acknowledged. Advice on mercury contamination and artisanal gold mining in Tanzania was kindly provided by Mr Gray L Mwakalukwa (Commissioner for Minerals); Professor Yunus D Mgaya

References (64)

  • R. Nakagawa et al.

    Change and behaviour of residual mercury in paddy soils and rice of Japan

    Chemosphere

    (1998)
  • E. Sindayigaya et al.

    Copper, zinc, manganese, iron, lead, cadmium, mercury and arsenic in fish from Lake Tanganyika, Burundi

    Sci. Total Environ.

    (1994)
  • P. van Straaten

    Human exposure to mercury due to small scale gold mining in northern Tanzania

    Sci. Total Environ.

    (2000)
  • P. van Straaten

    Mercury contamination associated with small-scale gold mining in Tanzania and Zimbabwe

    Sci. Total Environ.

    (2000)
  • J.D. Appleton

    A study of the extent of mercury and related chemical pollution along the Naboc River, Monkayo, Davao del Norte; Hijo River, Apokon ore processing site; and their neighbouring areas (rice fields and banana plantations). A report for UNIDO Project DP/PHI/98/00511-02

    (2000)
  • Appleton JD, Taylor H, Lister TR, Smith B, Drasch G, Boese-O'Reilly S. Final Report for an Assessment of the...
  • S. Asano et al.

    Acute inorganic mercury vapor inhalation poisoning

    Pathology International

    (2000)
  • D. Barltrop

    The prevalence of pica

    Am. J. Dis. Child.

    (1966)
  • H. Barth

    Explanatory notes on the 1:500,000 provisional map of the Lake Victoria goldfields, Tanzania

    Geol. Jahrb., Reihe B Reg. Geol. Ausl.

    (1990)
  • J. Bicknell

    Pica: a childhood symptom

    (1975)
  • G. Borg

    The Geita gold deposit in NW Tanzania: geology, ore petrology, geochemistry and timing of events

  • S. Böse-O'Reilly et al.

    Medical investigation of 250 people living in the Rwamagasa area, Geita District, United Republic Of Tanzania. Final report, UNIDO Project EG/GLO/01/G34

    (2004)
  • E.J. Calabrese

    How much soil do young children ingest: an epidemiological study

    Regul. Toxicol. Pharmacol.

    (1989)
  • E.J. Calabrese et al.

    Adult soil ingestion estimates

  • E.J. Calabrese et al.

    Evidence of soil pica behaviour and quantification of soil ingested

    Human Exp. Toxicol.

    (1991)
  • L.M. Campbell et al.

    A review of mercury in Lake Victoria, East Africa: implications for human and ecosystem health

    J. Toxicol. Environ. Health, Part B. Crit. Rev.

    (2003)
  • L. Coetzee et al.

    Metal concentrations in Clarias gariepinus and Labeo umbratus from the Olifants and Klein Olifants River, Mpumalanga, South Africa: zinc, copper, manganese, lead, chromium, nickel, aluminium and iron

    Water SA

    (2002)
  • M. Cooper

    Pica

    (1957)
  • S. Davies

    Quantitative estimates of soil ingestion in normal children, between the ages of 2 and 7 years: population-based estimates using aluminium, silicon and titanium as soil tracer elements

    Arch. Environ. Health

    (1990)
  • D.H. Eccles

    Field guide to the freshwater fishes of Tanzania. FAO species identification sheets for fishery purposes

    (1992)
  • Environment Agency

    Soil guideline values for inorganic mercury contamination

    (2002)
  • Cited by (122)

    View all citing articles on Scopus
    View full text