Dietary exposure assessment for arsenic and mercury ...

Environmental Pollution 227 (2017) 552e559

Contents lists available at ScienceDirect

Environmental Pollution journal homepage: www.elsevier.com/locate/envpol

Dietary exposure assessment for arsenic and mercury following submarine tailings placement in Ratatotok Sub-district, North Sulawesi, Indonesia*,** Keith Bentley a, *, Amin Soebandrio b a b

Centre for Environmental Health, Canberra, ACT 2606, Australia Eijkman Institute for Molecular Biology, Ministry for Research, Technology and Higher Education, Jakarta, Indonesia

a r t i c l e i n f o

a b s t r a c t

Article history: Received 31 January 2017 Received in revised form 22 April 2017 Accepted 27 April 2017

The Mesel gold mine in the Ratatotok Sub-district operated between 1996 and 2004 with tailings disposal via an engineered submarine tailings placement (STP) into Buyat Bay. This operation raised concerns of increased levels of arsenic (As) and mercury (Hg) associated disease in the local communities from consumption of seafood contaminated with anthropogenic As and Hg. This report uses the dietary exposure to As and Hg, from local fishermen and market-purchased Codex “as consumed” and environmental fish results from the pre-mining baseline (1990e1995), the mine operational (1996e2004) and post-closure monitoring (2007e2016) to examine the potential health effects. The Ratatotok Subdistrict consumers total As average daily intake from fish was between 152 and 317 mg/day (adults) and 58 and 105 mg/day (infants). The average daily intake of inorganic arsenic (Asi) from the dietary staples fish and rice and drinking water consumption was 77 mg/day (adults) and 35 mg/day (infants) at Buyat Pantai and 39 mg/day (adults) and 19 mg/day (infants) at Ratatotok township. Fish consumption contributed 8.2% (adults) and 6.5% (infants) to total daily Asi intake. Average Hg intake from fish consumption, exceeded the FAO WHO PTWI for methylmercury (MeHg) for all age and gender groups at Buyat Pantai 4.6 mg/kg bw/wk (adults) and 7.3 mg/kg bw/wk (infants) and for the infants at Buyat village and Ratatotok township (2.5 and 2.8 mg/kg bw/wk respectively). The Manado City consumers had average intakes below the MeHg PTWI. The Hg exceedances resulted from the high fish consumption in coastal communities and not elevated levels of Hg in fish. Hg exposure levels from the pre-mining baseline, Mesel STP operation and post-closure monitoring, confirmed that exceedances were unrelated to the tailings deposited into Buyat Bay. © 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords: Arsenic Mercury Seafood Pollution Submarine tailings disposal Dietary intake

1. Introduction PT Newmont Minahasa Raya (PT NMR) operated a 700,000 tonne per annum open pit gold mine at Mesel in the Buyat River watershed in the Ratatotok Sub-district (RS) of North Sulawesi between 1996 and 2004. Approximately 4.5 million m3 of detoxified tailings, containing As and Hg were discharged through a

* This work reports the dietary exposure to As and Hg from fish consumption in Ratatotok Sub-district communities, potentially exposed to elevated levels from submarine placement of gold tailings residues. **

This paper has been recommended for acceptance by Joerg Rinklebe. * Corresponding author. E-mail address: [email protected] (K. Bentley).

submarine tailings pipeline at a depth of 82 m, some 800 m from the shoreline of Buyat Bay (BB). Pre-mining baseline assessment confirmed that the BB eco-system was not contaminated by Hg residues from the extensive historical small-scale gold recovery in the Totok River watershed (Kinhill, 1991; Kambey et al., 2001; Limbong et al., 2005). The two watersheds are separated by the Ratatotok Peninsula (Fig. 1). In the Indonesian coastal, rural and subsistence fishing communities, marine fish is the principal accessible and affordable source of animal protein. The Mesel mine and STP were perceived by the local communities as having resulted in an elevated prevalence of As- and Hg-associated diseases from contamination of the local seafood, both during the mine operation and following closure and cessation of tailings discharge. The Buyat Pantai (BP) beachfront community at BB (population about 200) having traditional

http://dx.doi.org/10.1016/j.envpol.2017.04.081 0269-7491/© 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

K. Bentley, A. Soebandrio / Environmental Pollution 227 (2017) 552e559

553

Fig. 1. Map of the main features in the Ratatotok Sub-district study area.

fishing as their main occupation also lost their source of income as a result of their fish being unsaleable. In 2005, more than half of the BP community left the area to re-settle in Central Sulawesi. The impact on other communities was less traumatic, but at Buyat Village (BV) population 450 and Ratatotok township (RT) population 1600, economic hardship from a loss of marine tourism and the local fishing industry, compounded existing occupational and public health and environmental pollution concerns resulting from the artisanal gold mining in the Totok River watershed. In 2007, an Independent Scientific Panel (ISP) was established by Newmont Indonesia, together with the Government of Indonesia, with the broad objective of evaluating the impact of the marinedeposited tailings on community health and the environment. The ISP conducted a 10-year physical and chemical oceanography, marine ecology and artisanal fishermen and market-purchased fish and environmental fish tissue monitoring program for the BB embayment and reference locations between 2007 and 2015 (ISP, 2016). The purposes of the present study were to assess: (i) the dietary exposure of Indonesian rural adults (by gender), adolescents and infants to total and inorganic As from consumption of fish and other food staples and drinking water; and (ii) the dietary exposure of these groups to Hg (as MeHg) from fish consumption. The highest concentration of dietary total As occurs in fish and other seafood, mostly present as the non- or low-toxicity organoarsenicals including arsenobetaine and arsenosugars (ATSDR, 2007). Contributions to Asi intake in the Asian diet is primarily

from the dietary staples seafood (in high fish-consuming groups), cereal grains, particularly rice and consumption of drinking water. Between 50% and 70% of the total As in rice is present as the Asi species (US FDA, 2013). Typical of areas associated with gold mineralisation and volcanic regions, the level of Asi in the RS bore waters is somewhat elevated (ANDAL, 1994; Apte et al., 2004). Hg is actively bio-accumulated in fish and other seafood, with concentrations dependent on the particular species, habitat, maturity, weight, length, diet and the levels in the local environment. The concentration of Hg in other food products is generally below the limit of detection. In fish and other seafood, Hg bioaccumulates as the more toxic methylated species, with generally 70e85% of the total tissue Hg present as MeHg. In risk assessments, many authors and jurisdictions assume that 100% of the total Hg is MeHg (Health Canada, 2008). This study examined detailed results for total As and Hg levels in marine fish tissue from BB and reference locations. This included pre-mining baseline data from 1991 to 1995, the Mesel gold mine operational monitoring from 1996 to 2004 and the ISP monitoring between 2007 and 2015. The pre-mining baseline data and that from the Mesel gold mine monitoring has been provided by Newmont Indonesia (Kinhill, 1991; Rescan, 1995, 1996; PT NMR, Quarterly reports). The results for the 2007e2011 ISP BP fisherman and local market fish dietary studies and the 2007e2015 As and Hg in BB and the Pulau Nanas (PN) reference location environmental fish surveys are from the final report of the ISP (ISP, 2016).

554


2. Methods

2.2. Consumption of fish and other seafood

The dietary As and Hg exposure assessments require detailed data on their concentrations in edible fish tissue and quantitative results for the amounts consumed by age and gender. Comparison of Hg exposure with the FAO/WHO Provisional Tolerable Weekly Intake (PTWI) health guidelines also requires age and gender weight-for-age results data for the local consumers.

Fish and seafood availability and consequent seasonal consumption patterns were examined using three 24-hour dietary recall surveys between October 2007 and May 2008. The wet and dry season patterns are impacted by the North Sulawesi seasonal monsoon, particularly for the artisanal fishermen who used outrigger canoes. Each survey used a stratified two-stage method, with the village as the first level and the household as the second level of sampling. BP and one sub-village from BV (Buyat-1) and RT (Ratatotok South) were randomly selected. All of the residents aged over 6 months in the 23 households at BP were selected, together with a similar number at Buyat-1 and Ratatotok South, using proportionate random sampling following determination of the total number of households. The Multiple-Pass four-stage dietary survey approach using pretrialled questionnaires, direct weighed food intake measurements and pictorial representation of portion size against a recognised scale was adopted, with the food intake during the previous 24 h recorded. The portion size estimates were converted to gram intakes. The methodology also investigated whether the consumption on the recorded day was typical and if there were any village or household food security issues (Guenther et al., 1998; EFCOSUM, 2001). For each survey 24-hour dietary recall was implemented twice, scheduled to represent weekday and weekend consumption, the mean daily intake calculated as: [2 (weekend-intake) þ 5 (week day-intake)]/7.

2.1. As and Hg in Ratatotok Sub-district fish Results reported for total As and Hg concentrations in fish tissue have used analytical results from artisanal fishermen and marketpurchased fish and environmental monitoring. Summary As and Hg mean, standard deviation, median and 90th percentile (P90) results in fish from BB and the 2007e2015 PN are shown in Table 1. The database includes unpublished consultant results from baseline data (1991e1995) and the PT NMR operational period (1996e2004). These results have been provided by PT NMR from the original environmental impact assessment and the PT NMR Quarterly reports to the Indonesian Mines Department (Kinhill, 1991; ANDAL, 1994; Rescan, 1995). The detailed As and Hg dietary fish tissue results for the BB fishermen and market-sourced fish from BV, RT and Manado City, conducted by the ISP between 2007 and 2011, and the 2007e2015 environmental fish tissue database, have been presented in the final report of the ISP (ISP, 2016). Detailed laboratory analytical procedures are available for the data generated between 1996 and 2015. All samples were digested using standardised methods (US EPA, Method 200.3, 1991). Five grams of the wet flesh sample was weighed out and digested with a mixed nitric-hydrochloric acid for 2 h at 90e95 C. This digested sample was then diluted to 50 mL with deionised water and shaken well. The final digestate was then diluted five-fold with deionised water and analysed for As by inductively coupled plasma atomic emission spectroscopy (US EPA, Method 6010b, 1996). Hg analysis was by cold vapour atomic absorption spectroscopy (US EPA Method 245.7, 2005). Laboratory quality assurance and quality control included laboratory replicate samples, blank samples and the National Research Council Canada Certified Reference Materials DORM-2 (dogfish muscle) and DORM-3 (fish protein). The two analytical laboratories routinely participated in external QA/QC Proficiency Testing and Inter-laboratory comparisons with consistent good performance.

2.3. Bodyweight-for-age for consumers There were no age and gender bodyweight-for-age results available for rural communities in North Sulawesi. Weight-for-age results used in the Hg risk assessment calculations were from a socio-economically comparable rural population in Indonesia: all adults (18 þ years) 55.9 kg, adult males 57.8 kg, adult females 53.6 kg, pre-adolescents (10e12 years) 28.3 kg and infants (2e4 years) 11.7 kg (Indonesian Ministry of Health (2014). 2.4. Dietary exposure estimates for human consumers JECFA, in considering the intake of organic As in high fishconsuming populations, concluded that there were no reports of ill effects from consuming intakes of about 50 mg/kg bw day,

Table 1 Total As and Hg in fish tissue from the Ratatotok Sub-district and Manado City, North Sulawesi (all values mg/kg ww). Database

Group

n

Mean ± sd

54 54 54 54 25 315 328 331

2.01 1.71 1.76 2.12 1.14 0.64 3.13 5.78

54 54 54 54 25 315 328 331

0.233 0.105 0.141 0.043 0.204 0.174 0.167 0.169

Median

P90

1.58 1.29 1.35 1.31

4.18 3.71 3.42 4.60 e 1.40 6.97 12.94

Total As 2007e2011 Market fish

1991e1995 Pre-mining 1996e2004 PT NMR Mesel 2007e2015 ISP post-closure

Buyat Pantai fishermen Buyat market Ratatotok market Manado fish market Buyat Bay Buyat Bay Buyat Bay Pulau Nanas reference

± ± ± ± ± ± ± ±

1.56 1.53 1.23 1.82 0.93 1.74 3.87 10.71

0.17 1.47 2.22

Total Hg 2007e2011 Market fish

1991e1995 Pre-mining 1996e2004 PT NMR Mesel 2007e2015 ISP post-closure

Buyat Pantai fishermen Buyat market Ratatotok market Manado fish market Buyat Bay Buyat Bay Buyat Bay Pulau Nanas reference

± ± ± ± ± ± ± ±

0.250 0.174 0.152 0.036 0.304 0.334 0.208 0.199

0.137 0.050 0.070 0.031 0.081 0.091 0.090

0.568 0.253 0.434 0.089 e 0.329 0.459 0.409


equivalent to an adult (60 kg bodyweight) intake of 3000 mg/day. Health-based guideline values for Asi, such as the FAO/WHO PTWI and reference doses have been withdrawn and estimates of health effects are presented as comparisons with national dietary intakes. For example, WHO provides population-based mean and upper percentile exposure estimates of daily intakes of Asi from a number of regions and countries (FAO/WHO JECFA, 2011; Table 7, p 33). The total As intake (mg/day) was calculated by multiplying the average As concentration in fish tissue (mg/kg ww) from Table 1, by the annual per capita consumption for the location and age and gender group (kg/year) and dividing by 365 to obtain daily intake. The fraction absorbed has been assumed as 100%. The upper percentile P90 values have been included, to estimate the daily intake for consumers who preferentially eat species with a high As content. Hg in fish is largely present in the MeHg form and total dietary intakes are almost exclusively through the consumption of fresh and marine fish and other seafood. Consistent with the approach adopted by a number of national regulatory agencies, the health risk assessment has allocated 100% of total Hg intake to the MeHg form and used the PTWI of 1.6 mg/kg bw/wk (FAO/WHO JECFA, 2007). The total Hg exposure was calculated by multiplying the average Hg concentration in fish tissue (mg/kg ww) from Table 1, by the annual per capita consumption for the location and age and gender group (kg/year) and dividing by 52 body weight of the target group. The percentage of Hg as MeHg and the fraction absorbed were both assumed to be 100%. For comparison with the PTWI, the total exposure was converted to mg/kg bw/wk.

555

increase in the consumption of dried fish. Total fish consumption by weight, after adjusting the dried fish data, was comparable for all seasons. Marine and dried fish and fresh water fish consumption for the three surveys for the adult (by gender), adolescent, child and infant groups are described in Appendix A, Table A1. Mean total annual per capita fish consumption by age and gender has been derived with the dried fish results converted to the ww equivalent by adopting an average moisture content of 80% (Ravichandran et al., 2011). The infant and child consumption values are derived from the all villages data in Appendix A, Table A1. Upper percentile intakes have not been derived, with all consumers having high levels of fish consumption. The summary results are detailed in Table 2. A 12-month fish availability survey, identified only minor contributions to total seafood consumption by cephalopod molluscs (2.2%), crustaceans (0.3%) and bivalve molluscs (0.3%). Seaweed consumption was negligible at all of the study communities. Some crustaceans and molluscs, had higher mean total As levels, but were only minor contributors to the total dietary intakes and were not considered further.

3.3. Dietary intake assessment for As and Asi The WHO Tolerable Daily Intake (TDI) and FAO/WHO PTWI and reference doses for Asi have been withdrawn, due to concerns of epidemiological evidence of harmful carcinogenic effects at concentrations below the previously recommended intakes (FAO/WHO JECFA, 2011). Neither the Codex Alimentarius Commission nor the European Union currently has a standard for Asi in seafood. Food Standards Australia New Zealand has a standard for Asi levels in fish and crustaceans of 2 mg/kg ww (FSANZ, 2000). In the absence of a FAO/WHO PTWI, or broadly accepted reference dose, it is not possible to provide quantitative risk assessments for Asi. This study, which considers data sourced over the last 25 years, of necessity focuses on total As screening for exposures and comparison of these results with other national total As dietary intakes. For comparative purposes, Asi intake has been derived using a total As:Asi ratio of 50:1 for fin fish. This approach is consistent with recently published results indicating Asi as being about 1e2% of total As (UK COT, 2002; Baeyens et al., 2009; Stewart and Turnbull, 2015). The results for the mean and upper percentile P90 values for the 2007e2011 market fish surveys and the 1991e2015 environmental surveys are shown in Table 3. The total As intake (mg/day) was derived by multiplying the average As concentration in fish tissue (mg/kg ww), by the annual per capita consumption for the location and age and gender group (kg/year) and dividing by 365 to obtain daily intake. Table 3 has assumed an adult per capita fish consumption at Manado City of 30 kg/year and that the per capita consumption at PN was the same as that at the BP households. For the RT consumers, the average adult and infant daily intake of total As from market-purchased fish was between 152 and 58 mg/ day respectively. The Manado City adult and infant intakes were similar at 174 and 68 mg/day. The BP fishermen had average adult

3. Results and discussion 3.1. As and Hg in Ratatotok Sub-district fish The mean and P90 results for the concentrations of total As and Hg in the 2007e2011 BP fishermen and market-sourced fish study and the results from the 1991e2015 pre-mining baseline, Mesel mine operation and ISP post-closure monitoring are shown in Table 1. The overall range of means for total As was between 0.64 and 5.78 mg/kg wet weight (ww) and the P90 value range was 1.4e12.9 mg/kg ww. These results are