Artisanal and small-scale gold mining (ASGM) is the largest global anthropogenic mercury (Hg) source and is widespread in the Peruvian Amazon. While numerous studies have examined fish Hg content near ASGM, Hg accumulation in other commonly consumed animal- and plant-based foods from terrestrial environments is often overlooked. These data were collected in 2018 and 2019 to understand Hg exposure from food staples in Peru's Madre de Dios region. This dataset contains measurements of total Hg and methyl Hg content in locally sourced crops, fish, chicken meat, chicken feathers, and eggs from ASGM-impacted and upstream reference communities. Stable carbon and nitrogen isotope signatures from fish and chicken were also measured to characterize trophic position and magnification.

All data were collected from the Madre de Dios region of the Peruvian Amazon. Sampling sites are described in Marchese et al. (2024), and coordinates can be found in the data file entitled "Site.coordinates.csv".

Sampling, laboratory analysis, and quality control methodology are presented in Marchese et al. (2024). In brief, we collected crops, fish muscle, chicken meat, chicken eggs, and chicken feathers from areas heavily and minimally impacted by mining. The data were used to determine how geogenic and anthropogenic Hg accumulates in terrestrial and aquatic organisms that serve as regional food staples. Samples were collected during June through August in 2018 and 2019. These data are predominantly comprised of total Hg measurements for crops, chicken, and fish plus methyl Hg for crops and chicken. Stable carbon and nitrogen isotope data are presented for chicken feathers and fish tissue. Survey data from individuals who contributed chicken samples to the study are included. Coordinates for each community from which samples were obtained can be found in the .csv files.

The following is an excerpt describing sample collection, laboratory analysis, and quality control in crops, chicken muscle/organs, chicken feathers, chicken eggs, and fish muscle, as reported in Marchese et al. (2024):

"Fifty crop samples were collected in seventeen Madre de Dios communities from June to August 2018. We purchased crops from every community we passed while traveling 200km by boat along the Madre de Dios River from Shipetiari to Laberinto. Crops were purchased from markets after vendors confirmed they were grown locally. Crop samples were washed with ultrapure water, placed in individual plastic bags, and stored in a freezer. They were transported frozen and stored at -20 °C upon arrival to Duke University (Durham, NC, USA) until lyophilization and grinding. 

Fish samples represented a variety of species, trophic levels, and origins. Fifteen locally accessible and commonly consumed species were selected. From June to August 2018, we purchased 98 fish samples from aquaculture farms, markets, and fishermen in eight Madre de Dios communities. River-caught, locally farmed, and imported ocean fish were collected. 

Sourcing fish from sellers rather than the river directly ensures the samples appropriately represent available foods and differentiates these data from prior regional studies (Diringer et al., 2015; Martinez et al., 2018; Roach et al., 2013). Fish species and their origins were identified by vendors and confirmed using the Instituto de Investigaciones de la Amazonía Peruana (IIAP, Institute for Investigations of the Peruvian Amazon) guide to regional fish(García et al., 2018). Since the whole fish was available for purchase in most cases, total length (mouth to tail) and standard length (mouth to last vertebra, excluding tail) were measured. Next, small muscle filets (<5g fresh weight) were placed in plastic bags, stored in a cooler with ice packs, and transferred to a Credo Cube Series 20M during the remaining fieldwork. Samples were transported frozen to Duke University, skin was removed, and samples were stored at -20 °C until lyophilization and grinding. Fish samples were weighed before and after drying. Moisture content was calculated as follows:

% moisture =   (fresh weight-dry weight )/(fresh weight)  *100

From July to August 2019, we sampled feathers (n=38), eggs (n=16), and meat (n=8) from backyard chickens raised in a mining-impacted community (Boca Colorado) and an upstream community (Boca Manu). Fifteen commonly consumed chicken cuts were collected (liver, skin, leg, breast, intestine, gizzard, tail, back, thigh, neck, wing, heart, spleen, feet, fat). Multiple feathers were collected from the breast and back of each chicken. All internal tissue and feather samples were rinsed with ultrapure water, placed in individual plastic bags, and stored on dry ice inside a Credo Cube Series 20M until lyophilization at the IIAP Laboratorio de Mercurio y Química Ambiental (Mercury and Environmental Chemistry Laboratory, Madre de Dios). Lyophilized samples were transported to Duke University for analysis.

Surveys were performed during the Boca Manu and Boca Colorado chicken sample collections. Consenting participants were asked about their primary protein sources and how frequently they consume different chicken cuts. Participants who raised chickens were asked about their chickens' living conditions and diets.

Institutional Review Boards at Duke University (#2019-0530) and the University of Peru Cayetano Heredia (chicken: #104284) provided ethical clearance for this project. All participants provided verbal consent in the chicken collection survey. Survey data were deidentified for analysis. The Duke University Animal Care and Use Committee approved field methods involving live animals (#A161-19-07).

Total Hg content of dried, homogenized crop and fish muscle tissue was measured at Duke University via thermal decomposition, amalgamation, and atomic absorption spectrophotometry (EPA method 7473) on a Milestone Direct Mercury Analyzer (DMA-80) per methodology in Gerson et al.(Gerson et al., 2018). Samples were run in duplicate, and the average result was used for data analysis. If the relative percent difference (RPD) exceeded 10% between the two sample runs, samples were rerun in duplicate. Instrument calibration was performed using the Brooks Rand Instruments Total Mercury Standard (1.0 ng/L). Continuous calibration verification (CCV), quality control standard (QCS), matrix spikes (MS), and blank measurements were performed after every tenth sample. Standard reference materials for crops, NIST 1633c (coal fly ash) and NIST 2709a (San Joaquin Soil), had percent recoveries of 95% ± 3% SD (n= 39) and 94% ± 3% SD (n=8), respectively. Blanks contained <0.0037 mg/kg total Hg. Standard reference materials for fish tissue, NIST 1633c (coal fly ash) and DORM-4 (fish protein), had percent recoveries of 96% ± 3% SD (n=52) and 91% ± 7% SD (n=27), respectively. All blanks contained <0.0023 mg/kg  [total Hg]. The instrument detection limit for fish and crop total Hg analysis was 0.5 ng Hg. For fish, dry-weight (dw) total Hg was converted to fresh-weight (fw) total Hg as follows:

[total Hg fw]=([total Hg dw])/((sample dw)/(sample fw))

Chicken tissues and feathers were analyzed at the Biodiversity Research Institute (Portland, ME, USA) for total Hg following EPA method 7473 on a Milestone DMA-80. Samples were run in duplicate and rerun if the RPD exceeded 15%. For quality control, empty sample boats (blank boats, n=7) and blanks (no boat or sample, n=19) were run to test for residual Hg in the analyzer and the sample boats, respectively. All blanks and blank boats returned <0.002 mg/kg [total Hg]. Standard reference materials used for feathers and tissues were Dolt-5 (dogfish liver; 97% recovery ± 0.7 % SD, n=25) and CE 464 (tuna fish; 99% recovery ± 2% SD, n=25). The instrument detection limit for internal chicken tissue, egg, and feather total Hg analysis was 0.003 ng Hg.

Crops and internal chicken tissues were analyzed for MeHg at Syracuse University (Syracuse, NY, USA) on a Tekran 2500 spectrometer per the biomass method in Gerson et al.(Gerson et al., 2017). 20 mg samples were digested in glass vials using 3 mL potassium hydroxide (25% w/w) in methanol at 55 °C for 48h and frozen (Hall and St. Louis, 2004; Hintelmann and Nguyen, 2005; Horvat et al., 1993). Following digestion, samples underwent direct aqueous ethylation with sodium tetraethylborate, purge and trap, and cold vapor atomic fluorescence spectrometry (CVAFS, EPA Method 1630). Calibration, CCV, MS, ongoing precision recovery, method detection limit, and blank samples were incorporated into the run. Standard reference materials for chicken tissues were ERM-CE477 mussel tissue (butyltins; 109.1% recovery ± 3.54% SD, n=3), TORT-2 (lobster hepatopancreas; 98.8% recovery ± 11.01% SD, n=4), and DORM-4 (fish protein; 104.1% recovery ± 6.15% SD, n=4). The method detection limit for chicken tissues was 0.0007 ng Hg, and all blanks had <0.0006 mg/kg   [MeHg]. The method detection limit for crops was 0.002 ng MeHg, and all blanks had <0.0056 mg/kg MeHg.

MeHg is reported as absolute values and percent MeHg. For samples with detectable MeHg, percent MeHg is defined as follows:

%MeHg=  ([MeHg dw])/([total Hg dw])×100

Different instrument detection limits led to MeHg values higher than total Hg in some samples. To avoid skewed representations of MeHg proportions, all MeHg percentages exceeding 100% were rounded down to 100%.

Stable carbon and nitrogen isotope analyses were performed on ground feather samples and dried fish muscle tissue at the University of California-Davis Stable Isotope Facility (SIF; Davis, CA, USA) and the Duke Environmental Stable Isotope (DEVIL) Laboratory (Durham, NC, USA), respectively. δ¹³C and δ¹⁵N signatures were determined via mass spectrometry using an Elementar Micro Cube elemental analyzer interfaced to a PDZ Europa 20-20 isotope ratio mass spectrometer for feather samples and two Thermo-Finnigan Delta Plus XL continuous flow mass spectrometer systems for fish muscle tissue. For quality control, samples were interspersed with several replicates of internal laboratory standards, which were calibrated against international reference materials. δ¹³C and δ¹⁵N values were normalized to correct for changes in instrument environment, parameters, and gas drift. Stable isotope ratios are expressed in δ notation in parts per thousand (‰) relative to Vienna PeeDee Belemnite standards for ¹³C and atmospheric N₂ for ¹⁵N, as follows:

δX = [(R_sample-R_standard) - 1] * 1000 (‰)

where X represents ¹³C or ¹⁵N and R is ¹³C/¹²C or ¹⁵N/¹⁴N."

References:

Diringer, S.E., Feingold, B.J., Ortiz, E.J., Gallis, J.A., Araújo-Flores, J.M., Berky, A., Pan, W.K.Y., Hsu-Kim, H., 2015. River transport of mercury from artisanal and small-scale gold mining and risks for dietary mercury exposure in Madre de Dios, Peru. Environmental Sciences: Processes and Impacts 17, 478–487. https://doi.org/10.1039/c4em00567h

García, C.R., Riveiro, H., Flores, M.A., Mejia de Loayza, J.E., Angulo, C.A.C., Castro, D., Estivals, G., García, A., Nolorbe, C., Dávila, G., Núñez, J., Mariac, C., Duponchelle, F., Renno, J.-F., 2018. Peces de consumo de la amazonía peruana.

Gerson, J.R., Driscoll, C.T., Demers, J.D., Sauer, A.K., Blackwell, B.D., Montesdeoca, M.R., Shanley, J.B., Ross, D.S., 2017. Deposition of mercury in forests across a montane elevation gradient: Elevational and seasonal patterns in methylmercury inputs and production. J Geophys Res Biogeosci 122, 1922–1939. https://doi.org/10.1002/2016JG003721

Gerson, J.R., Driscoll, C.T., Hsu-Kim, H., Bernhardt, E.S., 2018. Senegalese artisanal gold mining leads to elevated total mercury and methylmercury concentrations in soils, sediments, and rivers. Elementa: Science of the Anthropocene 6. https://doi.org/10.1525/elementa.274

Hall, B.D., St. Louis, V.L., 2004. Methylmercury and Total Mercury in Plant Litter Decomposing in Upland Forests and Flooded Landscapes. Environ Sci Technol 38, 5010–5021. https://doi.org/10.1021/es049800q

Hintelmann, H., Nguyen, H.T., 2005. Extraction of methylmercury from tissue and plant samples by acid leaching. Anal Bioanal Chem 381, 360–365. https://doi.org/10.1007/s00216-004-2878-5

Horvat, M., Bloom, N.S., Liang, L., 1993. Comparison of distillation with other current isolation methods for the determination of methyl mercury compounds in low level environmental samples. Anal Chim Acta 281, 135–152. https://doi.org/10.1016/0003-2670(93)85348-N

Martinez, G., McCord, S.A., Driscoll, C.T., Todorova, S., Wu, S., Araújo, J.F., Vega, C.M., Fernandez, L.E., 2018. Mercury contamination in riverine sediments and fish associated with artisanal and small-scale gold mining in Madre de Dios, Peru. Int J Environ Res Public Health 15, 1–15. https://doi.org/10.3390/ijerph15081584

Roach, K.A., Jacobsen, N.F., Fiorello, C. V, Stronza, A., Winemiller, K.O., 2013. Gold Mining and Mercury Bioaccumulation in a Floodplain Lake and Main Channel of the Tambopata River, Perú. J Environ Prot (Irvine, Calif) 4, 51–60. https://doi.org/10.4236/jep.2013.41005