Heavy Metal Contamination in Birds from Protected Regions in the Amazon: Heavy metals in birds from the Amazon

Abstract

The extraction of crude oil and gold has substantially increased heavy metal contamination in the environment, yet the study of wildlife exposure and biological response to this pollution remains nascent even in the most biodiverse places in the world. We present a survey of heavy metal exposure in the feathers of wedge‐billed woodcreepers (Glyphorynchus spirurus), a resident neotropical bird found within protected regions of the Amazon near oil and gold extraction sites. Our results show elevated heavy metal contamination in samples collected from protected areas proximate to known oil and gold extraction. Surprisingly, several samples from remote reference sites also displayed elevated levels of various heavy metals, suggesting a background of natural deposition or complex heavy metal contamination in the environment from anthropogenic sources. These results highlight the need to understand the ecological and biological impacts of increased heavy metal exposure on wildlife across space and time, including remote regions of the world purportedly untouched by these human‐mediated stressors. Toward this goal, historical and contemporary data from native bird populations may provide crucial indicators for heavy metal contamination and exposure in wildlife and human communities. Environ Toxicol Chem 2024;43:2601–2607. © 2024 The Author(s). Environmental Toxicology and Chemistry published by Wiley Periodicals LLC on behalf of SETAC.

INTRODUCTION

Over the last century, anthropogenic activities have redistributed and magnified heavy metal concentrations in environments worldwide (Carson, 1962; Mohammed et al., 2011; Nriagu, 1996). Rapid technological and methodological advancements for geological resource extraction beginning in the Industrial Revolution have considerably increased the release of heavy metals, resulting in widespread environmental contamination far beyond that driven by natural geological processes (Mohammed et al., 2011; Vareda et al., 2019). The global demand for oil and gold has accelerated the liberation of heavy metals that co‐occur in rock ores, leading to contamination in the air, water, and soil of surrounding regions (Vareda et al., 2019). In addition, the extraction or refinement of these resources commonly requires the use of heavy metal reagents, further contaminating the environment through the extensive production of gaseous vapors and polluted wastewater (Fashola et al., 2016; Hansen et al., 2019; Mohammed et al., 2011). Although various heavy metals act as essential trace elements in biological functions, the resultant mixtures of concentrated heavy metal pollutants from these mining practices are highly toxic and persist in the environment long after release (Mohammed et al., 2011). Understanding the extent and impact of this contamination has become vitally important as oil and gold mining practices often occur disproportionately within or near biodiversity hotspots, jeopardizing unique wildlife including endemic, threatened, and endangered species with chronic toxic exposure.

Although oil and gold mining practices are a substantial source of industrial and economic support for many countries, the lasting biological impacts of persistent heavy metal contamination on organismal health and ecosystem function remain understudied (Finer et al., 2008; Sayers et al., 2023). This includes the Amazon rainforest of South America, home to many culturally rich Indigenous communities and one of the planet's most biodiverse regions (Bass et al., 2010; Myers, 1988). Despite being an important hotspot of diversity in need of protection, the discovery of crude oil in the Amazon forests of Ecuador in the mid‐1900s led to the extraction of countless barrels of oil, releasing what has been estimated as billions of liters of crude oil through spills and contaminated wastewater in the process (Durango‐Cordero et al., 2018; Finer et al., 2008; Kimerling, 1990; Lessmann et al., 2016; Pellegrini et al., 2020; Rivera‐Parra et al., 2020; Sawyer, 2004; Sebastián et al., 2001). Local health reports from exposed Indigenous groups fueled community lobbying for reparations and environmental remediation, yet ecological exposure and detailed physiological responses to contamination in this region remain mostly undocumented (Hurtig & Sebastián, 2005; Sebastián et al., 2001). Similarly, gold exploration and mining beginning in the mid‐1850s have also exposed Indigenous communities and wildlife to heavy metals in the Amazonian rainforests of French Guiana (Douine et al., 2018; Fréry et al., 2001; Malm et al., 1995). The use of mercury (Hg) in small artisanal mining cheapened the extraction process; however, surveys of the Indigenous communities surrounding gold mining sites displayed substantial Hg contamination from consuming fish exposed to mining wastewater (Fréry et al., 2001). Gold mining in French Guiana and oil extraction in Ecuador continue today, despite the paucity of studies on the characterization and impact of heavy metal contamination from these anthropogenic sources.

To gain a better understanding of the extent and magnitude of heavy metal contamination within undersurveyed regions of the Amazon, we assessed the heavy metal content of feathers collected from a common and widespread forest understory bird across areas with historical and contemporary oil and gold mining in the protected rainforests of Ecuador and French Guiana. Feathers have been a widely used nondestructive tool for the biomonitoring of heavy metals deposited from the blood throughout the growth of an organism (Abdullah et al., 2015; Burger & Gochfeld, 2000). As resident understory birds, wedge‐billed woodcreepers (Glyphorynchus spirurus) serve as a useful bioindicator for local exposure, absorbing heavy metals from soil, water, and their insectivorous diet throughout their life span (Milá et al., 2000). The comparison of heavy metal content in these sentinels of contamination found within protected areas proximate to oil and gold mining can uncover unexpected scenarios of pollution and highlight the need for future studies to understand the ecological and biological impacts of exposure.

METHODS

We analyzed heavy metal concentrations in a collection of outermost tail feathers (retrices) from wedge‐billed woodcreepers (G. spirurus) captured by mist net across protected sites in Ecuador and French Guiana between the years 2000 and 2009 (Martínez‐Renau et al., 2022). Study sites in Ecuador were identified by spatially overlaying protected regions according to the United Nations’ Environmental World Conservation Monitoring Centre (accessed August 2022) with gas and oil data prior to sampling from the Global Energy Monitor database for Latin America (August 2022 release) and characterizing feather collection sites in protected regions by their kernel density estimation values and distance from oil activities in ArcGIS and R (ArcGIS, Ver. 10.0, 2010; R, Ver. 4.4.1, 2021; Supporting Information, Figures S1, S2). Due to the geologic history and topographic nature of the Andes, past and present oil extraction has been concentrated in the eastern lowlands of Ecuador (Figure 1A; Supporting Information, Figure S1). As protected areas in the Andean highlands and foothills, feathers collected from San Rafael Falls (n = 6; latitude −0.1037, longitude −77.5813), Hollin River (n = 18; latitude −0.689, longitude −77.727), Sangay (n = 21, latitude −2.113571, longitude −78.370998), and Miazal (n = 6, latitude −2.636, longitude −77.798) were sampled as reference sites devoid of nearby reported oil activities (Figure 1A; Supporting Information, Figure S2). Heavy metal concentrations were compared to samples taken from Ecuador's Yasuní National Park and Biosphere Reserve (n = 20; latitude −0.714352, longitude −76.347389), a protected area in the lowlands and home to the research‐intensive Tiputini Biodiversity Station near oil mining at the large active Ishpingo‐Tambococha‐Tiputini oil block and the historic Chevron Texaco oil fields and spills (Figure 1A; Supporting Information, Figure S2; Bass et al., 2010). Suggestive heavy metal contamination from gold mining was similarly mapped and assessed across protected regions in French Guiana using data from the French Republic Camino Digital Mining Cadastre (accessed November 2023; Supporting Information). Heavy metal exposure was also measured in feather samples collected from Nouragues National Nature Reserve (n = 24; latitude 4.0855, longitude −52.6818), a protected area surrounded by several licensed and unlicensed gold mines in comparison to the Andean reference sites free from local gold extraction (Figure 1B; Fréry et al., 2001).

The suite of heavy metals chosen for our analyses were based off known pollutants from oil and gold extraction and refinery, which can vary by the geologic formation and mineral composition of a particular landscape (Mohammed et al., 2011). In general, crude oil extraction and refinery processes have been known to produce extensive amounts of wastewater containing heavy metals such as zinc (Zn), manganese (Mn), iron (Fe), aluminum (Al), copper (Cu), barium (Ba), strontium (Sr), nickel (Ni), chromium (Cr), lead (Pb), selenium (Se), cadmium (Cd), cobalt (Co), Hg, vanadium (V), arsenic (As), and uranium (U; Chaudhuri, 1978; Dittert et al., 2010; Erickson et al., 1954; Groudeva et al., 2001; Hansen et al., 2019; Mere et al., 2022; Osuji & Onojake, 2004; Pedrozo‐Peñafiel et al., 2019; Saleh et al., 2021; Wake, 2005; Wilhelm & Bloom, 2000). In gold mining, Hg is often used to facilitate the separation of other metals naturally found in gold ores such as Zn, Mn, Fe, Al, Cu, Ni, Cr, Pb, Se, Cd, Co, antimony (Sb), V, As, and U (Abdul‐Wahab & Marikar, 2012; Jasiak et al., 2021; Khamkhash et al., 2017; Malm et al., 1995; Winde et al., 2019; Zupunski et al., 2023). Although Al, Ba, Se, and Sr are not considered heavy metals with densities >5 g/cm³, they often co‐occur in the ore and have been associated with environmental pollution (Abdul‐Wahab & Marikar, 2012). As such, samples were analyzed for 23 metals, including Al, As, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Se, Sr, molybdenum (Mo), silver (Ag), Cd, tin (Sn), Sb, cesium (Cs), Ba, Hg, thallium (Tl), Pb, and U.

In the field, dry feathers were stored in paper envelopes and sealed in plastic containers until final storage in a −20 °C freezer. Prior to analysis, a single feather sample for each individual was weighed and rinsed with acetone for 30 s, followed by a 30‐s rinse in double‐distilled H₂O and air‐drying according to standard microwave digestion protocols (Aloupi et al., 2020). Concentrations for all 23 metals were determined at Dartmouth's Trace Element Analysis Core using inductively coupled plasma–mass spectrometry as part of their trace heavy metal spectrum analysis workflow (8900 Triple Quadrupole). All samples (n = 95) were sent across two runs, each passing Dartmouth's Trace Element Analysis Core's multiple calibration and quality assurance standard tests (90%–110% recovery) according to the US Environmental Protection Agency's Hazardous Waste Test Methods SW 846 (Supporting Information). Sample concentrations below the method detection limit for each metal were labeled as “baseline detection low” and considered to have no detectable metal concentrations in our analyses (Supporting Information).

All statistical analyses were conducted in RStudio with tidyverse and rstatix packages (R, Ver 4.4.1, 2021; Kassambara, 2023; Wickham & Grolemund, 2017). For all tests, we used an alpha threshold of 0.05 to reject the null hypothesis, and significance values below 0.0001 were described as p < 0.0001. After scaling the data for each metal, a Shapiro‐Wilk test on the mean metal load per sample confirmed a lack of normality in the data (W = 0.75715, p < 0.0001), with unequal variances across reference and test groups (F = 8.5715, df = 43/50, p < 0.0001), prompting nonparametric analyses. The weight of the feather analyzed (mean = 0.006 g, SD = 0.002 g) was not considered as an interacting factor in our tests based off the best‐fit one‐way model of the mean metal load with the lowest Akaike information criterion. Kruskal‐Wallis and Mann‐Whitney U tests were used to calculate the rank sum of each metal concentration across the reference and test groups. A Wilcoxon rank‐sum test was subsequently performed for pairwise analyses with a Holm‐Bonferroni p adjustment for multiple comparisons. These tests were used to compare concentrations in protected areas, including reference sites in the Andean highlands and foothills of Ecuador, samples from the environmentally protected area of Tiputini near oil activity, and the protected site of Nouragues, French Guiana, near gold mining.

RESULTS

The metal content in wedge‐billed woodcreeper feathers from protected reference sites in the Andean highlands and foothills of Ecuador (n = 51) was compared to those sampled from protected lowlands in Ecuador near oil extraction sites (Tiputini, n = 20) and similarly to samples from protected areas in French Guiana near gold mining sites (Nouragues, n = 24). In the comparison of the mean load of all metals measured within an individual feather, samples collected from the predicted contaminated sites of Tiputini, Ecuador, and Nouragues, French Guiana, were significantly greater than samples collected from the reference sites along the Andean highlands and foothills (Kruskal‐Wallis, χ² = 29.612, df = 2, p < 0.0001; Supporting Information, Figure S3). Separately, the mean metal load was significantly higher in both the Tiputini samples (W = 665, p = 0.04828) and the Nouragues samples (W = 149, p < 0.0001) compared to the reference samples (Supporting Information, Figure S3 and Table S1).

Samples collected from Tiputini, Ecuador, near oil activities displayed significantly higher concentrations of Mn (W = 792, p = 0.0003), Fe (W = 735, p = 0.0041), Cu (W = 705, p = 0.0129), Ba (W = 750, p = 0.0022), Ni (W = 836, p < 0.0001), Cd (W = 790, p = 0.0003), Co (W = 799, p = 0.0002), V (W = 789, p = 0.0003), and U (W = 663, p < 0.0001) than in the reference sites according to Mann‐Whitney tests. Barium and all heavy metals elevated at this site are known to be associated with oil activities (Figure 2; Chaudhuri, 1978; Dittert et al., 2010; Erickson et al., 1954; Groudeva et al., 2001; Hansen et al., 2019; Mere et al., 2022; Osuji & Onojake, 2004; Pedrozo‐Peñafiel et al., 2019; Saleh et al., 2021; Wake, 2005; Wilhelm & Bloom, 2000). Surprisingly, samples from the purported reference sites, far from sources of known pollution and presumed to contain minimal contamination, displayed significantly higher concentrations of several heavy metals, including Sb (W = 94, p < 0.0001), Tl (W = 306, p = 0.0070), and Cs (W = 122, p < 0.0001), as well as known byproducts of oil mining, including Se (W = 90, p < 0.0001) and heavy metals such as Zn (W = 245, p = 0.0007) and As (W = 330, p = 0.0026; Figure 2; Chaudhuri, 1978; Dittert et al., 2010; Erickson et al., 1954; Groudeva et al., 2001; Hansen et al., 2019; Mere et al., 2022; Osuji & Onojake, 2004; Pedrozo‐Peñafiel et al., 2019; Saleh et al., 2021; Wake, 2005; Wilhelm & Bloom, 2000).

Samples collected from Nouragues, a protected site in French Guiana near known active gold mines, displayed a significantly elevated level of Se (W = 2, p < 0.0001) and heavy metals such as Zn (W = 40, p < 0.0001), Mn (W = 307, p < 0.0005), Fe (W = 130, p < 0.0001), Cu (W = 34, p < 0.0001), Ni (W = 16, p < 0.0001), Cr (W = 145, p < 0.0001), Pb (W = 241, p < 0.0001), Co (W = 331, p = 0.0001), Hg (W = 115, p < 0.0001), Cs (W = 392, p = 0.0126), U (W = 306, p < 0.0001), and V (W = 294, p < 0.0003) in comparison to the Andean reference sites according to Mann‐Whitney tests (Figure 2). All of these metals are known to be associated with gold mining and processing except Cs, which may be uniquely present in the local ore (Abdul‐Wahab & Marikar, 2012; Jasiak et al., 2021; Khamkhash et al., 2017; Malm et al., 1995; Winde et al., 2019; Zupunski et al., 2023). Similarly, the reference sites unexpectedly displayed significantly higher levels of Ba (W = 894, p = 0.0013), Sr (W = 1047, p < 0.0001), and heavy metals such as Sn (W = 952, p = 0.0001), Cd (W = 817, p = 0.0202), Sb (W = 1001, p < 0.0001), Tl (W = 881, p = 0.0012), and As (W = 828, p = 0.0010). Free from local gold extraction, the comparatively higher levels of Cd, Ba, Sr, and As may be associated with natural deposition or distant oil activities at these reference sites (Figure 2; Abdul‐Wahab & Marikar, 2012; Jasiak et al., 2021; Khamkhash et al., 2017; Malm et al., 1995; Winde et al., 2019; Zupunski et al., 2023).

DISCUSSION

Results show significantly higher levels of several metals in the feathers of a common nonmigratory understory bird residing throughout protected areas in Ecuador and French Guiana subjected to current and historical oil and gold mining practices (Figure 2). In comparing Andean reference sites to protected sites in Ecuador near oil mining, we found elevated concentrations of Ba and heavy metals with known associations to oil extraction processes. Similarly, Se and all but one of the heavy metals with elevated concentrations from protected sites in French Guiana near gold mining are associated with gold extraction. However, we also uncovered higher concentrations of several metals in the reference sites, suggesting a background of natural geologic deposition or possibly a larger geographic footprint of distant oil mining activities than previously expected.

Understanding the ecological sources and impact of this heavy metal contamination will require further interdisciplinary research involving information on the geologic resources and history of each region, as well as fine‐scale details on the extent and methods for oil and gold extraction at each site (Mohammed et al., 2011). A consensus of this information can inform novel models of heavy metal mobilization and accumulation in complex yet understudied tropical systems (Burger, 1997; Sayers et al., 2023). These dynamic models can uncover the profiles and range of contamination, identifying the sources of elevated heavy metals in the Andean reference sites and clarifying the limitations of comparisons to heavy metal concentrations from gold mining in French Guiana (Burger, 1997; Sayers et al., 2023). For example, concentrated heavy metals from the local ore and extractive reagents often leech into the soil and underground water systems from the large quantities of wastewater produced during oil and gold extraction (Kimerling, 1990; Pellegrini et al., 2020; Zhang & Wang, 2020). This contamination can be amplified by the toxic plume of smoke and acid rain resulting from the combustion of oil spills and gas flares (Durango‐Cordero et al., 2018). Because heavy metals cannot be naturally degraded, their dynamic movement and availability in the environment can depend on a variety of ecological and biological factors, such as microclimate, habitat structure, chemical biotransformation by microorganisms or endogenous enzymes, and biomagnification in organisms within different ecological niches and across feeding guilds (Durango‐Cordero et al., 2019; Mohammed et al., 2011). These factors can ultimately influence the concentration of heavy metals in the environment, as well as the absorption, accumulation, and bioavailability of heavy metals in the tissues of exposed organisms (Sayers et al., 2023; Sharma & Agrawal, 2005).

The functional significance of elevated heavy metal concentrations in the birds of the present study can only be broadly deduced from previous analyses of the biochemical uptake and multisystem effects of these pollutants in other species. In general, heavy metals at high concentrations have been found to competitively bind to macromolecules important for metabolism and homeostasis, often leading to neurological pathologies in response to chronic low‐level toxicity (Sharma & Agrawal, 2005). In addition, exposure to a toxic mixture of multiple heavy metals in higher quantities can cause synergistic dysfunction and other diseases like cancers (Sebastián et al., 2001). However, toxicity thresholds for each metal varies across species, life history, and tissue, making it difficult for comparative studies in wildlife (Abdullah et al., 2015; Burger & Gochfeld, 2000; Sayers et al., 2023). Metal concentrations documented in wild insectivorous birds have characterized baseline concentrations of Cd (0.1–2 μg/g), Ni (0.31–2.7 μg/g), Ba (3.910 μg/g), Mn (43.8 μg/g), Cu (6.47–54 μg/g), and Zn (119.5–264 μg/g) well below the concentrations reported in the test sites of Tiputini and Nouragues of the present study (Burger & Gochfeld, 1997, 2000; Durkalec et al., 2022; Janssens et al., 2001). Yet, thresholds for toxicity and biological effects should be independently studied in this system since extrapolation across multidimensional geographic, ecological, and biological factors is prone to inaccuracy, particularly in lesser studied tropical systems (Burger & Gochfeld, 1997; Sayers et al., 2023; US Environmental Protection Agency, 2007).

As the first study of its kind in this region and species, the varying concentrations of heavy metals in feathers sampled across this landscape of exposure provide a baseline for subsequent analyses on the effects and impact on organisms and their environment. Future research further utilizing historical collections and the addition of novel biological and environmental sampling can help to uncover the direct sources and extent of exposure, the ecological mechanisms governing past and future contamination dynamics, and the biological consequences in wildlife and human communities. Ultimately, this information will be leveraged to better monitor and preserve natural areas from the cascading impacts of heavy metal exposure on the ecosystem.

Supporting Information

The Supporting Information is available on the Wiley Online Library at https://doi.org/10.1002/etc.5984.

Acknowledgments

Thomas B. Smith provided funding from the University of California's Center for Tropical Research for feather preparation, lab work, and field sample collection in Ecuador. Borja Milá provided funding for lab work from Museo Nacional de Ciencias Naturales, Consejo Superior de Investigaciones Científicas; and fieldwork in French Guiana was funded through a Nouragues Travel Grant from the Centre National de la Recherche Scientifique (CNRS, France). For help with field sampling in Ecuador, we thank G. Castañeda, J. Chaves, J. F. Freile, T. Santander, B. T. Ryder, S. Tomassi, and J. McCormack. For field sampling in French Guiana, we thank CNRS Guyane, P. Gaucher, C. Thébaud, and I. de la Hera. We acknowledge the Global Energy Monitor organization for access to the global gas and oil network database for Latin America (Portal Energético para América Latina), the Dartmouth Trace Element Analysis Lab for data collection, and J. Devillechabrolle (Conservatrice de la Réserve Naturelle des Nouragues) for providing comments on the manuscript.

Conflict of Interest

The authors declare no conflict of interest.

Author Contribution Statement

Yeraldi Loera: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Project administration; Visualization; Writing—original draft; Writing—review & editing. Cristian Gruppi: Data curation; Investigation; Methodology; Project administration; Resources; Writing—review & editing. Kelly Swing: Investigation; Resources; Validation; Writing—review & editing. Shane C. Campbell‐Staton: Formal analysis; Software; Supervision; Visualization; Writing—review & editing. Borja Milá, Thomas B. Smith: Conceptualization; Data curation; Formal analysis; Funding acquisition; Investigation; Methodology; Project administration; Resources; Supervision; Validation; Writing—review & editing.

Data Availability Statement

All raw data files and code can be found in the open‐access Harvard Dataverse repository: Loera, Yeraldi, 2024, “Heavy metal contamination in birds from protected regions in the Amazon Data,” https://doi.org/10.7910/DVN/OEE6BJ, Harvard Dataverse, V1.

REFERENCES

Author notes