Mercury contamination is an invisible threat to declining migratory shorebirds along the East Asian-Australasian Flyway

The East Asian-Australasian Flyway (EAAF) is one of the world’s most important migratory flyways, stretching from the Russian Arctic and Alaska to southern Australia and New Zealand, supporting approximately 8 million migratory shorebirds annually. Many species have experienced steep population declines, with over 20 species now globally threatened. Climate change and habitat loss or modification are major contributors, but environmental contamination is an additional threat because shorebirds rely heavily on wetlands, which are subject to trace elements and persistent organic pollutants. Even at remote northern breeding sites, volatile and semi-volatile pollutants are detected due to long-range atmospheric transport. Among pollutants, mercury (Hg) and its toxic forms can have severe impacts on birds, affecting foraging, migratory behaviour, endocrine function, and survival. Shorebirds are thought to be highly exposed to methylmercury through wetland food webs (e.g., via polychaetes) and because wetlands are contaminated through atmospheric deposition and runoff. The extreme migratory demands of some shorebirds (e.g., Bar-tailed Godwit) make them particularly vulnerable to added stressors. Large-scale assessment of Hg accumulation across the flyway is essential to identify contamination distribution and inform conservation strategies. Previous work has focused on breeding shorebirds in Alaska, northern Canada, and Greenland, with limited non-breeding studies and key stopover sites examined in other regions. The Asia-Pacific region accounts for ~49% of global Hg emissions, overlapping with many EAAF sites, yet Hg exposure along much of the flyway remains poorly characterized.

Prior studies have highlighted climate change and habitat loss as key drivers of shorebird declines and documented pollutant presence even at remote breeding sites through long-range transport (e.g., Alaska, Canadian Arctic). Mercury is a contaminant of particular concern for birds, with sublethal effects on foraging, migration, hormones, and survival demonstrated in multiple species. Polychaetes can vector methylmercury to shorebirds. Limited shorebird Hg studies exist from Alaska, northern Canada, Greenland, and select non-breeding or staging sites (e.g., Delaware Bay, France, Taiwan). The Asia-Pacific contributes ~49% of global Hg emissions, overlapping with EAAF migratory habitats. Recent studies in China have shown elevated heavy metals and methylmercury in estuaries and wetlands, and persistent organic pollutants and agricultural chemicals are widespread in coastal systems. Despite this, comprehensive, flyway-wide assessments integrating moulting ecology to assign exposure region have been lacking.

Study design and sites: Multiple field surveys were conducted across the EAAF between 2019 and 2022, spanning breeding (Arctic and temperate) and non-breeding regions (Yellow Sea, South China, Australia). Sampling sites included, among others, Yalujiang, Tiaozini, Chongming Dongtan, Yangkou, Mai Po (Hong Kong), Leizhou (Guangdong), and multiple sites in northern and southeastern Australia (e.g., 80 Mile Beach, Roebuck Bay, Corner Inlet, Port Phillip Bay, Limestone Coast sites). Species sampled included Common Redshank, Marsh Sandpiper, Black-tailed Godwit, and many others, totaling 33 species. Sampling and moult inference: Birds were captured with mist nets during non-breeding periods (January to early August) and additional feathers were collected from incidental mortalities/carcasses. Age classes (juvenile or adult) were assigned by capture location and plumage. Juvenile feathers (grown at breeding grounds) were used to represent Hg exposure at breeding sites; adult feathers (grown at non-breeding grounds) represented exposure at non-breeding/staging sites, unless active moult at capture indicated local growth. Target feather was the fourth primary covert (PCG), which moults with the 6th primary and reflects near-half depuration values for Hg; where necessary, the nearest most completely grown primary was collected. Feathers were stored in envelopes or zip-lock bags at room temperature. Trait data: For each species, habitat preference during non-breeding was categorized as coastal obligate (100% use) or non-coastal obligate (<100% use). Foraging stratum was classified by prevalence of foraging on/just below the water surface (surface <50% vs depth 50–79% vs 80–100%); diet was coarsely assigned as proportion of invertebrates based on literature. Mercury analysis: Whole feathers were rinsed with 1% acetone and deionised water, dried overnight, and analyzed using a Direct Mercury Analyzer (DMA-80; detection limit 0.005 ng) per US EPA Method 7473 at the State Key Laboratory of Environmental Geochemistry (Guiyang, China). Sample masses ranged 0.0010–0.0900 g. QA/QC included blanks, method blanks, certified reference materials (IAEA-086 human hair; DORM-4 fish protein), and duplicate feathers per batch (≤15 samples). Detection limit for feathers was 0.0008 mg/kg. Duplicate RPD was 12.13 ± 11.25% (n=66). Recoveries: IAEA-86, 93.40 ± 10.34% (n=40); DORM-4, 96.46 ± 5.83% (n=56). Risk assessment: Population-level risk categories followed prior Arctic shorebird Hg work, using bird blood toxicity benchmarks converted to feather-equivalent body Hg thresholds. Five categories: no apparent effect (<1.62 mg/kg), low (1.62–4.53 mg/kg), moderate (4.53–9.14 mg/kg), high (9.14–19.99 mg/kg), very high (≥20 mg/kg). Statistical analyses: THg in feathers was log-transformed (log2 for some analyses; log10 for mixed models). For individuals with clear life-stage inference (n=876), a linear mixed-effects model tested differences between breeding vs. non-breeding locations (fixed effect), with species and sampling year as random effects. For species with n≥15, one-way ANOVAs and estimated marginal means assessed interspecific differences. For individuals with clear moult origin (n=742), a candidate set of linear mixed-effects models examined fixed effects of Region, Habitat Preference, Diet, and Foraging Stratum, with species and year as random effects. Multicollinearity was evaluated via VIF (models retained with VIF<6). Models were ranked by AICc; ΔAICc≤2 identified top models, and unconditional model averaging estimated variable importance and coefficients. Estimated marginal means with 95% CIs were plotted. Analyses were performed in R with packages performance, ggplot2, lme4, MuMIn, and emmeans.

- Scope: 984 individuals from 33 shorebird species were analyzed across the full latitudinal span of the EAAF. - Overall THg: Mean 1.87 ± 2.27 mg/kg (median 1.28 mg/kg; range 0.11–38.63 mg/kg). - Breeding vs non-breeding exposure: Feathers from non-breeding grounds (n=428) had 2.09 ± 2.74 mg/kg (median 1.38; range 0.25–36.63), significantly higher than breeding-ground feathers from juveniles (n=448) at 1.51 ± 1.76 mg/kg (median 0.99; range 0.11–17.83); reported as nearly 14 times higher with F1,88 = 510.6, p < 0.0001. - Species variation: Large interspecific differences (e.g., Marsh Sandpiper 4.39 ± 5.35 mg/kg; median 3.40; range 0.35–38.63). Dunlin showed extensive risk: 93.24% of 148 samples within at-risk categories: 63.42% low, 8.54% moderate, 1.06% high, 1.68% very high. - Risk distribution: Over one-third of all samples exceeded toxicity benchmarks. Several Tringa species (e.g., Common Redshank, Marsh Sandpiper) showed elevated Hg in both juvenile (breeding) and adult (non-breeding) feathers, indicating year-round risk. Non-coastal obligate species frequenting freshwater habitats (e.g., Marsh Sandpiper, Common Redshank, Spotted Redshank, Pied Avocet, Black-tailed Godwit, Temminck's Stint, Sharp-tailed Sandpiper) may be at highest Hg exposure risk. - Spatial patterns (models): Region was the strongest predictor of feather THg. Model-averaged coefficients indicated higher Hg in Australia (estimate 0.16, p<0.01) and South China (0.24, p<0.01), with Yellow Sea elevated but marginal (0.13, p=0.07), relative to reference; temperate and Arctic regions were lower. Observed and modelled means both showed significantly higher Hg in South China vs Arctic and temperate (p<0.001). Some extremely high values in temperate samples (e.g., Common Redshank, Kentish Plover) influenced observed trends. - Other predictors: Habitat preference, foraging stratum, and diet showed weaker, non-significant effects after accounting for region in model averaging. - Hotspots: South China (e.g., Mai Po) and the Yellow Sea showed substantially higher feather Hg than breeding and Arctic regions.

This study provides the first intercontinental, flyway-wide assessment of Hg exposure in migratory shorebirds that explicitly integrates moult strategy to infer exposure regions. Findings demonstrate substantial Hg contamination across the EAAF, with more than a third of samples exceeding toxicity thresholds and clear spatial hotspots in South China and the Yellow Sea. Region emerged as the dominant predictor of feather Hg, indicating that local environmental contamination at moulting sites largely determines exposure. Elevated Hg in Tringa species (e.g., Common Redshank, Marsh Sandpiper) across both breeding and non-breeding seasons suggests persistent, year-round risk for some taxa, likely reflecting frequent use of contaminated freshwater and marsh/mangrove-edge habitats where methylmercury bioavailability can be high. The concurrence of Hg contamination with ongoing habitat loss and reduced food availability in critical migratory stopovers exacerbates conservation concerns. The results emphasize the need for targeted mitigation and biomonitoring at identified hotspots, and for evaluating additional contaminants (e.g., POPs, PCBs, PAHs, neonicotinoids) that may co-occur with Hg and interact to affect shorebird health and reproduction.

The study reveals pervasive and spatially heterogeneous Hg contamination in migratory shorebirds along the EAAF, with region of feather growth being the primary determinant of exposure. South China and the Yellow Sea are consistent hotspots, and non-coastal obligate species—particularly within the Tringa genus—appear most at risk. By leveraging moult ecology and a standardized feather-sampling protocol across nearly 1000 individuals and 33 species, this work establishes a robust baseline for flyway-wide contaminant monitoring. Conservation actions should prioritize pollution control and remediation in identified hotspots, alongside continued international collaboration to expand standardized biomonitoring. Future research should assess multi-contaminant exposures (e.g., POPs, industrial chemicals, pesticides), investigate mechanistic links between Hg and demographic performance, and refine species- and site-specific risk thresholds to guide management.

- Some individuals (n=108) lacked clear region-of-growth assignments and were omitted from regional modeling, potentially reducing power and representativeness. - Uneven sample sizes among species and regions and limited samples for certain taxa may constrain species-specific inferences and risk categorization. - Extreme values in a few temperate-origin samples (e.g., Common Redshank, Kentish Plover) influenced observed regional means. - Interaction effects among predictors were not evaluated (by study design), which may mask combined influences of habitat, diet, and foraging behavior. - Feather Hg reflects exposure during feather growth and does not capture temporal variation outside the moult window; reliance on moult inference may introduce misclassification in a subset of individuals with atypical moult timing. - Although QA/QC was rigorous, inherent variability in feather Hg (e.g., within-individual feather differences) and use of different feather types in a few cases may add measurement noise.