Introduction

Mercury (Hg) and methylmercury (MeHg) are highly neurotoxic substances that are widely distributed across the globe1,2,3. Since the Industrial Revolution, the background concentration of Hg in the environment has increased 3–5 times, and the MeHg concentration in Atlantic cod surged by as much as 23% between the 1970s and 2000s4. Additionally, overfishing driven by dietary changes has exacerbated this issue. Human exposure to Hg and MeHg is linked to long-term neurocognitive deficits, with significant impacts on children that persist into adulthood. The global economic burden of such exposure is estimated to exceed US$1,173 billion5,6. Given the continuous rise in Hg and MeHg levels, this problem is expected to escalate further in the future.

Rivers are recognized as hotspots for Hg transportation, methylation, and bioaccumulation and have consequently drawn significant global attention. Dam construction in rivers introduces several environmental changes, including hydrodynamic reduction, substance interception, and the creation of anoxic conditions. These alterations influence the spatial distribution and methylation of Hg7,8. For instance, studies have reported that MeHg production in reservoirs can lead to a tenfold increase in riverine MeHg levels and a 2.6-fold increase in concentrations within estuarine surface waters9. Sediment is also considered a crucial reservoir for Hg10, but the fate of Hg within sediments is highly complex and influenced by a variety of factors11,12. For example, Azaroff et al. (2019) found that high Hg content is often associated with the accumulation of organic matter and fine sediment particles13. Similarly, Kelly et al. (2018) demonstrated that Hg is primarily transported via fine particles, which subsequently settle in low-energy environments, leading to elevated Hg concentrations14. Liem-Nguyen et al. (2016) observed that the deposition of autochthonous organic matter in sediment zones enhances MeHg production15. Furthermore, Diez et al. (2016) highlighted the role of Hg methylation occurring in suspended particles and organic matter within the water column of Lake Geneva16.

Hg distribution and methylation in static reservoir systems have been extensively studied17. However, the effects of reservoir regulation on Hg distribution and methylation remain poorly understood18,19. Reservoir regulation, as an environmental factor, may further influence the fate of Hg by altering sediment activity. Specifically, the mobility of Hg in sediments plays a critical role in determining its bioavailability within aquatic ecosystems20,21. During reservoir regulation, sediment particle transport occurs, and variations in mobile sediment particle Hg and organic carbon (OC) levels may affect Hg methylation processes22,23. Vertical profiles of Hg species in sediments are easily disturbed by reservoir regulation and can exhibit significant variability within just a few millimeters24,25. Consequently, high spatial resolution is essential for understanding the Hg biogeochemical cycle during reservoir regulation processes 26. Sediment transport in reservoirs typically involves the mixing of bottom sediments with the overlying water column, which is often anoxic, within a depth of several centimeters27,28. This mixing is primarily driven by hydrodynamic disturbances, such as reservoir regulation and wave activity, which can result in several adverse environmental effects. On one hand, Hg redistribution during sediment transport is influenced by the competition between particles of varying grain sizes under hydrodynamic disturbances14,22. On the other hand, OC is repeatedly incorporated into sediment particles during transport29,30. Therefore, sediment transport during reservoir regulation can significantly impact Hg redistribution and methylation within sediment particles.

To test this hypothesis, we investigated the spatial patterns of sediment Hg and MeHg in both horizontal and vertical profiles using the diffusive gradients in thin films (DGT) technique in the Nuozhadu (NZD) Reservoir. The primary goals of this study were to (1) examine the vertical and spatial distributions of Hg species in sediment profiles and (2) clarify the transport potential of sediment-bound Hg and assess sedimentation processes under the influence of reservoir regulation.

Materials and methods

Study area

The Lancang River, located upstream of the Mekong River, flows through Myanmar after traversing Yunnan Province in Southwest China. The river has an annual mean discharge of 2,180 m3/s and a catchment area of 167,487 km2.As shown in Fig. 1, the Nuozhadu (NZD) Reservoir, one of the largest dams in the upper Mekong River, spans a 261.5 km stretch of the Lancang River’s main channel. It has a storage capacity of 23.7 × 109 m3 and a total catchment area of 1.5 × 105 km2.

Fig. 1
figure 1

Locations of the sampling sites and dams within the Nuozhadu (NZD) Reservoir.

Full size image

The NZD Reservoir is part of a cascade system, with the Dachaoshan (DCS) Reservoir located upstream and the Jinghong (JH) Reservoir downstream. The region’s climate is classified as temperate or subtropical monsoon, with annual temperatures ranging from 12 °C to 27 °C. The mean annual precipitation is approximately 1,399.6 mm, of which 80% occurs during the rainy season from May to October.

Sample collection

The field survey was conducted in October 2018 across an approximately 230 km valley of the Lancang River. Surface sediment samples were obtained at 10 sites from the reservoir tail to the dam front of the NZD Reservoir at a spacing of approximately 25 km. This arrangement accounts for the spatial heterogeneity in hydrodynamic and sedimentation conditions (Fig. 1, red circles). The surface sediment samples, denoted as S1 to S10, were collected with an Ekman grab sampler. At each sampling site, samples were collected in triplicate after complete homogenization. Overall, the 200.0 g sediment samples were stored in 50.0 mL tubes and frozen in the dark until Hg, MeHg, OC and grain size analyses were conducted. Separate 5.0 g sediment samples were stored in 5.0 mL Eppendorf (EP) tubes for hgcAB analysis.

Notably, sediment cores (φ12cm × 90 cm) were extracted with a Uwitec gravity corer near the central river at the reservoir tail, reservoir middle, and dam front sites of the NZD Reservoir in April 2019 (Fig. 1, yellow circles). The sediment cores were sealed with the intention of preventing interference by air and oxygen. The sediment cores were subsequently maintained upright, delivered to the laboratory and stabilized immediately for at least 24 h.

The DGT probes (Easy Sensor Ltd Nanjing, China) included a plastic base, 3-mercaptopropyl-functionalized silica resin (3 M, USA), 1.5% agarose gel as the diffusive layer, a plastic cover plate and a 0.13 mm-thick cellulose nitrate filter membrane (Whatman, 0.45 mm pore size). The reactivity of 3-mercaptopropyl-functionalized silica gel is 79%. Before deployment, the DGT probes were deoxygenated with N2 in Milli-Q water for 1 day. Notably, two DGT probes for Hg and MeHg were pushed into each sediment core smoothly and gently until the mark matched the sediment–water interface position. The displaced push-down water was obtained and preserved at a height of approximately 5.0 cm. The DGT probes were laid out for 24 h in the dark at each sample site. During deployment, the pH varied between 6.6 and 7.3, with the temperature remaining stable at 19.6 °C ± 0.8 °C, which suggests steady environmental conditions. Following retrieval, the probes were washed with Milli-Q water. Each resin gel was then segregated into 2.0 mm (vertical) × 6.0 mm (horizontal) slides in a clean laboratory for Hg and MeHg analysis.

The DGT technique was used to measure the DGT-Hg concentrations in both porewater and sediment. The DGT probes were deployed into the sediment cores, and the displaced push-down water was collected and preserved. The DGT-Hg concentration in porewater was determined by analyzing the resin gel exposed to the porewater, while the DGT-Hg concentration in sediment was determined by analyzing the resin gel exposed to the sediment matrix. This allowed us to distinguish between the mercury species present in the porewater and those adsorbed onto the sediment particles.

Sample analysis

All the sediments were dried at -60 °C for 7 days in the dark to remove water with a freeze dryer (Christ ALPHA 2–4 LD plus, Germany). A portion of the dried sediment was passed through a 0.15-mm mesh prior to Hg, MeHg and OC analysis. Hg in the sediment and resin gel was analyzed with a direct mercury analyzer (DMA-80, Milestone, US). The MeHg concentrations in the sediment and resin gel were obtained with a methylmercury analyzer (Tekran 2600, Tekran Inc., Canada).

For the analysis of OC, Hg, and MeHg, comprehensive quality assurance measures were employed. For OC measurement, the average recovery was 90% ± 6%, the relative percent difference (RPD) was less than 8%, and the method detection limit (MDL) was 0.1 mg g-1. For Hg and MeHg analysis, quality assurance was implemented using method blanks, blank spikes at five-sample intervals, matrix spikes, and duplicate analyses. Standard solutions for Hg (GBW07423-GSS-9) and MeHg (ERM-CC580) covered the full sample concentration range, with detection limits of 0.002 and 0.001 ng g-1, respectively. Blank spike recoveries were 88%-100% for Hg and 90%-107% for MeHg; matrix spike recoveries were 82%-95% for Hg and 84%-100% for MeHg. Duplicate analysis showed average relative standard deviations of 1.6%-5.9% for Hg and 0.8%-6.3% for MeHg. Labile Hg and MeHg in DGT extracts were analyzed by CVAFS, and similar quality assurance methods were applied for DGT analysis, with blank spike recoveries of 92%-99% for Hg and 91%-107% for MeHg, and matrix spike recoveries of 87%-94% for Hg and 89%-97% for MeHg. These quality control metrics confirm the reliability and accuracy of the analytical results.

The total OC content in the sediment was analyzed with an elemental profiler (Vario MACRO cube, Elementar Inc., Germany). The size of the sediment particles was determined via a Mastersizer 2000 laser grain size meter (clay: < 4.0 μm m; silt: 4.0–64.0 μm; sand: > 64.0 μm). Following grain size analysis, the sediment samples were distributed via an ultrasonic oscillator (PS-60A, Shenzhun Shenghuatai ultrasonic equipment Co., Ltd., China).

The quantitative polymerase chain reaction (qPCR) technique was employed to quantify the microbes involved in Hg methylation in the sediment samples. Several functional genes control the Hg methylation process, of which hgcAB in sulfate-reducing microorganisms is the most common biomarker for Hg methylation and was selected for qPCR analysis in this study. DNA extraction was performed with Powersoil DNA Isolation Kits following the manufacturer’s guidelines. qPCR analysis was performed with primers targeting hgcAB. Gene copies were enlarged and quantified within a Bio-Rad cycler containing the iQ5 real-time fluorescence detection mechanism and software (v 2.0; Bio-Rad). The total reactions were finished within a total volume of 20 μL and included 10 μL of SYBR® Premix Ex TaqTM (TOYOBO, Japan), 0.8 μL of BSA (3 mg mL–1, Sigma) (each primer: 0.5 mM), dual-distilled water and template DNA. For hgcA, the following qPCR program was adopted: 95 °C for 3 min; 95 °C for 15 s; 65 °C for 20 s (40 times); and extension for 4 min at 72 °C. A standard curve was generated via serial dilution (10–2 ~ 10–8) of the identified density plasmid DNA with target fragments. The total PCRs were performed in triplicate in 96-well plates (Bio-Rad) sealed with optical-quality sealing tape (Bio-Rad). In the absence of a DNA template, negative controls were included for each PCR operation.

Data processing

The DGT-Hg and MeHg contents in the sediment column were determined as Eq. 1:

$$C_{DGT} = \frac{M \times \Delta g}{{D \times A \times t}}$$
(1)

where M denotes the mass of Hg and MeHg accumulated on the resin gel, △g denotes the diffusive layer thickness, D denotes the diffusive coefficient of Hg of the gel, A denotes the region of the exposed surface, and t denotes the exposure time.

When applying Eq. (1) to the sediment column, it should be acknowledged that the concentration at the diffusive gel-water sediment solution interface may decrease over time due to the lack of stirring. However, the DGT technique is designed to estimate the diffusive flux of Hg under static conditions. The deployment time of the DGT probes (24 h) was optimized based on previous studies9 to ensure that the concentration changes during deployment were within a controllable range. This allows for the reliable estimation of the diffusive flux of Hg in the sediment column.

The diffusive flux (J) is the sum of the fluxes of sediment (Js) and overlying water (Jw). In accordance with Fick’s law, the diffusive flux of DGT-Hg was determined as Eq. 2:

$$J = J_{w} + J_{s} = - D_{w} \left( {\frac{\delta CDGT}{{\delta xw}}} \right)\left( {x = 0} \right) = \varphi D_{s} \left( {\frac{\delta CDGT}{{\delta xs}}} \right)\left( {x = 0} \right)$$
(2)

where Jw is the DGT-Hg flux between water and the sediment water interface, Js is the DGT-Hg flux from the sediment to the sediment water interface, and Ds and Dw denote the diffusive coefficients of Hg of the porewater and natural water, respectively. In accordance with the research of Gao et al. (2018), a sediment porosity (φ) of 0.9 was selected31. Moreover, (δCDGT/δxs)(x = 0) and (δCDGT/δxw)(x = 0) denote the DGT-Hg content gradients within the sediment and overlying water columns, respectively. Notably, gradients within 1 cm between the sediment and water interface were determined.

When deploying the DGT, two identical DGT were deployed in each sediment column, and the results presented are the average values measured by the two DGT plates. All analyses were conducted in SPSS 22.0, and any differences were considered statistically significant at p < 0.05.

Results

Sediment Hg in the NZD reservoir

In the NZD Reservoir, sediment Hg content exhibited a spatial pattern, first increasing and then slightly decreasing from the tail to the dam front, while sediment grain size (d50) showed a gradual decrease (Fig. 2A). In the reservoir tail area, sediment Hg content remained relatively stable. However, it increased from 77.12 ng g⁻1 at site S4 to 114.14 ng g⁻1 at site S8 in the reservoir middle area before decreasing to 85.68 ng g⁻1 at site S10 in the dam-front area. Similarly, sediment grain size (d50) remained unchanged in the reservoir tail area but decreased from 7.61 μm at site S4 to a minimum of 5.01 μm at site S8 in the reservoir middle area, followed by an increase to 6.33 μm at site S10 in the dam-front area. Grain size analysis revealed that silt dominated the sediments in the NZD Reservoir, with clay, silt, and sand accounting for 32.09%, 64.18%, and 3.73%, respectively. The proportion of clay in the sediments showed minimal variation in the reservoir tail area but increased slightly from 27.82% at site S4 to a maximum of 38.21% at site S8 in the reservoir middle area, before decreasing to 33.15% at site S10 in the dam-front area. Correlation analysis indicated that sediment Hg content was positively correlated with clay content (r2 = 0.66, p < 0.05) and negatively correlated with silt and sand contents (Fig. 2B).

Fig. 2
figure 2

Spatial pattern of the sediment Hg content and grain sizes in the NZD Reservoir. (A) Spatial modes of the sediment Hg content and different grain sizes in the NZD Reservoir. (B). Correlations between the sediment Hg content and the sand, silt, and clay contents in sediments. A grain size < 4 μm is classified as clay, a size of 4–64 μm is classified as silt, and a size > 64 μm is classified as sand. The error bars denote the standard deviations (DSs).

Full size image

In contrast to the reservoir tail and dam-front sediment columns, the reservoir middle sediment column exhibited greater spatial heterogeneity in DGT-Hg contents. The mean (± SD) DGT-Hg contents in porewater at the reservoir tail, reservoir middle, and dam-front sampling sites were 29.14 ± 3.93 ng L⁻1, 33.87 ± 8.68 ng L⁻1, and 29.66 ± 2.27 ng L⁻1, respectively. Similarly, the DGT-Hg contents in sediment at the reservoir tail, reservoir middle, and dam front sites were 33.99 ± 1.51 ng L⁻1, 34.87 ± 5.10 ng L⁻1, and 15.55 ± 2.72 ng L⁻1, respectively. The DGT-Hg contents at the reservoir tail and reservoir middle sites were more than double those at the dam-front site. The net Hg flux from sediment to the overlying water varied across sites, reaching 7.36, 3.25, and 3.69 ng d⁻1 cm⁻2 at the dam front, reservoir middle, and reservoir tail sites, respectively (Fig. 3B). Notably, the Hg flux at the reservoir tail site was approximately twice as high as that at the reservoir middle and dam front sites.

Fig. 3
figure 3

Sediment DGT-Hg content and Hg diffusion flux in the NZD Reservoir. (A) Sediment DGT-Hg content. (B) Diffusive flux of DGT-Hg throughout the sediment–water interface.

Full size image

Sediment MeHg and MeHg/Hg ratio in the NZD reservoir

In the NZD Reservoir, sediment MeHg and %MeHg values were lower at the reservoir tail but increased from 0.41 ng g⁻1 and 0.53% at site S4 to 0.68 ng g⁻1 and 0.74% at site S7, respectively. These values then slightly decreased to 0.44 ng g⁻1 and 0.60%, respectively, at the dam front (Fig. 4A). In contrast, sediment organic carbon (OC) content increased from 11.26 mg g⁻1 at site S1 in the reservoir tail to a maximum of 25.17 mg g⁻1 at site S8 in the reservoir middle, before slightly decreasing to 24.30 mg g⁻1 at the dam front. Additionally, the abundance of hgcAB genes, associated with Deltaproteobacteria, generally increased along the reservoir, from 19.14 × 105 gene copies per gram of sediment at the reservoir tail to 48.24 × 105 gene copies per gram at the dam front (Fig. 4B).

Fig. 4
figure 4

Spatial patterns of MeHg and %MeHg within the NZD Reservoir. (A) Sediment MeHg and %MeHg contents in the sediment of the NZD Reservoir. (B) Spatial patterns of hgcAB genes associated with Deltaproteobacteria abundance and organic carbon in the sediment of the NZD Reservoir. The error bars denote the DSs.

Full size image

The spatial variations in DGT-MeHg concentrations during the first sediment layer (0–60 mm depth) were significantly greater than those in the deeper layers and decreased rapidly below a depth of 60 mm (Fig. 5). Mean (± SD) DGT-MeHg concentrations in porewater during the first 60 mm were 0.25 ± 0.17 ng L-1, 0.59 ± 0.22 ng L-1, and 0.17 ± 0.14 ng L-1 at the reservoir tail, reservoir middle, and dam front sites, respectively.

Fig. 5
figure 5

DGT-MeHg concentration and %MeHg level in the sediment with depth calculated from the DGT measurements. (A) DGT-MeHg. (B) %MeHg. The sediment–water interface is shown by the dotted line at the zero depth.

Full size image

In the reservoir tail sediment column, DGT-MeHg concentrations slightly increased near the sediment water interface at an 8 mm depth, followed by a sharp decrease at the 10 mm depth. This trend repeated at depths of 10–16 mm, 18–24 mm, 42–52 mm, 54–88 mm, and 90–100 mm. In contrast, the dam front sediment column exhibited an opposite trend, with intermittent sharp increases followed by decreases near the bottom. Specifically, DGT-MeHg concentrations sharply increased at the 10-mm depth, slightly decreased until the 30-mm depth, and repeated this pattern at depths of 32–54 mm, 56–94 mm, and 96–100 mm . Meanwhile, the reservoir middle sediment column showed continuous fluctuations in DGT-MeHg concentrations throughout the profile.

The DGT-MeHg concentrations in the overlying water column were 0.05 ± 0.03 ng L-1, 0.55 ± 0.18 ng L-1, and 0.04 ± 0.02 ng L-1 at the reservoir tail, reservoir middle, and dam front sampling sites, respectively. Notably, the DGT-MeHg concentration at the reservoir middle site was more than 11 times higher than those at the dam front and reservoir tail sites. However, due to the insufficient linearity of the DGT-MeHg concentrations in the overlying water, the MeHg diffusion flux could not be calculated.

Discussion

Both Hg and MeHg in reservoir sediments are determined by hydrodynamics

The Hg content in the sediment of the NZD Reservoir was high at the middle of the reservoir, where fine particles settled because of the weak hydrodynamics. The sediment Hg content decreased in the reservoir tail and dam-front areas, where coarse-grained sediment is enriched. Hence, at the basin scale, the spatial variation in Hg levels was affected by hydrodynamics and grain size variables (Fig. 2A).

The NZD Reservoir is located in an environment with limited anthropogenic disturbance, in which sediment Hg is derived mainly from nonpoint sources. With dam construction in rivers, the flow velocity decreases, and Hg adsorbed onto particulates eventually settles onto the sediment32. However, hydrodynamic factors influence the geographic pattern of the Hg content13,14. During hydrodynamic reinforcement, sediment Hg can be released into the overlying water column and redistributed via competition between grain particles of different sizes20,22. In the resettlement process, coarse particles resorb limited Hg and reach a steady state relatively quickly, whereas fine particles may resorb more Hg but be transported across greater distances or become deposited in nearby areas after a prolonged period to reach a steady state in the unsteady flow process of the reservoir13,33. This process can function as a “sieve” for size grouped particles, thus affecting the spatial pattern of the Hg content in sediment. Hence, a remarkable positive relationship was observed between the sediment Hg and clay contents but not with the silt and sand contents (Fig. 2B). The decrease in the sediment Hg content between the reservoir tail and dam front can be attributed to water released from upstream via discharge and the dam outlet, which makes it comparatively difficult for fine sediment particles to settle in these areas.

Inside the NZD Reservoir, the sediment MeHg content increased from the reservoir tail to the reservoir middle and subsequently decreased to the dam front, exhibiting spatial patterns similar to those of the %MeHg and Hg contents in the sediment. However, the abundance of hgcAB genes and the OC content in the sediment increased between the reservoir tail and the dam front, exhibiting different spatial patterns from those of the MeHg and %MeHg contents in the sediment (Fig. 4).

Hg methylation might have occurred during the transport of mobile sediment particles and their redeposition in the area with weak hydrodynamics because fresh carbon was continuously reloaded in the water column, thereby providing a microanoxic environment for mobile sediment particles, which is ideal for Hg methylation16,22,34. Therefore, MeHg in the sediment comprises two parts, namely, MeHg from in situ production and MeHg from offsite accumulation. In the NZD Reservoir, the abundance of hgcAB and the OC content in the sediment increased between the reservoir tail and the dam front, indicating that MeHg production increased (Fig. 4). However, MeHg and %MeHg in the sediment were dominated by accumulated Hg. This finding indicates that mobile sediment particles are an important source of MeHg in sediments. Thus, the spatial distributions of both Hg and MeHg in sediments are determined by hydrodynamics.

Deposition and risk of exposure to Hg and MeHg during reservoir operation

The sediment column samples from different depths provide historical reservoir operation information35,36. Both DGT-Hg and DGT-MeHg exhibited periodic fluctuations along the vertical direction, but the fluctuations in DGT-MeHg were greater, indicating that the Hg methylation intensity varied during reservoir operation. Moreover, DGT-MeHg occurred at the middle and in the tail of the reservoir, with the first 60 mm exhibiting the largest spatial differences.

In this study, the DGT-Hg distribution in the sediment column of the NZD Reservoir can be clarified on the basis of the behavior of mobile sediment particles sorted by hydrodynamics37. At the reservoir tail and dam front sites of the NZD Reservoir, particles were carried out because of the changing hydrodynamic conditions due to the discharge of the upstream DCS Reservoir and the dam outlet of the NZD Reservoir, respectively. Hence, the DGT-Hg content in the sediment column exhibited relatively high spatial heterogeneity. At the reservoir middle site of the NZD Reservoir, considerable amounts of fine particles were deposited due to the weak hydrodynamics. Thus, a low spatial heterogeneity in DGT-Hg was observed in the reservoir middle sediment column.

The data revealed that Hg methylation is closely related to the characteristics of particles because the DGT-MeHg concentration was positively correlated with the %MeHg concentration in the bulk sediment but not with the DGT-Hg content. The irregular distribution of DGT-MeHg in the overlying water column supports the possibility that Hg methylation may occur in oxygen deficient zones of stratified water columns, whereas the grain size “sieve” phenomenon during reservoir regulation further affects the changes in DGT-MeHg in sediments. Moreover, the discontinuous change in the DGT-MeHg distribution within the sediment vertical profiles indicates that the DGT-MeHg exposure risk is influenced by human activities. The dam front zone is generally the deepest area of the reservoir, and it is particularly prone to particle resuspension and resettlement because of the operation height of turbines, which may prevent fine particle transport downstream. Thus, the DGT-MeHg concentration at the dam front sharply increased intermittently and then decreased at the bottom during reservoir regulation. However, in the reservoir tail sediment column, ripple transport caused the preferential burial of fine particles. Thus, in the reservoir tail sediment column, the DGT-MeHg concentration slightly increased from the sediment water interface to the bottom and then sharply decreased. Completely opposite characteristics of the vertical distribution were recorded compared with those at the dam front site. Compared with nonmobile sediment particles, mobile sediment particles exposed to water usually better reflect the MeHg exposure risk. Thus, the potential ecological implication for this finding is that the risk of exposure to MeHg increases during reservoir regulation.

Moreover, the DGT-MeHg concentrations during the first period (0–60 mm) were significantly greater than those during the second period but decreased rapidly below 60 mm. Two main mechanisms may explain this phenomenon. One hypothesis states that seasonal MeHg variations govern the process. For example, high seasonal variations in the reservoir sediment MeHg content have been obtained at most sites in the Northern Hemisphere, with minima in winter and peaks in summer19,38. Another hypothesis is that DGT-MeHg desorption is related to organic matter mineralization advancing more slowly since organic matter in deeper sediment is more recalcitrant39,40. We can accept the second assumption, since seasonal MeHg changes commonly occur synchronously with Hg changes. In addition, a seasonal change was observed in the region, in which the change in the NZD Reservoir was very small. However, our data and information do not support the first hypothesis.

Risk of exposure to MeHg during reservoir regulation

Hg resuspension, desorption, resorption, and deposition in mobile sediment particles are ongoing processes that occur in balance41. Under the change in hydrodynamics caused by reservoir regulation, fine particles can carry more Hg substances than coarse particles can, are resuspended first, and experience a prolonged period of exposure to fresh OC in the water column during transport. However, coarse particles exhibit the opposite trend. We attributed the differences in %MeHg in the DGT probes to mobile sediment particles being driven during transport. Our data revealed that Hg methylation is closely related to the characteristics of particles because the DGT-MeHg concentration was positively correlated with the %MeHg concentration in the bulk sediment but not with the DGT-Hg content. The irregular distribution of DGT-MeHg in the overlying water column supports the possibility that Hg methylation may have occurred in oxygen-deficient zones of stratified water columns, whereas the grain size “sieve” phenomenon during reservoir regulation may have further affected the changes in DGT-MeHg in the sediment.

Moreover, the discontinuous change in the DGT-MeHg distribution within the sediment vertical profiles indicates that the DGT-MeHg exposure risk is influenced by human activities. The dam-front zone is generally the deepest area of the reservoir, and it is particularly prone to particle resuspension and resettlement because of the operation height of turbines, which may prevent fine particle transport downstream. Thus, the DGT-MeHg concentration at the dam front sharply increased intermittently and then decreased at the bottom during reservoir regulation. However, in the reservoir tail sediment column, ripple transport caused the preferential burial of fine particles. Thus, in the reservoir tail sediment column, the DGT-MeHg concentration slightly increased from the sediment water interface to the bottom and then sharply decreased. Completely opposite characteristics of the vertical distribution were recorded compared with those at the dam front site. Compared with nonmobile sediment particles, mobile sediment particles exposed to water usually better reflect the MeHg exposure risk. Thus, the potential ecological implication for this finding is that the risk of exposure to MeHg increases during reservoir regulation.

Risk of exposure to Hg and MeHg during reservoir regulation

In the NZD Reservoir, DGT-Hg was released from the sediments into the overlying water, which indicates that sediments increasingly became the source of DGT-Hg in the overlying water. Within the overlying water column, the density gradient of DGT-derived MeHg is not suitable for linear testing, so the flux cannot be calculated. However, the density of DGT-generated MeHg within the overlying water is greater than that in the sediment at the corresponding point. These findings indicate the risk of MeHg release from sediments. Furthermore, the sediment Hg content in front of the dam and at the tail of the reservoir decreased, but there was greater release of DGT-Hg, indicating that hydrodynamic enhancement promoted Hg release from the sediment. The strong reduction conditions at the tail of the reservoir further facilitate the release of DGT-Hg, resulting in significantly greater release of DGT-Hg at the reservoir tail than in the other regions. The risk of Hg and MeHg among reservoirs is usually determined by their concentration in the water body42,43. Therefore, when studying the hazards of Hg and MeHg in reservoirs, the Hg and MeHg contents in sediments can serve as a reference, and more importantly, attention should be given to the tail area of reservoirs with strong hydrodynamics and reducing conditions.

Conclusion

Under the control of reservoir regulation, active transfer of sediment particles occurs along a reservoir within a canyon river, which may influence the spatial patterns of Hg in sediment. This study revealed that reservoir transport can facilitate local storage and methylation of Hg above its source point. The main conclusions are as follows:

  1. (1)

    Hydrodynamics control Hg and MeHg distribution: The spatial distributions of both Hg and MeHg in the sediments of the NZD Reservoir are driven primarily by hydrodynamic conditions and the particle size. In fine particles, which settle in weak-hydrodynamics zones, relatively high Hg and MeHg levels can accumulate, whereas coarse particles, which are enriched in the reservoir tail and dam front, exhibit reduced Hg and MeHg contents due to their limited adsorption capacity.

  2. (2)

    Risk of exposure to Hg and MeHg during reservoir operation: The spatial distribution of MeHg reflects both in situ production and offsite accumulation, whereas mobile sediment particles contribute significantly to MeHg accumulation. The hydrodynamic changes induced by reservoir operation affect the flux of Hg in sediments, particularly in reservoir tail areas. These hydrodynamic fluctuations increase the Hg exposure risk.