Introduction

The fission product 137Cs (t1/2 = 30.17 years) has been introduced into the marine environment by nuclear bombs, nuclear accidents, and nuclear power plants since the 1940s. The estimated 137Cs inventory of the global ocean is ~603 PBq (1 PBq=1015 Bq), accounting for over 60% of the total global fallout1. Additionally, 137Cs has high radiotoxicity and is easily absorbed by muscular tissues. Therefore, the environmental impact of 137Cs has become of great concern to the society since the Fukushima Dai–ichi Nuclear Power Plant (FDNPP) accident, which occurred in 20112.

137Cs is an excellent tracer that is widely applied to study environmental processes. For example, 137Cs is used to study currents, circulation and water mass movement in the oceanic environment3,4. The profile of 137Cs in sediment cores is typically used to reconstruct its depositional history5,6. In the terrestrial environment, 137Cs is an effective tracer in soil erosion studies7,8,9,10. In past decades, 137Cs has been widely investigated in order to evaluate its environmental behavior and impact3,4,11,12,13,14.

Accurate 137Cs measurement is very important for 137Cs related studies. The analytical tool conventionally used to measure 137Cs activity in environmental samples is γ–spectrometry. However, γ–spectrometry has several notable disadvantages, such as requiring a large–volume sample (e.g., ~100 L seawater), laborious pretreatment and a long counting time15. Therefore, compared to ordinary marine measurements, such as temperature, salinity and nutrients, very limited marine 137Cs datasets are available.

The present study focuses on the China Seas, which are located in the western North Pacific (WNP) and include the East China Sea (ECS), South China Sea (SCS), Yellow Sea (YS) and Bohai Sea. The China Seas interact intensely with the surrounding lands and are influenced by nutrients and terrestrial particulate matters discharged from large rivers: e.g., the Yangtze River, the Pearl River, and the Yellow River. Additionally, the China Seas have very complicated circulation patterns due to the influence of the East Asian Monsoon. For example, the surface current in the SCS is an anti–cyclonic gyre in summer while the weak southwest monsoon prevails16,17. In contrast, its circulation pattern is a cyclonic gyre in winter while the strong northeast monsoon prevails16,17. The Taiwan Current flows northward in summer, and the ECS coastal current flows southward in winter18 (Fig. 1). Additionally, the Kuroshio Current, a bifurcation of the North Equatorial Current (NEC), is an important control on the dynamic exchange between the China Seas and the WNP19. Across the Luzon Strait, the Kuroshio current exhibits a seasonal pattern, with a weaker intrusion into the SCS in summer than in winter20. The Kuroshio flows northeastward along Taiwan’s eastern coast throughout the year, with a branch entering the ECS at the northeast corner of Taiwan while the mainstream continues to flow northeast to the eastern Japanese coast21 (Fig. 1). In the YS, the Yellow Sea Cold Water Mass, which is the cold water located below the seasonal thermocline from spring to autumn, and tides, which are mixed diurnal or semi–diurnal22,23, are the most important hydrographic features22.

Figure 1
figure 1

Maps of: (a) seawater and sediment sampling sites in the China Seas. The red circles and blue squares represent surface seawater and surface sediment sample collection sites, respectively. The sampling locations of surface seawater and surface sediment are cited from previous studies (East China Sea5,13,24,25,26,31,44,45,49; South China Sea6,12,13,28,30,39,76; Yellow Sea26,27,29). (b) a schematic chart of the China Seas surface current: the Kuroshio, and the North Equatorial Current (NEC), represented by the blue dotted lines. In the South China Sea (SCS), the red dashed and black solid lines indicate the anticyclonic gyre in summer and the cyclonic gyre in winter, respectively. In the East China Sea (ECS), the red dashed and black solid lines represent the coastal currents in summer and winter, respectively. We modified this pattern of ocean circluation according to Hu et al.77. Acronyms include: BSCC: Bohai Sea Coastal Current, CDW: Changjiang Diluted Water, KC: Kuroshio Current, MZCC: Min–Zhe Coastal Current, MC: Mindanao Current, NEC: North Equatoral Current, TWC: Taiwan Warm Current, YSWC: Yellow Sea Warm Current. Figure (a) was prepared with Golden software surfer 10.0 (10.0 version, https://www.goldensoftware.com). Figure (b) was drawn using the free software Ocean Data View (ODV 5.1.2) (Schlitzer, R., Ocean Data View, https://odv.awi.de, 2018).

Full size image

To our knowledge, reports on the spatiotemporal distribution of 137Cs and its influencing factors in the China Seas as a whole are scarce, with most publications focusing on localized areas. For instance, the activity level and distribution of 137Cs in ECS seawater were investigated after the FDNPP accident24,25. Zhang et al.26 discussed the source and budget of 137Cs in the ECS covering a small area based on a limited 137Cs dataset. Hao et al.27 investigated the activity level of 137Cs in coastal seawater of the Bohai Sea. In the SCS, Yamada et al.12 and Zhou et al.28 investigated the horizontal and vertical distribution of 137Cs in the water column. Additionally, 137Cs in sediments of the ECS was investigated to calculate the apparent sedimentation rates5, while Zhuang et al.29 studied the impact of tides on the distribution of 137Cs in Bohai Sea sediment; in the SCS, 137Cs in sediments was used to trace the transport of terrestrial particles30. However, the limited study areas present just a glimpse of 137Cs behavior, which may lead to biased interpretations of 137Cs behavior in the China Seas. Expanded 137Cs datasets including both seawater and sediment are rarely reported, which limits our ability to fully understand the distribution of 137Cs and discuss its important influencing factors in the China Seas.

The objective of this study is to provide detailed insight into the distribution and budget of 137Cs in the China Seas, by compiling expanded 137Cs datasets that include both seawater and sediments and investigating the factors influencing them. This will also improve our general understanding of the fate of 137Cs in marine environments. Furthermore, information about the activity and budget of 137Cs also contributes important background information that can be used for future risk assessment in the China Seas, and for the study of 137Cs biogeochemistry in marginal seas. Finally, the prospect of future 137Cs studies in the China Seas is discussed.

Methods

Data source and treatment

We reviewed over 400 datasets of 137Cs in the China Seas covering the past 30 years, which include surface measurements and profiles in both sediment and seawater. Detailed sample information is shown in the Supplementary Information (SI). These 137Cs datasets were mainly extracted from bibliometric databases such as the Web of Science, Google Scholar and Scopus, using the keywords “East China Sea, South China Sea, and Yellow Sea” combined with the primary keyword “137Cs”. For ease of presentation and comparison, all 137Cs datasets are uniformly decay–corrected to January 1, 2020.

Results and discussion

Distribution of 137Cs in the China Seas

Seawater

Horizontal distribution of 137Cs

Over 200 datasets containing information on surface seawater 137Cs in the China Seas were combined12,13,25,26,27,28,31. The 137Cs activity in surface seawater of the China Seas varied from 0.03 to 1.80 Bq m−3, averaging 0.74 ± 0.34 Bq m−3 (Fig. 2). Spatially, the 137Cs activity in the SCS surface seawater varied from 0.47 to 1.80 Bq m−3, averaging 0.92 ± 0.28 Bq m−3. In the ECS, surface seawater 137Cs activity ranged from 0.10–1.37 Bq m−3, with a mean of 0.71 ± 0.27 Bq m−3. In the YS, 137Cs activity in surface water ranged from 0.03 to 0.90 Bq m−3, averaging 0.20 ± 0.20 Bq m−3. As shown in Fig. 2a, the horizontal distribution of 137Cs activities displayed three significant features: (1) the 137Cs activity in the Kuroshio was higher than in the China Seas (i.e., a declining trend from the SCS basin to the Luzon Strait); (2) the 137Cs activity gradually decreased from south to north along the China Seas with respect to latitude (i.e., the 137Cs activity in the SCS was comparable to that in the ECS, but higher than in the YS); (3) the 137Cs activity in the nearshore and the estuary (e.g., Yangtze River mouth) was lower than in the marginal sea. Overall, these patterns of 137Cs activity are influenced by the source, transportation and scavenging of 137Cs. Potential 137Cs sources in the China Seas are from the WNP, riverine inputs, and the deposition of global fallout. Outside of the Luzon Strait, the higher 137Cs activity in Kuroshio water compared to the China Seas indicates that the outer WNP is a potential 137Cs source to the China Seas. The 137Cs input flux from the WNP will be discussed in detail later. As shown in Fig. 3a, the low 137Cs activities appeared in low salinity zones, while the high 137Cs activities occurred in high salinity zones. This also indicates that the source of 137Cs potentially originates from the WNP and is derived from the Kuroshio current.

Figure 2
figure 2

The lateral distribution of 137Cs activity (Bq m−3) in seawater (a) and in sediment (Bq kg−1) (b) of the China Seas. 137Cs data are cited from previous studies (East China Sea5,13,24,25,26,31,44,45,49; South China Sea6,12,13,28,30,71; Yellow Sea14,26,27,29). This figure was drawn using the free software Ocean Data View (ODV 5.1.2) (Schlitzer, R., Ocean Data View, https://odv.awi.de, 2018).

Full size image
Figure 3
figure 3

The relationships between 137Cs activity and the salinity (a) and temperature (b) of surface seawater in the China Seas (data soured from previous studies12,13,25,26,27,28,31). This figure was prepared with Sigma–Plot professional 10.0 software.

Full size image

Fluvial input is an additional 137Cs source to the China Seas, due to the abundant detrital materials delivered by the rivers of China (e.g., the Pearl River, Yangtze River and Yellow River) that may carry 137Cs26,32. For example, the Yangtze River discharges a terrestrial particle flux of about 1.3 × 108 tons/year into the ECS33. However, terrestrial inputs have differing effects on 137Cs cycling. On the one hand, 137Cs in the river watershed originating from global fallout is bound to soil and sediment particles, which are discharged into the China Seas due to soil erosion. Based on the “fingerprint” of Pu, Xu et al.34 calculated the global fallout 239+240Pu contribution to the ECS using a two end–member mixing model. They found that riverine input accounts for ~80% of ECS inventories and direct deposition accounts for ~20%. Here, we assumed that 137Cs and 239+240Pu originating from global fallout have a similar chemical behavior35 (the riverine input flux of 137Cs is also calculated later in discussion). On the other hand, abundant particles facilitate rapid scavenging of 137Cs from the ECS water column, resulting in the lower 137Cs activity observed in the Yangtze River mouth compared to the shelf or basin of the ECS. This is also consistent with the 137Cs showing a stronger particle affinity in the low salinity or freshwater zones because the high salinity in seawater causes desorption of 137Cs36,37. The 137Cs activity in the China Seas gradually decreases with increasing latitude (e.g., from mid–latitude (SCS and ECS) to high–latitude (YS)), which is inconsistent with the 137Cs deposition flux of global fallout, namely, the deposition flux of 137Cs in high–latitudes is higher than that observed in low–latitudes since the atmospheric nuclear weapons testing in the early 1960s was mostly conducted in high latitude zones3. Such a difference is possibly caused by the different oceanic regimes. The SCS and ECS are deeper and most 137Cs is preserved in the water column. The shallower YS experiences strong hydrodynamic conditions (e.g., storms, stronger winds, and intensified waves), which result in the resuspension of particles and vigorous mixing. Therefore, the 137Cs in the YS is more readily scavenged from the water column and then deposited in the sediment.

Additionally, the relationship between temperature and 137Cs was examined based on a large number of field observations. We found no correlation (Fig. 3b), suggesting that the seasonal variation of 137Cs in the China Seas was minor. This may be related to the long half–life and residence time of 137Cs in the China Seas (see discussion below)38.

Vertical distribution of 137Cs

The 137Cs activities in 33 water column locations of the China Seas were collected in previous studies24,25,26,28,31,39, and their vertical distributions are plotted in Fig. 4.

Figure 4
figure 4

Sampling sites of 137Cs in the water column of the China Seas (a) and its vertical distributions in the South China Sea (b), East China Sea (c) and Yellow Sea (d). The water column 137Cs data are cited from previous studies24,25,26,28,31,39. This figure was prepared with the free software Ocean Data View (ODV 5.1.2) (Schlitzer, R., Ocean Data View, https://odv.awi.de, 2018) (a) and Sigma–Plot professional 10.0 software (b–d).

Full size image

The vertical distributions of 137Cs activity in the SCS are shown in Fig. 4b, displaying an initial increasing tendency with water depth and a maximum in the subsurface, followed by a more gradual decrease. This profile distribution of 137Cs agrees with those obtained elsewhere40. For example, in the Pacific Ocean and the Atlantic Ocean, the vertical pattern of 137Cs gradually increases from surface to subsurface (~150 m), which is followed by a slower decrease3,11,41,42. The high activity level of 137Cs in the SCS subsurface seawater (~150 m) is potentially related to the input of Kuroshio (i.e., the SCS seawater exchanging with the Kuroshio through the Luzon Strait) and the difference of particle removal across this depth layer (namely, higher removal above and lower removal below). The vertical distribution of suspended matter indicates suspended particulate matter (SPM) from the SCS shelf through lateral transport was noticeable in this depth range, which would enhance 137Cs scavenging43. In the middle deep layer, the concentration of SPM was low, and regeneration should be the dominant process because of SPM microbial decomposition during downward transport43.

In the ECS, the vertical distribution of 137Cs activity was similar to the pattern in the SCS, namely, a remarkable maximum usually occurred in the subsurface at depths ranging from 100–600 m (Fig. 4c). Wide depth range of the ECS subsurface maxima is potentially due to the relatively wide spatial coverage of the collected samples. For example, the spatial coverage of sampling extends from the estuary to the shelf and the basin. In the YS, the vertical pattern of 137Cs activity had no consistent distribution (Fig. 4d), which was related to the tidal effect (semi–diurnal and diurnal tides) and the sampling locations in the coastal area. Nevertheless, they overall showed the 137Cs activity in the YS surface seawater was higher than that in the bottom water.

Sediment

Horizontal distribution of 137Cs

The lateral distribution of 137Cs activity in the China Seas sediment is plotted in Fig. 2b based on compilation of over 205 datasets. The 137Cs activity of surface sediment in the China Seas varies widely from 0.06 to 5.55 Bq kg−1, averaging 1.44 ± 1.07 Bq kg−114,25,29,31,44,45. Overall, the horizontal distribution of 137Cs activity decreases from nearshore to offshore in the China Seas sediment. This pattern is mainly controlled by the 137Cs source, transportation, sedimentation rate, mineral composition and particle size. Given that 137Cs is mainly bound to riverine particles, it follows the deposition patterns of riverine particles. For example, high 137Cs activity is observed in the Min–Zhe coastal zone and the northwest corner of Taiwan in the ECS. The high 137Cs activity observed in the Min–Zhe coastal zone was caused by a large abundance of terrestrial particle–bound 137Cs. The terrestrial particulate matters entrained by Changjiang diluted water (CDW) in winter, and the resuspended sediments generated by the typhoons in summer, eventually join the northeastwardly Kuroshio in the northern Taiwan Strait46. Therefore, the terrestrial particulate matters and resuspended sediments create favorable conditions for the resulting high 137Cs activity in the northwest corner of Taiwan. In contrast, the distribution of 137Cs activity in the northern SCS shelf shows a decrease from nearshore to offshore, and high 137Cs activity is observed on both sides of the Pearl River Estuary (PRE). In the northern SCS, the Pearl River plume disperses southwestward in winter, but northeastward in summer47. The terrestrial particulate matter carried by the Pearl River plume is preferentially deposited along the dispersing direction of the plume and favors quick 137Cs removal26. It is worth noting that we cannot present the spatial distribution with a high resolution because of the limited number of samples available. Finally, the distribution of 137Cs activity in the YS is likely controlled by tides, ocean current, fluvial input, and resuspension. In the Yellow River estuary and the central YS, high 137Cs activity was observed, while low 137Cs activity was observed in the western YS. The Bohai sea, a semi–enclosed bay, connects Yellow River discharge upstream with the YS downstream. The estuarine circulation is mainly influenced by riverine influx and tides48. The semi–diurnal and diurnal tides around the Bohai bay are favorable for the resuspension of 137Cs in the shallow area. High 137Cs activity in the central YS is potentially related to the mineral composition of sediment (i.e., clay, silt and sand)26. Further examination of the relationship between 137Cs activity and mineral composition/mean size of sediment in the China Seas is shown in Fig. 5. The sediments are mainly composed of silt (2–63 μm), sand (>63 μm) and clay (<2 μm), of which range from 6.7%–73.8% of silt (averaging 46.9%±20.1%, n = 148), 0.2%–89.7% of sand (averaging 34.1%±21.9%, n = 148), and 3.3%–48.6% of clay (averaging 19.0% ± 9.5%, n = 148)26,30,45,49,50. By analyzing a large number of field observation data, the 137Cs activity showed linear positive correlations with the content of clay (R2 = 0.3247) and silt (R2 = 0.2027), and a negative correlation with the content of sand (R2 = 0.2834) (Fig. 5b–d). The mineral compositions of clay and silt mainly include illite, chlorite, kaolinite and smectite, which more easily adsorbs 137Cs. In contrast, the mineral composition of sand is mainly silicon dioxide, which is unfavorable for 137Cs adsorbed on the particles. Laboratory experiment also suggested the adsorption of 137Cs in sediment depends on the grain size and have reported this type of empirical relationship51. Here, the relationship between 137Cs activity and grain size was examined based on in–situ data, indicating the 137Cs activity exponentially decreased with increasing particle size (Fig. 6). The 137Cs is easily adsorbed and accumulated in the finer particles compared to the coarser particles51. Walling and Woodward (1992)52 suggested the 137Cs activity in the finer fractions of Jackmoor Brook catchment soil (Devon, UK) was several times higher than those in the coarser fractions. Indeed, the mineral composition and grain size in sediments of the China Seas has a significant influence on the distribution of 137Cs activity. For example, high percentage of clay minerals (>25%) and small grain size in sediments of the Min–Zhe coastal zone corresponds to high 137Cs activity26. Similar behavior is observed in the Yellow River estuary (percentage of clay >25%) and the central YS (percentage of clay >40%), where high clay content corresponds to high 137Cs activity26.

Figure 5
figure 5

Mineral composition of surface sediment in the China Seas (a) and its relationship with 137Cs activity (b–d). Data sourced from the previous studies14,26,30,45,49,50. This figure was prepared with Sigma–Plot professional 10.0 software.

Full size image
Figure 6
figure 6

The relationship between mean grain–size and 137Cs activity in surface sediments of the China Seas. Data sourced from the previous studies14,30,45,49,50. This figure was prepared with Sigma–Plot professional 10.0 software.

Full size image
Vertical distribution of 137Cs

The 137Cs activity in 25 sediment cores of the China Seas was collected from published papers (ECS–8 cores5,31,44,45; SCS–2 cores6; YS–15 cores29) and their profile patterns are plotted in Fig. 7.

Figure 7
figure 7

Sampling sites for analyzing 137Cs in sediment cores from the China Seas (a) and its vertical profiles in the South China Sea (b), East China Sea (c) and Yellow Sea (d). The sediment 137Cs data are cited from previous studies5,6,29,31,44,45. This figure was prepared with the free software Ocean Data View (ODV 5.1.2) (Schlitzer, R., Ocean Data View, https://odv.awi.de, 2018) (a) and Sigma–Plot professional 10.0 software (b–d).

Full size image

The activity level of 137Cs in the different sediment cores displays huge spatial variability with distance from the shore. In the nearshore, the 137Cs activities show large fluctuations with respect to core depth (e.g., the coastal YS; Fig. 7d). In the shelf zone, the 137Cs activity is well preserved in the sediment cores (e.g., the SCS and ECS shelf: Fig. 7b,c). The vertical distribution of 137Cs in the shelf zone shows an initial increase down core, a prominent maximum appears in the mid–layer, and then a marked decline further down core. Indeed, the variation of 137Cs activity with depth in well preserved sediment cores usually reflects the input and depositional history of 137Cs, which is widely used to reconstruct sedimentary chronology in the marine environment6,44,45. For example, utilizing the time marker of 137Cs fallout pulse input (i.e., the 137Cs peak concentration in sediment cores corresponding to the global fallout maximum, circa 1963), Wu et al.6 calculated the sedimentation rate in the northern SCS shelf to be ~0.328 cm yr−1, which agrees with the rate calculated by an another independent natural radionuclide–210Pbex (~0.337 cm yr−1). This indicates that biological perturbation in the northern SCS shelf is limited5. Therefore, the temporal change of 137Cs input in the sediment core could be reflected by the sedimentary record.

Here, we discuss the temporal variation of 137Cs in the northern SCS shelf (Fig. 8). It is of note that the influence of diffusion and mixing generally exists, although the dominant process on the shelf is sedimentation. Therefore, we cannot present the annual change of 137Cs. Nevertheless, we can discuss the source influences and the features of 137Cs on a ten–year timescale by subdividing the sedimentary record into four time periods: pre–1945, 1946–1965, 1965–1985, and post–1986. Pre–1945, there was no the input of anthropogenic radionuclide 137Cs into the earth’s environment. However, the 137Cs signal was traced in depths of the sediment core A8 corresponding to this period, which is possibly caused by the post depositional downcore mixing of 137Cs originating from the global fallout and the Pacific Proving Grounds (PPG) tests conducted in the early 1950s. Between 1946–1965, the 137Cs activity dramatically increased because of large–scale atmospheric nuclear weapon testing and test in the PPG. The highest 137Cs activity (1.97 ± 0.36 Bq kg−1) in 1960–1965 indicates maximum 137Cs deposition by global fallout in 1963, which is often used as a time marker for the study of sedimentary chronology. From 1965 to 1985, the 137Cs activities showed a slight decrease from 1.76 to 1.56 Bq kg−1. This decrease was not as sharp as expected as the global large scale atmospheric nuclear weapon testing had been banned in this period. This suggests a continuous input of 137Cs derived from the PPG via the North Equatorial Current and Kuroshio transport. Post–1986, the 137Cs activities showed a gradual decrease, which was inconsistent with the termination of atmospheric nuclear weapon testing during this period. The reason is similar as in the above discussion during the period of 1965–1985.

Figure 8
figure 8

Temporal variations of 137Cs activity in sediment core A8 from the northern South China Sea shelf. This figure was prepared with Sigma–Plot professional 10.0 software.

Full size image

Based on the above method (using the 137Cs time marker), the sedimentation rates in the ECS and the YS were also calculated. They ranged from 0.01 to 6.3 cm yr−1 (mean of 0.91 cm yr−1), which agreed with the results estimated with 210Pbex26,44,53,54. The lateral distribution of apparent sedimentation rates in the China Seas is plotted in Fig. 9. Overall, high sedimentation rates appeared in the river mouth and nearshore, and gradually decreased with increasing distance from the shoreline. For example, the very high apparent sedimentation rates observed in the estuary of Yellow River and Yangtze River can be attributed to large amounts of terrestrial particles discharged from the two major rivers. In the shelf, low sedimentation rates suggest the sedimentary process is dominant. In contrast, in the nearshore, riverine input of terrestrial particles is a major influencing factor on the sedimentation rates. For example, the distribution of sedimentation rates decreased southwards along the inner shelf and offshore, which is consistent with the dispersal of CDW carried terrestrial particles. The apparent sedimentation rates in the YS showed an increasing trend from east to west. Therefore, 137Cs was a great tracer for sedimentary chronology.

Figure 9
figure 9

The lateral distribution of apparent sedimentation rates (cm yr−1) in the China Seas. The sediment rate datasets are cited from previous studies5,6,29,31,44,45. This map was drawn using the free software Ocean Data View (ODV 5.1.2) (Schlitzer, R., Ocean Data View, https://odv.awi.de, 2018).

Full size image

The inventory of 137Cs

The inventory of 137Cs in the water column and the sediment cores is calculated by integrating the activity measured at each depth24,55. The 137Cs inventory of seawater and sediment in the China Seas was reviewed and their spatial distributions are plotted in Fig. 10a,b, respectively.

Figure 10
figure 10

The distribution of 137Cs inventories (Bq m−2) in seawater (a) and sediment (b) of the China Seas. 137Cs inventories are cited from previous studies5,6,14,24,26,28,29,31,45. Note that, the 137Cs inventory in sediment core of the SCS basin is calculated based on 239+240Pu inventory and 137Cs/239+240Pu activity ratio. This map was drawn using the free software Ocean Data View (ODV 5.1.2) (Schlitzer, R., Ocean Data View, https://odv.awi.de, 2018).

Full size image

In the ECS water column, the 137Cs inventory varied from 3 to 908 Bq m−2, averaging 88 Bq m−2, which is lower than that expected solely from global fallout within the same latitudinal zone (20°–35°N: ~750 Bq m−2)56. Indeed, the quantity was lower than that of the Atlantic Ocean (~1190 Bq m−2, 20°–35°N) and the Pacific Ocean (~1440 Bq m−2, 20°–35°N)3. The low 137Cs inventory was potentially due to the shorter turnover time of ESC water and rapid scavenging of 137Cs compared with the open ocean. The water turnover time of ECS is estimated to be less than 3 months24, which makes preserving 137Cs in the water column unfavorable in the ECS. Additionally, a large number of terrestrial particles discharged from the Yangtze River and the enhanced biomass in the ECS57,58 create favorable conditions for 137Cs scavenging in the water column59. Therefore, the bulk of 137Cs is preferentially deposited into sediments, resulting in the high 137Cs inventory of the ECS sediment5,31,45. For example, the 137Cs inventory in the ECS sediment exhibited a wide range of 39–3732 Bq m−2 (average = 970 Bq m−2), which is over twice what is predicted solely from global fallout (~750 Bq m−2)56. The distribution of 137Cs inventory in the ECS sediment decreased significantly away from the shore, indicating that the input of terrestrial particle is an important 137Cs source (see in discussion below). High 137Cs inventory also indicate elevated recent deposition of particles near the estuary, most likely, preferential deposition and accumulation. The distribution of apparent accumulation rates in the China Seas indicates a high sedimentation rate near the Yangtze River estuary with the value reaching up to 4 cm yr−1, which gradually decreased with distance from the shoreline5,53. Overall, the distribution of 137Cs inventory in the sediment of ECS is in agreement with the distribution of sedimentation rates.

It is worth noting that the reported 137Cs inventory in the upper 600 m in the northeastern SCS and the Luzon Strait (0–600 m) represents only a part of the whole water column28. We thus carefully examined the literature data concerning the profiles of the 137Cs activity in the Pacific Ocean and found that the spatial variation is limited where the typical percentage of the 137Cs inventory at a depth interval of 0–600 m accounted for 62.9–76.5% (average of 69.4%, n = 36) of the whole water column3,4,11,39. Given that the depth profiles in the northeastern SCS in the upper 600 m were rather similar to the Pacific, we extrapolated the upper 600 m inventory to the whole water column using the percentage partitioning in the Pacific Ocean. Accordingly, the 137Cs inventory in the northeastern SCS water column was calculated, varying from 49 to 1257 Bq m−2 (average = 867 Bq m−2)28,39, which is over two times what is predicted solely from global fallout (~412 Bq m−2, 10–20°N)56. This 137Cs inventory is significantly higher than that found in the ECS (~88 Bq m−2). Note that, in the northeastern SCS, the 137Cs inventory in the water column was higher than that found in the sediment core (48–401 Bq m−2, averaging 186 Bq m−2), which is different from the ECS. This indicates that, in the deep marginal sea (e.g., SCS), 137Cs largely resides in the water column, in contrast to the ECS where 137Cs is preferentially stored in the sediment. Therefore, for the deeper marginal sea or open ocean the 137Cs inventory of the water column positively correlates with water depth, and the 137Cs inventory of sediment cores negatively correlates with water depth.

The 137Cs inventory in the YS water column varied from 0.3 to 36 Bq m−2 (average = 7 Bq m−2)26, which is three orders of magnitude lower than that expected solely from global fallout within the same latitudinal zone (30°–50°N: ~1016 Bq m−2)56. The reason for the low 137Cs inventory in the YS water column is similar to that of the ECS, and depends on the input of terrestrial particles, water depth, the turnover time and quick of scavenging 137Cs as discussed above. Accordingly, this would lead to higher 137Cs inventory preserved in the YS sediment cores. The 137Cs inventory of the YS sediment exhibited a wide range, from 240 to 5071 Bq m−2, averaging 1736 Bq m−2 26,29, which is nearly twice that expected solely from global fallout within the same latitudinal zone (30°–50°N: ~1016 Bq m−2)56. This 137Cs inventory is also higher than those calculated for the ECS and the SCS, indicating the 137Cs carried by the terrestrial particles is more quickly and easily scavenged and deposited into the sediment.

Mean residence time of 137Cs in the China Seas

The effective environmental half–life of 137Cs is a good indicator for evaluating the environmental risk of 137Cs in the China Seas. In the marine environment, 137Cs is a good tracer to study water mass transport due to its high solubility in seawater2. In marine systems, the distribution coefficient (Kd) of 137Cs in sediment is reported to be ~2000, and 137Cs scavenging from the water column mainly depends on its diffusion and decay rate60. In general, 137Cs activity in surface seawater decreases exponentially with time. The temporal change of 137Cs activity in the ocean can be expressed by the following exponential function61,

$${R}_{t}={R}_{0}\exp (-at)$$
(1)

where Rt is the 137Cs activity during year t and R0 is the 137Cs activity at t = 1967.

A previous study roughly estimated the effective environmental half–life of 137Cs (TEF) in the SCS and the ECS based on limited in situ 137Cs data13. Here, the most comprehensive 137Cs dataset in the China Seas was collected in order to further improve the accuracy or reduce the uncertainty of this estimate. The temporal change of 137Cs activity in the China Seas surface seawater over the past 60 years, by expanding our dataset to include 137Cs datasets from previous studies12,13,25,26,28,31, is shown in Fig. 11a–c (SCS: Fig. 11a, ECS: Fig. 11b, YS: Fig. 11c). According to this expanded 137Cs dataset, fitted equations of 137Cs activity with respect to time are shown in Fig. 11a–c. The TEF in the China Seas was then determined as 15.4 ± 1.3 years for the SCS, 13.8 ± 1.1 years for the ECS, and 6.5 ± 0.5 years for the YS. These estimated values are slightly lower than previous results based on the more limited 137Cs datasets13,26. The longer TEF in the SCS indicates that a large amount of 137Cs is preserved in the water column, which is consistent with the higher 137Cs inventories found in the SCS compared to the ECS and YS. Our estimates were also lower than those calculated in the WNP at the same latitude (15–24 years)61 and in the coastal water of Japan (~18.7 years)62. Additionally, the estimated TEF values in the China Seas were significantly lower than the 137Cs half–life. TEF is related to the natural decay of 137Cs and subsequent marine processes, including vertical and horizontal water mass movements and particle scavenging. Note that estimated values are usually larger than the real values since the fraction of radioactive decay and scavenging activity are ignored. Nevertheless, they are indicative of current fallout deposition and terrestrial inputs, although the upper limit of 137Cs mean residence time (TM) is a more realistic indicator62. The TM could be expressed using the following formula62,

$$\frac{1}{{T}_{M}}=\frac{\mathrm{ln}(2)}{{T}_{EF}}-\frac{\mathrm{ln}(2)}{{T}_{R}}$$
(2)

where TEF is the effective environmental half–life and TR represents the radiological half–life (TR = 30.17 years). Using Eq. (2), TM in the China Seas was calculated as 45.6 ± 3.8 years for the SCS, 36.8 ± 3.1 years for the ECS, and 12.0 ± 1.0 years for the YS. Our estimated residence times will help to understanding the turnover time of 137Cs in the China Seas. Longer residence times indicate that 137Cs activity decreases more slowly with time. For example, the 137Cs residence time of SCS is longer than those in the ECS and the YS, indicating the activity and inventory of 137Cs in the former is higher than the latter, which agrees with the above field observation results. In the future, the activity level of 137Cs in the SCS would be still higher than that observed in the ECS and the YS with the assumption of no additional input of 137Cs. Our estimated residence times of 137Cs in the China Seas are slightly lower than those obtained in the Atlantic Ocean (~100 years)63 and the coastal waters of Japan (~60–70 years)62.

Figure 11
figure 11

Temporal variations of surface 137Cs activity in the South China Sea (a), East China Sea (b) and Yellow Sea (c) over the 1960–2020. 137Cs data are cited from previously published studies12,13,24,25,26,27,28,31 and the IAEA’s marine information system (MARIS) (https://maris.iaea.org/). This figure was prepared with Sigma–Plot professional 10.0 software.

Full size image

Budget of 137Cs in the China Seas

The activity level of 137Cs in water column of the China Seas depends on its sources and sinks. The input sources of 137Cs in the China Seas mainly include discharge from the Chinese nuclear power plants, the input by FDNPP, riverine input, direct deposition of global fallout and exchange with the Pacific Ocean. The output of 137Cs in the China Seas includes radioactive decay and burial. Until now, there have been no reports of 137Cs leakage from Chinese nuclear power plants in the China Seas. Previous studies confirmed the influence of FDNPP in the China Seas was minor13,24,26.

The temporal variation of 137Cs inventory in the water column of the China Seas is expressed by the following equation:

$$\frac{dW}{dt}={R}_{{137}_{Cs}}+{A}_{{137}_{Cs}}+{O}_{{137}_{Cs}}-{S}_{{137}_{Cs}}-\lambda \times W$$
(3)

where W represents the total 137Cs inventory in the water column (Bq), λ (0.023 yr−1) is the 137Cs decay coefficient. The last term of −λW in the right–hand of the equation is ignored when the decay–corrected 137Cs concentration was used to calculate W values. Here, all the 137Cs data is decay–corrected to January 1, 2020. R137Cs, A137Cs, O137Cs, S137Cs represent the riverine input flux of 137Cs (Bq yr−1), the 137Cs flux of global fallout (Bq yr−1), the net 137Cs exchange flux with the North Pacific and the 137Cs burial flux in surface sediment (Bq yr−1), respectively. With the assumption that the China Seas are in steady state, the variation of 137Cs should keep constant over time (i.e., dW/dt = 0). However, as discussed above, the temporal change of 137Cs activity in surface water showed an exponential decrease. Here, based on the distribution of 137Cs inventories in the water column (taking an average value of 137Cs inventories) and the surface area (as calculated in Google Earth) of the China Seas (detail in Table S3 in the SI), we calculate total water column 137Cs inventories of (28.05 ± 5.32) × 1014 Bq for the SCS, (1.40 ± 0.90) × 1014 Bq for the ECS, and (6.98 ± 5.34) × 1012 Bq for the YS. Using the above calculated mean residence time of 137Cs in the China Seas, the 137Cs fluxes in the water column are (6.15 ± 1.16) ×1013 Bq yr−1 for the SCS, (3.80 ± 2.45) × 1012 Bq yr−1 for the ECS, and (5.82 ± 4.45) ×1011 Bq yr−1 for the YS. Therefore, the total 137Cs flux in the water column of the China Seas is calculated to be (65.88 ± 14.50) × 1012 Bq yr−1.

The burial flux of 137Cs in sediment depends on the 137Cs activity, sedimentation rate and bulk density. Zhang et al.26 estimated the burial fluxes of 137Cs as 7.3 ± 2.3 Bq m−2 yr−1 for the ECS, 6.9 ± 1.6 Bq m−2 yr−1 for the YS and 22.4 ± 8.7 Bq m−2 yr−1 for the Bohai sea. Based upon their work, we use an expanded 137Cs dataset to further estimate the burial flux of 137Cs in the whole China Seas and determined fluxes of (5.62 ± 1.77) × 1012 Bq yr−1 for the ECS, (4.49 ± 1.31) × 1012 Bq yr−1 for the YS (including Bohai sea: (1.73 ± 0.67) × 1012 Bq yr−1). Note that previously there was no reported 137Cs inventory in the SCS basin. Nevertheless, based on the 239+240Pu inventory (~3.75 Bq m−2)64 in the SCS basin and the 239+240Pu/137Cs activity ratio of global fallout (0.0285 ± 0.0038, decay–corrected to January 1, 2020)34, we calculated the 137Cs inventory in the SCS basin to be ~75 Bq m−2, assuming a similar behavior between 137Cs and 239+240Pu. On the SCS shelf, the 137Cs inventory was estimated to be 207.9 ± 133.8 Bq m−2 (see details in Table S4 in the SI)6. The total inventory of 137Cs was thus calculated to be (3.96 ± 1.34) ×1014 Bq for the SCS. According to the distribution of sedimentation rates in the SCS, the burial flux of 137Cs was further calculated to be (6.95 ± 2.35) ×1012 Bq yr−1 for the SCS. The total burial flux of 137Cs in the China Seas was calculated to be (17.06 ± 5.43) ×1012 Bq yr−1. We point out that these estimates are subject to large uncertainties due to the limited collection of water column and sediment field data.

There are many rivers discharging a large quantity of terrestrial particulates into the China Seas as a result of soil erosion in the river drainage area. Riverine input is thus an important 137Cs source to the China Seas. Previous studies suggest the input of 137Cs from the Yangtze River and Yellow River is a major source to the ECS and YS, respectively5,26. The precise estimate of the total riverine input of 137Cs to the China Seas needs a large number of in–situ 137Cs data for each river. However, this is an extensive work to carry out, and currently only very limited 137Cs data in Chinese rivers are available. Here, we highlight the three major rivers in China: namely, the Yangtze River, Pearl River and Yellow River. The following equation can be used to roughly estimate the riverine input of 137Cs to estuaries with large drainage basin to estuarine area ratios65,66, and has been successfully applied in the ECS and the YS5,26.

$${I}_{d}={A}_{d}\times {I}_{f}\times {f}_{e}$$
(4)

where Ad is the area of the drainage basin, If is the 137Cs inventory in the soil of the river drainage basin and fe is the fraction of 137Cs inventory eroded each year from the watershed (fe = ln2/residence time of 137Cs in the watershed). The collective inventory of 137Cs in soil cores from river drainage basins around the China Seas are shown in Table S5 (SI). The average inventory of 137Cs was calculated based on the three major river systems flowing into the SCS (Pearl River), the ECS (Yangtze River) and the YS (Yellow River). The residence times of 137Cs in various global river drainage basins vary greatly, ranging from 800 to 4100 years5,26,67,68,69. Considering that the drainage area of the Yangtze River is among the largest in the world, taking the upper limit of residence time (4100 years) is reasonable26. The residence time of 137Cs in other rivers is shown in Table S6 (SI). Accordingly, the calculated riverine input of 137Cs is (1.66 ± 0.63) × 1011 Bq yr−1 for the SCS, (5.01 ± 0.83) × 1011 Bq yr−1 for the ECS, and (5.82 ± 1.72) × 1011 Bq yr−1 for the YS, resulting in a total riverine input of 137Cs into the China Seas of (12.49 ± 3.18) ×1011 Bq yr−1. The input of 137Cs discharged from the three major rivers (Yangtze River, Yellow River and Pearl River) is calculated to be (7.35 ± 1.69) ×1011 Bq yr−1, accounting for ~60% of the riverine input to the China Seas. These three major rivers account for ~70% of the terrestrial particulate matter discharged into the China Seas70,71,72,73, which is slightly higher than the 137Cs fraction contributed by riverine inputs (~60%).

It is well known that the large–scale global atmospheric nuclear weapons testing conducted in the 1950s and the early 1960s resulted in the worldwide deposition of 137Cs. The peak 137Cs deposition flux appeared in 196374, and then gradually decreased with time (with the exception of the additional depositional influence of the Chernobyl nuclear accident in 1986) due to the global ban of atmospheric nuclear weapons testing. At present, the 137Cs deposition of global fallout in the China Seas is mainly originating from the resuspension and transport of East Asian dust packaged 137Cs74. Through long–time series observation at the Japan Meteorological Station (36.05° N, 140.13° E), the deposition flux of 137Cs is calculated to be 0.22 ± 0.09 Bq m−2 yr−1 74. At another observation station in Shanghai (31.23° N, 121.40° E), the deposition flux of 137Cs is estimated to be 0.33 ± 0.20 Bq m−2 yr−1 49. For this study, we used the mean value (0.28 ± 0.08 Bq m−2 yr−1) at the two observation stations as the deposition flux of 137Cs in the China Seas, as the China Seas are located between them. Then we calculated the direct 137Cs deposition of global fallout to be (9.80 ± 2.80) × 1011 Bq yr−1 for the SCS, (2.16 ± 0.62) × 1011 Bq yr−1 for the ECS, and (1.34 ± 0.38) × 1011 Bq yr−1 for the YS. Our calculated result in the ECS is higher than the previously reported value26, since our study area is significantly larger than their area. The total 137Cs deposition flux of global fallout in the China Seas is calculated to be (13.30 ± 3.80) ×1011 Bq yr−1. Finally, according to the Eq. (3), the net 137Cs flux exchange between the China Seas and North Pacific is ~8.04 × 1013 Bq yr−1. The total 137Cs inventory in the China Seas is roughly estimated to be 5.4 × 1015 Bq, which is less than 1.0% of the 137Cs inventory in the global ocean. The 137Cs contribution to the China Seas from the oceanic input is estimated to be about 96.9%: the dominant 137Cs source. This result is consistent with the above discussion.

Overall, the total 137Cs inventories in water column and sediment core of China Seas are calculated to be (29.5 ± 6.3) × 1014 Bq and (17.6 ± 12.4) × 1014 Bq (data see in Tables S3 and S4 in the SI), accounting for 62.6% and 37.4%, respectively. In detail, the total 137Cs inventories in water column and sediment of the SCS are calculated to be (28.05 ± 5.32) × 1014 Bq and (3.96 ± 1.34) × 1014 Bq (data see in Tables S3 and S4 in the SI), respectively, accounting for 87.6% and 12.4%. This indicates that most of the 137Cs is well preserved in the SCS water column. In contrast, the percentage of the 137Cs inventory in water column and sediment of the ECS are about 20.6% and 79.4%, respectively, suggesting most 137Cs is deposited in the sediment. In the YS, the water column and sediments contain approximately 0.8% and 99.2% of the 137Cs inventory, respectively, suggesting that most of it is deposited in the sediment. Therefore, the bulk of 137Cs remains in the SCS water column, in contrast to the ECS and the YS where most of 137Cs is deposited in the sediments.

Additional work is needed to fully understand radio–cesium biogeochemistry and its fate in the environment. Accurate determination of radio–cesium isotopic composition could further help to identify its source; the 134Cs/137Cs isotopic ratio is widely used. However, this ratio is unavailable in the China Seas due to the decay of 134Cs to undetectable levels because of its short half–life (~2.06 years) and the absence of recent inputs. The 135Cs/137Cs isotopic ratio is a potential alternative chronometer–tracer to investigate 137Cs source contributions in the marine environment75. 135Cs is difficult to measure, complicating the study of its transport and fluctuations in the ocean. Therefore, developing methods for the determination of 135Cs and the 135Cs/137Cs isotopic ratio in the China Seas is of future research interest, which may help assess the environmental risk of Chinese nuclear power plants in the future. Lastly, understanding the biological speciation and transformation processes of 137Cs would also be useful for accurately evaluating its ecological impact.