Abstract
The mud deposit areas on continental shelf have developed in a relatively steady sedimentary condition with high sedimentation rates, thus rendering them ideal regions for exploring high-resolution paleo-sedimentary environment records. Since the sedimentary environment of the continental shelf is subject to the compound influence of multiple factors, the reconstruction of water salinity and temperature and comprehensive analysis of their response to global climate change remains challenging. Therefore, the present study characterized water salinity and temperature of the North Yellow Sea (NYS) using proxies, such as ratios of halogen elements of sediments and stable isotopic compositions of benthic foraminifera, and found the halogen elements of Cl in sediments was good proxy for water salinity. The evolutions of water salinity and temperature of the NYS for the past 3000 years were reconstructed, while the evolution stages and drivers of the sedimentary environment were explored. From approximately 3000 cal yr BP, the sea bottom salinity and temperature of the NYS were found to follow the same evolution trends predominantly consisting of the three stages: relatively low seawater salinity and temperature during 3000−2000 cal yr BP; relatively high seawater salinity and temperature during 2000−460 cal yr BP; and rapid changes of seawater salinity and temperature since 460 cal yr BP. The changes in seawater salinity and temperature of the NYS were largely driven by the East Asian winter monsoon (EAWM) as well as the influence of the Kuroshio Current. As the EAWM fluctuated, changes occurred in the flux of low temperature, low salinity coastal current water into the Yellow Sea, with concomitant changes in bottom temperature and salinity. Strengthening of the Kuroshio Current promoted the development of the Yellow Sea Warm Current, while the influx of warm, salty water into the Yellow Sea influenced the respective sea region. The abrupt global climate changes such as the “Little Ice Age” and the Medieval Warm Period first affected the East Asian monsoon, followed by the respective sea region.
Keywords
Introduction
Deep sea sediments, loess, stalagmites, and glaciers have become important paleoclimatic information archives in view of the environmental stability and have been widely used in paleoclimatic studies on the millennium timescale (An et al., 2005; Bond et al., 2001; Cheng et al., 2016; Jaccard et al., 2010; Wang et al., 2005; Zachos et al., 2001; Zhang et al., 2021). Besides these paleoclimatic carriers, continental shelf sediments with a high sedimentation rate are potential high-resolution information archives and have similarly been successfully applied in paleoclimatic and paleoenvironmental studies (Liu et al., 2010; Yang et al., 2015). Compared to deep sea sediment, continental shelf sediments are subject to influence of more factors. They are influenced not only by climatic and oceanic factors but also by the rivers entering the continental shelf (Liu et al., 2013; Wang et al., 2011). This in turn complicates the significance of traditional alternative indicators in palaeoceanographic studies, such as water temperature and salinity, as the environmental information obtained from a single indicator may have multiple or conflicting interpretations (Dong et al., 2019; Liu et al., 2010). As stable oxygen isotope (δ18O) present in foraminifera of the continental shelf sea region is affected by seawater salinity, it is not possible to obtain a direct estimate of seawater temperature (Kubota et al., 2010). Not only do there exist significant variations as well as significant fluctuations in the data of paleo-temperature indicators, such as Uk'37 (Nan et al., 2017; Pi et al., 2022) but also regional differences in the reconstructions of seawater temperature according to the Mg/Ca ratio and of seawater salinity according to the Sr/Ca ratio. This renders it challenging to obtain a uniform correction formula (Yu et al., 2019); thus, obtaining accurate seawater temperature and salinity information for a continental shelf area has become a key link in palaeoceanographic studies.
The East China continental shelf is one of the widest in the world. Since the last deglaciation, the continental shelf has received a great amount of sediments with global climate warming and the sea level rising, leading to the formation of the postglacial sedimentary systems distributing widely on the continental shelf (Liu et al., 2010; Qiao et al., 2011). Since approximately 7000 yr BP, the sea level has been the highest, tending toward stability, and multiple muddy sediments have been formed on the continental shelf (Hu et al., 2012; Qiao et al., 2017). These muddy sediments show a large degree of thickness, good continuity, and a high sedimentation rate, which are good information carriers for high-resolution sedimentary records (Alexander et al., 1991; Qin et al., 1987). Based on the evolution of the sedimentary environment in the eastern continental shelf of China and its response to climate change, Li et al. (2014) systematically expounded the sedimentary changes since the last glacial period. With the rise of sea level, the continental shelf area has developed specific sedimentary sequence, beginning at the fluvial sedimentary environment, then transitional environment, coastal environment, and shallow sea environment, including low sea level marine sequence, transgressive sea level marine sequence, and high sea level marine sequence. Previous studies have shown that the Yellow Sea Warm Current (YSWC) began to appear at about 6500 yr BP, followed by the entry of the entire eastern continental shelf of China into the modern sedimentary stage (Liu et al., 2008; Shi et al., 2019; Yuan and Hsueh, 2010). During the evolution of the Yellow Sea sedimentary environment in the postglacial period, previous studies speculated that changes occurred in seawater salinity when the sedimentary environment evolved from continental to shallow marine. These speculations were based on the extreme values of the stable oxygen isotope composition of the foraminifera in the sediment (Kim and Kennett, 1998; Xiang et al., 2008); however, paleo-salinity values have yet to be reconstructed effectively.
The North Yellow Sea is a semi-enclosed sea region and is the intermediate link by which substances enter the South Yellow Sea from the Bohai Sea, where the North Yellow Sea mud area and the wedge-shaped mud sediments around the Shandong Peninsula are developed (Hu et al., 2012; Yang and Liu, 2007). With the low temperature and low salinity of the Shandong Peninsula Coastal Current (SDCC) and the high temperature and high salinity of the YSWC passing through this region, any changes to them may be recorded in the sediments of this sea region. Therefore, the present study used core H205 taken from the North Yellow Sea mud area to reconstruct the paleo-salinity and temperature evolution of the North Yellow Sea since the Late-Holocene and discussed the sedimentary environmental changes and their associated drivers, based on a chronological framework established via AMS14C dating and various sedimentological indicators, such as elemental compositions of sediments and stable isotopic compositions of foraminiferal. The results would provide a comprehensive understanding of sedimentary environmental changes of the Yellow Sea during the postglacial period as well as useful insights into palaeoceanographic research on the continental shelf area.
Regional setting
The Yellow Sea is a semi-enclosed shallow marginal sea of the Western Pacific Ocean and is partially surrounded by the Chinese mainland and the Korean Peninsula. It is connected to the East China Sea at its southern end and the Bohai Sea via the Bohai Strait at its northwest end. The Yellow Sea is 870 km from north to south and 556 km wide from east to west, with a total area of approximately 38 × 104 km2 and an average water depth of 44 m. The Yellow Sea is divided into south and north by the line connecting Chengshan Jiao on the eastern Shandong Peninsula and Changshan-got on the western Korean Peninsula. The total area of the North Yellow Sea is approximately 7.10 × 104 km2, while the average water depth is 38 m. (Liu et al., 2004; Qin et al., 1989; Shi, 2014; Shi et al., 2021). The sediments in the Yellow Sea are widely distributed and characterized by patchy fine-grained sediments. They form the South Yellow Sea mud area and the North Yellow Sea mud area and wedge-shaped mud sediments along the Shandong Peninsula. These sediments are generally considered to have been deposited from the time of the high sea levels in the postglacial period (Hu et al., 2012; Shi et al., 2021; Yang and Liu, 2007). Fine-grained sediments originate from a mix of sources predominantly derived from the Yellow River and also from the Yangtze River (Lan et al., 2007).
With the development of the ocean current system in the Yellow Sea, YSWC and the Yellow Sea Coastal Current (YSCC) on the western and eastern sides appeared (Figure 1). Of these, YSWC originates from the Kuroshio Current and flows along the central region of the Yellow Sea toward the north and entering the Bohai Sea, and is characterized by high temperature and high salinity, with a current velocity of 5–10 cm/s (Bian et al., 2013; Naimie et al., 2001). On the western side, YSCC originates from the Bohai Sea, flows along the Shandong Peninsula and the Jiangsu coast from north to south, including SDCC, YSCC, and is characterized by low temperature and low salinity, with a relatively greater current velocity of approximately 20 cm/s (Qiao, 2012; Wei et al., 2011). The YSWC and YSCC on both sides constitutes the continental shelf circulation system of the Yellow Sea, which includes compensating currents and is strongly influenced by the East Asian monsoon (Yuan and Hsueh, 2010). It is manifested in winter and spring when the East Asian winter monsoon is strong that there is a strengthening of the Yellow Sea circulation system along with the strengthening of the warm current and coastal currents. In summer and autumn, the East Asian summer monsoon plays a dominant role, and the Yellow Sea circulation system weakens or disappears, while cold water masses occupy the northern part of the Yellow Sea (Song et al., 2009). The Yellow Sea circulation system is the major driving force for the long-distance transport of sediments in the area and is closely associated with the formation of fine-grained sediments in the Yellow Sea (Martin et al., 1993; Tana et al., 2017).

Location of the study area and core H205: (a) Mean annual sea surface temperature (SST) of the Western Pacific Ocean margin for 1955–2017 (Locarnini et al., 2018), the distribution of the major warm currents of the eastern China seas, and the location of cores used for comparison in this study—the northern slope core B-3GC of the Okinawa Trough (Jian et al., 2000), the South Yellow Sea core ZY2 (Wang et al., 2011), and stalagmites of Dongge Cave (Wang et al., 2005); (b) Mean annual sea bottom temperature (SBT) of the Western Pacific Ocean margin for 1955–2017, the East Asian monsoon, and the northernmost boundary of the modern East Asian summer monsoon (EASM) (Chen et al., 2015); (c) Water depths (based on ETOPO 1, https://www.ngdc.noaa.gov/mgg/global/) and fine-grained sediment distribution of the eastern China seas (gray shadow with numbers, 1: Bohai Sea mud area; 2: North Yellow Sea mud area; 3: South Yellow Sea mud area; 4: southwest of Cheju Island mud area; and 5: Yangtze river estuary and Zhejiang and Fujian coast mud area, Shi et al., 2022), circulation systems (Tsushima Warm Current (TC); Yellow Sea Warm Current (YSWC); Shandong Coastal Current (SDCC); Yellow Sea Coastal Current (YSCC); South Korean Coastal Current (SKCC), Hu et al., 2012; Shi et al., 2021; Wu et al., 2019), and location of core H205.
Materials and methods
Sampling
As the main study object, the sediment core H205 (122.08° E, 37.92° N) (Figure 1c) was collected via a gravity sampler aboard the “Dong Fang Hong 2” scientific research vessel in 2009. The length of the core was 240 cm, and the water depth where the core was collected was 40 m. The core location was just outside the North Yellow Sea mud area in southwest of the North Yellow Sea, which is in the pathway of the SDCC. There was no apparent disturbance nor missing data for the process of core collection. For further analyses, the core was subsampled at 0.25 cm intervals for continuous sampling for grain size analyses, 1 cm intervals at 0.5 cm thickness for geochemical analyses, and 5 cm intervals at 1 cm thickness for foraminiferal identification.
Sediment grain size and elemental analysis
The grain size composition was analyzed via a laser grain size analyzer. Since most of the sediments in the study area were derived from the Yellow River, the carbonate content of the Yellow River sediments ranged from 8% to 18% (Hu et al., 2011), which had a large impact on the sediments. Therefore, the following pretreatment method was chosen. First, 0.5–1 g of subsample was placed in a 100 mL beaker and preserved with 5 mL 30% H2O2 for 24 h to remove any organic matter. Next, 3 mL of dispersant (0.5 molL−1 sodium hexametaphosphate) was added, and the mixture was treated with ultrasound for 30 min to ensure adequate dispersion. Grain size analyses were conducted using a Mastersizer 2000 laser particle size analyzer (Malvern Instruments Ltd., UK) with the measurement range of 0.02–2000 μm and the resolution of 0.01 φ. The measuring error was within 3%. Grain size parameters were calculated using the McManus moment formula (Mcmanus, 1988).
The elemental concentrations were measured using an energy dispersive X-ray fluorescence (EDXRF) spectrometer (Liu and Fan, 2011). Approximately 4 g of subsample was dried and grinded to below 200 mesh (63 μm), and then was put in a polyethylene ring with a diameter is 32 mm and compacted. At the bottom of the ring is a special transparent film. The bottom film and sample were confirmed to be smooth for testing on the machine. The elemental concentrations were measured using a desktop polarization EDXRF spectrometer (SPECTRO XEPOS, SPECTRO Analytical Instruments, Germany). Quality control was accomplished using National Standard (GBW 07314, offshore sediment) and duplicate samples. Tests were repeated to ensure reliable results obtained. Relative standard deviations (RSDs) of the standards for most elements generally were within 2%.
Stable isotopic of benthic foraminifera
Sediment samples were obtained at an interval of 2–3 cm, from which foraminiferal samples were selected, and a mass spectrometry was used to determine the stable isotopic composition of the foraminifera. The sediment samples were first soaked in 5% H2O2 for 24 h, filtered (0.063 mm) and dried, followed by the selection of benthic foraminifer Protelphidium tuberculatum (d’Orbigny) for stable isotope δ18O testing. The analytical experiments were performed via a GV IsoPrime mass spectrometer of the Institute of Oceanology, Chinese Academy of Sciences, with a test standard error of 0.05‰, and the test results were corrected to the PDB standard by using the NBS19 standard (Chang et al., 2015).
Chronology
Mixed benthic foraminiferal from five different layers of core H205 were selected for AMS14C dating analysis. Approximately 20 g of the sediment samples were obtained, and after removal of organic matter, rinsing and drying, they were placed under a stereo microscope (OLYMPUS SZ 61), where approximately 5 mg of clean and completely mixed benthic foraminiferal were selected and sent to USA Beta Laboratory (Miami, Fl, USA) for analysis. All the AMS14C dates were calibrated to calendar ages using Marine20 calibration dataset (Heaton et al., 2020) and the delta R value was −139 ± 59 years (Nan et al., 2017; Southon et al., 2002). An age model was developed using a smooth spline based on the Bacon v2.3.9.1 code (Blaauw and Christen, 2011) in the software “R” (Figure 2).

Age-depth model (a) (The calibrated ages and error show the dating results. The gray shadow indicates the 95% confidence ranges), sediment grain size component (b), mean grain size (c), sensitive grain size component F2 (2.125–5.125 φ) (d), and changes in Cl, Br, and Al (e–g) with depth of core H205.
Results
Chronology, grain size and elemental composition
The chronology of core H205 based on five AMS14C data provided the continuous records of the past 3000 years (Table 1, Figure 2). The age-dating results were continuous without any reversal, which indicated that these AMS14C data were true and credible. The mean sedimentation rate for core H205 was estimated at 0.078 cm/yr.
AMS14C radiocarbon dates of core H205 from the North Yellow Sea.
Sediments in core H205 were composed of clayey silt in which the range of sand content was between 3.52% and 16.62% with an average value of 9.48%, the silt content ranged from 62.97% to 73.50% with a mean value of 68.20%, and the clay content ranged from 16.39% to 27.94% with an average value of 22.32% (Figure 2). The grain size compositions were stable in the vertical direction. The mean grain size of the core sediments ranged from 6.23 to 7.31 φ and the mean value was 6.78 φ. The mean grain size changed gently in the core. As the coarse-grained components showed an increasing trend from 235 to 130 cm, the mean grain size decreased. The mean grain size sharply increased from 130 to 50 cm, because the coarse-grained components reduced and the content of silt increased obviously. Sand content increased from 50 cm to the top, so mean grain size decreased again.
Using the grain size-standard deviation method (Boulay et al., 2003), sensitive grain size was obtained from the grain size analysis results. When the boundaries were set at 2.125 φ and 5.125 φ, the following three sensitive grain size intervals could be distinguished: <2.125 φ (F1); 2.125–5.125 φ (F2); and >5.125 φ (F3), with their average contents of 0.02%, 38.80%, and 61.18%, respectively, where the latter two were the major sensitive grain sizes (Supplemental Figure S1). Of this, the sensitive grain size component of F2 (2.125–5.125 φ) gradually rose at 235-130 cm, declined at 130-60 cm, and rose again at 60-0 cm, which showed a coarse-fine-coarse trend. According to previous studies, the sensitive grain size component of the coarse particles is a good proxy of the strength of the East Asian winter monsoon (Tian et al., 2019).
The Cl content of core H205 was in the range of 0.64%–1.23%, whereas the Br content was in the range of 0.0034%–0.0048 %. The mean values for the two elements were 0.98% and 0.0039% respectively, with both elements showing relatively consistent trends in the distribution (Figure 2). For more details, the Cl content increased gradually from a low value at 235-85 cm, decreased slowly at 85-5 cm, and displayed a sudden increase at the top 5-0 cm. The Br content remained relatively low and constant between depths of 235-160 cm before gradually increasing between depths of 160-70 cm. It then decreased gradually at 70-20 cm and increased significantly at the top 20-0 cm. The Al content of the core was stable with minor fluctuations and exhibited consistency with the grain size component of the core sediments.
Stable isotopic records of benthic foraminifera
The sediment core H205 was concentrated with benthic foraminifera, with an average abundance of 165 N/g and the variations were relatively greater in the middle of the core. Their abundance gradually decreased at 236-160 cm, then significantly increased at 160-30 cm, and sharply declined at 30 cm (Figure 3). Of this, the dominant species P. tuberculatum showed an average abundance of 98 N/g and an average content of 59.4%. P. tuberculatum proved to be the most dominant species in the habitats affected by the YSCC cold water and is representative of the typical species living in low temperature and low salinity environments (Fang et al., 2013; Mei et al., 2016).

The abundance of benthic foraminifera in core H205 (a), the percentage content of P. tuberculatum (b) and its stable isotopic δ18O content (c) with depth.
The average δ18O value was 2.05‰ with the range of 1.63‰–2.72‰. From the bottom to top of the core, the δ18O changes were characterized by an increase, followed by a decrease, subsequently tending toward relative stability. The δ18O value significantly rose at 236-188 cm, rapidly decreased at 188-153 cm, and remained relatively stable from 153 cm upwards (Figure 3).
Discussion
Reconstruction of bottom seawater salinity based on halogen elements compositions of sediments
As halogen elements in sediments, Cl and Br possessed similar sedimentary geochemical properties such that most of them occurred in the crystal lattice of evaporite minerals, likely came from the pore water in the sediments, and reflected the salinity characteristics of pore water. As a result of the weak diagenesis of the modern continental shelf sediments, the salinity of the pore water is closely associated with the bottom seawater salinity at the time of formation of the sediments (Fan and Yang, 2004). The core H205 sediments were predominantly consisted of silt and clay, with the large proportion of clay hindering the flow of pore water, thus preserving paleo-salinity information in the pore water of this core. Cl in the H205 sediment was not related to the mean grain size, and there was no apparent correlation existed between Br and the mean grain size of the sediment (Supplemental Figure S2). Thus, the influence of the grain size on the halogen elements contents was un-signified. To exclude the potential influence of the sedimentary mineral composition on Cl and Br, ratios of Cl/Al and Br/Al of the sediments were selected to calculate paleo-salinity. The paleo-salinity (S) was reconstructed based on the relationship between Cl/Al or Br/Al of the modern surface sediments collected in the Bohai Sea and Yellow Sea (Supplemental Table S1) and their corresponding bottom seawater salinity, yielding the following formula (Supplemental Figure S3):
Using equations (1) and (2), bottom seawater salinity since 3000 cal yr BP from sedimentary core H205 was estimated. According to the results of the Cl/Al and Br/Al-based salinity reconstructions (Figure 4a and b), slight differences were present in the derived salinity values, and the trends of changes in the core were almost the same.

The reconstructions of the bottom seawater salinity and temperature of the North Yellow Sea over the past 3000 years. (a) reconstructed bottom water salinity based on Cl/Al ratio of sediments; (b) reconstructed bottom water salinity based on Br/Al ratio of sediments; (c) reconstructed bottom water temperature based on δ18O composition of benthic foraminifera.
From Figure 4a, it is evident that SCl/Al has shown small fluctuations since 3000 cal yr BP, with an average salinity of 32.1‰ and a range of 31.8‰–32.3‰. The overall salinity was low with a gradual increase for 3000−1500 cal yr BP. The salinity reached high values at around 1500 cal yr BP and then gradually decreased during 1500−200 cal yr BP. Whereas salinity began to increase again after 200 cal yr BP at a gradually increasing rate until it reached its present level. Meanwhile, SBr/Al (Figure 4b) shows a similar trend with an average salinity of 32.1‰ and a range of 32.0‰–32.2‰. In contrast to SCl/Al, there is a significant increase since 460 cal yr BP.
In our another experiment, it was found that all Cl elements were basically removed after salt washing, with an average removal of 96.5%, while the average removal of Br was 78.4%. This shows that Cl is mainly stored in the pore water of the sediment, while Br is also stored in the other material in addition to the pore water and cannot be completely removed by salt washing (Supplemental Table S2). Ziegler et al. (2008) pointed out that Br was predominantly contained in marine organic material, and it could be used to estimate the organic content of the sediment by using the good relation between Br counts and total organic carbon. Therefore, the Cl/Al-based reconstruction results were taken as the salinity reconstruction results of the present core.
The salinity reconstruction based on the Cl/Al showed small fluctuations with mean value close to the modern seawater salinity. The δ18O composition of the foraminifera was used to further analyze the reliability. The value of δ18O in the upper sediments of the core was 1.93‰, and the mean bottom seawater temperature in this period corresponded to the modern seawater temperature value of 9.1°C (Locarnini et al., 2018, data from World Ocean Atlas 2018, http://www.nodc.noaa.gov/OC5/indprod.html). According to equation (3) for the paleowater temperature as well as to equation (4) for the relationship between the seawater δ18O and salinity values, the salinity based on δ18O was estimated to be 32.8‰. This estimate was relatively close to both the salinity value of 32.1‰ estimated from the SCl/Al and the bottom water salinity value of 32.3‰ of the North Yellow Sea over the past 55 years (data from WOA18). This in turn indicated that the salinity value estimated from the halogen elements was reliable.
Reconstruction of bottom seawater temperature based on δ18O of benthic foraminifera
The δ18O value of the foraminifera is significantly correlated with the water temperature, for which Shackleton (1974) proposed the following formula:
where T is the paleowater temperature (°C); δ18Oc is the δ18O value of the foraminifera (per mil vs VPDB); and δ18Ow is the δ18O value of seawater (per mil vs VPDB after conversion from VSMOW by subtracting 0.27‰, Bemis et al., 1998). From equation (3), the bottom seawater temperature could be reconstructed.
Since seawater δ18Ow is mostly influenced by changes in global ice sheets and local freshwater input, it shows a significant correlation with seawater salinity. The total amount of global ice has hardly changed since 7000 yr BP (Yang et al., 2019), and changes in the δ18Ow in the study area mainly originated from freshwater input from rivers and precipitation. Based on field survey data, previous studies have established the relationship between δ18Ow and salinity in the East China Sea and South Yellow Sea (Du et al., 2012) as follows:
Incorporating the SCl/Al value into equation (4) gives the δ18Ow value (converted to the relative VPDB standard). By combining the δ18Oc value according to equation (3), the bottom seawater temperature (T − δ18O) since the Late-Holocene could be obtained (Figure 4c). From this method, it was found that, since 3000 cal yr BP, the mean bottom seawater temperature of the North Yellow Sea has been 7.3°C with a range of 4.5°C–8.8 °C, and the core top temperature reconstructed was 7.8°C, similar to the average sea bottom temperature of 9.1°C for the North Yellow Sea over the past 55 years.
Paleoenvironmental evolution stages and their drivers in the Late-Holocene
Paleoenvironmental stage characteristics in the Late-Holocene
Based on the foraminifera abundance, sensitive grain size, the δ18O curve of foraminifera, and patterns and trends in the estimated temperature and salinity of the North Yellow Sea since 3000 cal yr BP, the environmental changes can be divided into the following three stages (Figure 5):
3000-2000 cal yr BP (1050 BCE-50 BCE): This stage was characterized by low water temperature, low salinity, and low benthic foraminiferal abundance. The average bottom water temperature was approximately 2.0°C lower than the current temperature, while salinity was 0.3‰ lower, with water temperature showing a gradually decreasing trend while salinity showing a gradually increasing trend. The abundance of benthic foraminifera was low during this period, while the dominant species P. tuberculatum, which is well adapted to cold water environments, was relatively high in content. The sensitive grain size component F2 was relatively high (Figure 5g), reflecting a relatively stronger hydrodynamic force.
2000−460 cal yr BP (50 BCE−1490 CE): This stage was characterized by high water salinity, high temperature and high foraminiferal abundance. The average salinity was 32.1‰ and peaked at approximately 1100 cal yr BP, slightly higher than the present seawater salinity. The average seawater temperature was 7.5°C, which was lower than the present seawater temperature. The foraminifera abundance showed a continuously increasing trend and peaked at 500 cal yr BP; however, the content of the dominant species P. tuberculatum did not increase in response to this and was even lower than the previous stage. There was a decrease followed by an increase in the content of the sensitive grain size component F2, with its relatively low content reflecting the relatively weak East Asian winter monsoon.
This stage covered the relatively warm period of the global climate and historical phenological records in China, as well as the relatively cold period of the phenological records. The water temperature and salinity of the Yellow Sea were significantly high during 1300−460 cal yr BP, corresponding to the relatively warm period of the Tang and Song dynasties in China (650–1350 CE, Zhu, 1973, Figure 5k) and roughly to the warm period in the Medieval period (900–1500 CE, Keigwin, 1996, Figure 5j).
460 cal yr BP onward (1490 CE−1950 CE): In the initial period of this stage, salinity and temperature were low, and there was a sharp decline in the foraminifera abundance. By 200the Tang and Song dynastiescal yr BP, the salinity and temperature sharply increased, while the foraminifera abundance remained low, and the content of the dominant species P. tuberculatum increased throughout the entire stage. The content of the sensitive grain size F2 increased significantly, indicating a relatively stronger East Asian winter monsoon during this stage compared to others. During the period of 450−150 cal yr BP (1490–1800 CE), the low salinity and low temperature well corresponded to the short-lived climatic events of the “Little Ice Age,” where global climate was relatively cold as widely recorded worldwide (Bartels-Jónsdóttir et al., 2006; Ji et al., 2021; Lei et al., 2017; Stuiver et al., 1995; Xiang et al., 2006). The North Atlantic had relatively high hematite content (Bond et al., 2001, Figure 5j), which corresponded to the relatively cold period between 1490–1900 CE, according to Chinese historical records. The temperature anomaly estimates indicated relatively low temperatures during this period (Ge et al., 2003; Ge et al., 2013), where a downward trend existed. Chinese historical records also indicated that this was the coldest period since 2 ka (Zhu, 1973). The seawater temperature and salinity have significantly increased since 200 cal yr BP, consistent with the rapid intensification of human activity since 1750 CE.

A comparison of environmental changes and their drivers in the North Yellow Sea since the Late-Holocene. (a) The abundance of foraminifera in core H205; (b) the dominant species P. tuberculatum and its percentage content in core H205; (c) the reconstructed bottom seawater salinity of the North Yellow Sea; (d) the reconstructed bottom seawater temperature of the North Yellow Sea; (e) Uk’37-SST of core ZY2 in the South Yellow Sea (Wang et al., 2011); (f) the Pulleniatina obliquiloculata content of core B-3GC in Okinawa Trough, representing sea surface temperature changes in the Kuroshio Current area (Jian et al., 2000); (g) the sensitive grain size of core H205; (h) the sensitive grain size changes of core ZY2 in the South Yellow Sea, representing the changes in intensity of the East Asian winter monsoon (Hu et al., 2012); (i) the stalagmite δ18O curve of Dongge cave, representing the changes in the East Asian summer monsoon (Wang et al., 2005); (j) the hematite-stained grains (HSG) content of the North Atlantic core MC52-VM29-191 (Bond et al., 2001). MWP: medieval warm period; LIA: Little Ice Age; and (k) temperature anomalies in eastern China (Ge et al., 2003; Ge et al., 2013), corresponding to China’s historical dynasties. The short bars at the top represent dating uncertainties.
Main drivers of the stage-specific evolution of sedimentary environment
To identify the main drivers of the stage-specific evolution of the sedimentary environment, the records of the Late-Holocene sedimentary curve of the North Yellow Sea were compared with those of the Uk’37-SST curve of the South Yellow Sea representing the YSWC fluctuations (Wang et al., 2011) (Figure 5e), the curve representing the Kuroshio Current SST changes (Jian et al., 2000) (Figure 5f), the sensitive grain size curve representing the East Asian winter monsoon changes (Hu et al., 2012) (Figure 5h), the δ18O stalagmite curve of Dongge cave representing the East Asian summer monsoon changes (Wang et al., 2005) (Figure 5i), and the temperature anomaly curve of eastern China (Ge et al., 2003; Ge et al., 2013) (Figure 5k). Since 3000 cal yr BP, the environmental changes in the North Yellow Sea have chiefly been controlled by the East Asian monsoon, followed by the water temperature changes stemming from the entry of the Kuroshio Current into the Yellow Sea (Figure 6), while inter-annual climate factors (e.g. ENSO) exerted no significant impact on the seawater environment.

Pearson’s correlation analysis for proxies and main drivers of sedimentary environment in the North Yellow Sea. The correlation is significant at the .05 level.
The East Asian monsoon fluctuations exhibited good consistency with the stage-specific environmental changes recorded by the core, and the East Asian winter monsoon curve showed a better consistency with the stage-specific fluctuations of the core (Figure 5h and i). In the first stage (3000−2000 cal yr BP), the water temperature and salinity were low, corresponding to a period of relatively strong East Asian winter monsoon and summer monsoon. In the second stage (2000−460 cal yr BP), the seawater temperature and salinity recorded by the core were relatively high, corresponding to a period when the East Asian winter monsoon was at its weakest, and the summer monsoon was also relatively weak.
The third stage (460 cal yr BP onward) was characterized by low water temperature and salinity which later showed a sharp increase, which well corresponded to the change in the East Asian winter monsoon and summer monsoon from strong to weak. The muddy deposits in the North Yellow Sea are similar to those in the continental shelf of eastern China, where the sediments are predominantly transported and deposited through the coastal current system, with transportation and deposition mostly taking place in winter, as a process summarized as “storage in summer and transport in winter” (Wang et al., 2014; Yang et al., 1992). Numerous studies have demonstrated that the coastal currents along the eastern China seas are driven by the East Asian winter monsoon, and the sediment sensitive grain size is an indicator of the strength of the winter monsoon (Tian et al., 2016). As the winter monsoon strengthened, SDCC strengthened, and its transport capacity rose. Thus, more low salinity and temperature seawater from the Bohai Sea entered the Yellow Sea. This in turn resulted in the records of relatively low water salinity and temperature in the North Yellow Sea in this period; however, the salinity and temperature recorded became relatively high when the situation was reversed.
In the past 3000 years, the core H205-recorded salinity and temperature showed synchronous changes. The changes of salinity and temperature and stage-specific divisions have been almost consistent with the fluctuations in the winter monsoon, which indicates that the East Asian winter monsoon is the main driver of the sedimentary environment in this sea region. Compared with the winter monsoon, which drives the coastal waters, the summer monsoon more significantly affects the coastal waters through rainfall changes. Previous studies have found that the δ18O values of both stalagmite and peat bog cellulose point to changes in the intensity of the summer monsoon (Hong et al., 2009; Wang et al., 2005) and are related to rainfall; that is, as the summer monsoon increases in intensity, rainfall increases and vice versa (Chen et al., 2015; Zhang et al., 2020). Increased rainfall results in an increased amount freshwater entering the coastal area through large rivers. The North Yellow Sea is a transitional sea region connecting the Bohai Sea and the South Yellow Sea, semi-enclosed, and easily influenced by changes in the flux of rivers entering the sea. Therefore, since 3000 cal yr BP, the three stages of the salinity fluctuations in the North Yellow Sea showed a certain degree of relationship with the summer monsoon fluctuations (Figures 5c, i and 6), which indicated that they may also be subjected to a certain degree of influence of the summer monsoon.
The changes in the temperature and salinity of the North Yellow Sea since 3000 cal yr BP were also influenced by fluctuations in the strength of the Kuroshio Current (Figure 5e and f). During the Quaternary period, the Kuroshio Current underwent significant fluctuations, whose strength could be reflected in P. obliquiloculata abundance and its stable isotope fluctuations. In other words, when the Kuroshio Current was strong, the abundance of P. obliquiloculata was high, whereas δ18O was low and vice versa (Jian et al., 2000). Previous studies have shown that the Kuroshio Current is able to spread through its derivative current, YSWC, into the South Yellow Sea (Ijiri et al., 2005; Wang et al., 2011). As the Kuroshio Current intensifies, high-temperature and high-salinity water enters the Yellow Sea and Bohai Sea through YSWC, thus resulting in corresponding temperature and salinity changes in these sea regions. Given the comparison of the H205-based bottom seawater temperature of the North Yellow Sea, the South Yellow Sea SST representing YSWC, and the P. obliquiloculata content representing the Kuroshio Current water temperature changes (Figure 5d), it was found that, since 3000 cal yr BP, the overall changes in the three curves were relatively consistent and showed a gradually increasing trend. This indicated that the temperature of the entire Yellow Sea accordingly rose as the Kuroshio Current strengthened. In particular, the core H205-based higher salinity and temperature of the North Yellow Sea in the second stage of the Late-Holocene (2000−460 yr BP) showed a better relationship with the increased Kuroshio Current intensity, and the significantly increased abundance of foraminifera, thus reflecting that the intensification of the Kuroshio Current caused foraminiferal to thrive (Figure 6). All these results suggest that the changes in the North Yellow Sea environment during the Late-Holocene were subject to the limiting effect of the Western Pacific boundary current.
The North Yellow Sea sedimentation also records the rapid global climate changes, with the global climate cooling period of the “Little Ice Age” being the most typical example (Figure 5j), by which a pronounced period of low temperature and low salinity was recorded in the core (Figure 5c). In addition, human-induced global warming has been a prominent climatic event since the Industrial Revolution, where the observed records of global average annual temperature in the period of 2006–2015 rose by approximately 0.85°C compared with that of 1850–1900 (Pachauri et al., 2014). This sharp temperature rise has also been well documented in core H205 of the North Yellow Sea; that is, the seawater temperature and salinity since 200 cal yr BP well corresponded to the rapid intensification of human activity since 1850 CE. The comparisons of the temperature, salinity, East Asian monsoon, and Kuroshio Current exponential curves of core H205 show that the Little Ice Age corresponds to the sudden increase of the East Asian winter monsoon, while the global warming period corresponds to the increased East Asian summer monsoon. This in turn indicates that the global climate change is mainly caused by the influence of alterations in the East Asian monsoon on the present sea region.
Conclusion
Based on the cores with stable sedimentation and accurate age dating, proxies of seawater salinity and temperature of the North Yellow Sea during the Late-Holocene were explored. The results showed that Cl in the sediments is a good salinity proxy, whereas the δ18O value of the foraminifera is a reasonable temperature proxy. The seawater salinity and temperature of the North Yellow Sea during the Late-Holocene since 3000 cal yr BP were subsequently reconstructed, and the drivers of the sedimentary environment changes were discussed.
Since 3000 cal yr BP, the average bottom seawater salinity of the North Yellow Sea has been 32.1‰ with the range from 31.8‰ to 32.3‰ and the average temperature was 7.3°C with the range from 4.5°C to 8.8°C, both of which are generally closer to the modern observations. The bottom seawater salinity and temperature have similar trends of evolution and have gone through the three stages: (1) a relatively low salinity and low temperature during 3000−2000 cal yr BP; (2) relatively high salinity and high temperature during 2000−460 cal yr BP, which was corresponding to the relatively warm period of the Tang and Song dynasties in China (650–1350 CE) and roughly to the Medieval Warm Period (900–1500 CE); and (3) low salinity and low temperature, followed by their rapid increase as of 460 cal yr BP, which corresponded well to the short-lived climatic events of the “Little Ice Age.”
Since the Late-Holocene, the seawater temperature and salinity of the North Yellow Sea have been predominantly affected by the East Asian winter monsoon as well as the Kuroshio Current. As fluctuations occurred in the East Asian winter monsoon, the flux of low-temperature and low-salinity water into the Yellow Sea was driven by coastal current changes, thus leading to concomitant fluctuations in the bottom seawater temperature and salinity in the study area. The strengthening of the Kuroshio Current promoted the development of the YSWC, and high-temperature and high-salinity water entered the Yellow Sea and affected the respective sea region. Meanwhile, the Little Ice Age corresponds to the sudden increase of the East Asian winter monsoon, while the global warming period corresponds to the increased East Asian summer monsoon. This also suggests that the sudden global climate changes exerted an impact via alterations in the East Asian monsoon.
Supplemental Material
sj-doc-1-hol-10.1177_09596836241254489 – Supplemental material for High-resolution water temperature and salinity evolutions and associated drivers of the North Yellow Sea during the Late-Holocene
Supplemental material, sj-doc-1-hol-10.1177_09596836241254489 for High-resolution water temperature and salinity evolutions and associated drivers of the North Yellow Sea during the Late-Holocene by Shiwen Zheng, Dejiang Fan, Judong Mao, Yonggang Jia and Zuosheng Yang in The Holocene
Footnotes
Author contribution(s)
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Natural Science Foundation of China (NSFC) (Grant Nos. 42176077, 41530966, 41831280). The sedimentary core was collected onboard of R/V “Dong Fang Hong 2” implementing the research cruise in 2009.
Data Availability Statement
Data sharing not applicable to this article as no datasets were generated or analyzed duringthe current study.
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
