To better understand the environmental significance of sediment grain size in continental shelf of the South China Sea (SCS), we carried out a detailed grain size study of sediments from the YJ Core, derived from the mud deposits of the northern SCS. Based on the grain size-standard deviation method, two sensitive grain size components were identified, namely, component 1 (8.2 ~ 9.3 μm) and component 2 (106.8 ~ 120.7 μm), respectively. The results indicate that the sensitive component 1 is likely to derive from fine-grained materials of the Pearl River. These fine-grained materials could be transported by the southwestward coastal current during the wet season, with the domination of the East Asian summer monsoon (EASM). Accordingly, the sensitive component 1 could be sensitive to climate change and has a great potential to reconstruct details of EASM variations. During the period of 7500-6800 cal yr BP, the sensitive component 1 may be controlled by both sea level change and EASM intensity. Besides, the curve of the sensitive component 1 in the YJ Core presents a strong EASM during the interval 6800-3500 cal yr BP and a weak EASM during the period of 3500-2000 cal yr BP, which is synchronous with other paleoclimate records in southern China. In the past 2000 years, the sensitive component 1 may reflect the increasing of human activities. It is essential to carry out more studies with higher resolution in mud areas to clarify a detailed historical evolution of EASM intensity over the whole Holocene.

The East Asian summer monsoon (EASM), an integral part of the global climate system, exerts a significant influence on regional hydrology, ecological environment, and societal stability [13]. Especially in the context of global warming, the increasing of frequency and magnitude of the extreme precipitation caused a catastrophic impact on regional economic activities and people’s lives. Therefore, in order to better understand the global climate evolution and predict the trend of future climate change, it is urgently needed to determine the behavior of the EASM and its underlying forcing mechanism. In recent decades, numerous studies have been conducted to investigate the evolution of the EASM during the Holocene [49]. These research archives were majorly derived from terrestrial records and deep-sea sediments. Furthermore, these studies generated inconsistent results and even contradictory trends [1013]. The reasons for these discrepancies are complex, but one of the most important reasons may be partly associated with the scarcity of high-resolution monsoon records in southern China, especially in the northern South China Sea (SCS).

As one of the proxies of geological records, the grain size parameters of sediments have been widely used in the reconstruction of paleoclimate and paleoenvironment [1417]. The use of grain size distributions in the SCS to determine the historic evolution of the EASM has been increasing [8, 18, 19], which supports the rationale for using the grain size parameters as indexes for paleoclimatic reconstruction. However, the material sources of offshore sediments are wide, and the dynamic process and transport medium are relatively complex. The sediments are always transported by coastal currents and buried in different area. Therefore, there may be deviations in the interpretation of sedimentary environment by the full sample size parameters. Various unmixing methods have been developed to separate different components from the full sample size data, in order to obtain relatively reliable paleoenvironmental change information [2024]. The components extracted by the grain size-standard deviation method can reflect the environmental information of the main controlling factors and are widely used in the reconstruction of paleoclimate and paleoenvironment in ocean and lake [14, 2528]. Besides, the paleoclimatic significance of grain size is poorly understood in the continental shelf sediments of the northern SCS. This limitation hinders our knowledge of the regional sedimentary process and paleoenvironment variation.

On the broad continental shelf area of the northern SCS, a large amount of terrigenous material input from the Pearl River are coupled with intense and stable coastal current, which are favorable for the formation of two mud areas, i.e., Pearl River proximal mud and distal mud [29, 30]. The annual runoff of the Pearl River is the second largest river in China. The Pearl River has the larger volume over other small rivers along the coast. Since the coastal currents in the west waters of the Pearl River Estuary (PRE) flow westward all the year round, the sediments of the nearshore mud area in the northern SCS mainly depended on the materials transported by the Pearl River with the southwestward coastal current [27, 3133]. Therefore, this mud area with a high deposition rate is very suitable for the study of high-resolution climate change during the Holocene. Furthermore, the northern SCS, which is closest to the moisture source of the EASM, is sensitive to climatic fluctuations, making it an excellent region for investigating the historic evolution of the EASM. In this study, the high-resolution sedimentary core (YJ Core), obtained from the Pearl River-derived distal mud deposits in the northern SCS, is used to explore the paleoclimatic significance of grain size components. Combining with a robust chronological framework, we try to reconstruct the paleoclimatic and paleoenvironmental changes during the Holocene.

The YJ Core (112° 8.08 E, 21° 31.44 N) was recovered from the Pearl River-derived distal mud areas in the northern SCS. The water depth is about 21 meters, and the core length is 6.1 meters (Figure 1). The sediments in the core display relatively uniform lithology. The core is composed mainly of yellow-brown clay (0-0.4 m) and gray clay (0.4-6.1 m), locally containing shells.

The age model of YJ Core was established using a combination of 210Pb and 137Cs in the upper 13 cm, and 18 AMS 14C dates from well-preserved shell samples. The age-depth model was constructed by using R (v3.2.1) [34] and Bacon v2.2 [35] software (Figure 2). For further details of the dating method and modelling approach, see Huang et al. [36].

For grain size analysis, all samples were weighed to the appropriate amount, pretreated with H2O2 (10%) to remove organic matter and with HCl (10%) to remove carbonates. Then, the samples were dispersed with 0.05 mol/L (NaPO3)6 on an ultrasonic vibrator for 10 min. The grain size distribution was measured by Malvern 2000 laser diffraction instrument with 100 bins ranging from 0.02 to 2000 μm. The test was performed in the Key Laboratory of Western China’s Environmental Systems, Lanzhou University. The relative error of repeated measurements is less than 3%.

The grain size-standard deviation method was chosen to obtain the sensitive grain size intervals with higher standard deviations within a sedimentary sequence [24]. The identification of a sensitive component is based on the relationship between the percentage of a given grain size range and its variability for a number of sediment samples. For instance, for n subsamples of a core, any grain size component will have n values, which can be statistically analyzed to determine its standard deviation:
(1)Si=i=1npip¯i2n1,
where Si is the standard deviation for the grain size class i, Pi is the frequency of occurrence for the grain size class i in the grain size distribution curve, and n is the number of samples.

The grain size-standard deviation method was chosen to obtain the sensitive grain size intervals with higher standard deviations within a sedimentary sequence [24]. As shown in Figure 3, the grain size-standard deviation analysis of the YJ Core shows that two sensitive grain size components can be observed based on two peaks of standard deviations at grain sizes of 8.2 ~ 9.3 μm and 106.8 ~ 120.7 μm. We designate sensitive component 1 and component 2 to represent the two sensitive grain size components in the YJ Core. The sensitive components 1 and 2 display opposite variation characteristics in terms of overall trends (Figure 4). The sensitive component 1 shows an increasing trend during 7500-5500 cal yr BP, followed by a downward trend, but then an upward trend from 2000 cal yr BP to the present. The sensitive component 2 displays a declining trend during 7500-5500 cal yr BP, followed by an increasing trend during the interval 5500-2000 cal yr BP. Since then, there is a decrease in the sensitive component 2 over the last 2000 years. As shown in Figure 4, the clay and silt fractions exhibit similar temoral pattern in terms of overall trends, which are opposite with the sand fraction and mean grain size (Mz). During the interval 7500-5500 cal yr BP, both the sand fraction and Mz exhibit clear decreasing trend, but the clay and silt fractions show an increasing trend. During the interval 5500-2000 cal yr BP, both the sand fraction and Mz display an upward trend, while the clay and silt fractions show a downward trend. During the last 2000 years, the sand fraction and Mz display an overall decreasing trend, but the clay and silt fractions exhibit an overall increasing trend.

4.1. Paleoclimatic Significance of Sediment Grain Size

Sediment grain size distribution has been widely used to determine provenance, transport dynamics, and the reconstruction of sedimentary environment [15, 37, 38]. Previous studies have successfully used a variety of grain size parameters to investigate the evolution of marine hydrological environment and regional climate change [19, 3942]. Several factors would affect the reliability of using grain size as a paleoclimate proxy, especially the sediment material sources. So it is very important to determine the material source of the YJ Core first. As over 80% of the annual sediment loading of the Pearl River in the wet season (from April to September) are deposited the PRE and the northern shelf of the SCS [31], two mud areas have been formed over the inner shelf, Pearl River proximal mud, and distal mud, respectively (Figure 1) [29, 30]. Liu et al. [33] found that the maximum cumulative rate in the mud area was equivalent to the magnitude of annual sediment discharge in the Pearl River, but much higher than that of other small rivers along the coast, indicating that the material in the mud area was mainly from the Pearl River. In addition, based on the estimation results of clay mineral content, it is found that about 79% of the sediments in the nearshore mud area are derived from the Pearl River [43]. The analysis of La-Th-Zr/10 and La-Th-Sc three-phase discrimination diagrams of YJ Core and surface sediments in the PRE shows that they have same provenance [44]. This is also supported by various observation results and numerical simulations [27, 32, 45, 46]. The Moyang River, which adjoins the YJ Core, is a small river. However, the runoff of the Pearl River is higher than Moyang River. In addition, the mean annual river sediment loadings of the Moyang River (80.0×104 t) is much smaller than that of the Pearl River (8735.0×104 t) [31]. All of these suggest that the long-distance transport materials from the Pearl River are the main source of sediments in the mud area, while the material input from Moyang River is relatively small. Based on observations, previous study has confirmed that the local rivers along the coast had small sediment influx and supplied mainly sandy materials [30]. The typhoon activity also generated coarse-grained component in the sediment, with prominent and sharp perturbations superimposed the long-term trend [4749]. However, the sensitive component 2 and sand content in the YJ Core did not show short-term large fluctuations, suggesting that the typhoon would have an extremely limited influence on coarse-grained component in the YJ Core. The Pearl River is likely the major source of the fine-grained matter for the distal mud [30, 33]. Furthermore, the hydrodynamic forcing in the northern inner shelf of the SCS is primarily associated with the southwestward coastal current, which plays a key role in sediment dispersal; the fine-grained matter from the Pearl River can be eventually transported to the distal mud deposit via the river plume or diluted water mass [31]. Therefore, we conclude that the sensitive component 1 is an indicator of the long-distance transport of fine-grained matter from the Pearl River, while the sensitive component 2 may indicate the short-distance transport coarse grains from the Moyang River. As over 80% of the annual sediment discharge of the Pearl River occurs in the wet season (from April to September) [31], the southwestward coastal current has an important influence on transporting sediments from the Pearl River to the YJ Core during the wet season. The Pearl River Basin is strongly influenced by the EASM. The warm and humid climate environment during summer is conducive to chemical weathering and generate finer particles [40, 50, 51]. Meanwhile, heavier monsoon precipitation also triggers faster physical erosion and greater fluvial input [5, 52, 53]. It is therefore that heavier monsoon precipitation triggers strong chemical weathering and produces greater erosion and eventually more fine-grained matter from the Pearl River supplying the YJ Core. We can reasonably speculate that the sensitive component 1 is likely to be sensitive indicator of the long-distance transport of fine-grained matter from the Pearl River which is associated with the EASM intensity. As shown in Figure 5, the sensitive component 1 is generally consistent with the chemical weathering indexes (Al/K and clay/feldspar) in the YJ Core in terms of overall trends [7, 44], indicating that chemical weathering is closely related to the monsoon rainfall intensity [33, 50, 54]. In addition, the sensitive component 1 exhibits similar temporal evolution with Ti concentrations, the environmental magnetic parameter anhysteresis remanent magnetization (ARM) [44, 55], and TOC concentrations in the YJ Core [36], which are believed to reflect changes in terrigenous input and fluvial discharge. This evidence confirms again that the terrestrial debris input by the river is the main source of fine grains in the sediment. The sensitive component 1 shows good consistency with linear sedimentation rate of the YJ Core in the overall trend [36]. Furthermore, our previous published studies have confirmed that these proxy indicators in the core can reliably indicate changes in the EASM [7, 37, 44, 55]. Therefore, we conclude that the sensitive component 1 can be used as an indicator of the EASM intensity.

4.2. Holocene Paleoclimate Changes Recorded by Sediment Grain Size Composition in the YJ Core

During the interval 7500-6800 cal yr BP, the sensitive component 1 has relatively low values, and the sensitive component 2 shows relatively high values (Figure 4). Meanwhile, the sea level rose rapidly during this period [56]. The low component 1 contents during this interval are consistent with the low concentrations of TOC and Ti in the YJ Core [36, 44], reflecting relatively weak sediment influx from the long-distance transport from the Pearl River. It appears to suggest that the sea level change has an important impact on the grain size distribution during this period. It is worth noting that the sensitive component 1 shows an increased trend during this interval, which is in accordance with the increasing trend of sediment discharge, as indicated by ARM, TOC, and Ti concentrations in the YJ Core. This suggests that other factors would affect sediment discharge other than the relative sea level. Previous studies have confirmed that the EASM was relatively strong during this period [6, 57, 58]. In addition, the chemical weathering intensity indicated by Al/K ratio in YJ Core was relatively strong [44]. These comparative analyses suggest that the strong EASM may be the main reason for the increasing trend of the sensitive component 1 and sediment discharge. Therefore, the variation of the sensitive component 1 would be collectively affected by the sea level and EASM intensity during the period of 7500-6800 cal yr BP.

During the period of 6800-3500 cal yr BP, the sensitive component 1 contents are relatively high, indicating that the long-distance transport fine-grained matter from the Pearl River makes a relatively great contribution to sediments in the YJ Core. But the sensitive component 2 contents are relatively low, reflecting a small contribution from the short-distance transport of relatively coarse grains from the Moyang River. Since the mean annual river sediment loading of the Pearl River is much larger than the Moyang River, the great contribution of Pearl River source leads to a high linear sedimentation rate in the YJ Core [36]. In terms of overall trends, the sensitive component 1 contents are generally consistent with temporal variation patterns in TOC and Ti concentrations [36, 44], suggesting large sediment discharge as a result of strong summer monsoon. Meanwhile, Al/K and clay/feldspar ratios in the YJ Core are relatively high [7, 44], indicating strong chemical weathering during this period. This similar temporal pattern was also found in other monsoon records. Geochemical parameters from Huguangyan Lake and Dahu peat suggested that the EASM was relatively strong during 6800-3500 cal yr BP (Figures 6(b) and 6(c)) [57, 58]. In addition, the strong EASM was also seen in high-resolution stalagmite δ18O from Dongge cave (Figure 6(d)) [6]. We speculate that strong monsoon precipitation triggers intense chemical weathering and generates more sediment discharge, leading to a higher concentration of fine-grained matter from the Pearl River in the sediments. Therefore, the sensitive component 1 records a relatively strong summer monsoon during this period.

During the period of 3500-2000 cal yr BP, the sensitive component 1 contents are relatively low, indicating that the clastic materials from long-distance transmission from the Pearl River are relatively reduced, while the sensitive component 2 contents are relatively high, reflecting that the coarser grain materials from near-distance transmission from the Moyang River are relatively increased. The low component 1 contents suggest weak sediment discharge, which is supported by the relatively low TOC and Ti concentrations in the YJ Core [36, 44]. It is worth noting that sediments in the YJ Core mainly derive from the Pearl River Basin. Although the coarse grains imported from the Moyang River are relatively increased, the annual sediment loading of the Moyang River was significantly different from that of the Pearl River. Therefore, the overall contribution of the coarse grains from the Moyang River relative to the YJ Core sediments is small, and the amount of fine-grained matter from the Pearl River is greatly reduced, which eventually lead to the lowest linear sedimentation rate of the YJ Core during this period [36]. In addition, Al/K and clay/feldspar ratios remain low values, indicating weak summer monsoon [7, 44]. The δ18O values in Dongge cave display a positive excursion, reflecting a weakening of the EASM during this period (Figure 6(d)) [6]. This weak summer monsoon is also supported by Δδ13C31–29 record from Huguangyan Lake [57] and δ13Corg record in Dahu peat [58]. Weak monsoon precipitation impedes chemical weathering and transports less terrigenous input and fluvial discharge, eventually generating a low concentration of fine-grained matter from the Pearl River. Therefore, these results suggest a relatively weak EASM during this period.

During the last 2000 years, the sensitive component 1 contents show an overall upward trend, and the sensitive component 2 contents display an overall decrease. The fluvial discharge and terrigenous input also exhibit overall increasing trend, as inferred from TOC and Ti concentrations in the YJ Core [36, 44]. Additionally, an overall increasing trend in the strength of chemical weathering is also found in Al/K and clay/feldspar ratios in the core [7, 44]. Our previous studies have confirmed that the increasing human activities, rather than climatic changes, may be the main factor controlling the change of various proxy indicators in the YJ Core [36, 55]. Along with the increase of population and productivity, the development level of the Pearl River Delta has been increasing, the primitive ecological environment has been continuously changed, and the fine-grained clastic materials from the Pearl River have increased [5, 59, 60], resulting in increasing of the sensitive component 1 content supplying the YJ Core.

We used detailed grain size analyses of high-resolution Holocene marine sediment records obtained from the Pearl River-derived mud in the northern SCS to trace variations in the EASM. According to the grain size-standard deviation method, two sensitive grain size components were identified in the YJ Core. The sensitive component 1 represents the long-distance transport of fine-grained matter from the Pearl River, while the sensitive component 2 indicates the short-distance transport of relatively coarse grains from the Moyang River. The sensitive component 1 is sensitive to climate change and could be used to reconstruct the Holocene evolution of the EASM. During the interval 7500-6800 cal yr BP, the sensitive component 1 may be jointly affected by the sea level change and EASM. The subsequent stage, from 6800 to 3500 cal yr BP, is characterized by a high content of the sensitive component 1 and a low proportion of the sensitive component 2, indicating that the relatively great contribution of fine grain matter from the Pearl River under the influence of a strong EASM. During the period of 3500-2000 cal yr BP, the grain size component 1 remains a relatively low value stage, indicating that a low proportion of fine grain matter input by long-distance transmission from the Pearl River resulting from a weak EASM. During the past 2000 years, the intensifying human activities, rather than climatic changes, may be the main factor controlling the change of the sensitive component 1.

The data for this paper are available at the online supplementary dataset.

The authors declare that they have no conflicts of interest.

This work was financially supported by the National Natural Science Foundation of China (42001078, 42006065); the Guangdong Natural Science Foundation of China (2020A1515010500); the Young Scholar of the South China Sea of Guangdong Ocean University (573118019); the Open Fund of the State Key Laboratory of Loess and Quaternary Geology, Institute of Earth Environment, Chinese Academy of Sciences (SKLLQG1908); the Innovative Team Project of Guangdong Universities (2019KCXTF021); and the Marine Science Research Team Project of Guangdong Ocean University (002026002004).

Exclusive Licensee GeoScienceWorld. Distributed under a Creative Commons Attribution License (CC BY 4.0).