Formation and preservation of greigite (Fe 3 S 4 ) in a thick sediment layer from the central South Yellow Sea

Formation and preservation of greigite (Fe 3 S 4 ) in a thick sediment layer from the central... Summary Sediments from continental shelves are sensitive to changes in both oceanic and terrestrial conditions, and, therefore, magnetic minerals in such sediments are affected strongly by depositional and diagenetic processes. Here, we investigated systematically an N-S transect of three sediment cores from the central South Yellow Sea (SYS) muddy area. Magnetic data indicate the presence of a horizontally distributed thick greigite-bearing layer. From an age model based on published magnetostratigraphy, accelerator mass spectrometry 14C dating ages, sedimentary characteristics and foraminiferal analysis, this layer was deposited within marine isotope stages (MIS) 17–13, following an enhanced sulphidic period over MIS 21–19 when the YS Warm Current and the associated YS Cold Water Mass were strong and where underlying sediments have higher total organic carbon, total sulphur and trace element molybdenum contents. Trace element cadmium enrichment in the greigite-bearing layers is documented for the first time, which indicates that weakly sulphidic (i.e. with trace levels of free H2S) conditions existed before greigite formed in a sulphidic environment during early diagenesis. It also indicates that subsequent conditions free of oxygen and H2S after greigite formation are more favourable for its preservation. We propose that organic matter supply was controlled over an extended period by moderate primary productivity. The combined effects of palaeoclimate and local tectonic subsidence were crucial for the formation and preservation of the identified greigite. In brief, our study improves understanding of the formation and preservation mechanisms of greigite in continental shelf sediments and reveals mid-Pleistocene palaeoenvironmental changes in the SYS. Sea level change, Environmental magnetism, Rock and mineral magnetism 1 INTRODUCTION The continental shelf provides a link between the land and deep oceans and plays an important role in geological processes, such as land–sea interaction, transfer and accumulation of terrigenous materials, and palaeoenvironmental evolution (McMillan 2002). With deposition at relatively shallow water depths (usually < 200 m), sediments in this region are highly sensitive to sea-level fluctuations, climate change and local tectonic subsidence (Yao et al. 2012). Large amounts of sediment carried by rivers are deposited on the continental shelf, which facilitates high-resolution studies of processes recorded by such sediments (e.g. Liu et al.2004). Generally, redox conditions of continental shelf sedimentary environments are influenced by multiple factors, including river and ocean hydrodynamics, which are controlled by climate and sea-level changes, terrigenous organic matter input, biological productivity, organic matter reactivity, etc. (Wang & Zhu 1989). All of these factors produce multiple features in continental shelf sedimentary archives. For example, complex processes are recorded by magnetic minerals in continental shelf sediments (e.g. Liu et al. 2004, 2016). Specifically, detrital and biogenic iron oxides (mainly magnetite, maghemite and/or haematite) can be altered to varying degrees during early diagenesis depending on redox conditions (Roberts 2015). In sulphate-reducing porewaters, a succession of iron sulphide phases will form, including mackinawite (FeS0.9), greigite (Fe3S4) and finally pyrite (FeS2) (Berner 1984). Authigenic greigite shares the same crystal structure as magnetite and is a strongly ferrimagnetic mineral that forms as a precursor to pyrite in sulphate-reducing sedimentary environments in association with chemical reactions driven by bacterial degradation of organic matter (Wang & Morse 1996). Although it was not first formally reported until 1964 in Miocene lake sediments in San Bernardino County, California (Skinner et al. 1964), its occurrence and importance remains underappreciated in most geochemical studies due to its potentially small concentrations as well as its metastable state with respect to pyrite (Kao et al. 2004). However, it has been extensively reported from globally distributed localities, where identification of its presence over the last 30 yr has been facilitated greatly by the development of modern superconducting rock magnetometers (Roberts et al. 2011). Largely independent of its palaeomagnetic importance, which includes the fact that greigite has been widely reported to record remagnetizations (e.g. Roberts & Weaver 2005; Sagnotti et al. 2010), it also has considerable potential as a palaeoenvironmental indicator, such as being a sensitive indicator of lake chemistry (e.g. Roberts et al. 1996; Reynolds et al. 1999; Frank et al. 2007) or marine environments (e.g. Fu et al. 2008). Widespread greigite preservation is likely if abundant reactive iron is available to react with limited dissolved sulphide during pyritization reactions (Kao et al. 2004). In the Yellow Sea (YS), which is a semi-enclosed sea resting on a typical continental shelf (Fig. 1), Lee & Jin (1995) discovered authigenic greigite in Holocene transgressive deposits off the southwestern Korean Peninsula (core 94P12 in Fig. 1b). They inferred that it probably formed in a reducing micro-environment around decaying organic matter. Subsequently, greigite was also detected magnetically in two post-middle-Holocene sedimentary sequences from the Korean Strait (cores SSDP102 and SSDP103 in Fig. 1b) and in a post-glacial muddy sediment core from the southeastern South YS (SYS, core YSDP103 in Fig. 1b) in studies of early diagenetic effects on magnetic minerals (Liu et al.2004, 2005). Recently, a thick (∼7 m) mid-Pleistocene greigite-bearing layer was identified via systematic magnetic analysis coupled with mineralogical observations in a sediment core (NHH01) from the central SYS (Fig. 1b, Liu et al.2014a,b). This is the oldest reported greigite in sedimentary strata in this region. However, the origin and distribution of this thick layer remains poorly constrained. In this study, we investigated systematically an N-S transect covering three long sediment cores in this area (Fig. 1b). We aim to outline the lateral distribution of this thick greigite-bearing layer, and its palaeoenvironmental significance. Figure 1. View largeDownload slide (a) Schematic map of the Yellow Sea (YS) and adjacent areas including major currents (modified after Mei et al.2016), with water depths in metres. Black arrows indicate paths of currents including the Kuroshio Current (KC), Tsushima Warm Current (TWC), Yellow Sea Warm Current (YSWC), Jiangsu Coastal Current (JCC), Yellow Sea Coastal Current (YSCC) and Korea Coastal Current (KCC). The blue dashed line indicates the spatial distribution of the Yellow Sea Cold Water Mass (YSCWM). CC and CS are abbreviations of Chengshan Cape and Changshangot, respectively. (b) Distribution of YS surface sediments (CSYSM, central South Yellow Sea muddy area) and adjacent seas (modified after Liu et al.2010), and locations of the studied cores (white dots) and other cores mentioned in the text (black dots). Figure 1. View largeDownload slide (a) Schematic map of the Yellow Sea (YS) and adjacent areas including major currents (modified after Mei et al.2016), with water depths in metres. Black arrows indicate paths of currents including the Kuroshio Current (KC), Tsushima Warm Current (TWC), Yellow Sea Warm Current (YSWC), Jiangsu Coastal Current (JCC), Yellow Sea Coastal Current (YSCC) and Korea Coastal Current (KCC). The blue dashed line indicates the spatial distribution of the Yellow Sea Cold Water Mass (YSCWM). CC and CS are abbreviations of Chengshan Cape and Changshangot, respectively. (b) Distribution of YS surface sediments (CSYSM, central South Yellow Sea muddy area) and adjacent seas (modified after Liu et al.2010), and locations of the studied cores (white dots) and other cores mentioned in the text (black dots). 2 GENERAL SETTING The YS is a broad and relatively shallow continental shelf on the western Pacific margin between the Chinese mainland and the Korean Peninsula (Fig. 1). It occupies a total area of 380 000 km2 and spans 870 km from north to south and 556 km from east to west. A line connecting Chengshan Cape (CC), Shangdong Peninsula and Changshangot (CS) on the Korean Peninsula divides the YS into two parts, the North Yellow Sea (NYS) and the SYS (Fig. 1a, Qin et al. 1989). Water depths in the NYS are mostly less than 60 m with an average of 38 m, while the SYS is mostly less than 100 m deep with an average water depth of 46 m. The SYS deepens toward the YS Trough, which is a topographic depression roughly defined by the 80-m isobath in the central SYS and extends NW–SE (Liu et al. 2010). With relatively stable subsidence over the Quaternary, sediments in the SYS reach up to a maximum thickness of ∼300 m (Qin et al. 1989; Liu et al. 2016). Although neither the Yellow River nor the Yangtze River currently empties into the YS directly (Fig. 1a), they are regarded to have governed sedimentation in major parts of this region during the Holocene (Yang et al. 2003). Based on mineralogical and geochemical studies on surface sediments, Lan et al. (2007) argued that these two large rivers provide the major source of sediments in the western and middle part of the YS, while rivers on the Korean Peninsula dominate in the east. At present, the Yellow River discharges ∼1.1 × 109 t yr−1 of sediment into Bohai Sea; 9−15 per cent of these (mainly fine-grained) sediments can be transported southward over long distances to be finally deposited in the central SYS (Alexander et al. 1991). In contrast, the Yangtze River annually empties ∼5 × 108 t of sediment into the southwestern YS and northern East China Sea (Milliman & Meade 1983), while Korean rivers discharge less than 3 × 107 t yr−1 into the eastern YS (Qin et al. 1989). The general circulation in the YS is currently dominated by the YS Warm Current (YSWC) and coastal currents (Liu et al. 2010). The northward YSWC is a branch of the Kuroshio Current (KC), which is an important Western Pacific boundary current that carries warm and salty water into the YS roughly along the YS Trough (Fig. 1a, Guan 1994). Instead, coastal currents with relatively low salinities along the Chinese and Korean coasts flow southward in winter and northward in summer, with the exception of the YS Coastal Current (YSCC) and Jiangsu Coastal Current (JSCC) which always flow southward during both summer and winter (Liu et al. 2010). In winter, the YSWC can intrude into the NYS or even the Bohai Sea because of the near homogeneity of the shelf water column (Guan 1994). However, in summer, the YS Cold Water Mass (YSCWM) is present in the central YS and intrudes southward (Fig. 1a), which hinders YSWC flow and makes it too weak to reach north of 35°N. This seasonal cold water eddy is considered to be responsible for the occurrence of the central SYS muddy area (CSYSM, Fig. 1b, Hu 1984). Although the warm current and cold water mass formed at around 6–7 kyr BP when sea level was close to the present level (e.g. Kim et al. 1999), the YSWC and YSCWM probably greatly affected sediment transportation and sedimentary environments even in the mid-Pleistocene (Mei et al. 2016). 3 MATERIALS AND METHODS 3.1 Core descriptions All of the studied sediment cores were recovered from the central SYS, within the CSYSM, which lies under the path of the present-day YSCWM (Fig. 1b). Water depths at the studied core locations shallow northward, with depths of 79 m for core EY02–2 (Zhuang et al. 2002), and 73 m for cores NHH01 (Liu et al. 2012) and DLC70–3 (Mei et al. 2013), respectively. More detailed information about the cores is given in Table 1. For the 125.64-m-long core NHH01 (∼1.06 Ma, Liu et al. 2014b), we only discuss the uppermost 71 m in this paper. Hereafter, NHH01 refers to the upper 71-m portion of the core. Table 1. Detailed information on the studied cores (from Zhuang et al. 2002; Liu et al.2012; Mei et al.2013). Core  Water depth (m)  Drilling depth (m)  Recovery rate (per cent)  Longitude/latitude  Drilling date  EY02–2  79.00  70.00  86.5  123°30΄E/34°30΄N  2001 Feb  NHH01  73.00  71.20  91.0  123°13΄E/35°13΄N  2009 Jun  DLC70–3  73.00  125.64  93.0  123°33΄E/36°38΄N  2009 Sep  Core  Water depth (m)  Drilling depth (m)  Recovery rate (per cent)  Longitude/latitude  Drilling date  EY02–2  79.00  70.00  86.5  123°30΄E/34°30΄N  2001 Feb  NHH01  73.00  71.20  91.0  123°13΄E/35°13΄N  2009 Jun  DLC70–3  73.00  125.64  93.0  123°33΄E/36°38΄N  2009 Sep  View Large Sediments in core EY02–2 are mainly clayey silt, sandy silt and silty sand (Fig. 2a, Ge et al. 2006). The sedimentary sequence in core NHH01 consists of silt and clay, with occasionally interbedded fine sand layers (Fig. 2f, Liu et al. 2014a), while sediments in core DLC70–3 are mainly silt, sandy silt and silty sand, with a nearly 5-m-thick sand layer in the intermediate section (Fig. 2k, Mei et al. 2016). Figure 2. View largeDownload slide General information on sedimentary sequences of the studied cores EY02–2 (Zhuang et al.2002; Ge et al.2006), NHH01 (Liu et al.2014b) and DLC70–3 (Mei et al.2016), respectively, including (a), (f) and (k) lithology, (b), (g) and (l) mean grain size (Mz) of the sediments in ϕ logarithmic units, (c), (h) and (m) abundance of benthic foraminifera (B.F.), (d), (i) and (n) magnetostratigraphy (the ages of geomanetic excursions are from Roberts (2008), which are 33 ka for M (Mono Lake) and 41 ka for L (Laschamp), respectively; B/M is the abbreviation of Matuyama–Brunhes boundary with an age of ∼780 ka), and (e), (j) and (o) relative sea-level (RSL) variations from Rohling et al. (2014), with the corresponding marine isotope stages (MIS) labeled in blue numbers. Figure 2. View largeDownload slide General information on sedimentary sequences of the studied cores EY02–2 (Zhuang et al.2002; Ge et al.2006), NHH01 (Liu et al.2014b) and DLC70–3 (Mei et al.2016), respectively, including (a), (f) and (k) lithology, (b), (g) and (l) mean grain size (Mz) of the sediments in ϕ logarithmic units, (c), (h) and (m) abundance of benthic foraminifera (B.F.), (d), (i) and (n) magnetostratigraphy (the ages of geomanetic excursions are from Roberts (2008), which are 33 ka for M (Mono Lake) and 41 ka for L (Laschamp), respectively; B/M is the abbreviation of Matuyama–Brunhes boundary with an age of ∼780 ka), and (e), (j) and (o) relative sea-level (RSL) variations from Rohling et al. (2014), with the corresponding marine isotope stages (MIS) labeled in blue numbers. 3.2 Age models and sedimentary history Radiocarbon ages used in the text were all obtained via accelerator mass spectrometer (AMS) 14C dating on mixed benthic foraminifera, and were converted to calibrated calendar (cal) ages by using the CALIB 7.0.4 program, an updated calibration curve, MARINE13 (Reimer et al. 2013), and a surface-ocean reservoir correction of ▵R = −100 ± 36 yr (Southon et al. 2002). Calendar ages are given with one standard deviation (1σ) uncertainty (Table 2). Table 2. List of AMS 14C dating on mixed benthic foraminifera of the studied cores. Core  Depth  Conventional age   Calendar ages (calyr BP)  References for    (m)  (yr)  Intercept  Range (1σ)  conventional age  EY02–2  1.05  8360 ± 60  7130  7036 − 7224  Zhuang et al. (2002)    2.00  10 600 ± 45  10 150  10 011 − 10 282  Zhuang et al. (2002)  NHH01  4.59  9230 ± 50  8220  8159 − 8286  Liu et al. (2012)    5.69  11 850 ± 65  11 450  11 374 − 11 518  Liu et al. (2012)    7.35  18 750 ± 130  22 310  22 159 − 22 469  This study  DLC70–3  0.04  9890 ± 50  9040  8931 − 9140  Mei et al. (2013)    2.24  10 200 ± 50  9340  9242 − 9441  Mei et al. (2013)    5.54  37 900 ± 350  40 030  39 772 − 40 278  Mei et al. (2013)  Core  Depth  Conventional age   Calendar ages (calyr BP)  References for    (m)  (yr)  Intercept  Range (1σ)  conventional age  EY02–2  1.05  8360 ± 60  7130  7036 − 7224  Zhuang et al. (2002)    2.00  10 600 ± 45  10 150  10 011 − 10 282  Zhuang et al. (2002)  NHH01  4.59  9230 ± 50  8220  8159 − 8286  Liu et al. (2012)    5.69  11 850 ± 65  11 450  11 374 − 11 518  Liu et al. (2012)    7.35  18 750 ± 130  22 310  22 159 − 22 469  This study  DLC70–3  0.04  9890 ± 50  9040  8931 − 9140  Mei et al. (2013)    2.24  10 200 ± 50  9340  9242 − 9441  Mei et al. (2013)    5.54  37 900 ± 350  40 030  39 772 − 40 278  Mei et al. (2013)  View Large For core EY02–2, two AMS 14C ages of ∼7100 and ∼10 200 cal yr BP were determined at depths of 1.05 and 2.00 m, respectively (Table 2, Zhuang et al. 2002). Magnetostratigraphic analysis defines the Matuyama–Brunhes boundary (B/M, ∼780 kyr) at 63.29 m, with an extrapolated basal age of ∼0.9 Ma (Fig. 2d, Ge et al. 2006). Sediments in core EY02–2 were largely deposited in littoral and neritic settings, which are referred to collectively as marine deposits and are interrupted by three fluvial layers (2.5 − 3.0, 58.7 − 61.8 and 68 − 70 m) based on benthic foraminiferal analysis and sedimentary characteristics (Figs 2a–c, Zhuang et al. 2002). The B/M in core NHH01 is located at a depth of 68.64 m (Fig. 2i, Liu et al. 2014b), which, combined with the AMS 14C ages (Table 2, Liu et al. 2012), suggest that the core records deposition from the mid-Pleistocene to Holocene. Based on benthic foraminiferal analyses and sedimentary characteristics (Figs 2f–h, Liu et al. 2014a,b), five fluvial layers are identified in the core, with depth ranges of roughly 14.9 − 15.9, 17.9 − 18.4, 33.1 − 34.5, 62.8 − 66.3 and 69.9 − 71 m, respectively; sediments from other parts of the core are mainly marine deposits. AMS 14C ages for core DLC70–3 indicate that sedimentation has not occurred since ∼9000 cal yr BP (Table 2, Mei et al. 2013), which indicates that the drilling location occurs in an area where Holocene sediments have been eroded. The B/M is recorded at 59.08 m, and the basal age in the core is estimated to be ∼0.9 Myr (Fig. 2n, Mei et al. 2016). Although the abundance of benthic foraminifera in some layers is low (<1 g−1, Fig. 2m), only three fluvial layers (roughly 28 − 33, 55 − 56 and 69 − 70 m) are suggested clearly from sedimentary characteristics (Figs 2k and l). Mei et al. (2016) proposed that the identified marine layers correlate to those in core NHH01. Under the constraints of B/M and AMS 14C ages, reconstructions of sedimentary history were tentatively carried out by correlating benthic foraminiferal abundance and sedimentary characteristic variations to the relative sea-level change (Figs 2e, j and o, Rohling et al. 2014). Taking core NHH01, for example (Figs 2f–j), the fluvial unit just above the B/M that is coeval with marine isotope stage (MIS) 19 (e.g. Lisiecki & Raymo 2005) should deposit in MIS 18, which can be further confirmed by the study of Liu et al. (2016). Considering that sediments in the central YS and its adjacent Bohai Sea are continuous at the 100-ka scale at least during interglacials (Yao et al. 2014; Liu et al. 2014b), the subsequent marine deposits with high foraminiferal abundance can correlate to sediments deposited in MIS 17, with the next to MIS 15. The AMS 14C ages in the upper section indicate that two thin negative inclination anomalies around 10 m are the records of Mono Lake (33 ka) and Laschamp (41 ka) excursions (e.g. Roberts 2008), respectively, and that the underlying foraminifera abundant layers should deposit in MIS 3. The last fluvial unit (14.9 − 15.9 m) can therefore be regarded as sediments deposited in MIS 4, with the underlying marine unit corresponding to MIS 5 and the penultimate fluvial unit to MIS 6, and so on. This procedure could give a reasonable chronology so far, especially to the focused layers (see below) which locate close to the B/M. 3.3 Rock magnetic experiments Hysteresis loops were measured on 107 samples for core EY02–2, 127 samples for core DLC70–3 and 67 supplementary samples in addition to the published 120 samples (Liu et al. 2014a,b) for core NHH01 at 0.4 − 1.0 m stratigraphic intervals using a Princeton Measurements Corporation vibrating sample magnetometer (VSM, MicroMag 3900). The magnetic field was cycled between +1.5 and −1.5 T, with a field increment of 5 mT and an averaging time of 300 ms. Coercivity (Bc), saturation remanence (Mrs) and saturation magnetization (Ms) were determined after paramagnetic slope correction. Specimens were then saturated magnetically again by applying a 1.5-T field. Finally, backfields were applied stepwise up to −1.0 T to determine the coercivity of remanence (Bcr). First-order reversal curves (FORCs) were measured for representative specimens using the VSM, with 100 FORCs measured at fields up to 1.0 T with an averaging time of 500 ms. FORC diagrams (Pike et al. 1999) were processed using the FORCinel software of Harrison & Feinberg (2008). In addition, magnetic susceptibilities (χ, mass-specific) of samples at parallel depths with all of the above samples were measured using a Kappabridge MFK1-FA system (Agico Ltd, Brno). Low-temperature measurements were conducted on splits (∼100 mg) of the representative samples using a Quantum Design Magnetic Property Measurement System (MPMS-5XL). Specimens were first cooled from room temperature to 20 K in zero field, then an isothermal remanent magnetization (IRM) was imparted in a 2.5 T field. After removal of the field, the IRM was thermally demagnetized by sweeping the temperature from 20 to 300 K at 5 K min−1. Cubic samples were also collected at parallel depths as the representative samples for thermal demagnetization of a three-component IRM following the procedure of Lowrie (1990). The specimens were magnetized in successively smaller fields along three mutually orthogonal axes using a 2-G Enterprises Pulse Magnetizer (2G660), with a 2.7 T induction along the z-axis (high-coercivity fraction), 0.5 T along the y-axis (medium-coercivity fraction) and 0.05 T along the x-axis (low-coercivity fraction), respectively. Subsequently, the specimens were subjected to progressive thermal demagnetization up to 690°C at 20 − 50°C intervals in a magnetically shielded (<300 nT) laboratory. These experiments were carried out in the Paleomagnetism and Geochronology Laboratory, Chinese Academy of Sciences, Beijing. 3.4 Previous data collections Prior to this study, attempts to research the initiation and evolution of the YSWC and the YSCWM in the mid-Pleistocene were conducted on core DLC70–3 by integrating multiple indices (Mei et al. 2016). Specifically, C37 alkenone contents (Fig. 3a) can reflect the change in seawater salinity of the SYS and track the YSWC; the distribution of two cold water benthic foraminifera species, B. frigida (Fig. 3b) and P. turberculatum (Fig. 3c) are related to the YSCWM, and the latter prefers to live in deeper water environments; high total organic carbon (TOC, Fig. 3d) and total sulphur (TS, Fig. 3e) contents in sediments usually indicate relatively high primary productivity in the surface or mixed layers of the water column and relatively weak oxidation degree of bottom waters; enrichment of trace element molybdenum (Mo, Fig. 3f), a redox-sensitive element, occurs only in a sulphide-rich (i.e. strongly sulphidic) sediments (Mei et al. 2016). In this study, another redox-sensitive element cadmium (Cd) which has been documented widely to be enriched in iron-reducing diagenetic (i.e. with trace levels of free H2S that are below typical chemical detection) environments (e.g. Rosenthal et al. 1995; Pattan & Pearce 2009) was measured. We used these geochemical parameters to investigate environmental conditions in which greigite is preferentially preserved. Details of measurement procedures for all of the above-mentioned indices can be found in Mei et al. (2016). Figure 3. View largeDownload slide Down-core changes in (a) C37alkenone content that is used to track the Yellow Sea Warm Current (YSWC), (b) and (c) relative content of two cold water benthic foraminiferal species that are indicative of the Yellow Sea Cold Water Mass (YSCWM), with (c) P. turberculatum preferring to live in a deeper water environment than (b) B. frigida, (d)–(f) total organic carbon (TOC) and total sulphur (TS) contents and trace element molybdenum (Mo) content, with high values indicating strongly sulphidic conditions, (g) trace element cadmium (Cd) content, with high values indicating weakly sulphidic conditions and (h) saturation isothermal remanent magnetization (SIRM). Results in (a)–(f) are reproduced from Mei et al. (2016). The green bars mark greigite-bearing layers, while the light blue bars denote layers deposited under the influence of strong YSWC and YSCWM that have undergone intense diagenesis. Figure 3. View largeDownload slide Down-core changes in (a) C37alkenone content that is used to track the Yellow Sea Warm Current (YSWC), (b) and (c) relative content of two cold water benthic foraminiferal species that are indicative of the Yellow Sea Cold Water Mass (YSCWM), with (c) P. turberculatum preferring to live in a deeper water environment than (b) B. frigida, (d)–(f) total organic carbon (TOC) and total sulphur (TS) contents and trace element molybdenum (Mo) content, with high values indicating strongly sulphidic conditions, (g) trace element cadmium (Cd) content, with high values indicating weakly sulphidic conditions and (h) saturation isothermal remanent magnetization (SIRM). Results in (a)–(f) are reproduced from Mei et al. (2016). The green bars mark greigite-bearing layers, while the light blue bars denote layers deposited under the influence of strong YSWC and YSCWM that have undergone intense diagenesis. 4 RESULTS 4.1 Rock magnetism Samples from depths of ∼50 m in all three cores have much enhanced Bc values (25–54 mT, Figs 4a, e and i) with relatively uniform Bcr/Bc ratios of ≈1.5 (Figs 4b, f and j) and ratios of Mrs/Ms ≈ 0.5 (Figs 4c, g and k). In addition, higher Mrs/χ ratios (Figs 4d, h and l) indicate single-domain (SD) magnetic behaviour and are consistent with those reported for sedimentary greigite-bearing samples (e.g. Snowball 1991; Roberts 1995). Figure 4. View largeDownload slide Variations of hysteresis parameters and their ratios along the studied cores, including (a), (e) and (i) coercivity (Bc), (b), (f) and (j) the ratio of the coercivity of remanence (Bcr) to Bc, (c), (g) and (k) the ratio of the saturation remanence (Mrs) to the saturation magnetization (Ms) and (d), (h) and (l) the ratio of Mrs to magnetic susceptibility (χ). The green bars mark the thick greigite-bearing layers discussed in the text, and some possible thinner greigite-bearing layers are marked with yellow bars. The blue characters denote the corresponding MIS for the thick greigite-bearing layers as indicated in Fig. 2. Figure 4. View largeDownload slide Variations of hysteresis parameters and their ratios along the studied cores, including (a), (e) and (i) coercivity (Bc), (b), (f) and (j) the ratio of the coercivity of remanence (Bcr) to Bc, (c), (g) and (k) the ratio of the saturation remanence (Mrs) to the saturation magnetization (Ms) and (d), (h) and (l) the ratio of Mrs to magnetic susceptibility (χ). The green bars mark the thick greigite-bearing layers discussed in the text, and some possible thinner greigite-bearing layers are marked with yellow bars. The blue characters denote the corresponding MIS for the thick greigite-bearing layers as indicated in Fig. 2. FORC diagrams for samples from these layers of cores EY02–2 (e.g. Figs 5b and c) and DLC70–3 (e.g. Figs 5j and k) have the same characteristics as those in core NHH01 (e.g. Figs 5f and g, Liu et al. 2014a), including closed concentric contours about a central peak, a negative region in the lower left-hand part of the diagram, a Bc peak of ∼60 mT, and significant magnetostatic interactions among magnetic grains suggested by vertical spread in the Bu direction and by the offset of the FORC distribution below the Bu = 0 axis. All of these features are typical of sedimentary SD greigite in natural environments (e.g. Roberts et al.2006, 2011). In contrast, FORC diagrams for other samples are typical of low-coercivity SD-PSD (pseudo-SD; e.g. Figs 5e, h, i and l) and/or PSD-MD (multidomain; e.g. Figs 5a and d) magnetic minerals (Roberts et al.2000, 2014; Muxworthy & Dunlop 2002). Figure 5. View largeDownload slide First-order reversal curve (FORC) diagrams for representative samples from the studied cores: (a)–(d) EY02–2, (e)–(h) NHH01 and (i)–(l) DLC70–3, respectively. Measurements were made in an applied field up to 1 T with 100 curves measured and an averaging time of 500 ms, and were processed using the FORCinel software of Harrison & Feinberg (2008) with a smoothing factor (SF) of 5. (e)–(g) FORC diagrams from Liu et al. (2014a). Figure 5. View largeDownload slide First-order reversal curve (FORC) diagrams for representative samples from the studied cores: (a)–(d) EY02–2, (e)–(h) NHH01 and (i)–(l) DLC70–3, respectively. Measurements were made in an applied field up to 1 T with 100 curves measured and an averaging time of 500 ms, and were processed using the FORCinel software of Harrison & Feinberg (2008) with a smoothing factor (SF) of 5. (e)–(g) FORC diagrams from Liu et al. (2014a). Temperature-dependent measurements can provide more directly diagnostic evidence of the presence of greigite. Detection of low-temperature magnetic transitions has also been used to identify magnetic minerals. For example, magnetite undergoes the Verwey transition (Tv) at around 120 K (Verwey 1939), haematite undergoes the Morin transition at 250–260 K (Morin 1950), monoclinic pyrrhotite (Fe7S8) undergoes the Besnus transition at 30–34 K (Rochette et al. 1990) and greigite lacks a low-temperature magnetic transition (Chang et al. 2009; Roberts et al. 2011). Low-temperature curves contain an obvious Verwey transition in most samples from the layers focused on here (e.g. Figs 6a, e, h, i and l), which indicates the presence of magnetite. Samples from the upper layers that should contain greigite also give rise to a detectable Verwey transition (e.g. Figs 6b, f and j). However, the Verwey transition is highly smeared in samples from the lower layers (e.g. Figs 6c, g and k), which indicates the minor presence of magnetite. In addition, the Tv temperature is consistently closed to 120 K (vertical dashed lines in Fig. 6), which indicates that the magnetite in the studied sediments is detrital rather than biogenic, which usually has a Tv of ∼100 K (Chang et al. 2016). Figure 6. View largeDownload slide Low-temperature curves of an IRM (mass specific) acquired at 20 K by applying a 2.5-T field and measured at 5 K intervals to 300 K during zero-field warming (blue lines) for representative samples from the studied cores (a)–(d) EY02–2, (e)–(h) NHH01 and (i)–(l) DLC70–3, respectively. The first derivative of magnetizations with respect to temperature (−▵IRM/−▵T) is also ploted on the right-hand y-axes (magenta lines). The dotted vertical lines indicate Tv peaks. (e)–(g) Low-temperature magnetization results from Liu et al. (2014a). Figure 6. View largeDownload slide Low-temperature curves of an IRM (mass specific) acquired at 20 K by applying a 2.5-T field and measured at 5 K intervals to 300 K during zero-field warming (blue lines) for representative samples from the studied cores (a)–(d) EY02–2, (e)–(h) NHH01 and (i)–(l) DLC70–3, respectively. The first derivative of magnetizations with respect to temperature (−▵IRM/−▵T) is also ploted on the right-hand y-axes (magenta lines). The dotted vertical lines indicate Tv peaks. (e)–(g) Low-temperature magnetization results from Liu et al. (2014a). Although the Curie temperature of greigite remains unknown but must exceed 350°C, it will undergo thermal decomposition below 400°C (Chang et al. 2008; Roberts et al. 2011). Thermal demagnetization of three-component IRMs for the studied layer indicates that the major magnetic carriers dominantly have medium to low coercivity with a significant unblocking temperature of ∼360°C and a minor component that unblocks fully at ∼585°C or even ∼680°C (e.g. Figs 7b and c, f and g and j and k). These and the above results indicate the dominant presence of greigite and traces of magnetite and/or haematite (especially in the upper sections) in the studied sediments. In contrast, IRMs of samples outside of this layer mainly unblock below 600°C, with some unblocking above 650°C (e.g. Figs 7a, d, e, h, i and l), which indicates the magnetic dominance of magnetite with some haematite. These results also confirm that magnetite is still the major magnetic mineral component in some samples (e.g. Fig. 7d), even if the Verwey transition is obscure (e.g. Fig. 6d), which can be caused by high degrees of oxidation and/or isomorphous substitution of other cations into the magnetite crystal lattice (Özdemir et al. 1993). Figure 7. View largeDownload slide Thermal demagnetization results for a three-component IRM for representative samples from cores: (a)–(d) EY02–2, (e)–(h) NHH01 and (i)–(l) DLC70–3, respectively. They were produced by magnetizing the cubic specimens in fields of 2.7, 0.5 and 0.05-T fields along the sample z-, y- and x-axes, respectively, to impart three mutually orthogonal IRMs, and then thermally demagnetizing them to 690°C at 20–50°C intervals. (e)–(g) Results are from Liu et al. (2014a). Figure 7. View largeDownload slide Thermal demagnetization results for a three-component IRM for representative samples from cores: (a)–(d) EY02–2, (e)–(h) NHH01 and (i)–(l) DLC70–3, respectively. They were produced by magnetizing the cubic specimens in fields of 2.7, 0.5 and 0.05-T fields along the sample z-, y- and x-axes, respectively, to impart three mutually orthogonal IRMs, and then thermally demagnetizing them to 690°C at 20–50°C intervals. (e)–(g) Results are from Liu et al. (2014a). Overall, the rock magnetic data suggest that greigite is the principal magnetic mineral in the layer at ∼50 m in cores EY02–2 and DLC70–3 as in core NHH01. The proportion of greigite in the magnetic mineral assemblages in the thick layers increases downward, as indicated by both FORC diagrams (e.g. Figs 5b and c, f and g and j and k) and temperature-dependent magnetization curves (e.g. Figs 6b and c, f and g and j and k; Figs 7b and c, f and g and j and k). In addition to the thick greigite-bearing layer, several thin greigite-bearing layers occur (yellow bars in Fig. 4), which cannot be correlated laterally among the three studied cores. 4.2 Trace elements Cd concentrations in core DLC70–3 range from 0.055 to 0.307 ppm, and average 0.140 ppm (Fig. 3g). Cd reaches marked maxima across the depth interval of 37.94–54.64 m (0.178 ppm on average), which corresponds to the thick greigite-bearing layer. Other thin greigite-bearing layers also have relatively high Cd concentrations (Fig. 3g). The covariant occurrence of Cd and greigite indicate that a weakly sulphidic diagenetic environment was established, which favoured greigite formation and preservation before pyritization reactions could be completed. We reach this conclusion based on the following observations: (1) Cd2+ has more rapid water exchange reaction kinetics than Fe2+, which results in CdS precipitating prior to FeS formation and subsequent pyrite formation (Morse & Luther 1999); (2) insoluble CdS occurs as a discrete phase rather than being associated with iron sulphides (Huerta-Diaz & Morse 1992) and (3) solid phase Cd in anoxic (i.e. strongly sulphidic) sediments is unstable and dissolved Cd diffuses upward and is deposited into sediments across the ferruginous/sulphidic boundary (Wei et al. 2005; see Canfield & Thamdrup (2009) and Roberts (2015) for terminology of diageneticzonations). 5 DISCUSSION Scanning electron microscopy observations have demonstrated that iron sulphides in core NHH01 occur as euhedral crystals either isolated and dispersed within clay matrix or in closely packed aggregates filling sediment voids (Liu et al.2014a), without being restricted to confined microenvironments like foraminifer tests where reducing conditions may develop locally (e.g. Rey et al. 2005). Therefore, the thick greigite-bearing sediment layer horizontally widespread in the central SYS indicates the long-term existence of sulphidic sedimentary environments in which the greigite formed. Generally, in the presence of abundant organic matter and porewater sulphide, pyritization reactions will proceed to completion, making greigite preservation less likely and pyrite preservation more likely (Berner 1984; Kao et al. 2004; Roberts & Weaver 2005). However, when organic matter supply is limited, coupled with high reactive iron concentrations, H2S can be removed so effectively that pyritization is inhibited, and the intermediate phase, greigite, will be preserved (Kao et al. 2004). The intermittent rather than continuous presence of greigite in the studied sediments (Fig. 4) is suggestive of time-varying diagenetic conditions known as non-steady-state diagenesis (Roberts 2015). Several forcing mechanisms, including hydrocarbon seepage, gas hydrate migration, tectonic deformation and associated fluid migration, large amplitude sea level changes on continental shelves, and submarine landslide events have been suggested to give rise to non-steady-state diagenesis that can induce intermittent greigite formation (Roberts & Weaver 2005). Tectonic deformation in this region was relatively weak in the mid-Pleistocene (Wan & Hao 2009), and gas hydrates and hydrocarbons have not been observed at such shallow depths in these sediments. Therefore, the first four mechanisms are unlikely to explain the documented greigite occurrences. Submarine landslide events are also unlikely considering the relatively flat seafloor of the YS (Qin et al. 1989). Large-amplitude sea-level changes on the continental shelf could result in variable bottom water oxygenation (e.g. Oda & Torii 2004; Roberts 2015). However, this alone does not explain why the thick greigite-bearing layer is only observed over a certain mid-Pleistocene interval when sea-level fluctuated cyclically (e.g. Fig. 2e, Rohling et al. 2014). Ge et al. (2003) observed that surface sediments with lower magnetic susceptibilities are mainly distributed in the CSYSM and suggested that a strong thermocline and pycnocline in the cold eddy prevents vertical mixing of water masses and results in a reducing diagenetic sedimentary environment in which iron sulphide minerals (such as pyrite) form. Greigite is only detected in the lower section of core YSDP103 (Fig. 1b), so that more complete pyritization has been invoked to explain the lack of greigite in the upper post-middle-Holocene (∼6 ka BP) section when the YSWC and its concomitant YSCWM formed and produced a more reducing environment (Liu et al. 2005). A recent study of core DLC70–3 (Fig. 1b) indicates that these hydrographic features were active as early as MIS 21 (Figs 3a–f, Mei et al. 2016). Sediments deposited during strong YSCWM conditions have characteristic low saturation IRM (SIRM) values (Fig. 3h), which indicates that detrital magnetic minerals in the sediments have been subjected to intense diagenesis, as reported by Ge et al. (2003) and Liu et al. (2005). In contrast, indices for the YSWC and YSCWM are not clearly observed in the greigite-bearing layers which have much higher SIRM values (Fig. 3). This should correspond to higher magnetic mineral concentrations resulting from greigite formation. It has been proposed that the availability of dissolved sulphide and iron are not the main constraints on pyrite formation and preservation of ferrimagnetic iron sulphides, but that it is the amount of available organic matter that acts as the controlling factor on diagenetic magnetic mineral formation in marine settings (Sagnotti et al. 2001; Liu et al. 2005). The relatively low TOC and TS contents of the greigite-bearing layers compared to the YSCWM layers (Figs 3d and e) indicate that more organic matter and sulphide were present in the latter to induce complete conversion to pyrite. Sulphate is one of the most abundant chemical compounds in seawater. Both the concentration of porewater sulphide and sulphur in sediments are controlled by organic matter availability since sulphate-reducing conditions enable incorporation of excess sulphur either as organic (e.g. biosynthetic) or inorganic (e.g. pyrite), in the presence of organic matter (Shi et al. 2016). These factors further indicate the dominant role of organic matter supply in magnetic mineral diagenesis. Based on published (Zhuang et al. 2002; Ge et al. 2006; Liu et al.2012, 2014b; Mei et al.2013, 2016) and reconstructed age models (Fig. 2), the thick greigite-bearing layers correspond to MIS 13–17 for cores EY02–2 and DLC70–3, and MIS 13–15 for core NHH01, respectively (Fig. 4), which succeeded a period with enhanced YSWC and YSCWM that was characterized by high primary productivity and strongly sulphidic conditions as indicated by elevated TOC, TS and Mo contents in MIS 19–21 (Figs 3d–f, Mei et al. 2016). In contrast, as mentioned above, trace element Cd enrichments in the greigite-bearing layer (Fig. 3g) indicate that the H2S released after CdS precipitation was limited and was exhausted by greigite formation. Otherwise, neither greigite nor greenockite (CdS) could be preserved, which suggests the potential of Cd as a promising indicator of greigite preservation in sediments. In addition, this further indicates a limited organic matter supply to the greigite-enriched layer. We propose that primary productivity provided the dominant control on sedimentary organic matter supply, which was vital to preservation of the thick greigite layer. YSCWM Influences should also be considered in relation to formation of the studied greigite-bearing layers. First, it could have prevented vertical mixing of water bodies and facilitated reducing conditions (e.g. Ge et al. 2003). Second, the discontinuous presence of greigite in core DLC70–3 compared to the other two cores (Fig. 4) is likely to be due to its location on the edge of the YSCWM deposition area (Fig. 1b), which made it fluctuate between reducing and oxic environmental conditions. Furthermore, the central location of core NHH01 in the CSYSM (Fig. 1b) is likely to have positioned it in a more strongly sulphidic environment during MIS 17–15, which resulted in a thinner greigite-bearing layer (Fig. 3). Lastly, greigite is not detected in coeval sediments in cores QC2 (Zhou & Ge 1990) and CSDP-1 (Liu et al. 2016) close to the CSYSM (Fig. 1b). These cores, along with results of Mei et al. (2016), could also indicate that the sphere of influence of the ancient cold eddy was no smaller than it is at present. A conceptual model is presented here to explain why the thick greigite-bearing layer only occurs over a certain stratigraphic interval. After a period with higher sea levels during MIS 21–19 (Mei et al. 2016), a large amplitude regression caused by tectonic uplift occurred during MIS 18 when seawater retreated from the central YS (Fig. 2, Liu et al. 2016). Subsequently, seawater covered the continental shelf again, but with lower sea levels than that in previous interglacials (MIS 21–19), which could have given rise to incursions of KC waters into the YS to form a relatively weak YSWC and YSCWM. The former maintained a moderate productivity level and, thus, a limited organic matter supply while the latter prevented vertical mixing of water bodies and facilitated an oxygen-free sedimentary environment, which together created sulphidic diagenetic conditions favourable for greigite formation and preservation. Substantial tectonic subsidence needed to have occurred to cause the relatively high sea level in MIS 16 when global sea level was in a marked lowstand (e.g. Fig. 2e, Rohling et al. 2014). In addition, an increasing number of sedimentary records from both hemispheres indicate that MIS 14 was the warmest mid-Pleistocene glacial period (e.g. Rackebrandt et al. 2011), which could be the principal reason for relatively high sea levels during this period in the YS. These conditions created ideal long-term conditions for greigite formation and preservation. 6 CONCLUSIONS By integrating rock magnetic measurements and trace element analyses on three long sediment cores from an N-S transect in the central SYS, we have obtained new insights into the distribution, origin and palaeoenvironmental implications of a thick greigite-bearing layer. The thick studied greigite-bearing layer is distributed laterally in the central SYS. The spatial extent of this layer could be as large as the present central SYS muddy area. Formation and preservation of the greigite was controlled by a limited organic matter supply that resulted from moderate primary productivity during MIS 17–13 when flows of the YSWC and the YSCWM were relatively weak. High Cd concentrations in the greigite-bearing layers indicate that conditions free of oxygen and H2S were more favourable for greigite preservation after its formation in sulphidic environments. This observation also suggests the potential of Cd as a promising index for indicating the presence of sedimentary greigite. Substantial local subsidence during MIS 16 and the unusually warm MIS 14 glacial should have maintained appropriate sea levels during the MIS 17–13, which created ideal material and ambient conditions for formation and preservation of the identified greigite. Therefore, we conclude that greigite has a widespread distribution over a considerable sediment thickness in early mid-Pleistocene sediments from the central SYS muddy area so that its presence was influenced by both local tectonic and global palaeoclimatic factors. ACKNOWLEDGEMENTS The data in this paper can be obtained by contacting the corresponding authors directly. We are grateful to Andrew Roberts and Daniel Rey and the Editors for their constructive comments and suggestions that greatly improved the manuscript. We have benefitted greatly from discussions with Xianwei Meng, Li Li and Hongmin Wang about redox sensitive elements. We thank the crew of R/V Kan407 for their assistance with coring. This work was jointly supported by the Natural Science Foundation of Shandong Province (grant no. ZR2016DB04), the Basic Scientific Fund for National Public Research Institutes of China (grant no. 2016Q06), the China Postdoctoral Science Foundation (grant no. 2016M592121), the National Natural Science Foundation of China (grant nos 41606075 and U1606401) and the China Geological Survey (grant nos GZH200800501 and 1212011220113). REFERENCES Alexander C.R., DeMaster D.J., Nittrouer C.A., 1991. Sediment accumulation in a modern epicontinental-shelf setting: the Yellow Sea, Mar. Geol. , 98, 51– 72. https://doi.org/10.1016/0025-3227(91)90035-3 Google Scholar CrossRef Search ADS   Berner R.A., 1984. Sedimentary pyrite formation: an update, Geochim. Cosmochim. Acta , 48, 605– 615. https://doi.org/10.1016/0016-7037(84)90089-9 Google Scholar CrossRef Search ADS   Canfield D.E., Thamdrup B., 2009. Towards a consistent classification scheme for geochemical environments, or, why we wish the term ‘suboxic’ would go away, Geobiology , 7, 385– 392. https://doi.org/10.1111/j.1472-4669.2009.00214.x Google Scholar CrossRef Search ADS PubMed  Chang L., Roberts A.P., Tang Y., Rainford B.D., Muxworthy A.R., Chen Q.W., 2008. Fundamental magnetic parameters from pure synthetic greigite (Fe3S4), J. geophys. Res. , 113, B06104, doi:10.1029/2007JB005502. https://doi.org/10.1029/2007JB005502 Google Scholar CrossRef Search ADS   Chang L., Roberts A.P., Rowan C.J., Tang Y., Pruner P., Chen Q.W., Horng C.-S., 2009. Low-temperature magnetic properties of greigite (Fe3S4), Geochem. Geophys.Geosyst. , 10, Q01Y04, doi:10.1029/2008GC002276. https://doi.org/10.1029/2008GC002276 Google Scholar CrossRef Search ADS   Chang L., Heslop D., Roberts A.P., Rey D., Mohamed K.J., 2016. Discrimination of biogenic and detrital magnetite through a double Verwey transition temperature, J. geophys. Res. , 121, 3– 14. https://doi.org/10.1002/2015JB012485 Google Scholar CrossRef Search ADS   Frank U., Nowaczyk N.R., Negendank J.F.W., 2007. Rock magnetism of greigite bearing sediments from the Dead Sea, Israel, Geophys. J. Int. , 168, 921−934. Fu Y.Z., von Dobeneck T., Franke C., Heslop D., Kasten S., 2008. Rock magnetic identification and geochemical process models of greigite formation in Quaternary marine sediments from the Gulf of Mexico (IODP Hole U1319A), Earth planet. Sci. Lett. , 275, 233– 245. https://doi.org/10.1016/j.epsl.2008.07.034 Google Scholar CrossRef Search ADS   Ge S.L., Shi X.F., Han Y.B., 2003. Distribution characteristics of magnetic susceptibility of the surface sediments in the southern Yellow Sea, Chin. Sci. Bull. , 48, 37– 41. https://doi.org/10.1007/BF02900938 Google Scholar CrossRef Search ADS   Ge S.L., Shi X.F., Zhu R.X., Liu Y.G., Yin P., Liu L.J., 2006. Magnetostratigraphy of borehole EY02-2 in the southern Yellow Sea and its paleoenvironmental significance, Chin. Sci. Bull. , 51, 855−865. Guan B.X., 1994. Pattern and structures of the current in Bohai, Huanghai (Yellow) and East China Seas, in Oceanology of China Seas , eds Zhou D., Liang Y.B., Zeng C.K., pp. 3−16. Kluwer Academic Publishing, Dordrecht. Harrison R.J., Feinberg J.M., 2008. FORCinel: an improved algorithm for calculating first-order reversal curve distributions using locally weighted regression smoothing, Geochem. Geophys.Geosyst. , 9, Q05016, doi:10.1029/2008GC001987. https://doi.org/10.1029/2008GC001987 Google Scholar CrossRef Search ADS   Hu D.X., 1984. Upwelling and sedimentation dynamics I. The role of upwelling in sedimentation in the Huanghai Sea and East China Sea−a description of general features, Chin. J. Oceanol. Limnol. , 2, 2−19 (in Chinese, with English summary). Huerta-Diaz M., Morse J.W., 1992. Pyritization of trace metals in anoxic marine sediments, Geochim. Cosmochim. Acta , 56, 2681– 2702. https://doi.org/10.1016/0016-7037(92)90353-K Google Scholar CrossRef Search ADS   Kao S.-J., Horng C.-S., Roberts A.P., Liu K.-K., 2004. Carbon–sulfur–iron relationships in sedimentary rocks from southwestern Taiwan: influence of geochemical environment on greigite and pyrrhotite formation, Chem. Geol. , 203, 153– 168. https://doi.org/10.1016/j.chemgeo.2003.09.007 Google Scholar CrossRef Search ADS   Kim D., Park B.K., Shin I., 1999. Paleoenvironmental changes of the Yellow Sea during the Late Quaternary, Geo-Mar. Lett. , 18, 189−194. Lan X.H., Wang H.X., Li R.H., Lin Z.H., Zhang Z.X., 2007. Major elements composition and provenance analysis in the sediments of the South Yellow Sea, Earth Sci. Front. , 14, 197– 203 (in Chinese, with English summary). https://doi.org/10.1016/S1872-5791(07)60033-5 Google Scholar CrossRef Search ADS   Lee C.H., Jin J.-H., 1995. Authigenicgreigite in mud from the continental shelf of the Yellow Sea, off the southwest Korean Peninsula, Mar. Geol. , 128, 11– 15. https://doi.org/10.1016/0025-3227(95)00099-K Google Scholar CrossRef Search ADS   Lisiecki L.E., Raymo M.E., 2005. A Pliocene-Pleistocene stack of 57 globally distributed benthic δ18O records, Palaeoceanography , 20, PA1003, doi:10.1029/2004PA001071. Liu J., Zhu R.X., Roberts A.P., Li S.Q., Chang J.-H., 2004. High-resolution analysis of early diagenetic effects on magnetic minerals in post-middle-Holocene continental shelf sediments from the Korea Strait, J. geophys. Res. , 109, B03103, doi:10.1029/2003JB002813. https://doi.org/10.1029/2003JB002813 Liu J., Zhu R.X., Li S.Q., Chang J.-H., 2005. Magnetic mineral diagenesis in the post-glacial muddy sediments from the southeastern South Yellow Sea: response to marine environmental changes, Sci. China Earth Sci. , 48, 134−144. Liu J., Satio Y., Kong X.H., Wang H., Wen Z., Yang Z.G., Nakashima R., 2010. Delta development and channel incision during marine isotope stages 3 and 2 in the western South Yellow Sea, Mar. Geol. , 278, 54– 76. https://doi.org/10.1016/j.margeo.2010.09.003 Google Scholar CrossRef Search ADS   Liu J.X., Shi X.F., Ge S.L., Li X.Y., Yao Z.Q., Wang X., 2012. Record of geomagnetic excursion during early Holocene in the central southern Yellow Sea Mud, Quat. Res. , 32, 663−669 (in Chinese, with English summary). Liu J.X., Shi X.F., Ge S.L., Liu Q.S., Yao Z.Q., Yang G., 2014a. Identification of the thick-layer greigite in sediments of the South Yellow Sea and its geological significances, Chin. Sci. Bull. , 59, 2764−2775. Liu J.X. et al.  , 2014b. Magnetostratigraphy of a greigite-bearing core from the South Yellow Sea: implications for remagnetization and sedimentation, J. geophys. Res. , 119, 7425– 7441. https://doi.org/10.1002/2014JB011206 Google Scholar CrossRef Search ADS   Liu J.X., Liu Q.S., Zhang X.H., Liu J., Wu Z.Q., Mei X., Shi X.F., Zhao Q.H., 2016. Magnetostratigraphy of a long Quaternary sediment core in the South Yellow Sea, Quat. Sci. Rev. , 144, 1– 15. https://doi.org/10.1016/j.quascirev.2016.05.025 Google Scholar CrossRef Search ADS   Lowrie W., 1990. Identification of ferromagnetic minerals in a rock by coercivity and unblocking temperature properties, Geophys. Res. Lett. , 17, 159– 162. https://doi.org/10.1029/GL017i002p00159 Google Scholar CrossRef Search ADS   McMillan A.A., 2002. Onshore Quaternary geological surveys in the 21st century—a perspective from the British Geological Survey, Quat. Sci. Rev. , 21, 889– 899. https://doi.org/10.1016/S0277-3791(01)00064-6 Google Scholar CrossRef Search ADS   Mei X., Zhang X.H., Li R.H., 2013. Distribution of late Quaternary benthic foraminifera in South Yellow Sea and its implication of paleo-water mass, Geol. Rev. , 59, 1024−1034 (in Chinese, with English summary). Mei X. et al.  , 2016. Evolution of the Yellow Sea warm current and the Yellow Sea cold water mass since the Middle Pleistocene, Palaeogeog. Palaeoclimatol. Palaeoecol. , 442, 48– 60. https://doi.org/10.1016/j.palaeo.2015.11.018 Google Scholar CrossRef Search ADS   Milliman J.D., Meade R.H., 1983. World-wide delivery of river sediment to the oceans, J. Geol. , 91, 1– 21. https://doi.org/10.1086/628741 Google Scholar CrossRef Search ADS   Morin F.J., 1950. Magnetic susceptibility of αFe2O3 and αFe2O3 with added Titanium, Phys. Rev. , 78, 819– 820. https://doi.org/10.1103/PhysRev.78.819.2 Google Scholar CrossRef Search ADS   Morse J.W., Luther G.W., 1999. Chemical influences on trace metal-sulfide interactions in anoxic sediments, Geochim. Cosmochim. Acta , 63, 3373– 3378. https://doi.org/10.1016/S0016-7037(99)00258-6 Google Scholar CrossRef Search ADS   Muxworthy A.R., Dunlop D.J., 2002. First-order reversal curve (FORC) diagrams for pseudo-single-domain magnetites at high temperature, Earth planet. Sci. Lett. , 203, 369– 382. https://doi.org/10.1016/S0012-821X(02)00880-4 Google Scholar CrossRef Search ADS   Oda H., Torii M., 2004. Sea-level change and remagnetization of continental shelf sediments off New Jersey (ODP Leg 174A): magnetite and greigite diagenesis, Geophys. J. Int. , 156, 443– 458. https://doi.org/10.1111/j.1365-246X.2004.02162.x Google Scholar CrossRef Search ADS   Özdemir Ö., Dunlop D.J., Moskowitz B.M., 1993. The effect of oxidation on the Verwey transition in magnetite, Geophys. Res. Lett. , 20, 1671– 1674. https://doi.org/10.1029/93GL01483 Google Scholar CrossRef Search ADS   Pattan J.N., Pearce N.J.G., 2009. Bottom water oxygenation history in southeastern Arabian Sea during the past 140ka: results from redox-sensitive elements, Palaeogeog. Palaeoclimat. Palaeoecol. , 280, 396– 405. https://doi.org/10.1016/j.palaeo.2009.06.027 Google Scholar CrossRef Search ADS   Pike C.R., Roberts A.P., Verosub K.L., 1999. Characterizing interactions in fine magnetic particle systems using first order reversal curves, J. appl. Phys. , 85, 6660– 6667. https://doi.org/10.1063/1.370176 Google Scholar CrossRef Search ADS   Qin Y.S., Zhao Y.Y., Chen L.R., Zhao S.L., 1989. Geology of the Yellow Sea , China Ocean Press, pp. 1−289 (in Chinese). Rackebrandt N., Kuhnert H., Groeneveld J., Bickert T., 2011. Persisting maximum Agulhas leakage during MIS 14 indicated by massive Ethmodiscus oozes in the subtropical South Atlantic, Paleoceanography , 26, 130– 137. https://doi.org/10.1029/2010PA001990 Google Scholar CrossRef Search ADS   Reimer P.J. et al.  , 2013. IntCal13 and Marine13 radiocarbon age calibration curves 0–50,000 years cal BP, Radiocarbon , 55, 1869– 1887. https://doi.org/10.2458/azu_js_rc.55.16947 Google Scholar CrossRef Search ADS   Rey D., Mohamed K.J., Bernabeu A., Rubio B., Vilas F., 2005. Early diagenesis of magnetic minerals in marine transitional environments: geochemical signatures of hydrodynamic forcing, Mar. Geol. , 215, 215– 236. https://doi.org/10.1016/j.margeo.2004.12.001 Google Scholar CrossRef Search ADS   Reynolds R.L., Rosenbaum J.C., van Metre P., Tuttle M., Callender E., Goldin A., 1999. Greigite as an indicator of drought − the 1912–1994 sediment magnetic record from White Rock Lake, Dallas, Texas, USA, J. Paleolimnol. , 21, 193– 206. https://doi.org/10.1023/A:1008027815203 Google Scholar CrossRef Search ADS   Roberts A.P., 1995. Magnetic properties of sedimentary greigite (Fe3S4), Earth planet. Sci. Lett. , 134, 227– 236. https://doi.org/10.1016/0012-821X(95)00131-U Google Scholar CrossRef Search ADS   Roberts A.P., 2008. Geomagnetic excursions: knowns and unknowns, Geophys. Res. Lett. , 35, L17307, doi:10.1029/2008GL034719. https://doi.org/10.1029/2008GL034719 Google Scholar CrossRef Search ADS   Roberts A.P., 2015. Magnetic mineral diagenesis, Earth-Sci. Rev. , 151, 1– 47. https://doi.org/10.1016/j.earscirev.2015.09.010 Google Scholar CrossRef Search ADS   Roberts A.P., Reynolds R.L., Verosub K.L., Adam D.P., 1996. Environmental magnetic implications of Greigite (Fe3S4) Formation in a 3 m.y. lake sediment record from Butte Valley, northern California, Geophys. Res. Lett. , 23, 2859– 2862. https://doi.org/10.1029/96GL02831 Google Scholar CrossRef Search ADS   Roberts A.P., Pike C.R., Verosub K.L., 2000. First-order reversal curve diagrams: a new tool for characterizing the magnetic properties of natural samples, J. geophys. Res. , 105, 28461– 28475. https://doi.org/10.1029/2000JB900326 Google Scholar CrossRef Search ADS   Roberts A.P., Weaver R., 2005. Multiple mechanisms of remagnetization involving sedimentary greigite (Fe3S4), Earth planet. Sci. Lett. , 231, 263– 277. https://doi.org/10.1016/j.epsl.2004.11.024 Google Scholar CrossRef Search ADS   Roberts A.P., Liu Q.S., Rowan C.J., Chang L., Carvallo C., Torrent J., Horng C.-S., 2006. Characterization of hematite (α-Fe2O3), goethite (α-FeOOH), greigite (Fe3S4), and pyrrhotite (Fe7S8) using first-order reversal curve diagrams, J. geophys. Res. , 111, B12S35, doi:10.1029/2006JB004715. https://doi.org/10.1029/2006JB004715 Google Scholar CrossRef Search ADS   Roberts A.P., Chang L., Rowan C.J., Horng C.-S., Florindo F., 2011. Magnetic properties of sedimentary greigite (Fe3S4): an update, Rev. Geophys. , 49, RG1002, doi:10.1029/2010RG000336.. https://doi.org/10.1029/2010RG000336 Google Scholar CrossRef Search ADS   Roberts A.P., Heslop D., Zhao X., Pike C.R., 2014. Understanding fine magnetic particle systems through use of first-order reversal curve diagrams, Rev. Geophys. , 52, 557– 602. https://doi.org/10.1002/2014RG000462 Google Scholar CrossRef Search ADS   Rochette P., Fillion G., Mattéi J.-L., Dekkers M.J., 1990. Magnetic transition at 30–34 Kelvin in pyrrhotite: insight into a widespread occurrence of this mineral in rocks, Earth planet. Sci. Lett. , 98, 319– 328. https://doi.org/10.1016/0012-821X(90)90034-U Google Scholar CrossRef Search ADS   Rohling E.J., Foster G.L., Grant K.M., Marino G., Roberts A.P., Tamisiea M.E., Williams F., 2014. Sea-level and deep-sea-temperature variability over the past 5.3 million years, Nature , 508, 477– 482. https://doi.org/10.1038/nature13230 Google Scholar CrossRef Search ADS PubMed  Rosenthal Y., Lam P., Boyle E.A., Thomson J., 1995. Authigenic cadmium enrichments in suboxic sediments: precipitation and postdepositional mobility, Earth planet. Sci. Lett. , 132, 99– 111. https://doi.org/10.1016/0012-821X(95)00056-I Google Scholar CrossRef Search ADS   Sagnotti L., Macri P., Camerlenghi A., Rebesco M., 2001. Environmental magnetism of Antarctic late Pleistocene sediments and interhemispheric correlation of climatic events, Earth planet. Sci. Lett. , 192, 65– 80. https://doi.org/10.1016/S0012-821X(01)00438-1 Google Scholar CrossRef Search ADS   Sagnotti L., Cascella A., Ciaranfi N., Macrì P., Maiorano P., Marino M., Taddeucci J., 2010. Rock magnetism and palaeomagnetism of the MontalbanoJonico section (Italy): evidence for late diagenetic growth of greigite and implications for magnetostratigraphy, Geophys. J. Int. , 180, 1049– 1066. https://doi.org/10.1111/j.1365-246X.2009.04480.x Google Scholar CrossRef Search ADS   Shi X.F. et al.  , 2016. Sedimentary architecture of the Bohai Sea China over the last 1 Ma and implications for sea-level changes, Earth planet. Sci. Lett. , 451, 10– 21. https://doi.org/10.1016/j.epsl.2016.07.002 Google Scholar CrossRef Search ADS   Skinner B.J., Erd R.C., Grimaldi F.S., 1964. Greigite, the thiospinel of iron, a new mineral, Am. Mineral. , 49, 543−555. Snowball I.F., 1991. Magnetic hysteresis properties of greigite (Fe3S4) and a new occurrence in Holocene sediments from Swedish Lappland, Phys. Earth planet. Inter. , 68, 32– 40. https://doi.org/10.1016/0031-9201(91)90004-2 Google Scholar CrossRef Search ADS   Southon J., Kashgarian M., Fontugne M., Metivier B., Yim W.W.S., 2002. Marine reservoir corrections for the Indian Ocean and Southeast Asia, Radiocarbon , 44, 167– 180. https://doi.org/10.1017/S0033822200064778 Google Scholar CrossRef Search ADS   Verwey E.J.W., 1939. Electronic conduction of magnetite (Fe3O4) and its transition point at low temperatures, Nature , 144, 327– 328. https://doi.org/10.1038/144327b0 Google Scholar CrossRef Search ADS   Wan T.F., Hao T.Y., 2009. Cenozoic tectonics of the Yellow Sea and its oil-gas exploration, Geoscience , 23, 385−393 (in Chinese, with English summary). Wang Q.W., Morse J.W., 1996. Pyrite formation under conditions approximating those in anoxic sediments I. Pathway and morphology, Mar. Chem. , 52, 99– 121. https://doi.org/10.1016/0304-4203(95)00082-8 Google Scholar CrossRef Search ADS   Wang Q., Zhu E.Q., 1989. Marine Sedimentology , China Science Press, pp. 1−254 (in Chinese). Wei G.J., Li X.H., Liu Y., Shao L., Liang X.R., 2005. Transfer patterns of Mn, Cd and Mo in sediments during early diagenesis: evidences from sediment cores at ODP Site 1148, Geochimica , 34, 129−135 (in Chinese, with English summary). Yang S.Y., Jung H.S., Lim D.I., Li C.X., 2003. A review on the provenance discrimination of sediments in the Yellow Sea, Earth-Sci. Rev. , 63, 93– 120. https://doi.org/10.1016/S0012-8252(03)00033-3 Google Scholar CrossRef Search ADS   Yao Z.Q., Guo Z.T., Xiao G.Q., Wang Q., Shi X.F., Wang X.Y., 2012. Sedimentary history of the western Bohai coastal plain since the late Pliocene: implications on tectonic, climatic and sea-level changes, J. Asian Earth Sci. , 54−55, 192−202. Yao Z.Q. et al.  , 2014. Paleomagnetic and astronomical dating of sediment core BH08 from the Bohai Sea, China: implications for glacial–interglacial sedimentation, Palaeogeog. Palaeoclimat. Palaeoecol. , 393, 90– 101. https://doi.org/10.1016/j.palaeo.2013.11.012 Google Scholar CrossRef Search ADS   Zhou M.Q., Ge Z.S., 1990. The magnetostratigraphy of unconsolidated sediments in the southern Yellow Sea and adjacent areas, Mar. Geol. Quat. Geol. , 10, 21−33 (in Chinese, with English summary). Zhuang L.H., Chang F.M., Li T.G., Yan J., 2002. Foraminiferal faunas and Holocene sedimentation rates of core EY02-2 in the South Yellow Sea, Mar. Geol. Quat. Geol. , 22, 7−14 (in Chinese, with English summary). © The Author(s) 2017. Published by Oxford University Press on behalf of The Royal Astronomical Society. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Geophysical Journal International Oxford University Press

Formation and preservation of greigite (Fe 3 S 4 ) in a thick sediment layer from the central South Yellow Sea

Loading next page...
 
/lp/ou_press/formation-and-preservation-of-greigite-fe-3-s-4-in-a-thick-sediment-h6q4CbBF1B
Publisher
The Royal Astronomical Society
Copyright
© The Author(s) 2017. Published by Oxford University Press on behalf of The Royal Astronomical Society.
ISSN
0956-540X
eISSN
1365-246X
D.O.I.
10.1093/gji/ggx556
Publisher site
See Article on Publisher Site

Abstract

Summary Sediments from continental shelves are sensitive to changes in both oceanic and terrestrial conditions, and, therefore, magnetic minerals in such sediments are affected strongly by depositional and diagenetic processes. Here, we investigated systematically an N-S transect of three sediment cores from the central South Yellow Sea (SYS) muddy area. Magnetic data indicate the presence of a horizontally distributed thick greigite-bearing layer. From an age model based on published magnetostratigraphy, accelerator mass spectrometry 14C dating ages, sedimentary characteristics and foraminiferal analysis, this layer was deposited within marine isotope stages (MIS) 17–13, following an enhanced sulphidic period over MIS 21–19 when the YS Warm Current and the associated YS Cold Water Mass were strong and where underlying sediments have higher total organic carbon, total sulphur and trace element molybdenum contents. Trace element cadmium enrichment in the greigite-bearing layers is documented for the first time, which indicates that weakly sulphidic (i.e. with trace levels of free H2S) conditions existed before greigite formed in a sulphidic environment during early diagenesis. It also indicates that subsequent conditions free of oxygen and H2S after greigite formation are more favourable for its preservation. We propose that organic matter supply was controlled over an extended period by moderate primary productivity. The combined effects of palaeoclimate and local tectonic subsidence were crucial for the formation and preservation of the identified greigite. In brief, our study improves understanding of the formation and preservation mechanisms of greigite in continental shelf sediments and reveals mid-Pleistocene palaeoenvironmental changes in the SYS. Sea level change, Environmental magnetism, Rock and mineral magnetism 1 INTRODUCTION The continental shelf provides a link between the land and deep oceans and plays an important role in geological processes, such as land–sea interaction, transfer and accumulation of terrigenous materials, and palaeoenvironmental evolution (McMillan 2002). With deposition at relatively shallow water depths (usually < 200 m), sediments in this region are highly sensitive to sea-level fluctuations, climate change and local tectonic subsidence (Yao et al. 2012). Large amounts of sediment carried by rivers are deposited on the continental shelf, which facilitates high-resolution studies of processes recorded by such sediments (e.g. Liu et al.2004). Generally, redox conditions of continental shelf sedimentary environments are influenced by multiple factors, including river and ocean hydrodynamics, which are controlled by climate and sea-level changes, terrigenous organic matter input, biological productivity, organic matter reactivity, etc. (Wang & Zhu 1989). All of these factors produce multiple features in continental shelf sedimentary archives. For example, complex processes are recorded by magnetic minerals in continental shelf sediments (e.g. Liu et al. 2004, 2016). Specifically, detrital and biogenic iron oxides (mainly magnetite, maghemite and/or haematite) can be altered to varying degrees during early diagenesis depending on redox conditions (Roberts 2015). In sulphate-reducing porewaters, a succession of iron sulphide phases will form, including mackinawite (FeS0.9), greigite (Fe3S4) and finally pyrite (FeS2) (Berner 1984). Authigenic greigite shares the same crystal structure as magnetite and is a strongly ferrimagnetic mineral that forms as a precursor to pyrite in sulphate-reducing sedimentary environments in association with chemical reactions driven by bacterial degradation of organic matter (Wang & Morse 1996). Although it was not first formally reported until 1964 in Miocene lake sediments in San Bernardino County, California (Skinner et al. 1964), its occurrence and importance remains underappreciated in most geochemical studies due to its potentially small concentrations as well as its metastable state with respect to pyrite (Kao et al. 2004). However, it has been extensively reported from globally distributed localities, where identification of its presence over the last 30 yr has been facilitated greatly by the development of modern superconducting rock magnetometers (Roberts et al. 2011). Largely independent of its palaeomagnetic importance, which includes the fact that greigite has been widely reported to record remagnetizations (e.g. Roberts & Weaver 2005; Sagnotti et al. 2010), it also has considerable potential as a palaeoenvironmental indicator, such as being a sensitive indicator of lake chemistry (e.g. Roberts et al. 1996; Reynolds et al. 1999; Frank et al. 2007) or marine environments (e.g. Fu et al. 2008). Widespread greigite preservation is likely if abundant reactive iron is available to react with limited dissolved sulphide during pyritization reactions (Kao et al. 2004). In the Yellow Sea (YS), which is a semi-enclosed sea resting on a typical continental shelf (Fig. 1), Lee & Jin (1995) discovered authigenic greigite in Holocene transgressive deposits off the southwestern Korean Peninsula (core 94P12 in Fig. 1b). They inferred that it probably formed in a reducing micro-environment around decaying organic matter. Subsequently, greigite was also detected magnetically in two post-middle-Holocene sedimentary sequences from the Korean Strait (cores SSDP102 and SSDP103 in Fig. 1b) and in a post-glacial muddy sediment core from the southeastern South YS (SYS, core YSDP103 in Fig. 1b) in studies of early diagenetic effects on magnetic minerals (Liu et al.2004, 2005). Recently, a thick (∼7 m) mid-Pleistocene greigite-bearing layer was identified via systematic magnetic analysis coupled with mineralogical observations in a sediment core (NHH01) from the central SYS (Fig. 1b, Liu et al.2014a,b). This is the oldest reported greigite in sedimentary strata in this region. However, the origin and distribution of this thick layer remains poorly constrained. In this study, we investigated systematically an N-S transect covering three long sediment cores in this area (Fig. 1b). We aim to outline the lateral distribution of this thick greigite-bearing layer, and its palaeoenvironmental significance. Figure 1. View largeDownload slide (a) Schematic map of the Yellow Sea (YS) and adjacent areas including major currents (modified after Mei et al.2016), with water depths in metres. Black arrows indicate paths of currents including the Kuroshio Current (KC), Tsushima Warm Current (TWC), Yellow Sea Warm Current (YSWC), Jiangsu Coastal Current (JCC), Yellow Sea Coastal Current (YSCC) and Korea Coastal Current (KCC). The blue dashed line indicates the spatial distribution of the Yellow Sea Cold Water Mass (YSCWM). CC and CS are abbreviations of Chengshan Cape and Changshangot, respectively. (b) Distribution of YS surface sediments (CSYSM, central South Yellow Sea muddy area) and adjacent seas (modified after Liu et al.2010), and locations of the studied cores (white dots) and other cores mentioned in the text (black dots). Figure 1. View largeDownload slide (a) Schematic map of the Yellow Sea (YS) and adjacent areas including major currents (modified after Mei et al.2016), with water depths in metres. Black arrows indicate paths of currents including the Kuroshio Current (KC), Tsushima Warm Current (TWC), Yellow Sea Warm Current (YSWC), Jiangsu Coastal Current (JCC), Yellow Sea Coastal Current (YSCC) and Korea Coastal Current (KCC). The blue dashed line indicates the spatial distribution of the Yellow Sea Cold Water Mass (YSCWM). CC and CS are abbreviations of Chengshan Cape and Changshangot, respectively. (b) Distribution of YS surface sediments (CSYSM, central South Yellow Sea muddy area) and adjacent seas (modified after Liu et al.2010), and locations of the studied cores (white dots) and other cores mentioned in the text (black dots). 2 GENERAL SETTING The YS is a broad and relatively shallow continental shelf on the western Pacific margin between the Chinese mainland and the Korean Peninsula (Fig. 1). It occupies a total area of 380 000 km2 and spans 870 km from north to south and 556 km from east to west. A line connecting Chengshan Cape (CC), Shangdong Peninsula and Changshangot (CS) on the Korean Peninsula divides the YS into two parts, the North Yellow Sea (NYS) and the SYS (Fig. 1a, Qin et al. 1989). Water depths in the NYS are mostly less than 60 m with an average of 38 m, while the SYS is mostly less than 100 m deep with an average water depth of 46 m. The SYS deepens toward the YS Trough, which is a topographic depression roughly defined by the 80-m isobath in the central SYS and extends NW–SE (Liu et al. 2010). With relatively stable subsidence over the Quaternary, sediments in the SYS reach up to a maximum thickness of ∼300 m (Qin et al. 1989; Liu et al. 2016). Although neither the Yellow River nor the Yangtze River currently empties into the YS directly (Fig. 1a), they are regarded to have governed sedimentation in major parts of this region during the Holocene (Yang et al. 2003). Based on mineralogical and geochemical studies on surface sediments, Lan et al. (2007) argued that these two large rivers provide the major source of sediments in the western and middle part of the YS, while rivers on the Korean Peninsula dominate in the east. At present, the Yellow River discharges ∼1.1 × 109 t yr−1 of sediment into Bohai Sea; 9−15 per cent of these (mainly fine-grained) sediments can be transported southward over long distances to be finally deposited in the central SYS (Alexander et al. 1991). In contrast, the Yangtze River annually empties ∼5 × 108 t of sediment into the southwestern YS and northern East China Sea (Milliman & Meade 1983), while Korean rivers discharge less than 3 × 107 t yr−1 into the eastern YS (Qin et al. 1989). The general circulation in the YS is currently dominated by the YS Warm Current (YSWC) and coastal currents (Liu et al. 2010). The northward YSWC is a branch of the Kuroshio Current (KC), which is an important Western Pacific boundary current that carries warm and salty water into the YS roughly along the YS Trough (Fig. 1a, Guan 1994). Instead, coastal currents with relatively low salinities along the Chinese and Korean coasts flow southward in winter and northward in summer, with the exception of the YS Coastal Current (YSCC) and Jiangsu Coastal Current (JSCC) which always flow southward during both summer and winter (Liu et al. 2010). In winter, the YSWC can intrude into the NYS or even the Bohai Sea because of the near homogeneity of the shelf water column (Guan 1994). However, in summer, the YS Cold Water Mass (YSCWM) is present in the central YS and intrudes southward (Fig. 1a), which hinders YSWC flow and makes it too weak to reach north of 35°N. This seasonal cold water eddy is considered to be responsible for the occurrence of the central SYS muddy area (CSYSM, Fig. 1b, Hu 1984). Although the warm current and cold water mass formed at around 6–7 kyr BP when sea level was close to the present level (e.g. Kim et al. 1999), the YSWC and YSCWM probably greatly affected sediment transportation and sedimentary environments even in the mid-Pleistocene (Mei et al. 2016). 3 MATERIALS AND METHODS 3.1 Core descriptions All of the studied sediment cores were recovered from the central SYS, within the CSYSM, which lies under the path of the present-day YSCWM (Fig. 1b). Water depths at the studied core locations shallow northward, with depths of 79 m for core EY02–2 (Zhuang et al. 2002), and 73 m for cores NHH01 (Liu et al. 2012) and DLC70–3 (Mei et al. 2013), respectively. More detailed information about the cores is given in Table 1. For the 125.64-m-long core NHH01 (∼1.06 Ma, Liu et al. 2014b), we only discuss the uppermost 71 m in this paper. Hereafter, NHH01 refers to the upper 71-m portion of the core. Table 1. Detailed information on the studied cores (from Zhuang et al. 2002; Liu et al.2012; Mei et al.2013). Core  Water depth (m)  Drilling depth (m)  Recovery rate (per cent)  Longitude/latitude  Drilling date  EY02–2  79.00  70.00  86.5  123°30΄E/34°30΄N  2001 Feb  NHH01  73.00  71.20  91.0  123°13΄E/35°13΄N  2009 Jun  DLC70–3  73.00  125.64  93.0  123°33΄E/36°38΄N  2009 Sep  Core  Water depth (m)  Drilling depth (m)  Recovery rate (per cent)  Longitude/latitude  Drilling date  EY02–2  79.00  70.00  86.5  123°30΄E/34°30΄N  2001 Feb  NHH01  73.00  71.20  91.0  123°13΄E/35°13΄N  2009 Jun  DLC70–3  73.00  125.64  93.0  123°33΄E/36°38΄N  2009 Sep  View Large Sediments in core EY02–2 are mainly clayey silt, sandy silt and silty sand (Fig. 2a, Ge et al. 2006). The sedimentary sequence in core NHH01 consists of silt and clay, with occasionally interbedded fine sand layers (Fig. 2f, Liu et al. 2014a), while sediments in core DLC70–3 are mainly silt, sandy silt and silty sand, with a nearly 5-m-thick sand layer in the intermediate section (Fig. 2k, Mei et al. 2016). Figure 2. View largeDownload slide General information on sedimentary sequences of the studied cores EY02–2 (Zhuang et al.2002; Ge et al.2006), NHH01 (Liu et al.2014b) and DLC70–3 (Mei et al.2016), respectively, including (a), (f) and (k) lithology, (b), (g) and (l) mean grain size (Mz) of the sediments in ϕ logarithmic units, (c), (h) and (m) abundance of benthic foraminifera (B.F.), (d), (i) and (n) magnetostratigraphy (the ages of geomanetic excursions are from Roberts (2008), which are 33 ka for M (Mono Lake) and 41 ka for L (Laschamp), respectively; B/M is the abbreviation of Matuyama–Brunhes boundary with an age of ∼780 ka), and (e), (j) and (o) relative sea-level (RSL) variations from Rohling et al. (2014), with the corresponding marine isotope stages (MIS) labeled in blue numbers. Figure 2. View largeDownload slide General information on sedimentary sequences of the studied cores EY02–2 (Zhuang et al.2002; Ge et al.2006), NHH01 (Liu et al.2014b) and DLC70–3 (Mei et al.2016), respectively, including (a), (f) and (k) lithology, (b), (g) and (l) mean grain size (Mz) of the sediments in ϕ logarithmic units, (c), (h) and (m) abundance of benthic foraminifera (B.F.), (d), (i) and (n) magnetostratigraphy (the ages of geomanetic excursions are from Roberts (2008), which are 33 ka for M (Mono Lake) and 41 ka for L (Laschamp), respectively; B/M is the abbreviation of Matuyama–Brunhes boundary with an age of ∼780 ka), and (e), (j) and (o) relative sea-level (RSL) variations from Rohling et al. (2014), with the corresponding marine isotope stages (MIS) labeled in blue numbers. 3.2 Age models and sedimentary history Radiocarbon ages used in the text were all obtained via accelerator mass spectrometer (AMS) 14C dating on mixed benthic foraminifera, and were converted to calibrated calendar (cal) ages by using the CALIB 7.0.4 program, an updated calibration curve, MARINE13 (Reimer et al. 2013), and a surface-ocean reservoir correction of ▵R = −100 ± 36 yr (Southon et al. 2002). Calendar ages are given with one standard deviation (1σ) uncertainty (Table 2). Table 2. List of AMS 14C dating on mixed benthic foraminifera of the studied cores. Core  Depth  Conventional age   Calendar ages (calyr BP)  References for    (m)  (yr)  Intercept  Range (1σ)  conventional age  EY02–2  1.05  8360 ± 60  7130  7036 − 7224  Zhuang et al. (2002)    2.00  10 600 ± 45  10 150  10 011 − 10 282  Zhuang et al. (2002)  NHH01  4.59  9230 ± 50  8220  8159 − 8286  Liu et al. (2012)    5.69  11 850 ± 65  11 450  11 374 − 11 518  Liu et al. (2012)    7.35  18 750 ± 130  22 310  22 159 − 22 469  This study  DLC70–3  0.04  9890 ± 50  9040  8931 − 9140  Mei et al. (2013)    2.24  10 200 ± 50  9340  9242 − 9441  Mei et al. (2013)    5.54  37 900 ± 350  40 030  39 772 − 40 278  Mei et al. (2013)  Core  Depth  Conventional age   Calendar ages (calyr BP)  References for    (m)  (yr)  Intercept  Range (1σ)  conventional age  EY02–2  1.05  8360 ± 60  7130  7036 − 7224  Zhuang et al. (2002)    2.00  10 600 ± 45  10 150  10 011 − 10 282  Zhuang et al. (2002)  NHH01  4.59  9230 ± 50  8220  8159 − 8286  Liu et al. (2012)    5.69  11 850 ± 65  11 450  11 374 − 11 518  Liu et al. (2012)    7.35  18 750 ± 130  22 310  22 159 − 22 469  This study  DLC70–3  0.04  9890 ± 50  9040  8931 − 9140  Mei et al. (2013)    2.24  10 200 ± 50  9340  9242 − 9441  Mei et al. (2013)    5.54  37 900 ± 350  40 030  39 772 − 40 278  Mei et al. (2013)  View Large For core EY02–2, two AMS 14C ages of ∼7100 and ∼10 200 cal yr BP were determined at depths of 1.05 and 2.00 m, respectively (Table 2, Zhuang et al. 2002). Magnetostratigraphic analysis defines the Matuyama–Brunhes boundary (B/M, ∼780 kyr) at 63.29 m, with an extrapolated basal age of ∼0.9 Ma (Fig. 2d, Ge et al. 2006). Sediments in core EY02–2 were largely deposited in littoral and neritic settings, which are referred to collectively as marine deposits and are interrupted by three fluvial layers (2.5 − 3.0, 58.7 − 61.8 and 68 − 70 m) based on benthic foraminiferal analysis and sedimentary characteristics (Figs 2a–c, Zhuang et al. 2002). The B/M in core NHH01 is located at a depth of 68.64 m (Fig. 2i, Liu et al. 2014b), which, combined with the AMS 14C ages (Table 2, Liu et al. 2012), suggest that the core records deposition from the mid-Pleistocene to Holocene. Based on benthic foraminiferal analyses and sedimentary characteristics (Figs 2f–h, Liu et al. 2014a,b), five fluvial layers are identified in the core, with depth ranges of roughly 14.9 − 15.9, 17.9 − 18.4, 33.1 − 34.5, 62.8 − 66.3 and 69.9 − 71 m, respectively; sediments from other parts of the core are mainly marine deposits. AMS 14C ages for core DLC70–3 indicate that sedimentation has not occurred since ∼9000 cal yr BP (Table 2, Mei et al. 2013), which indicates that the drilling location occurs in an area where Holocene sediments have been eroded. The B/M is recorded at 59.08 m, and the basal age in the core is estimated to be ∼0.9 Myr (Fig. 2n, Mei et al. 2016). Although the abundance of benthic foraminifera in some layers is low (<1 g−1, Fig. 2m), only three fluvial layers (roughly 28 − 33, 55 − 56 and 69 − 70 m) are suggested clearly from sedimentary characteristics (Figs 2k and l). Mei et al. (2016) proposed that the identified marine layers correlate to those in core NHH01. Under the constraints of B/M and AMS 14C ages, reconstructions of sedimentary history were tentatively carried out by correlating benthic foraminiferal abundance and sedimentary characteristic variations to the relative sea-level change (Figs 2e, j and o, Rohling et al. 2014). Taking core NHH01, for example (Figs 2f–j), the fluvial unit just above the B/M that is coeval with marine isotope stage (MIS) 19 (e.g. Lisiecki & Raymo 2005) should deposit in MIS 18, which can be further confirmed by the study of Liu et al. (2016). Considering that sediments in the central YS and its adjacent Bohai Sea are continuous at the 100-ka scale at least during interglacials (Yao et al. 2014; Liu et al. 2014b), the subsequent marine deposits with high foraminiferal abundance can correlate to sediments deposited in MIS 17, with the next to MIS 15. The AMS 14C ages in the upper section indicate that two thin negative inclination anomalies around 10 m are the records of Mono Lake (33 ka) and Laschamp (41 ka) excursions (e.g. Roberts 2008), respectively, and that the underlying foraminifera abundant layers should deposit in MIS 3. The last fluvial unit (14.9 − 15.9 m) can therefore be regarded as sediments deposited in MIS 4, with the underlying marine unit corresponding to MIS 5 and the penultimate fluvial unit to MIS 6, and so on. This procedure could give a reasonable chronology so far, especially to the focused layers (see below) which locate close to the B/M. 3.3 Rock magnetic experiments Hysteresis loops were measured on 107 samples for core EY02–2, 127 samples for core DLC70–3 and 67 supplementary samples in addition to the published 120 samples (Liu et al. 2014a,b) for core NHH01 at 0.4 − 1.0 m stratigraphic intervals using a Princeton Measurements Corporation vibrating sample magnetometer (VSM, MicroMag 3900). The magnetic field was cycled between +1.5 and −1.5 T, with a field increment of 5 mT and an averaging time of 300 ms. Coercivity (Bc), saturation remanence (Mrs) and saturation magnetization (Ms) were determined after paramagnetic slope correction. Specimens were then saturated magnetically again by applying a 1.5-T field. Finally, backfields were applied stepwise up to −1.0 T to determine the coercivity of remanence (Bcr). First-order reversal curves (FORCs) were measured for representative specimens using the VSM, with 100 FORCs measured at fields up to 1.0 T with an averaging time of 500 ms. FORC diagrams (Pike et al. 1999) were processed using the FORCinel software of Harrison & Feinberg (2008). In addition, magnetic susceptibilities (χ, mass-specific) of samples at parallel depths with all of the above samples were measured using a Kappabridge MFK1-FA system (Agico Ltd, Brno). Low-temperature measurements were conducted on splits (∼100 mg) of the representative samples using a Quantum Design Magnetic Property Measurement System (MPMS-5XL). Specimens were first cooled from room temperature to 20 K in zero field, then an isothermal remanent magnetization (IRM) was imparted in a 2.5 T field. After removal of the field, the IRM was thermally demagnetized by sweeping the temperature from 20 to 300 K at 5 K min−1. Cubic samples were also collected at parallel depths as the representative samples for thermal demagnetization of a three-component IRM following the procedure of Lowrie (1990). The specimens were magnetized in successively smaller fields along three mutually orthogonal axes using a 2-G Enterprises Pulse Magnetizer (2G660), with a 2.7 T induction along the z-axis (high-coercivity fraction), 0.5 T along the y-axis (medium-coercivity fraction) and 0.05 T along the x-axis (low-coercivity fraction), respectively. Subsequently, the specimens were subjected to progressive thermal demagnetization up to 690°C at 20 − 50°C intervals in a magnetically shielded (<300 nT) laboratory. These experiments were carried out in the Paleomagnetism and Geochronology Laboratory, Chinese Academy of Sciences, Beijing. 3.4 Previous data collections Prior to this study, attempts to research the initiation and evolution of the YSWC and the YSCWM in the mid-Pleistocene were conducted on core DLC70–3 by integrating multiple indices (Mei et al. 2016). Specifically, C37 alkenone contents (Fig. 3a) can reflect the change in seawater salinity of the SYS and track the YSWC; the distribution of two cold water benthic foraminifera species, B. frigida (Fig. 3b) and P. turberculatum (Fig. 3c) are related to the YSCWM, and the latter prefers to live in deeper water environments; high total organic carbon (TOC, Fig. 3d) and total sulphur (TS, Fig. 3e) contents in sediments usually indicate relatively high primary productivity in the surface or mixed layers of the water column and relatively weak oxidation degree of bottom waters; enrichment of trace element molybdenum (Mo, Fig. 3f), a redox-sensitive element, occurs only in a sulphide-rich (i.e. strongly sulphidic) sediments (Mei et al. 2016). In this study, another redox-sensitive element cadmium (Cd) which has been documented widely to be enriched in iron-reducing diagenetic (i.e. with trace levels of free H2S that are below typical chemical detection) environments (e.g. Rosenthal et al. 1995; Pattan & Pearce 2009) was measured. We used these geochemical parameters to investigate environmental conditions in which greigite is preferentially preserved. Details of measurement procedures for all of the above-mentioned indices can be found in Mei et al. (2016). Figure 3. View largeDownload slide Down-core changes in (a) C37alkenone content that is used to track the Yellow Sea Warm Current (YSWC), (b) and (c) relative content of two cold water benthic foraminiferal species that are indicative of the Yellow Sea Cold Water Mass (YSCWM), with (c) P. turberculatum preferring to live in a deeper water environment than (b) B. frigida, (d)–(f) total organic carbon (TOC) and total sulphur (TS) contents and trace element molybdenum (Mo) content, with high values indicating strongly sulphidic conditions, (g) trace element cadmium (Cd) content, with high values indicating weakly sulphidic conditions and (h) saturation isothermal remanent magnetization (SIRM). Results in (a)–(f) are reproduced from Mei et al. (2016). The green bars mark greigite-bearing layers, while the light blue bars denote layers deposited under the influence of strong YSWC and YSCWM that have undergone intense diagenesis. Figure 3. View largeDownload slide Down-core changes in (a) C37alkenone content that is used to track the Yellow Sea Warm Current (YSWC), (b) and (c) relative content of two cold water benthic foraminiferal species that are indicative of the Yellow Sea Cold Water Mass (YSCWM), with (c) P. turberculatum preferring to live in a deeper water environment than (b) B. frigida, (d)–(f) total organic carbon (TOC) and total sulphur (TS) contents and trace element molybdenum (Mo) content, with high values indicating strongly sulphidic conditions, (g) trace element cadmium (Cd) content, with high values indicating weakly sulphidic conditions and (h) saturation isothermal remanent magnetization (SIRM). Results in (a)–(f) are reproduced from Mei et al. (2016). The green bars mark greigite-bearing layers, while the light blue bars denote layers deposited under the influence of strong YSWC and YSCWM that have undergone intense diagenesis. 4 RESULTS 4.1 Rock magnetism Samples from depths of ∼50 m in all three cores have much enhanced Bc values (25–54 mT, Figs 4a, e and i) with relatively uniform Bcr/Bc ratios of ≈1.5 (Figs 4b, f and j) and ratios of Mrs/Ms ≈ 0.5 (Figs 4c, g and k). In addition, higher Mrs/χ ratios (Figs 4d, h and l) indicate single-domain (SD) magnetic behaviour and are consistent with those reported for sedimentary greigite-bearing samples (e.g. Snowball 1991; Roberts 1995). Figure 4. View largeDownload slide Variations of hysteresis parameters and their ratios along the studied cores, including (a), (e) and (i) coercivity (Bc), (b), (f) and (j) the ratio of the coercivity of remanence (Bcr) to Bc, (c), (g) and (k) the ratio of the saturation remanence (Mrs) to the saturation magnetization (Ms) and (d), (h) and (l) the ratio of Mrs to magnetic susceptibility (χ). The green bars mark the thick greigite-bearing layers discussed in the text, and some possible thinner greigite-bearing layers are marked with yellow bars. The blue characters denote the corresponding MIS for the thick greigite-bearing layers as indicated in Fig. 2. Figure 4. View largeDownload slide Variations of hysteresis parameters and their ratios along the studied cores, including (a), (e) and (i) coercivity (Bc), (b), (f) and (j) the ratio of the coercivity of remanence (Bcr) to Bc, (c), (g) and (k) the ratio of the saturation remanence (Mrs) to the saturation magnetization (Ms) and (d), (h) and (l) the ratio of Mrs to magnetic susceptibility (χ). The green bars mark the thick greigite-bearing layers discussed in the text, and some possible thinner greigite-bearing layers are marked with yellow bars. The blue characters denote the corresponding MIS for the thick greigite-bearing layers as indicated in Fig. 2. FORC diagrams for samples from these layers of cores EY02–2 (e.g. Figs 5b and c) and DLC70–3 (e.g. Figs 5j and k) have the same characteristics as those in core NHH01 (e.g. Figs 5f and g, Liu et al. 2014a), including closed concentric contours about a central peak, a negative region in the lower left-hand part of the diagram, a Bc peak of ∼60 mT, and significant magnetostatic interactions among magnetic grains suggested by vertical spread in the Bu direction and by the offset of the FORC distribution below the Bu = 0 axis. All of these features are typical of sedimentary SD greigite in natural environments (e.g. Roberts et al.2006, 2011). In contrast, FORC diagrams for other samples are typical of low-coercivity SD-PSD (pseudo-SD; e.g. Figs 5e, h, i and l) and/or PSD-MD (multidomain; e.g. Figs 5a and d) magnetic minerals (Roberts et al.2000, 2014; Muxworthy & Dunlop 2002). Figure 5. View largeDownload slide First-order reversal curve (FORC) diagrams for representative samples from the studied cores: (a)–(d) EY02–2, (e)–(h) NHH01 and (i)–(l) DLC70–3, respectively. Measurements were made in an applied field up to 1 T with 100 curves measured and an averaging time of 500 ms, and were processed using the FORCinel software of Harrison & Feinberg (2008) with a smoothing factor (SF) of 5. (e)–(g) FORC diagrams from Liu et al. (2014a). Figure 5. View largeDownload slide First-order reversal curve (FORC) diagrams for representative samples from the studied cores: (a)–(d) EY02–2, (e)–(h) NHH01 and (i)–(l) DLC70–3, respectively. Measurements were made in an applied field up to 1 T with 100 curves measured and an averaging time of 500 ms, and were processed using the FORCinel software of Harrison & Feinberg (2008) with a smoothing factor (SF) of 5. (e)–(g) FORC diagrams from Liu et al. (2014a). Temperature-dependent measurements can provide more directly diagnostic evidence of the presence of greigite. Detection of low-temperature magnetic transitions has also been used to identify magnetic minerals. For example, magnetite undergoes the Verwey transition (Tv) at around 120 K (Verwey 1939), haematite undergoes the Morin transition at 250–260 K (Morin 1950), monoclinic pyrrhotite (Fe7S8) undergoes the Besnus transition at 30–34 K (Rochette et al. 1990) and greigite lacks a low-temperature magnetic transition (Chang et al. 2009; Roberts et al. 2011). Low-temperature curves contain an obvious Verwey transition in most samples from the layers focused on here (e.g. Figs 6a, e, h, i and l), which indicates the presence of magnetite. Samples from the upper layers that should contain greigite also give rise to a detectable Verwey transition (e.g. Figs 6b, f and j). However, the Verwey transition is highly smeared in samples from the lower layers (e.g. Figs 6c, g and k), which indicates the minor presence of magnetite. In addition, the Tv temperature is consistently closed to 120 K (vertical dashed lines in Fig. 6), which indicates that the magnetite in the studied sediments is detrital rather than biogenic, which usually has a Tv of ∼100 K (Chang et al. 2016). Figure 6. View largeDownload slide Low-temperature curves of an IRM (mass specific) acquired at 20 K by applying a 2.5-T field and measured at 5 K intervals to 300 K during zero-field warming (blue lines) for representative samples from the studied cores (a)–(d) EY02–2, (e)–(h) NHH01 and (i)–(l) DLC70–3, respectively. The first derivative of magnetizations with respect to temperature (−▵IRM/−▵T) is also ploted on the right-hand y-axes (magenta lines). The dotted vertical lines indicate Tv peaks. (e)–(g) Low-temperature magnetization results from Liu et al. (2014a). Figure 6. View largeDownload slide Low-temperature curves of an IRM (mass specific) acquired at 20 K by applying a 2.5-T field and measured at 5 K intervals to 300 K during zero-field warming (blue lines) for representative samples from the studied cores (a)–(d) EY02–2, (e)–(h) NHH01 and (i)–(l) DLC70–3, respectively. The first derivative of magnetizations with respect to temperature (−▵IRM/−▵T) is also ploted on the right-hand y-axes (magenta lines). The dotted vertical lines indicate Tv peaks. (e)–(g) Low-temperature magnetization results from Liu et al. (2014a). Although the Curie temperature of greigite remains unknown but must exceed 350°C, it will undergo thermal decomposition below 400°C (Chang et al. 2008; Roberts et al. 2011). Thermal demagnetization of three-component IRMs for the studied layer indicates that the major magnetic carriers dominantly have medium to low coercivity with a significant unblocking temperature of ∼360°C and a minor component that unblocks fully at ∼585°C or even ∼680°C (e.g. Figs 7b and c, f and g and j and k). These and the above results indicate the dominant presence of greigite and traces of magnetite and/or haematite (especially in the upper sections) in the studied sediments. In contrast, IRMs of samples outside of this layer mainly unblock below 600°C, with some unblocking above 650°C (e.g. Figs 7a, d, e, h, i and l), which indicates the magnetic dominance of magnetite with some haematite. These results also confirm that magnetite is still the major magnetic mineral component in some samples (e.g. Fig. 7d), even if the Verwey transition is obscure (e.g. Fig. 6d), which can be caused by high degrees of oxidation and/or isomorphous substitution of other cations into the magnetite crystal lattice (Özdemir et al. 1993). Figure 7. View largeDownload slide Thermal demagnetization results for a three-component IRM for representative samples from cores: (a)–(d) EY02–2, (e)–(h) NHH01 and (i)–(l) DLC70–3, respectively. They were produced by magnetizing the cubic specimens in fields of 2.7, 0.5 and 0.05-T fields along the sample z-, y- and x-axes, respectively, to impart three mutually orthogonal IRMs, and then thermally demagnetizing them to 690°C at 20–50°C intervals. (e)–(g) Results are from Liu et al. (2014a). Figure 7. View largeDownload slide Thermal demagnetization results for a three-component IRM for representative samples from cores: (a)–(d) EY02–2, (e)–(h) NHH01 and (i)–(l) DLC70–3, respectively. They were produced by magnetizing the cubic specimens in fields of 2.7, 0.5 and 0.05-T fields along the sample z-, y- and x-axes, respectively, to impart three mutually orthogonal IRMs, and then thermally demagnetizing them to 690°C at 20–50°C intervals. (e)–(g) Results are from Liu et al. (2014a). Overall, the rock magnetic data suggest that greigite is the principal magnetic mineral in the layer at ∼50 m in cores EY02–2 and DLC70–3 as in core NHH01. The proportion of greigite in the magnetic mineral assemblages in the thick layers increases downward, as indicated by both FORC diagrams (e.g. Figs 5b and c, f and g and j and k) and temperature-dependent magnetization curves (e.g. Figs 6b and c, f and g and j and k; Figs 7b and c, f and g and j and k). In addition to the thick greigite-bearing layer, several thin greigite-bearing layers occur (yellow bars in Fig. 4), which cannot be correlated laterally among the three studied cores. 4.2 Trace elements Cd concentrations in core DLC70–3 range from 0.055 to 0.307 ppm, and average 0.140 ppm (Fig. 3g). Cd reaches marked maxima across the depth interval of 37.94–54.64 m (0.178 ppm on average), which corresponds to the thick greigite-bearing layer. Other thin greigite-bearing layers also have relatively high Cd concentrations (Fig. 3g). The covariant occurrence of Cd and greigite indicate that a weakly sulphidic diagenetic environment was established, which favoured greigite formation and preservation before pyritization reactions could be completed. We reach this conclusion based on the following observations: (1) Cd2+ has more rapid water exchange reaction kinetics than Fe2+, which results in CdS precipitating prior to FeS formation and subsequent pyrite formation (Morse & Luther 1999); (2) insoluble CdS occurs as a discrete phase rather than being associated with iron sulphides (Huerta-Diaz & Morse 1992) and (3) solid phase Cd in anoxic (i.e. strongly sulphidic) sediments is unstable and dissolved Cd diffuses upward and is deposited into sediments across the ferruginous/sulphidic boundary (Wei et al. 2005; see Canfield & Thamdrup (2009) and Roberts (2015) for terminology of diageneticzonations). 5 DISCUSSION Scanning electron microscopy observations have demonstrated that iron sulphides in core NHH01 occur as euhedral crystals either isolated and dispersed within clay matrix or in closely packed aggregates filling sediment voids (Liu et al.2014a), without being restricted to confined microenvironments like foraminifer tests where reducing conditions may develop locally (e.g. Rey et al. 2005). Therefore, the thick greigite-bearing sediment layer horizontally widespread in the central SYS indicates the long-term existence of sulphidic sedimentary environments in which the greigite formed. Generally, in the presence of abundant organic matter and porewater sulphide, pyritization reactions will proceed to completion, making greigite preservation less likely and pyrite preservation more likely (Berner 1984; Kao et al. 2004; Roberts & Weaver 2005). However, when organic matter supply is limited, coupled with high reactive iron concentrations, H2S can be removed so effectively that pyritization is inhibited, and the intermediate phase, greigite, will be preserved (Kao et al. 2004). The intermittent rather than continuous presence of greigite in the studied sediments (Fig. 4) is suggestive of time-varying diagenetic conditions known as non-steady-state diagenesis (Roberts 2015). Several forcing mechanisms, including hydrocarbon seepage, gas hydrate migration, tectonic deformation and associated fluid migration, large amplitude sea level changes on continental shelves, and submarine landslide events have been suggested to give rise to non-steady-state diagenesis that can induce intermittent greigite formation (Roberts & Weaver 2005). Tectonic deformation in this region was relatively weak in the mid-Pleistocene (Wan & Hao 2009), and gas hydrates and hydrocarbons have not been observed at such shallow depths in these sediments. Therefore, the first four mechanisms are unlikely to explain the documented greigite occurrences. Submarine landslide events are also unlikely considering the relatively flat seafloor of the YS (Qin et al. 1989). Large-amplitude sea-level changes on the continental shelf could result in variable bottom water oxygenation (e.g. Oda & Torii 2004; Roberts 2015). However, this alone does not explain why the thick greigite-bearing layer is only observed over a certain mid-Pleistocene interval when sea-level fluctuated cyclically (e.g. Fig. 2e, Rohling et al. 2014). Ge et al. (2003) observed that surface sediments with lower magnetic susceptibilities are mainly distributed in the CSYSM and suggested that a strong thermocline and pycnocline in the cold eddy prevents vertical mixing of water masses and results in a reducing diagenetic sedimentary environment in which iron sulphide minerals (such as pyrite) form. Greigite is only detected in the lower section of core YSDP103 (Fig. 1b), so that more complete pyritization has been invoked to explain the lack of greigite in the upper post-middle-Holocene (∼6 ka BP) section when the YSWC and its concomitant YSCWM formed and produced a more reducing environment (Liu et al. 2005). A recent study of core DLC70–3 (Fig. 1b) indicates that these hydrographic features were active as early as MIS 21 (Figs 3a–f, Mei et al. 2016). Sediments deposited during strong YSCWM conditions have characteristic low saturation IRM (SIRM) values (Fig. 3h), which indicates that detrital magnetic minerals in the sediments have been subjected to intense diagenesis, as reported by Ge et al. (2003) and Liu et al. (2005). In contrast, indices for the YSWC and YSCWM are not clearly observed in the greigite-bearing layers which have much higher SIRM values (Fig. 3). This should correspond to higher magnetic mineral concentrations resulting from greigite formation. It has been proposed that the availability of dissolved sulphide and iron are not the main constraints on pyrite formation and preservation of ferrimagnetic iron sulphides, but that it is the amount of available organic matter that acts as the controlling factor on diagenetic magnetic mineral formation in marine settings (Sagnotti et al. 2001; Liu et al. 2005). The relatively low TOC and TS contents of the greigite-bearing layers compared to the YSCWM layers (Figs 3d and e) indicate that more organic matter and sulphide were present in the latter to induce complete conversion to pyrite. Sulphate is one of the most abundant chemical compounds in seawater. Both the concentration of porewater sulphide and sulphur in sediments are controlled by organic matter availability since sulphate-reducing conditions enable incorporation of excess sulphur either as organic (e.g. biosynthetic) or inorganic (e.g. pyrite), in the presence of organic matter (Shi et al. 2016). These factors further indicate the dominant role of organic matter supply in magnetic mineral diagenesis. Based on published (Zhuang et al. 2002; Ge et al. 2006; Liu et al.2012, 2014b; Mei et al.2013, 2016) and reconstructed age models (Fig. 2), the thick greigite-bearing layers correspond to MIS 13–17 for cores EY02–2 and DLC70–3, and MIS 13–15 for core NHH01, respectively (Fig. 4), which succeeded a period with enhanced YSWC and YSCWM that was characterized by high primary productivity and strongly sulphidic conditions as indicated by elevated TOC, TS and Mo contents in MIS 19–21 (Figs 3d–f, Mei et al. 2016). In contrast, as mentioned above, trace element Cd enrichments in the greigite-bearing layer (Fig. 3g) indicate that the H2S released after CdS precipitation was limited and was exhausted by greigite formation. Otherwise, neither greigite nor greenockite (CdS) could be preserved, which suggests the potential of Cd as a promising indicator of greigite preservation in sediments. In addition, this further indicates a limited organic matter supply to the greigite-enriched layer. We propose that primary productivity provided the dominant control on sedimentary organic matter supply, which was vital to preservation of the thick greigite layer. YSCWM Influences should also be considered in relation to formation of the studied greigite-bearing layers. First, it could have prevented vertical mixing of water bodies and facilitated reducing conditions (e.g. Ge et al. 2003). Second, the discontinuous presence of greigite in core DLC70–3 compared to the other two cores (Fig. 4) is likely to be due to its location on the edge of the YSCWM deposition area (Fig. 1b), which made it fluctuate between reducing and oxic environmental conditions. Furthermore, the central location of core NHH01 in the CSYSM (Fig. 1b) is likely to have positioned it in a more strongly sulphidic environment during MIS 17–15, which resulted in a thinner greigite-bearing layer (Fig. 3). Lastly, greigite is not detected in coeval sediments in cores QC2 (Zhou & Ge 1990) and CSDP-1 (Liu et al. 2016) close to the CSYSM (Fig. 1b). These cores, along with results of Mei et al. (2016), could also indicate that the sphere of influence of the ancient cold eddy was no smaller than it is at present. A conceptual model is presented here to explain why the thick greigite-bearing layer only occurs over a certain stratigraphic interval. After a period with higher sea levels during MIS 21–19 (Mei et al. 2016), a large amplitude regression caused by tectonic uplift occurred during MIS 18 when seawater retreated from the central YS (Fig. 2, Liu et al. 2016). Subsequently, seawater covered the continental shelf again, but with lower sea levels than that in previous interglacials (MIS 21–19), which could have given rise to incursions of KC waters into the YS to form a relatively weak YSWC and YSCWM. The former maintained a moderate productivity level and, thus, a limited organic matter supply while the latter prevented vertical mixing of water bodies and facilitated an oxygen-free sedimentary environment, which together created sulphidic diagenetic conditions favourable for greigite formation and preservation. Substantial tectonic subsidence needed to have occurred to cause the relatively high sea level in MIS 16 when global sea level was in a marked lowstand (e.g. Fig. 2e, Rohling et al. 2014). In addition, an increasing number of sedimentary records from both hemispheres indicate that MIS 14 was the warmest mid-Pleistocene glacial period (e.g. Rackebrandt et al. 2011), which could be the principal reason for relatively high sea levels during this period in the YS. These conditions created ideal long-term conditions for greigite formation and preservation. 6 CONCLUSIONS By integrating rock magnetic measurements and trace element analyses on three long sediment cores from an N-S transect in the central SYS, we have obtained new insights into the distribution, origin and palaeoenvironmental implications of a thick greigite-bearing layer. The thick studied greigite-bearing layer is distributed laterally in the central SYS. The spatial extent of this layer could be as large as the present central SYS muddy area. Formation and preservation of the greigite was controlled by a limited organic matter supply that resulted from moderate primary productivity during MIS 17–13 when flows of the YSWC and the YSCWM were relatively weak. High Cd concentrations in the greigite-bearing layers indicate that conditions free of oxygen and H2S were more favourable for greigite preservation after its formation in sulphidic environments. This observation also suggests the potential of Cd as a promising index for indicating the presence of sedimentary greigite. Substantial local subsidence during MIS 16 and the unusually warm MIS 14 glacial should have maintained appropriate sea levels during the MIS 17–13, which created ideal material and ambient conditions for formation and preservation of the identified greigite. Therefore, we conclude that greigite has a widespread distribution over a considerable sediment thickness in early mid-Pleistocene sediments from the central SYS muddy area so that its presence was influenced by both local tectonic and global palaeoclimatic factors. ACKNOWLEDGEMENTS The data in this paper can be obtained by contacting the corresponding authors directly. We are grateful to Andrew Roberts and Daniel Rey and the Editors for their constructive comments and suggestions that greatly improved the manuscript. We have benefitted greatly from discussions with Xianwei Meng, Li Li and Hongmin Wang about redox sensitive elements. We thank the crew of R/V Kan407 for their assistance with coring. This work was jointly supported by the Natural Science Foundation of Shandong Province (grant no. ZR2016DB04), the Basic Scientific Fund for National Public Research Institutes of China (grant no. 2016Q06), the China Postdoctoral Science Foundation (grant no. 2016M592121), the National Natural Science Foundation of China (grant nos 41606075 and U1606401) and the China Geological Survey (grant nos GZH200800501 and 1212011220113). REFERENCES Alexander C.R., DeMaster D.J., Nittrouer C.A., 1991. Sediment accumulation in a modern epicontinental-shelf setting: the Yellow Sea, Mar. Geol. , 98, 51– 72. https://doi.org/10.1016/0025-3227(91)90035-3 Google Scholar CrossRef Search ADS   Berner R.A., 1984. Sedimentary pyrite formation: an update, Geochim. Cosmochim. Acta , 48, 605– 615. https://doi.org/10.1016/0016-7037(84)90089-9 Google Scholar CrossRef Search ADS   Canfield D.E., Thamdrup B., 2009. Towards a consistent classification scheme for geochemical environments, or, why we wish the term ‘suboxic’ would go away, Geobiology , 7, 385– 392. https://doi.org/10.1111/j.1472-4669.2009.00214.x Google Scholar CrossRef Search ADS PubMed  Chang L., Roberts A.P., Tang Y., Rainford B.D., Muxworthy A.R., Chen Q.W., 2008. Fundamental magnetic parameters from pure synthetic greigite (Fe3S4), J. geophys. Res. , 113, B06104, doi:10.1029/2007JB005502. https://doi.org/10.1029/2007JB005502 Google Scholar CrossRef Search ADS   Chang L., Roberts A.P., Rowan C.J., Tang Y., Pruner P., Chen Q.W., Horng C.-S., 2009. Low-temperature magnetic properties of greigite (Fe3S4), Geochem. Geophys.Geosyst. , 10, Q01Y04, doi:10.1029/2008GC002276. https://doi.org/10.1029/2008GC002276 Google Scholar CrossRef Search ADS   Chang L., Heslop D., Roberts A.P., Rey D., Mohamed K.J., 2016. Discrimination of biogenic and detrital magnetite through a double Verwey transition temperature, J. geophys. Res. , 121, 3– 14. https://doi.org/10.1002/2015JB012485 Google Scholar CrossRef Search ADS   Frank U., Nowaczyk N.R., Negendank J.F.W., 2007. Rock magnetism of greigite bearing sediments from the Dead Sea, Israel, Geophys. J. Int. , 168, 921−934. Fu Y.Z., von Dobeneck T., Franke C., Heslop D., Kasten S., 2008. Rock magnetic identification and geochemical process models of greigite formation in Quaternary marine sediments from the Gulf of Mexico (IODP Hole U1319A), Earth planet. Sci. Lett. , 275, 233– 245. https://doi.org/10.1016/j.epsl.2008.07.034 Google Scholar CrossRef Search ADS   Ge S.L., Shi X.F., Han Y.B., 2003. Distribution characteristics of magnetic susceptibility of the surface sediments in the southern Yellow Sea, Chin. Sci. Bull. , 48, 37– 41. https://doi.org/10.1007/BF02900938 Google Scholar CrossRef Search ADS   Ge S.L., Shi X.F., Zhu R.X., Liu Y.G., Yin P., Liu L.J., 2006. Magnetostratigraphy of borehole EY02-2 in the southern Yellow Sea and its paleoenvironmental significance, Chin. Sci. Bull. , 51, 855−865. Guan B.X., 1994. Pattern and structures of the current in Bohai, Huanghai (Yellow) and East China Seas, in Oceanology of China Seas , eds Zhou D., Liang Y.B., Zeng C.K., pp. 3−16. Kluwer Academic Publishing, Dordrecht. Harrison R.J., Feinberg J.M., 2008. FORCinel: an improved algorithm for calculating first-order reversal curve distributions using locally weighted regression smoothing, Geochem. Geophys.Geosyst. , 9, Q05016, doi:10.1029/2008GC001987. https://doi.org/10.1029/2008GC001987 Google Scholar CrossRef Search ADS   Hu D.X., 1984. Upwelling and sedimentation dynamics I. The role of upwelling in sedimentation in the Huanghai Sea and East China Sea−a description of general features, Chin. J. Oceanol. Limnol. , 2, 2−19 (in Chinese, with English summary). Huerta-Diaz M., Morse J.W., 1992. Pyritization of trace metals in anoxic marine sediments, Geochim. Cosmochim. Acta , 56, 2681– 2702. https://doi.org/10.1016/0016-7037(92)90353-K Google Scholar CrossRef Search ADS   Kao S.-J., Horng C.-S., Roberts A.P., Liu K.-K., 2004. Carbon–sulfur–iron relationships in sedimentary rocks from southwestern Taiwan: influence of geochemical environment on greigite and pyrrhotite formation, Chem. Geol. , 203, 153– 168. https://doi.org/10.1016/j.chemgeo.2003.09.007 Google Scholar CrossRef Search ADS   Kim D., Park B.K., Shin I., 1999. Paleoenvironmental changes of the Yellow Sea during the Late Quaternary, Geo-Mar. Lett. , 18, 189−194. Lan X.H., Wang H.X., Li R.H., Lin Z.H., Zhang Z.X., 2007. Major elements composition and provenance analysis in the sediments of the South Yellow Sea, Earth Sci. Front. , 14, 197– 203 (in Chinese, with English summary). https://doi.org/10.1016/S1872-5791(07)60033-5 Google Scholar CrossRef Search ADS   Lee C.H., Jin J.-H., 1995. Authigenicgreigite in mud from the continental shelf of the Yellow Sea, off the southwest Korean Peninsula, Mar. Geol. , 128, 11– 15. https://doi.org/10.1016/0025-3227(95)00099-K Google Scholar CrossRef Search ADS   Lisiecki L.E., Raymo M.E., 2005. A Pliocene-Pleistocene stack of 57 globally distributed benthic δ18O records, Palaeoceanography , 20, PA1003, doi:10.1029/2004PA001071. Liu J., Zhu R.X., Roberts A.P., Li S.Q., Chang J.-H., 2004. High-resolution analysis of early diagenetic effects on magnetic minerals in post-middle-Holocene continental shelf sediments from the Korea Strait, J. geophys. Res. , 109, B03103, doi:10.1029/2003JB002813. https://doi.org/10.1029/2003JB002813 Liu J., Zhu R.X., Li S.Q., Chang J.-H., 2005. Magnetic mineral diagenesis in the post-glacial muddy sediments from the southeastern South Yellow Sea: response to marine environmental changes, Sci. China Earth Sci. , 48, 134−144. Liu J., Satio Y., Kong X.H., Wang H., Wen Z., Yang Z.G., Nakashima R., 2010. Delta development and channel incision during marine isotope stages 3 and 2 in the western South Yellow Sea, Mar. Geol. , 278, 54– 76. https://doi.org/10.1016/j.margeo.2010.09.003 Google Scholar CrossRef Search ADS   Liu J.X., Shi X.F., Ge S.L., Li X.Y., Yao Z.Q., Wang X., 2012. Record of geomagnetic excursion during early Holocene in the central southern Yellow Sea Mud, Quat. Res. , 32, 663−669 (in Chinese, with English summary). Liu J.X., Shi X.F., Ge S.L., Liu Q.S., Yao Z.Q., Yang G., 2014a. Identification of the thick-layer greigite in sediments of the South Yellow Sea and its geological significances, Chin. Sci. Bull. , 59, 2764−2775. Liu J.X. et al.  , 2014b. Magnetostratigraphy of a greigite-bearing core from the South Yellow Sea: implications for remagnetization and sedimentation, J. geophys. Res. , 119, 7425– 7441. https://doi.org/10.1002/2014JB011206 Google Scholar CrossRef Search ADS   Liu J.X., Liu Q.S., Zhang X.H., Liu J., Wu Z.Q., Mei X., Shi X.F., Zhao Q.H., 2016. Magnetostratigraphy of a long Quaternary sediment core in the South Yellow Sea, Quat. Sci. Rev. , 144, 1– 15. https://doi.org/10.1016/j.quascirev.2016.05.025 Google Scholar CrossRef Search ADS   Lowrie W., 1990. Identification of ferromagnetic minerals in a rock by coercivity and unblocking temperature properties, Geophys. Res. Lett. , 17, 159– 162. https://doi.org/10.1029/GL017i002p00159 Google Scholar CrossRef Search ADS   McMillan A.A., 2002. Onshore Quaternary geological surveys in the 21st century—a perspective from the British Geological Survey, Quat. Sci. Rev. , 21, 889– 899. https://doi.org/10.1016/S0277-3791(01)00064-6 Google Scholar CrossRef Search ADS   Mei X., Zhang X.H., Li R.H., 2013. Distribution of late Quaternary benthic foraminifera in South Yellow Sea and its implication of paleo-water mass, Geol. Rev. , 59, 1024−1034 (in Chinese, with English summary). Mei X. et al.  , 2016. Evolution of the Yellow Sea warm current and the Yellow Sea cold water mass since the Middle Pleistocene, Palaeogeog. Palaeoclimatol. Palaeoecol. , 442, 48– 60. https://doi.org/10.1016/j.palaeo.2015.11.018 Google Scholar CrossRef Search ADS   Milliman J.D., Meade R.H., 1983. World-wide delivery of river sediment to the oceans, J. Geol. , 91, 1– 21. https://doi.org/10.1086/628741 Google Scholar CrossRef Search ADS   Morin F.J., 1950. Magnetic susceptibility of αFe2O3 and αFe2O3 with added Titanium, Phys. Rev. , 78, 819– 820. https://doi.org/10.1103/PhysRev.78.819.2 Google Scholar CrossRef Search ADS   Morse J.W., Luther G.W., 1999. Chemical influences on trace metal-sulfide interactions in anoxic sediments, Geochim. Cosmochim. Acta , 63, 3373– 3378. https://doi.org/10.1016/S0016-7037(99)00258-6 Google Scholar CrossRef Search ADS   Muxworthy A.R., Dunlop D.J., 2002. First-order reversal curve (FORC) diagrams for pseudo-single-domain magnetites at high temperature, Earth planet. Sci. Lett. , 203, 369– 382. https://doi.org/10.1016/S0012-821X(02)00880-4 Google Scholar CrossRef Search ADS   Oda H., Torii M., 2004. Sea-level change and remagnetization of continental shelf sediments off New Jersey (ODP Leg 174A): magnetite and greigite diagenesis, Geophys. J. Int. , 156, 443– 458. https://doi.org/10.1111/j.1365-246X.2004.02162.x Google Scholar CrossRef Search ADS   Özdemir Ö., Dunlop D.J., Moskowitz B.M., 1993. The effect of oxidation on the Verwey transition in magnetite, Geophys. Res. Lett. , 20, 1671– 1674. https://doi.org/10.1029/93GL01483 Google Scholar CrossRef Search ADS   Pattan J.N., Pearce N.J.G., 2009. Bottom water oxygenation history in southeastern Arabian Sea during the past 140ka: results from redox-sensitive elements, Palaeogeog. Palaeoclimat. Palaeoecol. , 280, 396– 405. https://doi.org/10.1016/j.palaeo.2009.06.027 Google Scholar CrossRef Search ADS   Pike C.R., Roberts A.P., Verosub K.L., 1999. Characterizing interactions in fine magnetic particle systems using first order reversal curves, J. appl. Phys. , 85, 6660– 6667. https://doi.org/10.1063/1.370176 Google Scholar CrossRef Search ADS   Qin Y.S., Zhao Y.Y., Chen L.R., Zhao S.L., 1989. Geology of the Yellow Sea , China Ocean Press, pp. 1−289 (in Chinese). Rackebrandt N., Kuhnert H., Groeneveld J., Bickert T., 2011. Persisting maximum Agulhas leakage during MIS 14 indicated by massive Ethmodiscus oozes in the subtropical South Atlantic, Paleoceanography , 26, 130– 137. https://doi.org/10.1029/2010PA001990 Google Scholar CrossRef Search ADS   Reimer P.J. et al.  , 2013. IntCal13 and Marine13 radiocarbon age calibration curves 0–50,000 years cal BP, Radiocarbon , 55, 1869– 1887. https://doi.org/10.2458/azu_js_rc.55.16947 Google Scholar CrossRef Search ADS   Rey D., Mohamed K.J., Bernabeu A., Rubio B., Vilas F., 2005. Early diagenesis of magnetic minerals in marine transitional environments: geochemical signatures of hydrodynamic forcing, Mar. Geol. , 215, 215– 236. https://doi.org/10.1016/j.margeo.2004.12.001 Google Scholar CrossRef Search ADS   Reynolds R.L., Rosenbaum J.C., van Metre P., Tuttle M., Callender E., Goldin A., 1999. Greigite as an indicator of drought − the 1912–1994 sediment magnetic record from White Rock Lake, Dallas, Texas, USA, J. Paleolimnol. , 21, 193– 206. https://doi.org/10.1023/A:1008027815203 Google Scholar CrossRef Search ADS   Roberts A.P., 1995. Magnetic properties of sedimentary greigite (Fe3S4), Earth planet. Sci. Lett. , 134, 227– 236. https://doi.org/10.1016/0012-821X(95)00131-U Google Scholar CrossRef Search ADS   Roberts A.P., 2008. Geomagnetic excursions: knowns and unknowns, Geophys. Res. Lett. , 35, L17307, doi:10.1029/2008GL034719. https://doi.org/10.1029/2008GL034719 Google Scholar CrossRef Search ADS   Roberts A.P., 2015. Magnetic mineral diagenesis, Earth-Sci. Rev. , 151, 1– 47. https://doi.org/10.1016/j.earscirev.2015.09.010 Google Scholar CrossRef Search ADS   Roberts A.P., Reynolds R.L., Verosub K.L., Adam D.P., 1996. Environmental magnetic implications of Greigite (Fe3S4) Formation in a 3 m.y. lake sediment record from Butte Valley, northern California, Geophys. Res. Lett. , 23, 2859– 2862. https://doi.org/10.1029/96GL02831 Google Scholar CrossRef Search ADS   Roberts A.P., Pike C.R., Verosub K.L., 2000. First-order reversal curve diagrams: a new tool for characterizing the magnetic properties of natural samples, J. geophys. Res. , 105, 28461– 28475. https://doi.org/10.1029/2000JB900326 Google Scholar CrossRef Search ADS   Roberts A.P., Weaver R., 2005. Multiple mechanisms of remagnetization involving sedimentary greigite (Fe3S4), Earth planet. Sci. Lett. , 231, 263– 277. https://doi.org/10.1016/j.epsl.2004.11.024 Google Scholar CrossRef Search ADS   Roberts A.P., Liu Q.S., Rowan C.J., Chang L., Carvallo C., Torrent J., Horng C.-S., 2006. Characterization of hematite (α-Fe2O3), goethite (α-FeOOH), greigite (Fe3S4), and pyrrhotite (Fe7S8) using first-order reversal curve diagrams, J. geophys. Res. , 111, B12S35, doi:10.1029/2006JB004715. https://doi.org/10.1029/2006JB004715 Google Scholar CrossRef Search ADS   Roberts A.P., Chang L., Rowan C.J., Horng C.-S., Florindo F., 2011. Magnetic properties of sedimentary greigite (Fe3S4): an update, Rev. Geophys. , 49, RG1002, doi:10.1029/2010RG000336.. https://doi.org/10.1029/2010RG000336 Google Scholar CrossRef Search ADS   Roberts A.P., Heslop D., Zhao X., Pike C.R., 2014. Understanding fine magnetic particle systems through use of first-order reversal curve diagrams, Rev. Geophys. , 52, 557– 602. https://doi.org/10.1002/2014RG000462 Google Scholar CrossRef Search ADS   Rochette P., Fillion G., Mattéi J.-L., Dekkers M.J., 1990. Magnetic transition at 30–34 Kelvin in pyrrhotite: insight into a widespread occurrence of this mineral in rocks, Earth planet. Sci. Lett. , 98, 319– 328. https://doi.org/10.1016/0012-821X(90)90034-U Google Scholar CrossRef Search ADS   Rohling E.J., Foster G.L., Grant K.M., Marino G., Roberts A.P., Tamisiea M.E., Williams F., 2014. Sea-level and deep-sea-temperature variability over the past 5.3 million years, Nature , 508, 477– 482. https://doi.org/10.1038/nature13230 Google Scholar CrossRef Search ADS PubMed  Rosenthal Y., Lam P., Boyle E.A., Thomson J., 1995. Authigenic cadmium enrichments in suboxic sediments: precipitation and postdepositional mobility, Earth planet. Sci. Lett. , 132, 99– 111. https://doi.org/10.1016/0012-821X(95)00056-I Google Scholar CrossRef Search ADS   Sagnotti L., Macri P., Camerlenghi A., Rebesco M., 2001. Environmental magnetism of Antarctic late Pleistocene sediments and interhemispheric correlation of climatic events, Earth planet. Sci. Lett. , 192, 65– 80. https://doi.org/10.1016/S0012-821X(01)00438-1 Google Scholar CrossRef Search ADS   Sagnotti L., Cascella A., Ciaranfi N., Macrì P., Maiorano P., Marino M., Taddeucci J., 2010. Rock magnetism and palaeomagnetism of the MontalbanoJonico section (Italy): evidence for late diagenetic growth of greigite and implications for magnetostratigraphy, Geophys. J. Int. , 180, 1049– 1066. https://doi.org/10.1111/j.1365-246X.2009.04480.x Google Scholar CrossRef Search ADS   Shi X.F. et al.  , 2016. Sedimentary architecture of the Bohai Sea China over the last 1 Ma and implications for sea-level changes, Earth planet. Sci. Lett. , 451, 10– 21. https://doi.org/10.1016/j.epsl.2016.07.002 Google Scholar CrossRef Search ADS   Skinner B.J., Erd R.C., Grimaldi F.S., 1964. Greigite, the thiospinel of iron, a new mineral, Am. Mineral. , 49, 543−555. Snowball I.F., 1991. Magnetic hysteresis properties of greigite (Fe3S4) and a new occurrence in Holocene sediments from Swedish Lappland, Phys. Earth planet. Inter. , 68, 32– 40. https://doi.org/10.1016/0031-9201(91)90004-2 Google Scholar CrossRef Search ADS   Southon J., Kashgarian M., Fontugne M., Metivier B., Yim W.W.S., 2002. Marine reservoir corrections for the Indian Ocean and Southeast Asia, Radiocarbon , 44, 167– 180. https://doi.org/10.1017/S0033822200064778 Google Scholar CrossRef Search ADS   Verwey E.J.W., 1939. Electronic conduction of magnetite (Fe3O4) and its transition point at low temperatures, Nature , 144, 327– 328. https://doi.org/10.1038/144327b0 Google Scholar CrossRef Search ADS   Wan T.F., Hao T.Y., 2009. Cenozoic tectonics of the Yellow Sea and its oil-gas exploration, Geoscience , 23, 385−393 (in Chinese, with English summary). Wang Q.W., Morse J.W., 1996. Pyrite formation under conditions approximating those in anoxic sediments I. Pathway and morphology, Mar. Chem. , 52, 99– 121. https://doi.org/10.1016/0304-4203(95)00082-8 Google Scholar CrossRef Search ADS   Wang Q., Zhu E.Q., 1989. Marine Sedimentology , China Science Press, pp. 1−254 (in Chinese). Wei G.J., Li X.H., Liu Y., Shao L., Liang X.R., 2005. Transfer patterns of Mn, Cd and Mo in sediments during early diagenesis: evidences from sediment cores at ODP Site 1148, Geochimica , 34, 129−135 (in Chinese, with English summary). Yang S.Y., Jung H.S., Lim D.I., Li C.X., 2003. A review on the provenance discrimination of sediments in the Yellow Sea, Earth-Sci. Rev. , 63, 93– 120. https://doi.org/10.1016/S0012-8252(03)00033-3 Google Scholar CrossRef Search ADS   Yao Z.Q., Guo Z.T., Xiao G.Q., Wang Q., Shi X.F., Wang X.Y., 2012. Sedimentary history of the western Bohai coastal plain since the late Pliocene: implications on tectonic, climatic and sea-level changes, J. Asian Earth Sci. , 54−55, 192−202. Yao Z.Q. et al.  , 2014. Paleomagnetic and astronomical dating of sediment core BH08 from the Bohai Sea, China: implications for glacial–interglacial sedimentation, Palaeogeog. Palaeoclimat. Palaeoecol. , 393, 90– 101. https://doi.org/10.1016/j.palaeo.2013.11.012 Google Scholar CrossRef Search ADS   Zhou M.Q., Ge Z.S., 1990. The magnetostratigraphy of unconsolidated sediments in the southern Yellow Sea and adjacent areas, Mar. Geol. Quat. Geol. , 10, 21−33 (in Chinese, with English summary). Zhuang L.H., Chang F.M., Li T.G., Yan J., 2002. Foraminiferal faunas and Holocene sedimentation rates of core EY02-2 in the South Yellow Sea, Mar. Geol. Quat. Geol. , 22, 7−14 (in Chinese, with English summary). © The Author(s) 2017. Published by Oxford University Press on behalf of The Royal Astronomical Society.

Journal

Geophysical Journal InternationalOxford University Press

Published: Apr 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off