Historical reconstruction of organic carbon decay and preservation in sediments on the East China Sea shelf


  • Xinxin Li,

    1. Department of Oceanography, Texas A&M University, College Station, Texas, USA
    2. Now at Geochemical and Environmental Research Group, Texas A&M University, College Station, Texas, USA
    Current affiliation:
    1. Institute for Geophysics, University of Texas at Austin, Austin, Texas, USA
    Search for more papers by this author
  • Thomas S. Bianchi,

    Corresponding author
    1. Department of Oceanography, Texas A&M University, College Station, Texas, USA
    2. Now at Department of Geological Sciences, University of Florida, Gainesville, Florida, USA
    Current affiliation:
    1. Now at The Water Institute of the Gulf, Baton Rouge, Louisiana, USA
    • Now at The Water Institute of the Gulf, Baton Rouge, Louisiana, USA
    Search for more papers by this author
  • Mead A. Allison,

    1. Institute for Geophysics, University of Texas at Austin, Austin, Texas, USA
    2. Now at The Water Institute of the Gulf, Baton Rouge, Louisiana, USA
    Current affiliation:
    1. Key Laboratory of Marine Chemistry Theory and Technology, Ministry of Education, Ocean University of China, Qingdao, China
    Search for more papers by this author
  • Piers Chapman,

    1. Department of Oceanography, Texas A&M University, College Station, Texas, USA
    Search for more papers by this author
  • Guipeng Yang

    1. Key Laboratory of Marine Chemistry Theory and Technology, Ministry of Education, Ocean University of China, Qingdao, China
    Search for more papers by this author

Corresponding author: T. S. Bianchi, Department of Geological Sciences, University of Florida, Gainesville, FL 32611, USA. (tbianchi@ufl.edu)


[1] Sediment cores were collected from the East China Sea inner shelf in 2010 to study the decay and preservation of organic carbon (OC). The highest sediment mass accumulation rate (0.61 ± 0.20 g cm−2 yr−1), derived from 210Pb, was found near the river mouth and decreased alongshore to the south (0.17 ± 0.004 g cm−2 yr−1), and in an offshore direction (0.31 ± 0.08 g cm−2 yr−1). Average total OC content was higher at inner shelf stations (0.52%) than those offshore (0.38%). The δ13C was more depleted at nearshore (−23.49‰ to −21.97‰) than offshore (−22.49‰ to −21.60‰) stations. Principal component analysis indicated that terrestrial OC, as indicated by lignin-phenols (Λ8) values, was preserved in sediment closer to the coast (0.22–0.44), while offshore sediment was more composed of lignin-poor (0.12–0.24) degraded OC that was likely hydrodynamically sorted. Marine-derived OC, as indicated by plant pigments, was significantly more abundant in the sediment mixed layer than the underlying accumulation layer. Historical flooding events were detected in Λ8 profiles in two of the six cores located at midshelf stations. Despite the magnitude of the 2010 flood in East China, we did not see any signature of this event with the chemical biomarker in these two cores. This may suggest that reduced sediment loading due to recent dam construction may have greatly decoupled river inputs with sediment loading to shelf sediment. The total OC standing stock since 1900 was approximately 1.62 ± 1.15 kg C m−2, about one tenth of all the middle and lower lakes in the Changjiang catchment basin. This work further supports the need for more research to better understand how the reduced inputs of fluvial input of sediments from Chinese rivers (due to river diversions and dams) affect carbon cycling in the East China Sea.

1 Introduction

[2] Sediment in continental margins receives inputs of organic carbon (OC) from both terrestrial and marine sources [Berner, 1982; Bianchi et al., 2007; Hedges and Keil, 1995; Hedges and Oades, 1997]. Studies have shown that more than 80% of the OC preserved in the marine sediment occurs in these regions [Berner, 1982; Hedges and Keil, 1995] especially in large delta-front estuaries [Paerl et al., 2006] characterized by rapid sediment accumulation [Bianchi and Allison, 2009, and references therein]. For example, of the total amount of approximately 200 Tg terrestrially derived particulate OC delivered to the ocean by rivers each year, 58 (±17) Tg C yr−1 gets buried in continental margins, of which about 40–50% is in deltaic sediment while the rest is distributed along the shelves and margins [Blair and Aller, 2012; Burdige, 2007]. Hence, preservation of OC in large delta-front estuaries likely plays an important role in the carbon sequestration and thus global carbon cycling [Bauer and Druffel, 1998; Berner, 1989; Ludwig et al., 1996].

[3] Increased nutrient inputs to large delta-front estuaries from anthropogenic activities may cause higher primary production or eutrophication [Nixon, 1995]. Sediment in large delta-front estuaries has been used to reconstruct anthropogenic effects on changes in storage and degradation of OC such as from the Gulf of Mexico [Allison et al., 2000; Goñi and Hedges, 1995; Goñi et al., 2006a, 2006b; Sampere et al., 2011a; Zhao et al., 2012], Chesapeake Bay [Zimmerman and Canuel, 2000], the Amazon-Guianas coast [Aller and Blair, 2006], the Washington coast [Prahl et al., 1994], the Pearl River estuary [Yang et al., 2011a], and the Arctic [Goñi et al., 2000]. As one of the largest delta-front estuaries in the world, the Changjiang estuary is characterized by mobile mud belts along the coast [Liu et al., 2007] with high sediment accumulation rate and nutrient loading, as well as rapid turnover of particulate OC and dissolved OC [Cai and Dai, 2004]. Similar to the Mississippi River estuary, the Changjiang River has experienced extensive anthropogenic alterations, particularly increased nutrient loading [e.g., Zhao et al., 2012] that influences marine primary production in the coastal plume region [Bianchi et al., 2010; Chen et al., 2009; Gong et al., 2006; Rabalais et al., 2007; Yu et al., 2012]. In addition, the Changjiang catchment basin has experienced significant changes from damming, recent climate change, and other land use changes. These processes have reduced the overall sediment load to the East China Sea and enhanced erosion processes [P. Li et al., 2012; Wang et al., 2008]. Perhaps the most significant change has been the construction of the Three Gorges Dam, which became fully operational in 2009 [New and Xie, 2008]. It has recently affected the delivery of water, sediment, and terrestrial OC to the Changjiang estuary and associated East China Sea [Li et al., 2011; X. Li et al., 2012].

[4] The aforementioned alterations have resulted in considerable interests in this dynamic and globally important delta-front estuary. For example, some studies have focused on changes in primary and secondary production [Furuya et al., 2003]; sources of OC to sediment in the lower Changjiang River estuary, and East China Sea [Hu et al., 2012; Xing et al., 2011; Zhang et al., 2007; Zhu et al., 2008]; paleoclimate and paleoproductivity changes in the Changjiang watershed [Yang et al., 2011a]; and biomarker studies that have examined OC inputs and linkages with eutrophication and hypoxia [Li et al., 2011; Yu et al., 2011; Zhao et al., 2012]. These studies have greatly enhanced our understanding of how sediment, nutrient, and terrestrial OC delivery and primary production in the Changjiang estuary and the East China Sea being altered in the Anthropocene. Unfortunately, the linkages between historical OC source inputs, decay dynamics, and effects of physical forcing events (e.g., flooding and tropical cyclones) on the decay and preservation of OC in the Changjiang estuary and adjacent East China Sea are still not well understood.

[5] In this study, the working hypothesis was that spatial and temporal variability in downcore chemical biomarker profiles in the Changjiang estuary and the East China Sea shelf sediments would largely be controlled by differences in sediment accumulation rates and the occurrence of large perturbation events, such as flooding and tropical cyclones. We also posit that anthropogenic changes in the watershed (e.g., dams) have affected the efficiency with which proxies can be used to reconstruct past changes in organic carbon loading to coastal sediments. To do this, we collected sediment cores and measured radionuclides (210Pb and 137Cs) to determine the sediment accumulation rates and chronology at different locations. We also analyzed sediments for bulk (δ13C and total OC) and chemical biomarker proxies (lignin-phenols, plant pigments) to determine changes in OC source inputs. Finally, we make some preliminary linkages between OC preservation in coastal sediments, with regional climate variability and anthropogenic activities in the drainage basin.

2 Materials and Methods

2.1 Site Description and Sample Collection

[6] The Changjiang River is the longest river in China. It flows 6300 km from the Tibetan Plateau to East China Sea and covers 1.8 × 106 km2 catchment area. The Changjiang River is ranked fourth and fifth, respectively, in terms of water (900 km3 yr−1) and sediment (470 Mt yr−1) discharge to the global ocean [Milliman and Farnsworth, 2011]. The primary sediment depositional pathway is along the inner shelf off the Changjiang estuary in the East China Sea. A complicated array of major currents limits the export of the riverine-delivered sediment to the outer shelf [Liu et al., 2006, 2007], e.g., the southward China Coastal Currents along the coast, the northward Taiwan Warm Current and Kuroshio Current on the outer shelf (Figure 1).

Figure 1.

Sampling locations of the six sediment cores in the East China Sea. The dashed line labels the 75 m isobath (CDW—Changjiang Diluted Water; CCC—China Coastal Currents; TWC—Taiwan Warm Current; KC—Kuroshio Current).

[7] Six sediment cores were collected along the inner shelf within the mobile mud belt using a stainless steel box corer onboard R/V Dong Fang Hong 2 in November 2010; these cores covered a broad spatial area of the inner East China Sea shelf sediments. Subcores (7 cm inner diameter) were collected from each box core. The cores were cut into 1 cm increments using a stainless steel cutter, packed in aluminum foil, and stored at −80°C freezer onboard. Samples were shipped to Texas A&M University with dry ice right after the cruise. Sediment samples were freeze-dried using a benchtop vacuum freezing dryer (Millrock Co.), except those for the porosity analysis. Freeze-dried samples were homogeneously ground with a mortar/pestle and stored under vacuum until analysis. Sampling stations and locations are listed and shown in Table 1 and Figure 1.

Table 1. Sampling Location, Linear Sediment Accumulation Rate (LSR), Mass Sediment Accumulation Rate (MAR), and Sediment Mixed Layer Depth Determined by 210Pb Sampled From the Inner Shelf of the East China Sea
IDSampling DateLongitude (°E)Latitude (°N)Water Depth (m)Mixed Layer Depth (cm)LSR (cm yr−1)MAR (g cm−2 yr−1)
ECS212/2/2010122.6529.903700.56 ± 0.130.61 ± 0.20
ECS2a12/2/2010122.8429.775250.40 ± 0.090.46 ± 0.15
ECS2b12/2/2010123.0329.646340.26 ± 0.050.31 ± 0.08
ECS311/30/2010122.2428.823230.45 ± 0.270.45 ± 0.38
ECS411/29/2010121.5628.011980.20 ± 0.080.21 ± 0.12
ECS511/27/2010120.3626.284850.18 ± 0.0030.17 ± 0.004

2.2 Radionuclides

[8] Porosities for the depth core intervals were calculated prior to freeze-drying from water content (wet-dry weight). Sample intervals for radiotracer analysis were sealed in 40 mm diameter Petri dishes and allowed to sit for a minimum of 21 days to allow 210Pb activities to ingrow to secular equilibrium. Downcore activities of particle-reactive radiotracers (210Pb and 137Cs) were determined by counting for 1–2 days on a Canberra low-energy intrinsic germanium gamma spectrometer, with 2000 mm2 geometry. Total 210Pb activity was determined from the 46 keV photopeak, and supported 210Pb activities were determined by using averaged activities of the 226Ra daughters 214Pb (295 and 352 keV) and 214Bi (609 keV). The 137Cs activities were determined from the 661.6 keV photopeak and were derived to check the sediment accumulation rates derived from Pb. Detector efficiencies for this geometry were calculated using a natural sediment standard (IAEA-300 Baltic Sea sediment) and detector backgrounds were determined using Petri blanks at the energies of interest.

[9] A constant rate of radionuclides supply model was used which assumes a steady state input activity [Turner and Delorme, 1996]. Below a surficial mixed layer of homogenous excess 210Pb activity created by biophysical mixing, an exponential best fit (e.g., decay) was applied to calculate the linear sedimentation rates (LSR in cm yr−1) [Allison et al., 2007] and mass accumulation rate (MAR in g cm−2 yr−1) using an average porosity of 0.60 and a particle bulk density for siliciclastic sediment of 2.55 g cm−3 [Berner, 1980].

[10] The surface mixed layer of the sediment cores was assumed to be deposited in the sampling year (2010), since there is an annual cycle of sediment loading and reworking [Zhao et al., 2012] on the East China Sea shelf. In the sediment accumulation layer, a chronology was calculated based on a depth-age relationship derived from the LSR model for each core. We were able to track as far back to the year of 1880 (ECS5) across all these cores. While highly energetic deltaic shelves like the East China Sea have a variability of sediment flux that influences sediment deposition rate on seasonal to annual timescales, the relatively high regression (R2 from 0.56 to 0.99) fit to the 210Pb data, and the relatively good agreement between the 210Pb- and 137Cs-derived accumulation rates, suggests these episodic changes are averaged sufficiently in the sediment record to utilize these results, as in other systems [Nittrouer et al., 1979], as decadal-scale averages of sediment accumulation. This method has been applied in prior accumulation studies in the East China Sea as well [e.g., Liu et al., 2007].

2.3 Bulk Carbon, Nitrogen, and Stable Isotopes

[11] About 40 mg of ground sediment was placed in silver capsules and acidified with 12 N HCl vapor for 8 h in a desiccator in order to get rid of the inorganic carbon, mainly in the form of carbonate. Samples were then dried at 60°C and carefully crimp sealed in tin capsules for bulk parameter analysis [Harris et al., 2001]. Carbon (C) and nitrogen (N) analyses were conducted on an elemental analyzer interfaced to a continuous flow isotope ratio mass spectrometer in the Stable Isotope Geosciences Facility at Texas A&M University. The final isotopic values are reported in the δ notation relative to the international standards:

display math

where RSTD is either the 13C/12C ratio of the Vienna Pee Dee Belemnite for carbon or the 15N/14N ratio of air for nitrogen, respectively. Rsample is the ratio for the sediment samples. The precision for δ13C and δ15N is better than ±0.1‰ and ±0.4‰.

2.4 Lignin-Phenols Analyses

[12] The CuO oxidation process under a basic environment was utilized to extract lignin-phenols from sediment samples according to the method of Hedges and Ertel [1982] as modified by Goñi and Hedges [1992]. Briefly, homogenized sediment containing approximately 3–5 mg of OC was transferred to stainless steel reaction vials with 330 mg (±4 mg) CuO and 2 N NaOH. The mixtures were purged by nitrogen in a glove box before sealing and digestion at 150°C for 3 h. Reaction products were neutralized and then extracted with three successive 3 mL aliquots of diethyl ether (peroxides removed), filtered through combusted glass fiber, dried under N2, and converted to trimethylsilyl derivatives using bis-(trimethylsilyl)-trifluoroacetamide. Lignin-phenol derivatives were analyzed with an Agilent 6890-Gas Chromatograph/5973-Mass Spectrometric Detector.

[13] Quantification of lignin-phenols was based on the response factors derived from a mixed standard calibration curve containing known amounts of 12 lignin reaction products of interest as well as the internal standard ethyl vanillin. Eight lignin-phenol oxidation monomer products, C (ferulic acid + cinnamic acid), V (vanillin + acetovanillone + vanillic acid), and S (syringealdehyde + acetosyringone + syringic acid), were quantified and used as molecular indicators for source and diagenetic state of terrestrial vascular plant tissue [Hedges and Mann, 1979]. Other compounds were also derived after cupric oxidation and quantified, including the p-hydroxybenzenes (P: p-hydroxybenzaldehyde, p-hydroxyacetophenol, p-hydroxybenzoic acid) and 3,5-dihydroxybenzoic acid (3,5 Bd) which were derived from both soils and marine sources [Goñi and Hedges, 1995]. The variability of triplicate analysis for the sum of lignin-phenols was less than 15%, while that for individual compounds ranged from 5% to 25%.

[14] Lambda-8 (Λ8) is used as an indicator of terrestrial vascular plant material defined as the total weight in milligrams of the sum of C, V, and S phenols, normalized to 100 mg of organic carbon [Hedges and Parker, 1976]. Λ6 is a similar index but does not include the C phenols. The acid to aldehyde ratios of both V and S phenols, (Ad/Al)v, (Ad/Al)s, were used as indicators of the degradation state of lignin prior to burial due to microbial oxidation of the propyl side chains [Barrick and Hedges, 1981; Crawford, 1981; Hedges et al., 1988]. Since angiosperm plants only produce S phenols, and nonwoody vascular plants only produce C phenols, the C/V and S/V phenols were plotted as indicators of woody/nonwoody and gymnosperm/angiosperm sources of OC [Hedges and Mann, 1979]. P/(S + V) is used as a specific brown rot degradation pathway of lignin side chains [Dittmar and Lara, 2001].

[15] The lignin-phenol vegetation index (LPVI) is applied to provide the plant sources of OC [Sánchez-García et al., 2009; Tareq et al., 2004]. The 3,5 Bd is used as a common product of soil degradation process, and the 3,5 Bd/V indicated the input and humification of soil OC sorbed on fine particles [Farella et al., 2001; Houel et al., 2006; Louchouarn et al., 1999].

2.5 Plant Pigments

[16] Plant pigment biomarkers were first ultrasonically extracted with acetone from sediment according to the methods of Bianchi et al. [1995], as modified by Chen et al. [2001]. After centrifugation, the supernatant was blown to dryness under a N2 stream. The concentrated residue was re-dissolved in 100 μL of acetone for further analysis by high-performance liquid chromatography (HPLC). Samples were analyzed with a Waters HPLC 996 Photodiode array detector and Shimadzu RF 535 Fluorescence detector (excitation set at 440 nm and emission set at 660 nm) on a reversed-phase Grace Adsorbosphere C18 column (5 µm, 250 × 4.6 mm inside diameter) using the gradient flow described by Chen et al. [2001]. The qualification and quantification of each plant pigment were performed by comparison of retention time, absorption spectra, and the area of the peak in each sample chromatogram to those of the standards. trans-β-apo-8′-carotenal was used as an internal standard in the extraction. Pigment standards were purchased from the DHI Water and Environment Company, Denmark. They were run individually and mixed to determine retention times, spectra, and the response factors. The detection limit on this instrument was approximately 1 nM g−1 OC assuming an OC content of approximately 1% sediment dry weight.

2.6 Statistical Analyses

[17] The Origin 8.6 software was used in the data analysis and graphing process. As multivariate, statistical, and exploratory analysis methods, principal component analysis (PCA) and K means cluster analysis were performed on a matrix of geochemical data to determine which variables analyzed above significantly explain the relative characteristics of the sedimentary OC. Observations containing one or more underdetected or missing values were replaced by a random number between zero and the limit of detection before PCA. Statistical differences were performed using one-way analysis of variance. Statistically significant differences are discussed within the 95% confidence interval.

3 Results

3.1 Radionuclides and Chronology

[18] According to long-standing interpretation of 210Pb downcore profiles [Nittrouer et al., 1979], a surface mixed layer of homogenous activity is present when biological downmixing, or physical suspension and redeposition occur on time scales relatively short compared to the half-life of 210Pb (22.3 years). The depth of this layer (Table 1 and Figure 2a) was utilized as a marker for determining the upper limit for the sediment accumulation calculation and as an indicator of the intensity of biophysical mixing. The depths of mixed layer varied significantly depending on the locations of the stations with ranges from 0 at ECS2 to 8 cm at ECS4 with no significant patterns along the coast and offshore due to the dynamic feature of the East China Sea (Table 1).

Table 2. Bulk Parameters of the Six Sediment Cores Sampled in 2010 in the East China Seaa
  %TNδ15N (‰)%TOCδ13C (‰)C/NPorosity
  1. a

    C/N is the Atomic Molar Ratio.


[19] 210Pb profiles were used to propagate the chronology with an assumption of the constant rate of radionuclides supply model. The chronology was further confirmed by the 137Cs peak and downcore biomarker data especially by the pigment profiles (e g., Figure 2b) that indicated a surface mixed layer on top of each core. The LSR shown in Table 1 decreased from the river mouth region (ECS2, 0.56 ± 0.13 cm yr−1), southward (ECS5, 0.18 ± 0.00 cm yr−1), and also from nearshore (ECS2, 0.56 ± 0.13 cm yr−1) to the offshore (ECS2b, 0.26 ± 0.05 cm yr−1) direction. The average MAR of 0.37 ± 0.17 g cm−2 yr−1 (Table 1) is comparable to approximately 0.3 g cm−2 yr−1 for the shelf cited by Huh and Su [1999]. This distribution pattern follows previous studies of the dispersal system that Changjiang-derived sediment is remobilized from the river mouth and moved offshore and downdrift by prevailing coastal and tidal currents [Huh and Su, 1999; Liu et al., 2007]. This alongshore-offshore trend has also been observed in other large delta-front estuaries, such as the Mississippi-Atchafalaya [Corbett et al., 2006; Neill and Allison, 2005] and the Amazon systems [Kuehl et al., 1986].

Figure 2.

Example of (a) radionuclides and (b) pigment profiles at station ECS2b (samples missing at 11 and 24 cm for pigment analysis). The excess 210Pb, chlorophyll a and its degradation products (pheophytin a and pheophorbide a) showed a surface mixed layer and sediment accumulation layer labeled by the dashed line. Linear sediment accumulation rate was derived from the exponential decay of 210Pb in the sediment accumulation layer.

3.2 Bulk Carbon/Nitrogen, Stable Isotopes, and Porosity

[20] The average total OC (TOC) (%) content ranged from 0.37 ± 0.07 at offshore station ECS2b to 0.61 ± 0.08 at ECS5 in the south (Table 2). The OC concentration showed no obvious increases with time at ECS2 and ECS2a, but a slight increase from the bottom to the top of the cores at stations ECS2b, ECS3, ECS4, and ECS5 (data not shown here). Stations ECS2a and ECS2b had significantly lower TOC (%) than the other stations that were closer to the coastline.

[21] Total nitrogen (TN %) and porosity profiles covaried and were positively correlated with TOC (%) (%TOC = 0.009 + 6.78× %TN, R2 = 0.91; %TOC = 0.39 + 0.41 × Porosity, R2 = 0.71). The average TN (%) content was 0.06 ± 0.02 at ECS2, ECS2a, ECS2b, and ECS4 and slightly higher at ECS3 (0.08 ± 0.02) and ECS5 (0.09 ± 0.01). The porosity was lower at stations closer to the river mouth such as ECS2 (0.59 ± 0.03), ECS2a (0.56 ± 0.03), and ECS2b (0.55 ± 0.05), and higher at farther south stations with 0.63 ± 0.07 at ECS3, 0.61 ± 0.07 at ECS4, and 0.63 ± 0.04 at ECS5. This was likely due to the differences in the LSR and the extent of compaction in different areas in the ECS.

[22] The average C/N ratios were significantly different between the coastal stations and offshore stations (Table 2a) . For instance, C/N ratios in sediment at nearshore stations, ECS2 (8.87 ± 0.45), ECS3 (8.35 ± 0.61), ECS4 (8.18 ± 0.69), and ECS5 (7.89 ± 0.21), were higher than stations ECS2a (7.34 ± 0.18) and ECS2b (7.76 ± 0.55), which were farther offshore. The δ13C (‰) showed more depleted average values at stations closer to the coast, −23.49 ± 1.46 at ECS2, −22.84 ± 1.27 at ECS3, −21.97 ± 0.46 at ECS4, and −22.52 ± 0.51 at ECS5, compared to offshore stations such as ECS2a and ECS2b which had more enriched average values (−21.60 ± 0.43, −22.49 ± 1.37, respectively), indicative of greater contributions from the marine-derived OC. There were no significant differences in average δ15N among different sediment cores; the lowest average value was 5.28 ± 0.25 at ECS2, while the highest was 6.10 ± 0.30 at ECS4.

Table 3. Concentration and Ratios of the CuO Oxidation Products of the Six Sediment Cores Sampled in 2010 in the East China Sea From the Inner Shelf of the East China Seaa
  Λ8 (mg/100 mg OC)Λ6 (mg/100 mg OC)(Ad/Al)v(Ad/Al)sC/VS/VP/(S + V)VSPC3,5 Bd (mg/100 mg OC)3,5 Bd/V∑8 (mg/g dry)LPVI
  1. a

    Λ8 = C + V + S, normalized to 100 mg of organic carbon while Λ6 does not include C phenols; C = ferulic acid + cinnamic acid; V = vanillin + acetovanillone + vanillic acid; S = syringealdehyde + acetosyringone + syringic acid; P = p-hydroxybenzaldehyde + p-hydroxyacetophenol + p-hydroxybenzoic acid; 3,5 Bd = 3,5–dihydroxybenzoic acid; (Ad/Al)v, (Ad/Al)s are the acid to aldehyde ratios of both V and S phenols: LPVI = lignin-phenol vegetation index.

Table 4. Plant Pigment Concentrations (nmol g−1 OC) in the Six Sediment Cores Sampled in 2010 in the East China Seaa
  1. a

    SCEs are Sterol Chlorin Esters and CCEs are Carotenol Chlorin Esters (CCEs).

 Clorophyll C219-ButanoyloxyfucoxanthinFucoxanthinNeoxanthin19-HexanoyloxyfucoxanthinPrasinoxanthinPheophorbide aDiadinoxanthinAlloxanthinDiatoxanthinZeaxanthinCantaxanthinChlorophyll bChlorophyll aPheophytin aα-caroteneβ-caroteneCCESCE

[23] Apparent decay rate and preservation efficiency were calculated to study the OC decay dynamics. For the OC deposited in the sediment mixed layer, apparent decay rate was calculated from the exponential decrease in the sediment cores with time based on one G-model (Table 5) [Chen et al., 2005]. Preservation efficiency was calculated as the mean OC concentration in the sediment accumulation layer divided by the upper value in the sediment mixed layer [Sampere et al., 2011b] considering an average porosity of 0.60 (Table 2). Sediment TOC in this study did not show significant decay (Table 5) with TOC MAR ranging from 9.08 ± 7.03 g m−2 yr−1 at ECS4 to 28.43 ± 12.96 g m−2 yr−1 at ECS2 using the average TOC content of each core (Table 6). Correspondingly, the TOC preservation efficiencies were between 71 and 99% (Table 5).

Table 5. Decay and Preservation of TOC and Biomarkers of the Six Sediment Cores Sampled in 2010 in the East China Sea
  1. a

    CuO products such as Λ8 did not show apparent decay. Apparent degradation rate constants were derived from the slope of ln(Ct/C0) versus year (one G-model), while C0 is the initial concentration in the surface sediment and C is the concentration at year t [Chen et al., 2005; Sampere et al., 2011b].

  2. b

    Preservation efficiency (PE) of representative biomarkers in the six sediment cores sampled in 2010 in the East China Sea (inline image) [Sampere et al., 2011b]. na: not applicable.

Apparent Degradation Rate Constantsa
TOCnana−0.04 ± 0.01−0.05 ± 0.04na0.34 ± 0.01
Chlorophyll bna2.10 ± 0.72na0.00 ± 0.00nana
Chlorophyll a0.80 ± 1.290.30 ± 0.100.11 ± 0.030.12 ± 0.160.29 ± 0.19na
Pheophytin ana0.10 ± 0.03nana0.17 ± 0.110.49 ± 0.01
Pheophorbidea0.09 ± 0.030.10 ± 0.030.06 ± 0.010.02 ± 0.020.10 ± 0.070.17 ± 0.00
Fucoxanthin0.44 ± 0.150.49 ± 0.170.49 ± 0.130.37 ± 0.50na0.47 ± 0.01
Diatoxanthin0.01 ± 0.000.45 ± 0.150.24 ± 0.060.17 ± 0.220.18 ± 1.17na
Zeaxanthin0.12 ± 0.04na0.06 ± 0.020.10 ± 0.14nana
Canthaxanthin0.11 ± 0.04na0.32 ± 0.080.11 ± 0.15nana
Neoxanthin0.59 ± 0.200.04 ± 0.010.40 ± 0.100.40 ± 0.531.37 ± 0.880.13 ± 0.00
Prasinoxanthin0.41 ± 0.141.33 ± 0.450.27 ± 0.230.24 ± 0.321.57 ± 1.010.51 ± 0.01
Diadinoxanthin0.39 ± 0.131.24 ± 0.420.35 ± 0.090.44 ± 0.591.18 ± 0.760.31 ± 0.01
Alloxanthin0.44 ± 0.152.92 ± 1.000.60 ± 0.16na1.81 ± 1.170.10 ± 0.00
SCE0.06 ± 0.010.01 ± 0.000.03 ± 0.01na0.02 ± 0.020.03 ± 0.00
CCE0.01 ± 0.00-0.01 ± 0.000.01 ± 0.010.01 ± 0.01-
Preservation Efficiencyb
Chlorophyll a0.210.420.320.210.060.52
Table 6. Mass Accumulation Rate of Selective Parameters in Six Sediment Coresa
  1. a

    Standard deviation calculation is based on the ranges of TOC values in each downcore profile.

TOC (g m−2 yr−1)28.43 ± 12.9618.21 ± 8.3211.27 ± 3.9922.94 ± 27.659.08 ± 7.0310.36 ± 0.33
TN (g m−2 yr−1)3.71 ± 1.692.88 ± 1.321.69 ± 0.603.21 ± 3.871.26 ± 0.971.54 ± 0.05
Λ8 (mg m−2 yr−1)1106.89 ± 504.79434.38 ± 198.55134.33 ± 47.49896.74 ± 580.93441.80 ± 341.87228.60 ± 7.39
Chlorophyll a (nmol m−2 yr−1)229.91 ± 104.85178.44 ± 81.56324.38 ± 114.69419.14 ± 403.2923.52 ± 18.20114.12 ± 3.69
Zeaxanthin (nmol m−2 yr−1)643.21 ± 293.34613.72 ± 280.53310.02 ± 109.61997.86 ± 827.82100.92 ± 78.10418.78 ± 13.54

3.3 Lignin-Phenols

[24] The Λ8 (mg 100 mg OC−1) values were slightly higher than Λ6 (mg 100 mg OC−1). This suggests that nonwoody plant detritus inputs were not important, as indicated by little or no C phenols (0.01 ± 0.01 to 0.02 ± 0.02) (Table 3), and/or that C phenols were preferentially degraded over V and S phenols [Hedges and Mann, 1979] before deposition in East China Sea sediment. For example, at station ECS2, Λ8 was 0.38 ± 0.05 while Λ6 was 0.36 ± 0.05. Although not significant, all the CuO oxidation products showed spatial and temporal differences among different cores. At stations near the coastline from the river mouth to the south, the average Λ8 values were higher (0.38 ± 0.05, 0.38 ± 0.13, and 0.44 ± 0.16 for ECS2, ECS3, and ECS4) than those stations farther offshore or south at stations ECS2a (0.24 ± 0.07), ECS2b (0.12 ± 0.02), and ECS5 (0.22 ± 0.04). Generally, the sediment Λ8 in the East China Sea was lower than that in the suspended particles in the lower reaches of the Changjiang River ranging from 0.90 ± 0.41 to 1.91 ± 0.33 [Yu et al., 2011].

[25] The (Ad/Al)v ratios were generally lower in nearshore compared to offshore stations. For example, ratios at nearshore stations ECS2 (0.29 ± 0.03), ECS3 (0.28 ± 0.09), and ECS4 (0.30 ± 0.04) were lower than those at ECS2a (0.34 ± 0.06), ECS2b (0.29 ± 0.05), and ECS5 (0.31 ± 0.05). While the (Ad/Al)s at ECS3 (0.23 ± 0.08), ECS4 (0.26 ± 0.08) were lower than those at ECS2a (0.28 ± 0.08). The average values of P/(S + V) in each core increased from ECS4 (0.21 ± 0.07), ECS3 (0.23 ± 0.07) along the coast to ECS2a (0.36 ± 0.07) and ECS2b (0.41 ± 0.10) offshore; a similar trend was observed for 3,5 Bd/V (Table 3).

[26] C/V and S/V ratios in a source plot indicated that the terrestrial OC was derived from a mixture of woody and nonwoody angiosperms but dominated by the woody angiosperms (Figure 3). This was confirmed by the LPVI values which were averaged from 24.37 ± 14.20 (ECS2a) to 69.84 ± 44.42 (ECS4) in all sediments (Table 3).

Figure 3.

Source plot of the sediment samples from the inner shelf of the East China Sea. A = woody angiosperm plants; a = nonwoody angiosperm plants; G = woody gymnosperm plants; g = nonwoody gymnosperm plants.

[27] Stations ECS3 and ECS4 showed several subsurface peaks in concentrations of Λ8 with a less degraded signature, e.g., lower P/(S + V) (3,5 Bd/V ratios are too low to show variability) (Table 3 and Figures 4d and 4e). Interestingly, these subsurface peaks corresponded to historical flooding events around the years 1936, 1954, 1983, and 1998 (Table 7 and Figure 4). In contrast, a more recent flooding event in 2010 did not result in a similar subsurface peak of Λ8 in these two cores as was shown in previous flooding vyears. In the other four cores, such variation was not captured in the sediment layers from the Λ8 profiles (Figures 4a–4c and 4f). The Λ8 profiles showed no apparent decay with extremely high preservation (Table 5) that the lignin MAR ranged from 229 ±7 mg m−2 yr−1 at the south station ECS5 to 1107 ± 505 mg m−2 yr−1 at ECS2 station (Table 5).

Figure 4.

Concentration of Λ8, Λ6 (per 100 mg of organic carbon), ratios of P/(S + V), and 3,5 Bd/V of the six sediment cores from the East China Sea (dashed lines label the mixed layer depth). The samples at 18 cm (ECS2) and 23 cm (ECS5) are missing.

Table 7. The Flood Events in the Changjiang River Catchment During 1840–2010a
YearFlood AreaPeak Discharge (m3 s−1)
  1. a

    1840–2000 data are from Shi et al. [2004].

  2. b

    Mean discharge in July [Gong et al., 2011].

1848Midstream and downstream
1849Midstream and downstream<92,600
1860Midstream and upstream92,500
1869Midstream and downstream
1870Midstream and upstream57,300
1889Midstream and downstream
1906Midstream and downstream 
1921Midstream and upstream76,400
1924Midstream and upstream36,900
1926Midstream and upstream29,900
1931The whole catchment50,000
1949Midstream and downstream
1954The whole catchment92,600
1969Midstream and downstream62,400
1980Midstream and downstream60,100
1995Midstream and downstream74,500
1998The whole catchment81,700
1999Midstream and downstream84,500
2010Midstream and downstream60,527b

3.4 Plant Pigments

[28] As biomarkers for marine phytoplankton, the average plant pigment concentrations did not show significant spatial patterns (Table 4). However, most plant pigments showed significantly higher average concentrations in the surface mixed layer than those in the underlying sediment accumulation layer (Table 5). For pigments deposited in sediment mixed layer, zeaxanthin, sterol chlorin esters (SCEs), and carotenol chlorin esters (CCEs) generally showed lower decay rates (0.06 ± 0.02 to 0.12 ± 0.04 year−1, 0.01 ± 0.00 to 0.06 ± 0.01 year−1, and 0.01 ± 0.01 year−1) than chlorophyll a (0.11 ± 0.03 to 0.80 ± 0.29 year−1) (Table 5). This is further supported by their higher preservation efficiency (21%–76%) than chlorophyll a (6–52%) at each station (Table 5). These decay rates are comparable to those found in the sediment accumulation layers from the Mississippi estuary in the northern Gulf of Mexico [Chen et al., 2003]. Compared to bulk sediment TOC, plant pigments represent only a small fraction of TOC deposited in the surface sediment [X. Li et al., 2012] with lower preservation efficiency of only 6 to 76% (Table 5). The pigment MAR was estimated to be 24 ± 18 to 419 ± 403 nmol m−2 yr−1 for chlorophyll a and 101 ± 78 to 998 ± 828 nmol m−2 yr−1 for zeaxanthin.

3.5 PCA and Cluster Analysis

[29] The final PCA model (Figure 5a) included 14 normalized variables. The variables that had similar slopes were not included as they explained similar characteristics of the TOC, such as TN (similar to TOC); most of the plant pigments located in the I quadrant, such as canthaxanthin (similar to diatoxanthin), chlorophyll b, 19-hexanoyloxyfucoxanthin (similar to chlorophyll a), diadinoxanthin (similar to pheophorbide a); and some CuO oxidation products such as (Ad/Al)s [similar to (Ad/Al)v], LPVI, C/V (similar to Λ8). Some of the samples that overlapped with others in the PCA plot were also excluded in Figure 5 such as ECS3-15 cm (overlapped with ECS3-18 cm), ECS4-28 cm (overlapped with ECS4-29 cm), etc.

Figure 5.

(a) PCA biplot and (b) K means cluster analyses of variables measured in this study.

[30] The PCA biplot showed a significant spatial-temporal pattern, suggesting the importance of the sedimentary OC distribution and preservation at spatial and temporal scales (Figure 5a). The first two axes explained 62.2% of the sedimentary OC (first axis 38.78%; second axis 23.46%). According to the absolute contribution of each variable to one axis (coordinate2/eigenvalue), the first axis mainly contained information as related to sediment depth (a contribution of −36% to PC1), fucoxanthin (36%), zeaxanthin (35%), diadinoxanthin (32%), diatoxanthin (27%), β-carotene (20%), CCEs (35%), and chlorophyll a (31%). The second axis was primarily composed of information related to C/N ratios (35%), TOC (46%), Λ8 (31%), δ13C (−23%), (Ad/Al)v (−34%), P/(S + V) (−33%), and 3,5 Bd/V (−20%).

[31] Most of the mixed layer samples (e.g., ECS4-1 cm, ECS3-2 cm, and ECS5-9 cm) were located in the I and IV quadrants, while most of the accumulation layer sediments (e.g., ECS2-20 cm, ECS3-30 cm, and ECS4-21 cm) were located in the II and III quadrants. Sediment samples from ECS2, ECS3, ECS4, and ECS5 were mostly located in the I and II quadrants and samples from ECS2a and ECS2b were located in the III and IV quadrants.

[32] In the nearshore stations (ECS2, ECS3, ECS4, and ECS5), there was more terrestrial OC presented as indicated by higher Λ8, C/N ratios associated with higher bulk TOC content, while in the offshore sediment (ECS2a and ECS2b), the OC was characterized by enriched δ13C (marine-derived OC) and degraded signature of terrestrial OC, e.g., higher ratios of (Ad/Al)v, 3,5 Bd/V and P/(S + V). Cluster analysis differentiated the data matrix into four different groups, similar to the PCA.

4 Discussion

4.1 Sediment Dynamics

[33] About 90% of the East China Sea sediment is derived from the Changjiang River [Liu et al., 2006]. Due to the counterclockwise direction of currents in the East China Sea, the mobile mud sediment is limited to the inner and midshelf overall [Liu et al., 2007; Xu et al., 2009]. These sediments, mainly composed of clay (32.1%) and silt (59.1%) particles [Youn et al., 2007], are hypopycnally dispersed southward along the coast [Zhu et al., 2011] for about 1000 km to the western Taiwan Strait. Substantial amounts of particulates delivered by the Changjiang River are buried in the continental shelf with rapid mud accumulation [Cai and Dai, 2004; Liu et al., 2007; Zhu et al., 2006]. This has important implications for the decay and preservation of OC in sediment of the East China Sea. For example, previous work has shown that more than 95% of terrestrial OC accumulated in the estuary was transported within 100 to 150 km offshore along the coast of East China Sea, with only a small percentage (i.e., <5%) being transported farther out of the shelf [Aller et al., 1985; Deng et al., 2006; Hu et al., 2012].

4.2 Decay and Preservation of Sedimentary OC Derived From Algal and Terrestrial Sources

[34] There was higher preservation efficiency for bulk TOC (71–99%) and TN (71–97%) than expected in these mobile muds in the East China Sea sediment. Other work has shown that mobile muds have some of the highest decay rates in sediment in part due to the frequent resuspension events [Aller, 1998; Blair and Aller, 2012; Keil et al., 1997]. However, some other mobile muds were shown to also have slower than expected decay rates such as off the Fly River in the Gulf of Papua (about 30% for terrestrial OC) [Goñi et al., 2006]. Some of the bulk materials in the East China Sea may be refractory with relatively old radiocarbon age. For example, ∆14C-TOC downcore measurements of the East China Sea sediment have been reported to be as old as 12,250 years BP [Wang and Li, 2007], considerably older than most shelf sediment at similar depth in large delta-front estuaries [Dickens et al., 2004; Gordon and Goñi, 2003; Mollenhauer et al., 2005]. Although not reported in this study, we and others have shown that black carbon [Wang and Li, 2007; X. Li et al., 2012] and more stable geochemical biomarkers such as hopanes, steranes, PAHs, and unresolved complex mixtures were abundant in the East China Sea sediment [Hu et al., 2012]. In addition, the deforestation and erosion processes due to the land use changes in the Changjiang drainage basin may cause an input of the older and refractory terrestrial OC into ECS. These sources may have also had more physical protection in complex sediment clay aggregates. Another factor that may have reduced the high decay capacity in these mobile muds is the fact that we collected the samples after a summer flood event in 2010 which likely remobilized part of the mixed layer [X. Li et al., 2012] where most efficient OC remineralization occurs. Thus, differences of TOC content between the mixed layer and the accumulation layer were likely to be underestimated, leading to an overestimate of preservation efficiency, compared to previous work that suggested 30% of OC is preserved in typical shelf sediment worldwide [e.g., Burdige, 2007]. As a result, the overall integrated standing stock of TOC in East China Sea (from 1900 to present) is estimated to be approximately 1.62 ± 1.15 kg C m−2 in this study which is about one tenth of the TOC stock of all the middle and lower lakes in the Changjiang catchment [Dong et al., 2012].

[35] Differences in the preservation efficiency of sediment TOC can be controlled by many pre- and post-depositional factors such as OC sources, redox, particle size, bio/geopolymerization, and sediment mixing, and accumulation rates [Burdige, 2006, and references therein]. Decay rates of marine phytoplankton- and terrestrial-derived OC showed significant differences across the East China Sea, based on biomarker profiles, as described below.

4.2.1 Marine Phytoplankton Inputs

[36] Algal-derived plant pigments deposited in the sediment mixed-layer had profiles that showed an exponential decrease with depth (data not shown here). The chlorophyll a apparent decay rate (0.11 ± 0.03 to 0.80 ± 0.29 year−1) is comparable to other large delta-front estuaries, where sediments were in highly mobile mud zones, such as the Mississippi/Atchafalaya estuaries, where chlorophyll a decay rates ranged from 0.01 to 0.54 year−1 [Sampere et al., 2011b]. More stable chlorophylls, e.g., CCE and SCE, have been used to track the historical eutrophication in the East China Sea due to their relatively greater stability [Zhao et al., 2012] with lower decay rates and higher preservation efficiency (Table 5). The existence of pheophorbide a, a common indicator of metazoan grazing [Jeffrey and Vesk, 1997], and pheophytin a, one of the most common chlorophyll a decay products, indicated two of the depositional decay pathways of the chlorophyll a [Zhao et al., 2012].

[37] Carotenoids such as zeaxanthin and fucoxanthin indicated that cyanobacteria and diatoms were common community groups in the East China Sea sediment (Table 4). Zeaxanthin has been used as a stable biomarker to reconstruct the historical cyanobacterial blooms in the Baltic Sea for the past approximately 8000 years [Bianchi et al., 2000]. As expected, zeaxanthin showed significantly higher preservation efficiency (22%–76%) than the more labile chlorophyll a (6%–52%) in East China Sea sediment. Fucoxanthin has also been used in reconstructing longer-term development of eutrophication-related blooms in the Baltic Sea proper [Bianchi and Engelhaupt, 2002]. However, zeaxanthin is chemically more stable than fucoxanthin, in part because of the presence of a 5,6′-epoxide group in fucoxanthin which makes it more susceptible to bacterial decay [Repeta and Gagosian, 1987; Bianchi and Engelhaupt, 2002]. Thus, differences in decay dynamics of algal-derived OC, which were very apparent in East China Sea sediment, were largely controlled by differential predepositional and postdepositional decay of different source inputs, as well as the changing spatial and temporal dynamics of mobile muds at different stations.

4.2.2 Terrestrial Inputs

[38] Lignin has been shown to be more resistant to degradation than marine phytoplankton-derived OC [Blair and Aller, 2012; Burdige, 2005; Haider, 1992; Zonneveld et al., 2010] due to its aromatic structure and nonhydrolyzable intramolecular bonds [Haider, 1992]. Terrestrially derived OC, based on lignin-phenols, had greater overall preservation than marine phytoplankton-derived material in East China Sea sediment (Table 5). However, different groups of lignin-phenols did show differential decay in East China Sea sediment. For example, the small amount of C phenols in sediment may in part have been due to their lower inputs of OC sources that contain C phenols or greater susceptibility over V and S phenols [Hedges and Parker, 1976]. However, it was more likely from preferential decay because previous studies [Yu et al., 2011; Zhu et al., 2008; Li et al., 2011] and this study have shown the existence of nonwoody gymnosperm and angiosperm plants sources in the East China Sea sediment. Lignin preservation efficiency was found to be more than 100% (Table 5) for vascular plant material deposited in the sediment mixed layer—this is not what we expected if we assume that the mixed layer should decompose OC efficiently. Once again, this calculation is likely biased and overestimated because of the effect of the flood events and the lack of information on the lignin flux to the sediment-water interface. However, there may have been some stabilization of lignin through bio/geopolymerization of lignin molecules during transport from the river to the large delta-front estuaries that can further protect lignin from degradation [Blair and Aller, 2012]. Furthermore, there is a significant fraction of lignin leached from plant litter in surface soils before and during their entry to aquatic systems [Hernes et al., 2007]. Subsequently, photodegradation and/or microbial degradation will occur during the transport from stream and rivers to the shelf [Opsahl and Benner, 1998], leaving more decay-resistant material to be deposited in shelf sediment as was shown from the insignificant variation from the lignin ratios with depth (Figure 4).

[39] To estimate lignin flux to the East China Sea, we used particulate OC flux which has decreased from >5 × 106 t yr−1 during the period of 1960–1980 to about 2 × 106 t yr−1 in 1997 due to decreasing sediment load within the Changjiang River [Zhu et al., 2006; Wu et al., 2007]. The estimated lignin flux decreased from >(9.55 ± 2.33) × 104 t yr−1 during the period of 1960–1980 to about (3.82 ± 0.93) × 104 t yr−1 in 1997 assuming an average Λ8 concentration of 1.91 ± 0.33 mg 100 mg−1 OC in the suspended particles in the downstream of the Changjiang River [Yu et al., 2011]. Assuming a sediment burial rate of 0.97 Gt yr−1 in the East China Sea [Deng et al., 2006], an average TOC (%) of 0.47 ± 0.29 and an average Λ8 (mg 100 mg−1 OC) of 0.30 ± 0.12 from this study, the lignin burial rate is (1.38 ± 0.72) × 104 t yr−1. Interestingly, this would imply that there was a significant amount of predepositional loss of lignin-phenol (85%) during sediment deposition and resuspension. This fits with the concept of frequent resuspension of materials in mobile muds belts which would enhance decay rates of OC. Another possibility may be the effects of priming, whereby labile algal exudates enhance the breakdown of terrestrially-derived POC during deposition and resuspension in the plume region. Bianchi [2011] has suggested that this process may be more important in coastal zones than previously thought.

[40] Decay and preservation of OC in East China Sea sediment have been affected by changes in redox and in particular the frequency and extent of hypoxia events [Middelburg and Levin, 2009] over the past 60 years [Li et al., 2011]. However, hypoxia in the Changjiang Estuary has rarely been documented, most likely due to an absence of high-resolution monitoring efforts [Chen et al., 2007]. Besides, the mechanisms controlling and fueling hypoxia are not clear [Li et al., 2011]; further work is needed to better understand historical OC cycling and hypoxic events in this area.

4.3 Spatial and Temporal Distribution of the Sedimentary OC

[41] Based on recent work, we have shown that marine-derived OC contributed the most to the TOC pool in surface sediment of the East China Sea [X. Li et al., 2012]. It has been reported that dissolved inorganic nitrogen (DIN) flux into the Changjiang Estuary increased from (261 ± 109) × 106 kg yr−1 in pre-1980 to (1385 ± 209) × 106 kg yr−1 in post-1990 [Zhao et al., 2012]. Increasing DIN inputs to the East China Sea are thought to be closely linked with coastal eutrophication, frequent harmful algal bloom events (HABs), and hypoxia events [Chen et al., 2012; Diaz, 2001; Wang, 2006; Wang et al., 2012]. Chlorophyll a concentrations induced by HABs in the inner shelf have been reported to be as high as 300 mg m−3 when dinoflagellates (notably Prorocentrum dentatum) dominate in spring and fall while diatoms (e.g., Skeletonema coastum) dominate in summer when the Changjiang floods [Zhang et al., 2006]. Recently, cyanobacteria has become an important group in the phytoplankton community in the East China Sea [Li et al., 2011], especially when Si became more limiting in the East China Sea due to the dam trapping effect [Li et al., 2011]. In fact, the Si:N ratio in the Changjiang river plume region decreased from 1.5 in 1998 to 0.4 in 2004 [Gong et al., 2006]. The flux of phytodetrital materials to the sediment is likely to be supported by the eutrophication and HABs in the water column. Our work shows that TOC had more enriched δ13C signatures in the sediment mixed layer (Figure 5), supporting the notion of greater inputs of phytodetrital materials in recent years. However, with increasing depth, marine-derived OC as indicated by plant pigments was relatively less abundant as indicated by high C/N ratios and Λ8 values (Figure 5). The decrease of plant pigment concentration with depth is in part from greater decay of algal-derived material, as discussed earlier, and from greater increase in primary production in recent years. Separating these differences can be quite difficult, and further work is clearly needed on this issue.

[42] Lignin-poor, more degraded [e.g., higher (Ad/Al)v and P/(S + V) ratios] terrestrial OC appeared to be hydrodynamically sorted with finer sediment [Hu et al., 2012; Kao et al., 2003; Zhu et al., 2011] and preserved in offshore sediment. It was also diluted by marine-derived OC as was indicated from a lack of correlation between TOC and lignin-phenol abundance. Conversely, lignin-rich, less degraded terrestrial OC were deposited nearshore close to the coast within the mobile mud belt region. This pattern of hydrodynamic sorting “freshness” of lignin near coast was also found in the Mississippi/Atchafalaya estuaries in the northern Gulf of Mexico [Bianchi et al., 2007] and the Washington margin (northwest U.S. coast) [Keil et al., 1994; Prahl et al., 1994]. The nearshore stations (ECS2, ECS3, ECS4, and ECS5) also showed higher TOC and higher pigment concentration as indicated from the PCA analyses with more enriched δ13C (marine-derived OC) than in the offshore stations (ECS2a and ECS2b) (Figure 5). Similar patterns along the East China Sea were also found using other terrestrial and marine biomarkers such as n-alkanes, C37 alkenones, brassicasterol, and dinosterol [Hu et al., 2012; Xing et al., 2011]. Both terrestrial [Burdige, 2005] and marine OC inputs [Gong et al., 2011] in nearshore sediment were higher than those in the offshore region.

[43] In addition to the hydrodynamic sorting effect and differences in decay rates between pigment and lignin-phenols, it appears that relocalization of surface mobile muds caused by tides, currents, and large-scale events like flooding, tropical cyclones, and monsoons are influencing the decay and preservation of OC. Significant remobilization of sediment during high energy (e.g., wind/wave) months was observed in the Mississippi estuary [Corbett et al., 2004]. Hurricane Ivan mobilized significant amounts of sediment and sedimentary OC offshore on the Louisiana shelf [Sampere et al., 2008]. In terms of the East China Sea as discussed earlier, the Changjiang River-derived sediment and terrestrial OC are distributed in the mobile mud belt along the inner shelf of the East China Sea as far south as Taiwan Strait [Liu et al., 2007]. Furthermore, in the East China Sea, mesomacrotidal (2.6–5 m) [Chai et al., 2006; Shi, 2004] effects on sediment remobilization and OC relocalization were likely greater than those in the microtidal (<1 m) Mississippi/Atchafalaya estuaries [Kjerfve et al., 2002]. Therefore, the permanent sediment accumulation occurs after several deposition and resuspension cycles along the coast [Corbett et al., 2004]. This likely affected the ability of using the deposition of terrestrial biomarkers in certain sediment cores to record historical flooding events (see section 'Reconstruction of Historical Flood Events'), hypothesizing that flooding events which induced higher sediment discharge could bring in more terrestrial-derived OC.

4.4 Reconstruction of Historical Flood Events

[44] Historical records of past flooding events in the river catchment can be recorded in the coastal marine sediment cores. Researchers have used radionuclides [Allison et al., 2000, 2005], and terrestrial biomarkers [Louchouarn et al., 1999] as tools to track past floods, hurricanes/monsoons, and changes in river discharge. We observed what we believe to be evidence for major historical flooding events in East China Sea sediment cores that could be dated back to 1880 from this work. Major flooding events have occurred in the Changjiang catchment over the past 150 years (Table 7) that have been linked in part to large-scale climatic effects (East Asian Monsoon; El Niño–Southern Oscillation, ENSO; and Pacific Decal Oscillation, PDO) [Qian, 1997; Tong et al., 2006]. Some of the worst flooding events occurred around 1936, 1954, 1983, and 1998 [Shi et al., 2004; Yu et al., 2009], mainly due to ENSO activities [Kim et al., 2012]. Water discharge was dramatically increased during these flooding events which also resulted in greater transport of sediment and terrestrial OC from the catchment to the East China Sea [Zhou et al., 2007]. For example, in cores ECS3 and ECS4, lignin-phenol concentrations of Λ8 and Λ6 varied with the variation of flooding years. Terrestrially derived OC (based on Λ8) showed increased concentration around peak flooding years 1936, 1954, 1983, and 1998 (Figure 4). The flooding events likely transported eroded and degraded soil-derived OC. Meanwhile, the higher river discharge also reduced the turnover time of the plant detritus delivery with less predepositional degradation [West et al., 2011] as was indicated from lower P/(S + V) ratios around the flooding years (Figure 4). However, for the very recent 2010 summer flood event in the Changjiang catchment [Luo et al., 2012], there were no Λ8 subpeaks in the two downcore profiles. We suggest that this recent sediment discharge and its terrestrially derived OC input is still in the mixed layer and has not yet found its way to the sediment accumulation layer. This is coupled with the overall reduction in sediment supply and accumulation as a result of the dam construction [Yang et al., 2011b]. More work is needed to determine if the apparent decoupling of flooding events in upper catchment and sediment in the East China Sea in more recent years is due to increased damming and/or climate change in the Anthropocene.

[45] Physical reworking and lateral transport in the East China Sea played an important role in OC decay and preservation in the sediment cores. In contrast to these patterns observed in nearshore cores at ECS3 and ECS4, the core at ECS5, farthest from the river mouth, did not record the flooding events discussed earlier but around years 1920 and 1945. This is likely due to the distance from the Changjiang River mouth; stations this far away are likely to be more temporally decoupled than stations closer to the mouth of the Changjiang River. However, stations ECS2, 2a, and 2b that were closer to the Changjiang River mouth also did not record the above mentioned flooding events although the lignin profiles seemed to show certain insignificant variations with depth corresponding to some flooding years indicated in Table 7. The complex effects from monsoons, currents, tides, and the “turbidity diluting” effects of the Changjiang water plume [P. Li et al., 2012] in the East China Sea could cause frequent sediment resuspension and lateral transport. These processes likely remobilized OC and disturbed the mixed-layer more at these stations than others. Although the mechanisms are not clear, we do know large delta-front estuaries are very dynamic sedimentary environments with spatial variability that is likely caused by small scale and nonlinear seasonal processes. For example, recent work has also shown a shift of the highest lignin concentration from the Changjiang mouth to further south in East China Sea partly due to the summer flood energy [X. Li et al., 2012]. A similar decoupling of OC preservation and monsoon variability was observed in the East China Sea that was attributed to the weak mixing gradient between fresh and marine waters [Yang et al., 2011a]. These processes would prevent the reconstruction of historical events over decadal or century scales [Yang et al., 2011a].

[46] Climate effects, such as the East Asian Monsoon, ENSO, and PDO, that have affected the precipitation in the Changjiang drainage basin likely also affect the delivery of terrestrial-derived OC to the coast. We did not find significant correlations between these climate indexes and the Λ8 profiles/sediment discharge in any sediment cores. We attribute this in part to human perturbations over the past decade (e.g., dam construction) which would tend to decouple such linkages in the sediment record. For example, Wang et al. [2008] concluded that the East Asian monsoon impacted the water and sediment discharge of the Changjiang River before 1950, but it was controlled more directly by human alterations of the system afterward. Longer cores would be needed to better examine the impact of monsoons on the terrestrial OC delivery and deposition. However, Tong et al. [2006] statistically analyzed teleconnections between the Changjiang flood/drought events and ENSO activities and found that the flood/drought cycles were longer than ENSO events. In contrast to the southern California (USA), which is located on the downstream side of the global large-scale atmospheric circulation, where sediment flux is positively correlated with ENSO [Inman and Jenkins, 1999], the East China Sea coast is located in the upstream side of the atmospheric circulation [Yang et al., 2004], where such correlations with precipitation and river discharge have been more difficult to find.

5 Conclusions

[47] The East China Sea sediment accumulation rate and mass accumulation rate followed the dispersal pattern of the mobile mud derived from the Changjiang River. Decay and preservation of OC showed apparent spatial and temporal variabilities in the East China Sea sediment. At nearshore stations, both terrestrial- and marine-derived OC had higher inputs than the offshore stations. The distribution of the terrestrial OC indicated by lignin-phenols was more input driven and more preserved along the coast, while lignin-poor, more degraded OC were hydrodynamically sorted and transported farther offshore likely with finer particles. The preservation of marine-derived OC was more decay driven, as most of the pigments deposited in sediment mixed layer were degraded and less preserved in the underlying sediment accumulation layer. Other factors, such as physical dynamics (wind, tide, and monsoon), lateral transport, and dam construction, played important roles in sediment accumulation, decay, and preservation of OC in East China Sea sediment. Due to the reduced sediment load in the East China Sea over the past 50 years, there has been a significant reduction (60%) in the lignin export to the East China Sea. The lignin burial rate is estimated to be (1.38 ± 0.72) × 104 t yr−1. The overall integrated standing stock of TOC in East China Sea (from 1900 to present) is approximately 1.62 ± 1.15 kg C m−2, about 1/10 of the TOC stock in all the middle and lower lakes in the Changjiang catchment.

[48] The detection of the flooding events using lignin-phenol biomarker in sediment cores was observed for older floods (e.g., 1936, 1954, 1983, and 1998) but not for the very recent flood (e.g., 2010). Further research on how watershed perturbations (e.g., flooding events and land use changes) are decoupled from OC storage in shelf sediments is needed in future work. This is especially important considering ongoing climate change effects and the continued decrease in sediment loads due to continued dam construction and water diversion projects in China. This work demonstrates that while large-scale events such as flooding are detectable in shelf sediments, there is a general enhancement of the decoupling between Chinese rivers and sediments in the East China Sea due to the dramatic reduction of sediment loading from the expansion of damming in recent years.


[49] This work was supported in part by funding from Texas A&M University Research Office, NOAA, the Chinese Scholarship Council, and the Chinese National Science Foundation (973 CHOICE-C project, 2009CB421200). The authors would like to acknowledge Minhan Dai, Weidong Zhai, and Deli Wang at Xiamen University for the cruise support. X Li would like to thank Qian Li, Xiaolong Lin, Jianbin Liu, Feifei Meng, Lei Wang, Songli Xu, and Honghai Zhang for the sample processing during the two field cruises and Zhao Xu for the help in data processing. The authors also acknowledge the crew onboard R/V Dong Fang Hong 2.