LIU Xiaojie, HAO Ting, FENG Lijuan, JI Yinli, WANG Qianqian, ZHANG Dahai, PAN Gang, GAO Xianchi, MENG Chunxia,*, and LI Xianguo,*

1) Key Laboratory of Marine Chemistry Theory and Technology (Ocean University of China), Ministry of Education,Qingdao 266100, China

2) College of Chemistry and Chemical Engineering, Ocean University of China, Qingdao 266100, China

3) State Key Laboratory of Marine Geology, Tongji University, Shanghai 200092, China

4) Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, China

5) School of Animal, Rural, and Environmental Sciences, Nottingham Trent University, Nottingham NG25 0QF, UK

Abstract The area of East China Sea (ECS) inner shelf is an important sink of suspended particulates from Yangtze River (YR)and materials originated from YR basin. In this study, the parameters of lignin and alkane biomarkers in surface sediment samples from ECS inner shelf were determined to distinguish the sources and to trace the transport of terrigenous organic matters in the region. Our dataset showed that total alkanes with carbon numbers from 10 to 38 (T-alkanes) were significantly correlated to both TAR(terrigenous/aquatic ratio) and HMW/LMW (the ratio of high molecular weight to low molecular weight alkanes) (r = 0.88, P ＜ 0.05 for both), indicating that the majority of T-alkanes was predominantly originated from terrestrial sources, and T-alkanes are important constituents of terrestrial organic matters in the study area. The area was probably affected by petroleum pollution to a certain degree,as indicated by the values of carbon preference index (CPI), odd-over-even carbon number predominance (OEP) and the ratio of pristane to phytane (Pr/Ph). The values of Pr/n-C17 and Ph/n-C18 suggested a strong reductive sedimentary condition in the region with no obvious biodegradation. The content of eight lignin phenols (Σ8) decreased from the coast to the open sea, indicative of riverine input and hydrodynamic transport of terrigenous organic matters. Lignin degradation parameters presented an increasing trend from the coastline toward the open sea. The lignin vegetation parameters and alkane index (AI) suggested the predominance of non-woody angiosperms in the YR basin. The obvious correlation between Λ8 (Σ8 normalized to 100 mg organic carbon), TAR and HMW/LMW reveals the significantly concurrent input of lignin and alkanes from terrestrial sources.

Key words terrigenous organic matter; biomarker; lignin; alkanes; surface sediment; East China Sea inner shelf

1 Introduction

Understanding the delivery, fate and cycling of terrigenous organic matters (TerrOMs) to the oceans is of great importance to understanding the global marine carbon cycling. The study of sources and transport of TerrOMs by using organic biomarkers is a major field for the application of organic geochemistry. Various biomarkers have been applied to marine sediments to evaluate the composition and provenance of TerrOMs as well as the fate of marine organic matters (MarOMs) since 1970s (Albaigéset al.,1984). However, the accuracy of single category parameters is still unsatisfactory. Previous works using different categories of biomarkers have certainly demonstrated the importance of the approach (e.g., Ishi-watariet al., 2005)for obtaining generally applicable quantitative parameters.

The inner shelf of East China Sea (ECS) is an important interface between continents and oceans that receive organic carbon (OC) inputs from both terrestrial and marine sources (Bianchi and Allison, 2009). The Yangtze River (YR) discharges 480 Mt yr-1of sediments and 2-5 Mt yr-1of organic matters to the ECS (Wuet al., 2004);twenty to thirty percent of them are finally buried in the ECS inner shelf. Therefore, the ECS inner shelf is an important sink of suspended particulates from YR and materials originated from YR basin. As a consequence, the area is characterized by a high sedimentation rate and productivity in the coastal environment (Liuet al., 2010;Chenet al., 2015), which is ideal for studying the cycling of terrigenous organic carbon (TerrOC). The terrestrial input in the area is also influenced by human activities and hydrological features in Zhejiang-Fujian coast mud area.It is of great significance to choose the area for the study of anthropogenic disturbance on marine organic geochemical processes.

As typical terrestrial biomarkers, both alkane and lignin have been widely used in estuarine sediments to investigate the sources and transport of TerrOC; and a variety of derived parameters have been used or proposed for this purpose (Prahlet al., 1994; Harriset al., 1996; Ohkouchiet al., 1997; Meyers and Lallier-vergés, 1999; Meyers, 2003; Pisaniet al., 2013). Some alkane derived parameters can also indicate the redox condition of surroundings, the maturity (aging) of organic matters and the pollution of petroleum. Lignin is essentially absent from the tissues of marine organisms and relatively resistant to degradation compared to other plant organic materials(Tommasoet al., 2007; Liet al., 2014; Chenet al., 2015;).Lignin phenols are indicative of certain plant tissue types and of the degradation processes of TerrOC (Tommasoet al., 2007). Therefore, lignin phenols are valuable tracers for studying the sources and transport of TerrOC from river watersheds onto the inner shelf and then into deeper marine areas (Tommasoet al., 2007; Kuzyket al., 2008).

Although relevant references have described and discussed the sources, transport and distribution of organic carbon with alkanes or lignin phenols in surface sediments, few of the studies have tackled the correlation of the two categories of parameters and their relevance. A multi-proxy approach could be helpful for a more comprehensive understanding and quantifying the relative contributions of terrestrial organic matters and the biogeochemical processes in the study area (Yaoet al., 2015).The correlation analysis for different categories of parameters can be corroborated and complement each other,and can synthetically analyze the sources and transport of TerrOC more accurately. In our study, we provided a synoptic view of TerrOC distribution in the surface sediments of the ECS inner shelf based on correlation study of alkane and lignin derived parameters. The study is aimed to gain an insight into the provenance, composition and fate of TerrOC in the ECS inner shelf.

2 Materials and Methods

2.1 Samples Collection

Eighteen surface sediment samples from the inner shelf of ECS (121°-124°E, 27°-32°N) were collected by using a Van Veen stainless steel grab sampler (Qingdao Orson,China) during cruises conducted byR/V Dong Fang Hong 2of Ocean University of China in April, 2010 and April,2011. The surface layers (0-2 cm) were cut by using a stainless steel bladder. All the samples were wrapped in aluminum foil, which had been burned in muffle at 450℃before use, and immediately stored at -20℃ until analysis. Sampling sites are shown in Fig.1.

Fig.1 Location of sampling sites in the ECS inner shelf and major water mass and currents in the region. Muddy areas are marked in gray (Modified after Wang et al. (2016)). YDW, Yangtze Diluted Water; ZFCC, Zhejiang-Fujian Coastal Current; TaWC, Taiwan Warm Current; KC, Kuroshio Current; TsWC, Tsushima Warn Current; YSCC, Yellow Sea Coastal Current; YSWC, Yellow Sea Warm Current; KCC, Korean Coastal Current.

2.2 Total Organic Carbon (TOC) and Mean Grain Size (MGS) Analysis

Coarse gravels and debris were picked out and discarded before samples were freeze-dried and ground. Prior to TOC analysis, the homogenized dry sediment samples were decalcified with 4 mol L-1HCl at room temperature for 24 h, and then rinsed three times with deionized water(centrifuged to remove the supernatant after each rinsing)and dried in an oven at 55℃. After the sample treatment process, TOC content was determined by using a Thermal Flash 2000 Elemental Analyzer with 30 mg of powdered HCl-treated samples (Zhaoet al., 2013). The standard deviation of TOC analysis was ± 0.02 wt% (n= 6).

Grain size analysis was performed by using a Mastersizer-2000 laser particle-size analyzer (Malvernb Instruments, UK) from 0.02 to 2000 μm with a 0.01 Φ size resolution (Haoet al., 2017). The measuring error was within 3%. Prior to these grain-size analyses, the samples were pretreated with 10% H2O2and 4 mol L-1HCl for 24 h each to remove the organic matter and biogenic carbonate, respectively. The grain sizes were divided into three groups:＜4 μm for clay, 4-63 μm for silt, and＞63 μm for sand,according to the Folk classification system. The MGS data of the samples in our study had been reported in our previous work (Duanet al., 2013), and were used only for correlation study in this work.

2.3 Sample Extraction, Separation and Analysis of Alkanes

The method for extraction of lipid biomarkers (includingn-alkanes) from sediment samples, both free and bounded fractions, was modified after Helleret al. (2012).A dry sediment sample (2.0 g) was added into a centrifugal tube with 0.05 g activated copper powder. Then a mixed solution of dichloromethane and methanol (with aV/Vratio of 9/1) was added, followed by Vortex mixing,ultrasonication and centrifugation. The supernatant was separated and combined for four repeated extraction cycles. The combined supernatant was evaporated to about 1-2 mL in a rotary vacuum evaporator and transferred to a reaction kettle to be saponified at 70-75℃ for 3 h with 15 mL 0.5 mol L-1KOH in methanol. After saponification,the mixture was extracted with n-hexane for 3 times to yield neutral lipids fraction, in which hydrocarbons and alcohols are included.

The combined extract from the saponified mixture was then concentrated to about 0.2-0.3 mL by a rotary vacuum evaporator and then transferred to a silica gel column (50 mm× Φ5 mm with silica gel of 100 mesh size). The targeted hydrocarbons, includingn-alkanes with carbon numbers of 10 to 38, pristane (Pr) and phytane (Ph), were eluted withn-hexane (5 mL) and analyzed.

The analysis was performed on a Shimadzu 2010Plus GC equipped with a flame ionization detector (FID) and a splitless injector. A DB-5 fused silica capillary column(30 m × 0.25 mm × 0.25 μm) was used, and nitrogen was taken as the carrier gas. The temperatures of injector and FID were set at 300℃. And the oven temperature was programmed from 80℃ (maintained for 2 min) to 300℃at 4℃min-1and held for 15 min.

2.4 CuO Oxidation and Analysis of Lignin Phenols

A conventional alkaline CuO oxidation of lignin in sediment samples (Hedges and Ertel, 1982; Go?i and Hedges, 1992; Have and Teunissen, 2001; Wanget al., 2005)was adopted with modifications made in our laboratory(Zhanget al., 2013). Briefly, in a glove box in N2atmosphere, a dry sediment sample of 1.00 g was charged into a Teflon reaction vessel with 0.5 g CuO powder and 0.05 g Fe(NH4)2(SO4)2·6H2O (used as an O2scavenger) in 15 mL 2 mol L-1NaOH solution. The reaction vessel was then capped in the glove box, and the oxidation was carried out at 170℃ for 3 h. After the reaction, the vessel was immediately cooled down to room temperature under running tap water. After adding recovery surrogates (transcinnamic acid and ethyl vanillin, to monitor the recoveries of lignin phenols), the mixture was transferred into a 30 mL centrifuge tube, ultrasonicated (for better mixing) for

15 min and centrifuged under 3000 r min-1for 10 min.The supernatant was transferred into another clean 50 mL centrifuge tube. The process was repeated twice to ensure complete transfer of the reaction products with the supernatant combined, which was then acidified with concentrated HCl to pH about 1, and transferred to a Cleanert PEP-SPE column (500 mg, Agela Technologies, China) forextraction of lignin phenols. The phenols were eluted with 5 mL ethyl acetate into a 5 mL vial, and was dried under a gentle N2stream and redissolved in acetonitrile for GC analysis.

Table 1 Lignin phenols analyzed in this work

An Agilent 6890 GC-FID system, coupled with on-column derivatization (Zhanget al., 2013) with BSTFA + 1%TMCS (N,O-bis-trimethylsilyl-trifluoroacetamide plus 1%trimethylchlorosilane), was adopted for analysis of lignin phenol products. A DB-1 column (30 m × 0.25 mm × 0.25 μm)was used for GC separation of derivatized pro- ducts with nitrogen as the carrier gas. The temperatures of injector and FID were set at 300℃. The oven temperature was programmed from 100℃ to 220℃ at 4℃min-1, and then ramped to 290℃ at 30℃min-1and held isothermally for 10 min.

The lignin phenols analyzed in this study are listed in Table 1.

2.5 Quality Assurance and Quality Control

Glassware and sodium sulfate (for dewatering on column chromatography) were heated at 450℃ for 3 h before experiments. Procedural blanks were processed in an identical procedure as any other samples (Sections 2.3 and 2.4) except without sediment for each batch of experiment, to check for any potential contamination from reagents and glassware. No quantifiable alkanes or lignin phenols were detected in these blanks. The quantification of alkanes and lignin phenols in sediment samples was therefore not blank corrected.

For the analysis ofn-alkanes, 1,3,5-tri-isopropyl benzene (TIPB) was used as the internal standard for quantitation. The response factors of individualn-alkanes relative to the TIPB standard were assumed to be 1.0 (Ishiwatariet al., 2005). The relative standard deviations for triplicated measurements ofn-alkanes were within 5%. For lignin phenols, the recovery rates of surrogates were 72.42± 5.44% and 72.86 ± 7.40% for ethyl vanillin and transcinnamic acid, respectively. All data for the lignin phenols were not corrected by the recoveries of surrogates because of the different functional groups between surrogates and target lignin phenols.

2.6 Statistical Analysis

Statistical analysis was conducted by using the software IBM SPSS Statistics. Both correlation coefficient andPvalue were considered for the study, and aPvalue less than 0.05 was considered as statistically significant.

3 Results

3.1 TOC and Mean Grain Size (MGS)

The TOC contents in surface sediments ranged from 0.16 to 0.95 mg (100 mg dw)-1with a mean value of 0.59± 0.18 mg (100 mg dw)-1(Table 2). The values of MGS ranged from 3.95 - 7.48 μm with a mean value of 6.58 ±0.96 μm (Table 2). The distributions of TOC and MGS were shown in Figs.2a and 2b. The highest value of TOC appeared at the intersection of Zhejiang-Fujian coastal current (ZFCC) and Taiwan warm current (TaWC), while the lowest value appeared at the offshore non-muddy area(Fig.2a). The distribution pattern of MGS was very much similar to that of TOC. There was a strong positive linear correlation (r= 0.91,P＜ 0.01) between the TOC and MGS,in consistence with those observed by Wanget al. (2016)and Duanet al. (2013) in the same area.

Table 2 The values of TOC, MGS and alkane parameters in surface sediments of the study area

Fig.2 Distributions of TOC (a), MGS (b), 3s (c), TAR (d) and HMW/LMW (e) in the ECS inner shelf.

3.2 Alkanes and Relevant Parameters

The values of total alkanes with carbon numbers from 10 to 38 (T-alkanes) ranged from 1.98 to 23.73 μg g-1dw(the mean value was 6.50 ± 5.40 μg g-1dw), with the highest value appeared at the area between Zhoushan and the surrounding islands (Fig.2c) where the hydrodynamic or/and sedimentary conditions could be beneficial to the deposition of terrestrial organic materials. The mean values of TAR (the ratio of long-chainn-alkanes (C27, C29and C31) to short-chainn-alkanes (C15, C17and C19)) and HMW/LMW (the ratio of high molecular weight alkanes to low molecular weight alkanes) were 0.94 - 24.91 and 0.95 -15.93, respectively, with similar spatial distribution patterns to that of T-alkanes (Figs.2c, 2d and 2e).Alkane index (AI) is often used to estimate the relative contribution of woody and non-woody plants, and to determine the type of maternal vegetation (Zhanget al., 2008).An enhanced value could be indicative of increased nonwoody vegetation in the area. The AI value in our study area was 0.50 - 0.59 with a mean value of 0.54 ± 0.03, indicating that the contribution of non-woody plants toterrestrialn-alkanes in this area was greater than that of woody plants. As shown in Fig.3a, higher AI values were mainly occurred near the mouth of the YR estuary and at the southeast part of the study area.

The CPI15-35(carbon preference index) value in our study area was 1.06 ± 0.09 (Fig.3b). The value of CPI21-35in our study area ranged from 0.90 to 1.59 with a mean value of 1.25 ± 0.18 (Table 2), supporting a conclusion that mixed biogenic organic matters (terrigenous and marine)contributed significantly to YR estuary and the inner shelf region along the Zhejiang-Fujian coast, with gradually weakening terrestrial predominance toward deep sea (Fig.3c).The mean value of OEP (odd-over-even carbon number predominance) in our study area was close to 1.0, indicative of possible petroleum contamination, especially in the area outside Hangzhou Bay, which we believe contribution from the bay is considerable.

3.3 Lignin Phenols and Relevant Parameters

Fig.3 Distributions of alkane parameters AI (a), CPI15-35 (b), CPI21-35 (c) and OEP (d) in the ECS inner shelf.

Σ8 is the sum of the 8 lignin phenols (syringyl, cinnamyl and vanillyl phenols, namely S, C and V phenols, respectively) normalized to 10 g dry weight sediment, while Λ8 is Σ8 normalized to 100 mg OC. The values of Σ8 and Λ8 were 0.57 - 1.12 mg (10 g)-1dw and 0.88 - 3.59 mg (100 mg)-1OC, with mean values of 0.79 ± 0.19 mg (10 g)-1dw and 1.49 ± 0.64 μg (100 mg)-1OC, respectively (Table 3).Fig.4a shows that the value of Σ8 decreases significantly with seaward distance from the shoreline, indicative of a terrestrially originated source of lignin and the influence of hydrodynamic transport from terrestrial sources.

The value of Pon/P (the ratio ofp-hydroxyacetophenon top-hydroxyl phenols) in our study area ranged from 0.28 to 0.37, with a mean value of 0.32 ± 0.02 (Table 3). Relatively high values of Pon/P were observed near Zhoushan islands, whilst low values were found in the southern part of the study area and off the YR estuary (Fig.4c). Low values were also found in the area farther away from Zhoushan islands.

The values of (Ad/Al)v (Ad/Al ratio for V phenols) were 0.61 - 1.17 with a mean value of 0.89 ± 0.17, suggesting that V phenols in this area experienced a relatively high degree of oxidative degradation. Our result is similar to a previous report in YR estuary (Yanget al., 2008). An increasing degradation degree towards the open sea is evident(Fig.5a). While the values of (Ad/Al)s (Ad/Al ratio for S phenols) were 0.20 - 0.42 with a mean value of 0.31 ± 0.07(Fig.5b). It seems that there exists a prominent negative correlation between (Ad/Al)v and (Ad/Al)s.

An enhanced value of P/(V+S) is indicative of increased demethylation or/and demethoxylation degradation of lignin. In our study area, the value was 0.46 ± 0.08, with a very similar distribution to that of (Ad/Al)v (Figs.5a and c), suggesting that the demethylation or/and demethoxylation degradation of lignin in an offshore area is also more pronounced than that along the coastal area. The value of 3,5-Bd/V is an indicator for lignin humification. It was 0.18± 0.04 in the study area, with a different distribution showing a high value area just off YR estuary, which we believe is an indication of land soil contribution to the se-dimentary TerrOMs in the area.

Table 3 The values of TOC and lignin parameters in surface sediments of ECS inner shelf

Fig.4 Distributions of Σ8 (a), Λ8 (b) and Pon/P (c) in the ECS inner shelf.

Fig.5 Distributions of degradation parameters (Ad/Al)v (a), (Ad/Al)s (b), P/(V+S) (c) and 3,5-Bd/V (d) in the ECS inner shelf. For easy comparison, the distribution of MGS (e) is also included.

All of the S/V values in our study area were above 0.76(Table 3), indicating that the vascular plants in the source area were dominated by angiosperm plants. While the value of C/V ranged from 0.19 to 0.36 with a mean value of 0.29 ± 0.05. The scattered cross-plot of S/V and C/V ratios(Fig.6) showed obviously that the vascular plants were dominated by non-woody angiosperms.

Fig.6 The scattered cross-plot of S/V and C/V ratios according to Bianchi et al. (2011) and Gong et al. (2017) for 18 sediment samples in ECS inner shelf.

LPVI can detect the vegetation changes as a single proxy and the ranges of LPVI were calculated based on the plant lignin data (Tareqet al., 2004) and were further modified by Tareqet al. (2011). The values of LPVI in our study area ranged from 11.95 to 136.20, with a mean value of 74.99 ± 32.10, suggesting the dominance of nonwoody angiosperms with non-negligible contribution from woody angiosperms.

4 Discussion

4.1 The Indication of Alkane Parameters for Organic Matters

4.1.1 Total alkanes and alkane derived terrestrial parameters

The relative contribution of terrigenousversusmarine sources can be evaluated by the TAR index, the ratio of long-chainn-alkanes (C27, C29and C31) to short-chain nalkanes (C15, C17and C19) (Bourbonniere and Meyers, 1996;Silliman and Schelske, 2003).These long-chainn-alkanes are typical of leaf waxes from higher land plants, and they are considered indicative of predominant contribution from TerrOC, while the short-chain alkanes are attributed to plankton or petrogenic sources (Belickaet al., 2004; Yunkeret al., 1995). HMW/LMW can eliminate the influence of deposition rate and particle size on the source identification of organic matters, and a high HMW/LMW n-alkane ratio is a signal that TerrOC is dominating over the marine source (Karlssonet al., 2011). These parameters T-alkanes, TAR and HMW/LMW are highly correlated each other (Figs.2c - 2e). Therefore, it seems that the majority of T-alkanes was predominantly originated from terrestrial sources.

4.1.2 The determination of petroleum pollution

CPI is routinely used to estimate mixing sources (terrigenous and marine) and degradation status of terrestrialn-alkanes (Clark and Blumer, 1967). Then-alkanes from epicuticular wax of higher plants show a pronounced odd carbon number predominance (CPI ＞ 1) (Eglinton and Hamilton, 1967; Simoneitet al., 1979), while fossil fuel and microbial hydrocarbons exhibit a CPI - 1 (Bray and Evans,1961). For homologues ofn-alkanes with more than 21 carbon atoms, an odd to even predominance (CPI21-35＞1)usually suggests a predominantly terrestrial input from vascular plant waxn-alkanes (Goudie, 1990). Combining with the values of odd-over-even carbon number predominance(OEP) and Pr/Ph in the following section, we propose that the area was possibly affected by oil pollution to a certain degree.

Petroleum hydrocarbons are often characterized by low OEP values close to 1.0, while the OEP of freshly produced hydrocarbons (‘young hydrocarbons’) is higher than 1.6(Villanuevaet al., 1997). Our OEP values were close to 1.0,suggesting the area was probably polluted by petroleum.However, we noticed two high-value regions: one near the YR estuary and another near Jiao River around 28.5°N(Fig.3d). The riverine nutrient inputs are apparently responsible for the high OEP values, which is corresponding to the high productivity in these two regions.

Pristane (Pr) and phytane (Ph) were found in all samples. The ratio of Pr to Ph (Pr/Ph) is often used as an indicator for petroleum hydrocarbons (Readmanet al., 2002;Orenet al., 2006). It is less than 1 for sediments contaminated by petroleum hydrocarbons; otherwise, it would be more than 3 (Steinhauer and Boehm, 1992; Dréauet al.,1997; Readmanet al., 2002). On the other hand, hexadecane (n-C16) is often considered as one of the representative compounds of petroleum hydrocarbons, which is rarely found in marine sediment without oil contamination (Blumeret al., 1971; Youngbloodet al., 1971). In our study area, Pr/Ph values of the 18 samples fell in the range of 0.16 - 0.41 (Fig.7a), and the average concentration of n-C16 was (0.25 ± 0.17) μg g-1dw. Both parameters, along with CPI and OEP discussed above, suggest that the area was petroleum-contaminated to a certain extent. It was reported that more than two thousands of oil spill incidents happened at the coastal water area of China in the past 20 years. Especially, severe pollution accidents were witnessed at least once every year in the ECS (Guo and Lei, 2003).

Fig.7 Distributions of isoprenoid hydrocarbon parameters Pr/Ph (a), Pr/n-C17 (b) and Ph/n-C18 (c) in the ECS inner shelf.

Pr/n-C17and Ph/n-C18are usually used as proxies for the sedimentary redox condition, the maturation and biodegradation degree of hydrocarbons (B?ckeret al., 2013; Schwarzbaueret al., 2013). Fig.8 is a cross-plot of the two proxies in the study area, showing that Pr/n-C17and Ph/n-C18values fell in the range of 0.18 - 0.48 and 0.28 - 0.99 with mean values of 0.31 ± 0.08 and 0.73 ± 0.17, respectively (Fig.8). According to Shanmugam (1985), the sedimentary condition in our study area is strongly reductive with no obvious biodegradation, which is very different from the depositional environment of rivers and lakes reported by Chattopadhyay and Dutta (2014) and B?ckeret al.(2013).

4.2 The Indication of Lignin Parameters for TerrOM

4.2.1 The degradation of lignin

The sources of P phenols are not exclusively terrestrial plants. Marine phytoplankton, bacteria and zooplanktonetc.are also contributors (Hedgeset al., 1988; Go?i and Hedges,1995). However, Pon is only derived from vascular plants.Therefore, Pon/P can be used as an indicator for determining the sources of P phenols and to estimate the contribution of terrestrial organic matters in sediments (Go?iet al.,1998; Go?iet al., 2000). The distribution pattern of Pon/P(Fig.4c) suggests that the terrestrial inputs from Hangzhou Bay area and Zhoushan islands are significant, while the addition of MarOMs might be responsible for the low values. Especially, the area off the YR estuary is famous for the high productivity because of the huge input of nutrients from YR, and thus contributes more MarOMs. It was also confirmed that the addition of MarOMs could result in the difference between the values and spatial distributions of Σ8 and Λ8 (Kuzyket al., 2008; Yaoet al., 2015).

Under the influence of hydrodynamic sorting, fine- grained sediment is preferentially transported farther from its sources (Kaoet al., 2003; Kuzyket al., 2008; Zhuet al.,2011). Owing to its longer residence time during its transportation before settled down onto the seafloor, the organic matter in fine-grained particles is often much degraded than that in coarser particles (Mitraet al., 2000; Gordon and Go?i, 2004; Liet al., 2013). Therefore, the degradation of lignin adsorbed on particles would be higher as it transported farther and time went on.

There are a variety of lignin phenol-related parameters for characterizing the degradation status of TerrOMs. The ratio of acid to aldehyde (Ad/Al) is indicative of oxidative degradation of lignin phenols (Bianchiet al., 1999; Sánchez-Garcíaet al., 2009; Jexet al., 2014). The P/(S+V) has been suggested as a proxy for side chain decomposition of lignin via demethylation or/and demethoxylation of V and S phenols by brown-rot fungi (Dittmar and Lara, 2001;Jexet al., 2014). 3,5-dihydroxy-benzoic acid (3,5-Bd) is thought to be a humification product of soil organic matters (Prahlet al., 1994; Louchouarnet al., 1999; Farellaet al., 2001; Houelet al., 2006), and the ratio of 3,5-Bd to total V phenols (3,5-Bd/V) is therefore indicative of the input of humificated soil components (Liet al., 2014). Based on these parameters, previous studies have shown that lignin is basically humificated in land soil or/and degraded during its transport in fluvial and marine systems (Hedges and Ertel, 1982; Hedgeset al., 1985; Ishiwatari and Uzaki, 1987; Oremet al., 1997; Dittmar and Lara, 2001;Liet al., 2014), whilst the humification and degradation is negligible after buried in sediment.

It was widely recognized that V phenols are existed in all higher vascular plants, while the majority of S phenols is derived from angiosperms only. It was also found that S phenols were more conducive to be adsorbed on fine particles (Tesiet al., 2007; Kuzyk,et al., 2008). Therefore,during the transportation of TerrOMs, more S phenols are preferentially transported to and deposited in offshore area because of hydrodynamic sorting, while V phenols are less affected. Consequently, the value of (Ad/Al)s in offshore area was lower than that near the coast, presenting a different distribution pattern from that of (Ad/Al)v. Therefore,compared with (Ad/Al)v, (Ad/ Al)s is not a good indicator for lignin oxidative degradation. A similar phenomenon was also reported by Tareqet al. (2011) and Ottoet al.(2005). In summary, hydrodynamic sorting exerts an important control in theland-river-seatransport process of TerrOMs, and thus has a great impact on degradation status of TerrOMs or lignin.

4.2.2 The vegetation provenance of lignin

The C/V and S/V ratios can provide information on the type of vegetation in sediment source regions (Jexet al.,2014). All vascular plants contain V phenols, while C phenols are mainly found in non-woody plants, and S phenols are essentially found only in angiosperm plants (Hedgeset al., 1988). Therefore, an S/V value about zero is indicative of gymnosperm plants while a value greater than 0.9 is for angiosperm plants. C/V ratios of less than 0.1 and greater than 0.2 are indicative of woody and non-woody plants, respectively (Hedges and Mann, 1979; Go?i and Hedges, 1992). In addition, LPVI can provide a unified single proxy for identifying vegetation provenance of lignin, as stated in Section 3.3.

A similar variation trend is observed for LPVI and S/V,as shown in Figs.9a and c, in accordance with the predominance of angiosperm plants in the study area; while the values of C/V had a remarkable correlation with degradation parameters (Ad/Al)V, P/(V+S) and 3,5-Bd/V (r= 0.96,0.86 and 0.81,P＜ 0.01, respectively). On the other hand, the values of C/V generally showed an increasing trend from coast to offshore areas (Fig.9b). We believe that the relative enrichment of C phenols in finer particles (Jiet al.,2019) is responsible for the distribution pattern of C/V and for its good correlation with degradation parameters. Lignin had been degrading and finally depositing during the transport process over time. The longer time it consumed,the higher degree of degradation it would be. As we stated above, hydrodynamic sorting preferentially transports finer particles farther with longer residence time. Since finer particles are enriched with C phenols, that leads to the increasing trend of C/V and degradation degree of lignin with increasing distance from the shoreline. We therefore infer that TerrOMs from non-woody vascular plants (with higher content of C phenols) adsorbed on finer particles are preferentially transported to the distal sediments, which is in general consistent with those reported by Bianchiet al. (1997) and Go?iet al. (1998). These results implied that the composition, quantities and fate of TerrOMs in offshore sediments are generally considered to be dominantly controlled by hydrodynamic sorting process, which alters the grain size distribution in surface sediments. Consequently, the associated TerrOMs could be remobilized and transported farther, leading to a divergent partitioning behavior for different lignin phenols.

Fig.9 Distributions of vegetation parameters S/V (a), C/V (b) and LPVI (c) in the ECS inner shelf. For easy comparison,the distribution of MGS (d) is also included.

4.3 The Relevance Between Lignin and Alkane Biomarkers

Both lignin and alkane can be used as indicators for the input of TerrOMs in marine system. We assumed that there existed a certain correlation among these parameters for comprehensively explaining the sources and transport of TerrOMs.

The correlation between TOC, MGS and some of the biomarker (lignin and alkane) parameters is shown in Table 4.Compared with n-alkane parameters, lignin parameters are more obviously correlated linearly with TOC and MGS.Σ8 is positively correlated with TOC (or MGS), while a negative correlation between Λ8 and TOC (or MGS) is evident. This might be an indication that lignin is one of the major contributors to TOC. It is widely recognized that in a certain range of particle size, the adsorption of TOC onto particles increases with decreasing particle size (Tesiet al., 2007, 2016; Liet al., 2014; Wanget al., 2015),which is also evident from the very strong correlation between TOC and MGS (Table 4) in our study. However, the normalization of lignin phenols content to TOC will lead to a reversed trend of Λ8.

Table 4 The correlation coefficient between lignin and alkane parameters

On the other side, the lignin degradation degree would be greater with increasing distance and time during the transport process, which preferentially delivers finer particles farther away from the shoreline and along the coastal currents because of hydrodynamic sorting. Therefore,negative relationships were observed between MGS and the lignin degradation parameters (Ad/Al)v, P/(V+S) and 3,5-Bd/v to a certain extent. As stated above, (Ad/ Al)Sis not a good indicator for oxidative degradation of lignin;we then did not include (Ad/Al)Sin the correlation analysis. Moreover, the area with higher values of MGS and TOC also had higher values of S/V and LPVI. This suggests that organic matters from non-woody angiosperms are more prone to be enriched in finer particles with higher values of MGS and TOC.

The positive correlation between Λ8 and TAR, HMW/LMW (r= 0.526 and 0.543,P＜ 0.05, respectively) might be an indication that lignin and long-chain alkanes were concurrently transported into the area from their terrestrial sources.

For vegetation parameters, both lignin-derived C/V (＞0.2) and alkane-derived AI (≥ 0.50) are indicative of predominant non-woody angiosperms in the provenance region. A weakly positive correlation was observed between AI and C/V (r= 0.323,P＜ 00.01), S/V (r= 0.304,P＜0.01) and LPVI (r= 0.249,P＜ 0.01), indicating that when the con- tribution of non-woody plants increases, the contribution of angiosperms also increases to a certain extent,which is in general consistent with the predominance of non-woody angiosperms in YR basin.

5 Conclusions

In this work, we studied the sources and transport of TerrOMs in the ECS inner shelf by determining lignin and n-alkane parameters and their relationship.

There was a strong positive linear correlation between TOC and MGS in our study area, indicating that the adsorption of TOC onto particles increases with decreasing particle size. The alkane parameters revealed that T-alkanes were predominantly originated from terrestrial sources,and were enriched in the area of Zhoushan islands. Terrestrially originated sources of lignin and the influence of hydrodynamic transport in marine system are evident from the distribution patterns of lignin parameters. The alkane derived AI and the lignin derived S/V and C/V suggested that the terrestrial vascular plants were dominated by nonwoody angiosperms, and the LPVI indicates that the contribution from woody angiosperms was also non-negligible. Furthermore, alkane derived parameters CPI, OEP and Pr/Ph showed that the study area had been contaminated by petroleum to a certain degree. Pr/n-C17and Ph/n-C18suggested that the sedimentary condition is strongly reductive with no obvious biodegradation while the organic matters were buried in sediment.

Hydrodynamic sorting, which alters the grain size distribution, exerts an important control in theland-river-seatransport process of terrigenous organic matter and consequently the composition, quantities and fate of TerrOMs in offshore sediments. As a result, organic matters from nonwoody vascular plants adsorbed on finer particles are preferentially transported to the distal sediments. And the degradation of lignin would be enhanced as it transported farther and time went on. Lignin and alkane derived parameters, including content, vegetation and degradation parameters, are correlated well each other to different extents,which is a suggestion that both of them were concurrently transported into the area from terrestrial sources.

Acknowledgements

This study was financially supported by the National Natural Science Foundation of China (Nos. 41276067, 4102 0164005) and the National Basic Research Program of China (973 Program, No. 2010CB428901). We thank the crews ofR/V Dong Fang Hong 2for sampling assistance.