1 Introduction

The Bohai Sea (BS) and Yellow Sea (YS) are typical semi-enclosed marginal shallow seas of China, with complex hydrodynamics and large inputs of land-based materials. The annual organic carbon flux from the Yellow River and other rivers to the YS and BS is (1210 ± 240) × 10⁴ ton (t) yr⁻¹, accounting for 79% of all organic carbon entering these seas (Liu et al., 2015a), while atmospheric deposition contributes < 2% (Qiao et al., 2017). In addition, the Yellow Sea Warm Current (YSWC) transports about 10⁶ t yr⁻¹ of Yangtze River sediments to the North Yellow Sea (NYS) (Gao et al., 1996). The tracing of sedimentary organic matter (SOM) sources improves understanding of the organic matter cycle in aquatic environments (Hedges et al., 1997). SOM is a complex mixture of marine and terrestrial organic compounds, and it is difficult to quantitatively distinguish its sources at the edge of the continental shelf. Normal (n-) alkanes are commonly applied as biomarkers given their widespread occurrence in marine and terrestrial environments, and can be preserved in marine sediments. Compositions and distributions of n-alkanes from different biological sources are generally different. In addition, compared with fatty acids and alkanols, structure of n-alkanes is relatively stable and has strong anti-degradation ability (Mead et al., 2005). Previous studies have shown that n-alkanes may be effective in characterizing sources of SOM in coastal marine systems (Xing et al., 2011; Wang et al., 2013). Long-chain n-alkanes derived from terrestrial higher plants are most abundant in C₂₇, C₂₉, and C₃₁ n-alkanes (Bray and Evans, 1961). The freshwater and marine non-emergent macrophytes and sphagnum mosses are enriched in mid-chain n-alkanes (Pancost et al., 2002; Mead et al., 2005; Mügler et al., 2008; Bush and McInerney, 2013). Short-chain n-alkanes with odd carbon predominance such as C₁₇ n-alkane are generally considered to be derived from aquatic algae and photosynthetic bacteria(Meyers and Ishiwatari, 1993; Silliman and Schelske, 2003; Liu et al., 2012). Additionally, petroleum-derived hydrocarbons also contribute short-chain n-alkanes, with no obvious odd/even predominance (Hostettler et al., 1999). The long-chain (≥C₂₄), n-alkanes have clear odd carbon-number predominance in the NYS, indicating predominant input of terrestrial higher plant material (Lu et al., 2011). Analyses of n-alkanes in sediments of the Yellow River Estuary indicate that SOM originates mainly from terrigenous inputs, while marine microorganisms contribute to short-chain (C₁₂–C₂₂) n-alkanes offshore (Wang et al., 2018). Compositional analysis of n-alkanes in surface sediments of the central South Yellow Sea (SYS) indicates that SOM is derived mainly from terrestrial higher plant input from the modern and old Yellow rivers, with the contribution of herbaceous to woody plants is comparable (Zhang et al., 2014).

Similar chemical n-alkane compositions have been found in different types of organisms, which may confuse their interpretation (Ficken et al., 2000; Mead et al., 2005; Sikes et al., 2009). However, stable carbon isotopic compositions of individual n-alkanes from different sources in marine sediments are generally distinctive and may therefore constrain their sources (Hayes et al., 1990; Mead et al., 2005; Ankit et al., 2017), with n-alkane compositions and stable carbon isotopic characteristics together having been used to identify SOM sources (Eglinton, 1969; Jeng and Huh, 2008; Hu et al., 2013). Previous studies have shown that stable carbon isotopic compositions of long-chain n-alkanes in surface soils of eastern China can be used as an indicator of C₃/C₄ plant proportions in overlying vegetation (Rao et al., 2008). The analysis of long-chain n-alkanes and their stable carbon isotopic compositions in sediments of Qinghai Lake indicates that δ¹³C values of C₃₁ n-alkanes are consistent with those of modern land plants around the lake, and can therefore be used as a reliable tracer of C₃/C₄ compositions of terrestrial vegetation (Liu et al., 2015b). A recent study found that δ¹³C values of organic matter indicate that terrestrial organic carbon from the Yellow River accumulates mainly at the river mouth and in two muddy areas around it (Sun et al., 2018a).

Until now, little has been known of the spatial distribution of n-alkane stable carbon isotopes in sediments of the BS and NYS. The aim of this study was to elucidate the stable carbon isotopic composition of SOM n-alkanes and the sources of n-alkanes in surface sediments of the BS and NYS.

2 Study Area and Methods 2.1 Study Area

The BS is a shallow, semi-enclosed epicontinental sea. About 90% of its sediment input is supplied by the surrounding rivers, especially the Yellow River. The YS is a shallow, semi-enclosed, continental margin West Pacific sea, with an area of 400, 000 km² and an average water depth of 44 m, joining the BS in the north and the East China Sea in the south. The YS is divided into southern and northern parts by the Shandong and Korean peninsulas. The overall topography of the NYS seabed is inclined southward. Water depths in the BS and NYS are generally < 60 m. In the study area, surface currents include coastal currents and the northwestward YSWC (Fig. 1). The YSWC is a branch of the Tsushima Current with warm and saline water. There is no major direct local riverine input to the YS although, over time, fine-grained riverine sediment can be resuspended and transported from the BS to the YS by coastal currents. Analyses of sediment sources indicate that fine-grained sediments within the NYS originate mainly in the modern and old Yellow Rivers (Alexander et al., 1991; Lim et al., 2007). From 1976 to 2005, runoff and sediments from the Yellow River averaged 140.36×10⁸ m³ yr⁻¹ and 3.31×10⁸ t yr⁻¹, respectively (Cui and Li, 2011), with most sedimentary materials being deposited in front edge of the delta and estuary, and finer grained sediment transported to the coastal and shelf areas outside the Yellow River estuary. Muddy sediments along the north coast of Shandong Peninsula are considered to be directly and indirectly from the Yellow River (Yang and Liu, 2007). In addition to riverine input, coastal erosion from the old Yellow River Delta also contributes to sediments (Hu et al., 1998).

2.2 Samples

Surface sediments (0–3 cm depth) were collected from 23 sites in the BS and NYS using a box corer deployed from R/V Dong Fang Hong 2 during a cruise sponsored by the National Natural Science Foundation of China (NSFC) in June 2011 (Fig. 1). Sediment samples were wrapped in aluminum foil and stored at −20℃ until analysis.

2.3 Analytical Methods 2.3.1 Lipid extraction and purification of n-alkanes

Freeze-dried, powdered, and homogenized sediment samples were extracted four times with dichloromethane/ methanol (DCM/MeOH; 3:1, v/v) with ultrasonication (15 min each time), after adding internal standards containing n-C₂₄D₅₀. Extracts of the samples were dried in a N₂ stream and hydrolyzed with 6% KOH in MeOH. Nonpolar fractions containing n-alkanes were separated using activated silica gel column chromatography with elution by n-hexane, and dried in a N₂ stream.

To accurately measure the δ¹³C values of individual n-alkanes, n-alkanes need to be further purified. Zeolite molecular sieve is a commonly used and high recovery method for n-alkanes purification. The extracted n-alkanes were transferred to a column with AgNO₃-Silica gel and molecular sieve, and then eluted with n-hexane to ensure the components of the inner wall of the AgNO₃-Silica gel glass column were completely transferred into the molecular sieve. After elution of about 1.5 mL, the upper AgNO₃-Silica gel column was removed and the molecular sieve column was baked in an oven at 40℃ for > 12 h. Subsequently, the zeolite power was transferred to a 10 mL Teflon bottle for digestion with HF to release n-alkanes. A preheating Pasteur column (6 mm, i.d. × 2 cm) filled with Na₂SO₄ was used to remove residual HF before the n-alkanes were extracted with n-hexane (four times) and then was dried in a gentle N₂ stream pending instrumental analyses. The average recovery rate of long-chain (≥ C₂₆), mid-chain (C₂₁–C₂₅), short-chain (C₁₅–C₂₀) n-alkanes in all samples was 63%, 53%, 57%, respectively.

2.3.2 n-alkanes analysis

The n-alkane compositions were determined by an Agilent 6890N gas chromatography, with chromatographic separation on an HP-1 capillary column (50 m × 0.32 mm i.d.× 0.17 μm film thickness, J & W Scientific) using H₂ as a carrier gas (1.2 mL min⁻¹). Samples were injected in splitless mode with an injector temperature of 300℃. Oven temperature was programmed from 80℃ to 200℃ at 25℃ min⁻¹, 200℃ to 250℃ at 3℃ min⁻¹, 250℃ to 300℃ at 1.8℃ min⁻¹, 300 to 310℃ at 5℃ min⁻¹, and holding at 310℃ for 5 min. Quantification of compounds was performed by peak area integration in FID GC (Agilent 6890N) relative to the internal standards. The average relative standard deviation in concentrations was < 10%.

The average chain length (ACL; Cranwell et al., 1987), the terrigenous/aquatic ratio (TAR; Bourbonniere and Meyers, 1996), the P_mar-aq (odd mid-chain alkanes/odd mid- and long-chain alkanes; Ficken et al., 2000; Mead et al., 2005) of n-alkanes were calculated as follows:

${\rm{ACL}} = \frac{{21 \times {{\rm{C}}_{21}} + 23 \times {{\rm{C}}_{23}} + 25 \times {{\rm{C}}_{25}} + 27 \times {{\rm{C}}_{27}} + 29 \times {{\rm{C}}_{29}} + 31 \times {{\rm{C}}_{31}} + 33 \times {{\rm{C}}_{33}}}}{{{{\rm{C}}_{21}} + {{\rm{C}}_{23}} + {{\rm{C}}_{25}} + {{\rm{C}}_{27}} + {{\rm{C}}_{29}} + {{\rm{C}}_{31}} + {{\rm{C}}_{33}}}}, $

(1)

${\rm{TAR}} = \frac{{{{\rm{C}}_{{\rm{27}}}} + {{\rm{C}}_{{\rm{29}}}}{\rm{ + }}{{\rm{C}}_{{\rm{31}}}}}}{{{{\rm{C}}_{{\rm{15}}}} + {{\rm{C}}_{{\rm{17}}}}{\rm{ + }}{{\rm{C}}_{{\rm{19}}}}}}, $

(2)

${{\rm{P}}_{{\rm{mar - aq}}}} = \frac{{{{\rm{C}}_{{\rm{23}}}} + {{\rm{C}}_{{\rm{25}}}}}}{{{{\rm{C}}_{{\rm{25}}}} + {{\rm{C}}_{{\rm{27}}}}{\rm{ + }}{{\rm{C}}_{{\rm{29}}}}{\rm{ + }}{{\rm{C}}_{{\rm{31}}}}}}.$

(3)

2.3.3 Stable carbon isotopic composition (δ¹³C) analysis

Gas chromatography isotope ratio mass spectrometry (GC-IRMS; on an HP 6890 GC coupled with a Thermo Delta-V system.) was used to measure stable carbon isotopic compositions of n-alkanes. Chromatographic separation was achieved using a DB-1MS capillary column (60 m× 0.32 mm i.d. × 0.25 μm film thickness, J & W Scientific). The GC oven temperature was programmed from 60℃ to 200℃ at 15℃ min^–1, 200℃ to 250℃ at 4℃ min^–1, 250℃ to 300℃ at 1.8℃ min^–1, 300 to 310℃ at 5℃ min^–1, and holding at 310℃ for 5 min. The authentic standard was analyzed under the same conditions after every seven samples. The standard deviation for duplicate analysis of the standard was 0.3‰. Isotopic ratios were expressed as δ¹³C values (per mil) relative to the Vienna Pee Dee Belemnite (VPDB).

3 Results 3.1 Composition of n-Alkanes and the Hydrocarbon Indices

The GC-FID chromatograms of n-alkanes showed that n-alkanes were effectively purified after using the molecular sieve (Fig. 2). Total n-alkane contents (∑C_15–35) ranged from 456 to 3837 ng g⁻¹ (average = 1897 ng g⁻¹). The contents of long-chain n-alkanes (∑C_25–35) ranged from 267 to 2826 ng g⁻¹ (average = 1300 ng g⁻¹). In addition, the average percentage of long-chain, mid-chain, short-chain n-alkanes in samples was 58%, 27%, 14%, respectively. Furthermore, the total n-alkane contents of samples from muddy areas (average = 2666 ng g⁻¹) were significantly higher than those from non-muddy areas (average = 1683 ng g^–1). The ACL values varied between 26.1 and 28.9 (Fig. 4a). The values of TAR and P_mar-aq ranged from 3.4 to 25.7, from 0.2 to 0.7 (Figs. 4b, 4c), respectively.

3.2 δ¹³C Values of n-Alkanes in Surface Sediments

Compound-specific average δ¹³C values of n-alkanes in surface sediments were shown in Fig. 3, with average individual values for C₁₇–C₃₁ of −30.1‰ ± 0.5‰, −28.7‰ ± 0.4‰, −29.8‰ ± 0.6‰, −28.4‰ ± 0.3‰, −29.9‰ ± 0.5‰, −29.4‰ ± 0.6‰, −30.4‰ ± 0.3‰, −29.9‰ ± 0.5‰, −30.2‰ ± 0.5‰, −30.1‰ ± 0.5‰, −30.8‰ ± 0.5‰, −30.8‰ ± 0.8‰, −31.9‰ ± 0.6‰, −31.7‰ ± 1.1‰, and −32.1‰ ± 1.0‰, respectively. In both the BS and NYS, δ¹³C values of mid-chain n-alkanes (C₂₁–C₂₃) varied within a narrow range, while those of short- and long-chain n-alkanes were more variable. Furthermore, for short- and mid-chain n-alkanes, δ¹³C values of even-carbon-numbered cases were more positive than those of odd-carbon-numbered cases.

4 Discussion 4.1 Long-Chain n-Alkanes

The contents of long-chain n-alkanes were relatively high and exhibited a strong odd carbon predominance in C₂₇, C₂₉, and C₃₁ homologues (Fig. 3), consistent with terrestrial higher plant sources. The ACL describes the average number of carbon atoms in odd carbon n-alkanes in higher plants (Cranwell et al., 1987). The ACL values of BS and NYS surface sediments ranged from 26.1 to 28.9 (average = 27.5). ACL value of about 29 in sediments near the Yellow River estuary suggests an origin of terrestrial higher plants (Fig. 4a). The relative contribution of terrestrial n-alkanes to marine sediments can be assessed using the TAR index. TAR values of BS and NYS surface sediments ranged from 3.4 to 25.7, with an average value of 14.1 (Fig. 4b). This indicates a predominance of terrigenous n-alkanes input (Ankit et al., 2017). Furthermore, compositional analysis of n-alkanes in surface sediments of the BS and NYS also indicates that long-chain n-alkanes are derived mainly from terrestrial higher plant input (Cao, et al., 2018). Hence, Long-chain n-alkanes in the study areas were thus mainly derived from such plants.

The δ¹³C values of long-chain n-alkanes produced by C₃ and C₄ plants typically range from −31.0‰ to −39.0‰ and −18.0‰ to −25.0‰, respectively (Collister et al., 1994; Schefuẞ et al., 2003). Modern terrestrial higher plants from eastern China are characterized by n-alkane δ¹³C values of −21.9‰ to −34.8‰, −25.3‰ to −36.1‰, and −22.9‰ to −36.7‰ for C₂₇, C₂₉, and C₃₁ components (Rao et al., 2008), consistent with our corresponding average δ¹³C values of −30.8% ± 0.5‰, −31.9% ± 0.6‰, and −32.1% ± 1.0‰ (Fig. 3), respectively, and indicating that long-chain n-alkanes are mainly derived from terrigenous sources. Generally, odd-carbon-numbered long-chain n-alkanes are somewhat ¹³C-enriched than those of even-carbon-numbered long-chain n-alkanes in terrestrial higher plants (Chikaraishi and Naraoka, 2003). However, our results showed δ¹³C values of even-carbon-numbered long-chain n-alkanes (C_26–30) were more positive than those of odd-carbon-numbered long-chain n-alkanes (C_27–31) in the study area (Fig. 3). This implies there may be different sources of even-carbon-numbered long-chain n-alkanes. A previous study reported ¹⁴C ages for n-C₂₉₊₃₁ alkanes (Δ¹⁴C = −288‰ to −612‰) of 2670 to 7550 ¹⁴C yr, which differ markedly from those of strongly ¹⁴C-depleted n-C_26+28+30+32 alkanes (Δ¹⁴C = −700‰ to −961‰) ages of 9600 to 26050 ¹⁴C yr for Yellow River suspended particulate matter, implying ancient organic carbon inputs (Tao et al., 2015). This may indicate that even-carbon-numbered long-chain n-alkanes in the BS and NYS are derived from ancient organic carbon.

Weighted mean average δ¹³C of long-chain n-alkanes from sediment samples were determined to calculate the changes in biomass of C₃ and C₄ plants in historical periods (Kuang et al., 2013). A binary end-member mixing model was used to estimate the relative contributions of long-chain n-alkanes from C₃ and C₄ plants (Garcin et al., 2014), with δ¹³C values of −36.0‰ and −21.0‰ being used as end-members for these plants, respectively (Collister et al., 1994; Zhang et al., 2003). Calculations were performed as follows:

$ M = ({{\rm{ \mathit{ δ} }}^{13}}{{\rm{C}}_{27}} \times {{\rm{C}}_{27}} + {{\rm{ \mathit{ δ} }}^{13}}{{\rm{C}}_{29}} \times {{\rm{C}}_{29}} + {{\rm{ \mathit{ δ} }}^{13}}{{\rm{C}}_{31}} \times {{\rm{C}}_{31}})/({{\rm{C}}_{27}} + {{\rm{C}}_{29}} + {{\rm{C}}_{31}}) = \left({ - 36.0‰} \right) \times X + \left({ - 21.0‰} \right) \times (100\% - X), $

(4)

where M is the weighted mean average δ¹³C value of long-chain n-alkanes, and X is the C₃ contribution (%).

End-member estimations for the BS and NYS indicated that terrestrial C₃ plants were dominant n-alkane sources, with relative contributions of 64%–79% (Fig. 4d). This is consistent with the predominance of C₃ plants in north China, with a previous study having shown that δ¹³C values of n-alkanes in aerosols near the north China coast have terrestrial C₃ plant origins with the C₄ contribution being negligible (Guo et al., 2006). Moreover, soil organic matter δ¹³C values in a N–S section (34–52˚N) through central and eastern Asia indicate that vegetation in the area comprises mainly C₃ plants (Feng et al., 2008). Records of δ¹³C values in surface soils of northeast China indicate that the abundance of C₄ plants is relatively high in warm periods and almost exclusively C₃ plants exist in cold periods (Sun et al., 2018b). Previous studies have shown that δ¹³C values of dominant C₃ plants in the Chinese Loess Plateau range from −30.7‰ to −22.6‰, with average value of 27.2‰ (Zheng and Shangguan, 2007) and −27.1% ± 2.4‰ (n = 39; Liu et al., 2005). Both δ¹³C values of total organic carbon and long-chain n-alkanes derived from terrestrial higher plants show minor variations among surface soil samples from northern China, indicating the major contributor is from local grasses with a uniform C₃ photosynthetic pathway (Rao et al., 2011). It is likely, therefore, that long-chain n-alkanes in BS and NYS surface sediments are mainly derived from terrestrial higher plants, particularly C₃ plants.

Furthermore, recent studies have also shown that δ¹³C values of n-alkanes in gymnosperms are heavier than those in angiosperms (Diefendorf et al., 2011; Lane, 2017; Zhao et al., 2018). And angiosperm δ¹³C values generally decrease with increasing chain length of n-alkanes, while gymnosperm values increase (Bush and McInerney, 2009). It is clear here that δ¹³C values of long-chain n-alkanes decrease with increasing chain length (Fig. 3). Average δ¹³C values of C₂₉ and C₃₁ n-alkanes are −31.9% ± 0.6‰ and −32.1% ± 1.0‰, respectively, similar to values for herbaceous plants in the modern Yellow River drainage basin (−31.1‰ to −31.5‰ for C₂₉ n-alkanes, and −31.3‰ to −32.6‰ for C₃₁ n-alkanes in dust episode periods, Guo et al., 2006). This suggests that the contribution of C₃ angiosperms to the sedimentary long-chain n-alkanes is greater. This is consistent with the predominance of angiosperms in the last glacial period and Holocene on the Chinese Loess Plateau (Li et al., 2016).

4.2 Mid-Chain n-Alkanes

C₂₁, C₂₃, and C₂₅ n-alkanes are mainly contributed by aquatic plants. Previous studies have shown that the δ¹³C values of mid-chain n-alkanes in aquatic emergent macrophytes range from −28.6‰ to −31.2‰ (Chikaraishi and Naraoka, 2003; Mead et al., 2005). Although non-emergent marine macrophytes can also produce mid-chain n-alkanes, their δ¹³C values are relatively heavy, ranging from −13.0‰ to −22.0‰ (Ficken et al., 2000; Jaffé et al., 2001). In the NYS, there was little difference between stable carbon isotopic compositions of samples from muddy and non-muddy areas: average δ¹³C values of mid-chain n-alkanes (C₂₁, C₂₃, and C₂₅) in non-muddy areas were −29.8‰, −30.3‰, and −30.2‰, respectively, and those in muddy areas were −29.7‰, −30.3‰, and −30.2‰, respectively. This also applied to the BS, where average δ¹³C values were −29.7‰, −30.4‰, and −30.2‰, respectively, indicating that stable carbon isotope compositions of mid-chain n-alkanes in the BS and NYS were similar. The narrow range of these values may be due to there being a common source for BS and NYS sediments, namely the Yellow River (Bi et al., 2010). The δ¹³C values of C₂₁, C₂₃, and C₂₅ n-alkanes fell within the range of values for the corresponding n-alkanes in aquatic emergent macrophytes, with sediment mid-chain n-alkanes in the study area thus being mainly derived from such plants. Furthermore, the Pmar-aq index provides a measure of the relative contributions of aquatic non-emergent/emergent plants and terrestrial vegetation, with values of < 0.25 corresponding to terrigenous plants, 0.3–0.6 to aquatic emergent plants, and > 0.6 to aquatic non-emergent macrophytes in coastal marine environments (Ficken et al., 2000; Mead et al., 2005). The Pmar-aq values ranged from 0.2 to 0.7 (average = 0.4) in the study area (Fig. 4c). We concluded, therefore, that mid-chain n-alkanes were mainly derived from aquatic emergent macrophytes in the BS and NYS.

4.3 Short-Chain n-Alkanes

Short-chain n-alkanes are generally considered as being derived from microorganisms and marine algae. Those produced by marine planktonic algae are mainly C₁₅, C₁₇, and C₁₉ n-alkanes with odd carbon predominance, while even-carbon-numbered short-chain n-alkanes (C₁₆, C₁₈, and C₂₀) are derived from marine bacteria or petroleum hydrocarbons (Gogou et al., 2000; Wang and Fingas, 2006). Short-chain n-alkanes in marine sediments are predominantly C₁₇, indicating the major contribution of algae and photosynthetic bacteria (Han and Calvin, 1969), while even-carbon-numbered (C_16–22) n-alkanes in marine sediments are mainly attributable to non-photosynthetic bacteria (Jeng and Huh, 2008). Most of sediments in the BS and NYS exhibited an even-carbon-number preference in the range of n-C₁₆ to n-C₂₂ (Fig. 3), indicating that these short-chain n-alkanes could be from non-photosynthetic bacterial sources. The values of n-C₁₈/n-C₁₇ can be used to compare the relative contributions of n-alkanes from petroleum-derived n-alkanes and natural n-alkanes from algae and photosynthetic bacteria. Here, the calculated n-C₁₈/n-C₁₇ values of surface sediments are higher than 1 at all stations, indicating that short-chain n-alkanes are affected by petroleum pollution to some degree. Extremely depleted △¹⁴C values (−932‰ to −979‰) for short-chain n-alkanes (C₁₆ and C₁₈) were found in BS and YS sediments, suggesting a predominant input from sedimentary rocks (organic carbon) or petroleum products (Tao et al., 2016). The average δ¹³C value of short-chain n-alkanes, δ¹³C₁₇, δ¹³C₁₈, and δ¹³C₁₉, is −30.1% ± 0.5‰, −28.7% ± 0.4‰, and −29.8% ± 0.6‰, respectively (Fig. 3). Previous studies show that the δ¹³C_C17 values of cyanobacteria vary from −34.0‰ to −36.0‰ (Kristen et al., 2010), while δ¹³C₁₇ and δ¹³C₁₉ values of petroleum hydrocarbons are about −30.6‰ and −31.0‰, respectively (Li et al., 2009). The average δ¹³C value of algae in Laizhou Bay is −20.5‰ (Cai and Cai, 1993). Our results showed δ¹³C values of short-chain n-alkanes were relatively lighter than those of algae, possibly due to biodegradation of bacteria and input of petroleum hydrocarbons or other sources.

5 Conclusions

The relative inputs of terrestrial and marine organic matter were assessed using n-alkane. Terrigenous plants are the main source of n-alkanes in BS and NYS sediments. Long-chain n-alkanes in sediments were mostly derived from terrestrial sources with some contribution from biogenic and/or petroleum sources. The average δ¹³C values of long-chain n-C₂₇, n-C₂₉, and n-C₃₁ alkanes are −30.8% ± 0.5‰, −31.9% ± 0.6‰, and −32.1% ± 1.0‰, respectively, within the range of n-alkanes δ¹³C values of terrestrial C₃ plants. A hydrocarbon source distribution derived using a binary end-number mixing model based on δ¹³C values of long-chain n-alkanes indicates that organic matter in BS and NYS sediments is mainly sourced from C₃ plants, particularly angiosperms. The relative contribution of C₃ plants decreases from estuary to ocean. δ¹³C values of mid-chain n-alkanes in surface sediments indicate that mid-chain n-alkanes are mainly of aquatic emergent macrophyte origin. δ¹³C₁₇, δ¹³C₁₈ and δ¹³C₁₉ values, n-C₁₈/n-C₁₇ ratios indicate that short-chain n-alkanes in BS and NYS sediments have complex sources including petroleum pollution and bacterial action.

Acknowledgements

This work was financially supported by the Ministry of Science and Technology of People's Republic of China (No. 2016YFA0600904), and the National Natural Science Foundation of China (No. 41476058).