Supply of terrigenous organic matter from tidal flat to the marine environment: An example of neritic source rocks in the Eocene Pinghu Formation, Xihu Depression, East China Sea Shelf Basin

1. Introduction

The Meso-Cenozoic marine source rocks on the west coast of the Pacific Ocean are characterized by significant inputs of terrestrial organic matter (Doust and Noble, 2008; Zhang et al., 2013, 2021; Zhu et al., 2020; Ding et al., 2021). The lower Eocene Pinghu Formation contains the dominance of the terrigenously-dominated marine shales in the Xihu Depression of the East China Sea Shelf Basin (Jiang et al.,2011). The semi-closed bay environment, where the source rocks with a dominant Type III-II₂ organic matter were primarily deposited, received a significant input of higher plant organic matter (Su et al., 2015; Cai et al., 2019; Diao et al., 2019).

Terrigenous organic matter was assumed to be primarily transported by rivers into the lacustrine/marine environment (Saller et al., 2006; Deniau et al., 2010; Abbassi et al., 2014; Dai et al., 2015; Deng et al., 2019; Zhang et al., 2020). However, when the Eocene Pinghu Formation was deposited, and coals tended to deposit in tidal flats along the coast rather than in the delta region (Yu, 2020) . The anomalously high total organic carbon (TOC) levels (>6% on average) in the samples of tidal sand ridge suggested that coal-bearing sedimentary debris was likely transferred from the shore to the ocean by tidal currents. Although ex-situ coals transported in from tidal flats were found in several marine strata (Li et al., 2015), there is not any compelling evidence supporting this envisage in the Xihu Depression.

Plynofacies analysis was defined by Tyson (1995) as the palynological study of depositional environments and hydrocarbon source rock potential based on the total assemblage of particulate organic matter. It is an interdisciplinary approach because it considers both the entire organic content and the palynomorphs in palynological slides (Carvalho et al., 2006). The particles are considered to be sedimentary fragments from higher plants, algae, and some bacteria that indicate the original depositional settings and their biological sources. Thus, plynofacies analysis can shed light on the origins of organic materials in neritic source rocks.

The lithology of the Eocene Pinghu Formation changed from coal-bearing strata to neritic shales when the depositional environment changes from land to sea. Coal-bearing source rocks, including mudstones, carbonaceous mudstones, and coals, were primarily deposited in tidal flat and delta environments. Due to the fast alterations of the transitional facies, coal seams exhibit narrow single-layer thickness, multiple layers, and unstable lateral migration (Wang et al., 2020). In general, coal-bearing source rocks in tidal flats are thicker and have a greater potential for hydrocarbon generation than those in deltas (Yu, 2020). To study the origins of terrigenous organic matter in neritic layers, mudstone samples from the delta, the tidal flat, and the neritic environment from the Pinghu Formation in the Xihu Depression were comparatively analyzed. This work proposes an effective classification model based on discriminant analysis and plynofacies analysis data. To support the discriminant results, source-related biomarker ratios were used. This study provides a credible understanding of the origins of terrigenous organic matter in neritic source rocks in Meso-Cenozoic shelf basins along the west coast of the Pacific Ocean.

2. Geological settings

The East China Sea Shelf Basin is located offshore from eastern mainland China (Fig. 1a). The Xihu Depression lies between the Yushan-Haijiao Uplift to the west and the Diaoyudao Uplift to the east (Fig. 1b), and it has a surface area of approximately 5.0×10⁴ km² (Chen, 1998; Cai et al., 2019; He, 2020). Until 2020, about 8 000×10⁸ m³ of natural gases and about 3 000×10⁴ t of crude oils have been discovered in this depression. The coal-bearing Eocene Pinghu Formation has been assumed to be the main source rocks (Zhu et al., 2012; Yu, 2020).

Figure  1.  Present map of the regional geographical features and location of the East China Sea Shelf Basin (a), structural divisions of the Xihu Depression (modified after Diao et al., 2019) showing the well locations from which samples were obtained (b), NW−SE structural section across the Xihu Depression (A−B, profile line marked in Fig. 1a, modified from Cheng et al., 2020) (c), and generalized stratigraphic column of the Xihu Depression (modified after Cai et al., 2019) (d). Sy.: system; Sr.: series.

DownLoad: Full-Size Img PowerPoint

During the Paleocene−Eocene, the East China Sea Shelf Basin rifted, then was compressed during the Oligocene−Miocene, and eventually subsided from the Pliocene to the present (Zhang et al., 2015; Wang et al., 2017; Zhu et al., 2020). The depression develops a series of tectonic units (Fig. 1b), including the slope belts on the east and west, two sub-depression, and a central inverted structural belt in the middle. The Hangzhou Slope, the Pinghu Slope, and the Tiantai Slope are located on the western slope from north to south. Strata in the Xihu Depression consists of the sediments in the Eocene Baoshi Formation to the Quaternary Donghai Formation, with a maximum thickness of 10 000 m (Fig. 1c).

During the rifting period, this basin was filled with continental, transitional, and marine deposits (Fig. 1d; Fu et al., 2003). There was an entire sea level eustacy during the deposition of the Pinghu Formation. From bottom to top, the Pinghu Formation is subdivided into the Ping-5 (E₂p₅), Ping-4 (E₂p₄), Ping-3 (E₂p₃), Ping-2 (E₂p₂), and Ping-1 (E₂p₁) members. Due to erosion and loss of the E₂p₁ Member in the top section of the slope zone, the E₂p₁ Member was united with the E₂p₂ Member and they were renamed as the E₂p_1-2 member (Yu, 2020).

During the depositional period between E₂p₅ and E₂p₄, the sea level rose and the delta receded (Figs 2a and b; Chen, 1998). The water in the depression center (where the Well T-1 is located) gradually deepened to form a bay. The tidal flat environment formed mainly in the southern part of the Pinghu Slope (where the Wells P-2 and P-3 are located) and the Tiantai Slope, while the delta dominated the northern part of the Pinghu Slope (where the Well P-1 is located) and the Hangzhou Slope. As the sea receded between the E₂p₃ to E₂p_1-2 depositional periods (Figs 2c and d), the central inversion tectonic belt area turned into tidal flats (He, 2020). The delta progressively expanded in size, and the Tiantai area transitioned from a neritic to a tidal flat environment, with the southern part of the Pinghu Slope remaining dominated by tidal flat and the delta area gradually expanding in the northern half of the Pinghu Slope and the Hangzhou Slope.

Figure  2.  Sedimentary facies of E₂p₅ (a), E₂p₄ (b), E₂P₃ (c), and E₂p_1-2 (d) of the Pinghu Formation in the western Xihu Depression (modified after Yu, 2020) and the locations of the sampling wells.

DownLoad: Full-Size Img PowerPoint

3. Samples and methods

3.1 Samples

Cutting samples were collected from four wells (Fig. 1b). Well P-1 is located in the north part of the Pinghu Slope. The Pinghu Formation in this area is comprised of sandstones, mudstones, and interbedded coals in a delta sedimentary environment (Xie et al., 2013). Delta coal-bearing mudstones were primarily found in the delta’s interdistributary bay and underwater interdistributary bay. Deltas were not as well-developed in the south part of the Pinghu Slope as in the north part. The south part of the Pinghu Slope was primarily tidal flats during the deposition of the Pinghu Formation. Wells including P-2 and P-3 uncovered rich coals and carbonaceous shales that were frequently interbedded with mudstones, siltstones, and sandstones. Well T-1 was drilled into the marine stratigraphy of the E₂p₄ and E₂p₃ Formations. The lithology of the marine strata includes mudstones and interbedded siltstones, in which there is no coal or carbonaceous shale. From the later depositional period of the E₂p₃ to the E₂p_1-2, seawater receded and the environment of the depression center changed to tidal flat.

3.2 TOC measurements and Rock-Eval pyrolysis

Total organic matter measurements and Rock-Eval pyrolysis were performed on 48 cutting samples that were obtained from the Well T-1 at depths of 3 575 m to 4 828 m (Supplementary Table S1). Twenty-eight of the 48 samples are tidal flat mudstones, and the rest 20 are neritic mudstones. For the determination of TOC and Rock-Eval, cutting samples were powdered to 200 mesh. In the pretreatment procedure, TOC samples were treated with dilute hydrochloric acid to remove inorganic carbon and then rinsed with deionized water more than 50 times to completely remove residual hydrochloric acid. The pretreated samples were combusted in a high-temperature oxygen stream, with the carbon dioxide measured using an infrared detector. The Rock-Eval analysis was conducted by heating the samples in a nitrogen environment from 300℃ to 600℃ in a Rock-Eval VI instrument. The amounts of free hydrocarbons (S₁), potential hydrocarbons (S₂), and the temperature of maximum generation (T_max) were measured.

3.3 Plynofacies analysis

Sedimentary organic matters (SOM) were classified using plynofacies analysis, based on their morphology, structures, origins, and other characteristics of the plynofacies. It is a holistic approach that deals with all aspects of SOM and is an effective method for investigating organic matter sources and the sedimentary environment (Tyson, 1995; Carvalho et al., 2006; Garcia et al., 2011; Cai et al., 2020).

Plynofacies analysis was conducted at the State Key Laboratory of Marine Geology of Tongji University in Shanghai, China. A total of 89 cutting samples (Supplementary Table S2) were obtained from Wells P-1 (19 samples), P-2 (45 samples), and T-1 (25 samples). The samples were weighed and then crushed to 80 mesh. To determine the absolute amounts, lycopodium pollen tablets (10 069 lycopodium pollen per tablet) made at Lund University, Sweden, were added to the samples as an internal standard. Calcium carbonates are removed using dilute hydrochloric acid, and silica is removed using hydrofluoric acid. Then the suspended debris is floated using heavy liquor (specific gravity of 2.2). All suspended materials were collected to produce thin sections for microscopic observation. The modified classification nomenclature scheme for sedimentary organic debris in the study of Tyson (1995) was used. Under a transmission microscope, the contents of charcoals, woods, degraded woods, cuticles, exinites, spores, fungal issues, and amorphous organic materials (AOM) were counted.

3.4 Discriminant analysis (DA)

Stepwise DA was used to identify the resemblance of neritic source rock samples to deltaic and tidal flat coal-bearing source rock samples. DA is a subset of supervised machine learning methods that have found widespread application in a variety of fields (Zhang, 1991; Kotsiantis, 2007; Zhang et al., 2019). DA uses pre-classified known samples (i.e., deltaic and tidal flat source rock samples) as learning samples to construct functions that best discriminate across classes and then applies the resulting functions to categorize unknown samples (i.e., neritic source rock samples).

There are generally two kinds of theoretical basis, namely Bayesian and Fischer’s discriminant rules. In Bayesian statistics, a posterior probability is the revised or updated likelihood of an event occurring after additional information is considered. The posterior probability calculated from DA using the Bayesian algorithm is used to classify samples quantitatively (Zhang and Liao, 1998; Zhang et al., 2014; Zhou et al., 2021). A high posterior probability suggests that the sample of the neritic source rock is highly similar to the source. The DA, which is based on Fischer’s rule, is a linear projection of the point groups in multivariate space that classifies samples based on their distance from known class centers (Wang, 2017). For a detailed description of the DA approach, refer to Zhang et al. (2019).

3.5 Gas chromatography-mass spectrometry analysis

The gas chromatography-mass spectrometry (GC-MS) analyses of aliphatic fractions of 17 cutting samples from Wells P-1 (6 samples), P-2 (2 samples), P-3 (3 samples), and T-1 (6 samples) were performed at the State Key Laboratory of Heavy Oil Processing, China University of Petroleum (Beijing) on a gas chromatograph-mass spectrometer of Agilent 7890-5975c. GC was performed using an HP-5MS capillary column (60 m×0.25 mm×0.25 m) with helium as the carrier gas (1.0 mL/min). The temperature was set at 50℃ for 1 min, then programmed at 20℃/min to 120℃, at 4℃/min to 250℃, then at 3℃/min to 310℃, and finally held for 30 min. The mass spectrometer was operated in electron impact (EI) ionization mode and selected ion mode at an electron energy of 70 eV, detecting ion mass-to-charge ratios (m/z) in the saturation fraction of m/z 123, 191, and 217.

4. Results

4.1 Total organic carbon, Rock-Eval pyrolysis, and source rock quality evaluation

The whole section of the Pinghu Formation in Well T-1 was systematically sampled, and 48 cutting samples were collected for analyses of TOC and Rock-Eval pyrolysis. The results of this analysis are listed in Supplementary Table S1. Twenty-eight of the total 48 samples were collected from the upper Pinghu Formation. These mudstone samples reached a maximum depth of 4 212 m and were deposited in tidal flat environments. TOC values range from 0.63% to 67.80% in tidal flat coal-bearing mudstones, while S₁+S₂ values ranged from 0.74 mg/g to 164.56 mg/g. There is a positive correlation between TOC and S₁+S₂ (Fig. 3). Source rock quality of twenty-eight tidal flat mudstone samples was evaluated according to the evaluation criteria of coal-bearing mudstones specified by the Chinese industry-standard SY/T 5735−2019 Geochemical Method for Source Rock Evaluation. Six of the samples are coals (TOC>40%), nine are carbonaceous shales (TOC=6%−40%), and five are graded as excellent (TOC=3%−6%). Seventy-one percent (20/28) of all samples were graded as good or better quality, indicating that the samples from the tidal flat are rich in organic components. The TOC and S₁+S₂ values of the twenty marine mudstone samples were lower than those of the coal-bearing mudstone samples. TOC values range from 0.43% to 7.16% in marine mudstone samples, with S₁+S₂ values between 0.16 mg/g and 12.66 mg/g. The evaluation criteria for marine mudstones and coal-derived mudstones were different. According to the evaluation criteria for marine mudstones, the quality of seven samples from Well T-1 is excellent (TOC>2%), four are of good quality (TOC=1%−2%), and the proportion of samples in good-excellent quality is 55%. Four tidal sand ridge samples have exceptionally high TOC values (>6%), especially the coal-bearing mudstones (Fig. 3).

Figure  3.  The plots of S₁+S₂ versus TOC showing the richness of organic matter for tidal flat and neritic source rock samples of the Pinghu Formation in Well T-1 (a) and neritic source rock samples subdivided into the tidal sand ridge and neritic shelf mud micro-facies (b). Quality evaluation of coal-bearing mudstones (a) and marine mudstones (b) according to Chinese industry-standard SY/T 5735-2019 Geochemical Method for Source Rock Evaluation.

DownLoad: Full-Size Img PowerPoint

The results of Rock-Eval pyrolysis data demonstrate that kerogen in the coal-bearing mudstones and neritic mudstones are Types Ⅱ₂ (Fig. 4). It indicates that the majority of the organic material in neritic mudstone is derived from terrestrial higher plants, with a minor input of aquatic organisms including algae. All coal-bearing source rocks and neritic source rocks are above the high thermal maturity stage (vitrinite reflectance, R_o>0.5%), in which some neritic source rocks have entered the gas-forming stage (R_o>1.3%) due to their deeper burial depth.

Figure  4.  The plots of hydrocarbon index (HI) versus T_max of source rock samples of the Pinghu Formation in Well T-1.

DownLoad: Full-Size Img PowerPoint

4.2 Plynofacies analysis

4.2.1   Total abundance of SOM

The total abundance of SOM was defined as the number of SOM debris particles per gram of each source rock sample. The abundance of SOM in 19 delta source rock samples ranged from 30×10⁴ debris/g to 318×10⁴ debris/g, with an average of 136×10⁴ debris/g. The average SOM abundance of the 45 tidal flat samples was similar to that of the delta samples at 133×10⁴ debris/g. However, the abundance of SOM in the neritic samples was significantly lower, with an average value of 49×10⁴ debris/g for 25 samples.

4.2.2   Compositions of SOM

SOM was classified by Tyson (1995) into charcoal, woods, undegraded woods, cuticles, exinites, spores, fungal issues, and amorphous organic matter (AOM). Charcoal, wood, and undegraded wood are all formed from land plant woody tissues. The high abundance of coals and woody matter represents a nearby, abundant input of plant debris (Deniau et al., 2010). Cuticles, exinites, and spores are non-woody tissues of land plants and have significant potential as a source of liquid hydrocarbons (Tyson, 1995). Abundant cuticles and exinites indicate sediment transport. It is believed that fungal tissues play a critical role in the breakdown of mangrove detritus and marine biodegradation (Jones, 1988). AOM is formed predominantly from aquatic organisms such as algae (Cai et al., 2020).

Percentages of charcoal, woods, undegraded woods, cuticles, exinites, spores, fungal tissues, and AOM in SOM were displayed (Supplementary Table S2). The average proportions of each component in the SOM for samples from the delta, tidal flat, and neritic facies are in Fig. 5. The delta samples have the highest abundances of charcoal and woods (including the undegraded woods), with an average value of 80%. The least amount of charcoal and woods with an average percentage of about 65%, is found in the neritic samples. The proportion of cuticles, exinites, and spores varies in the opposite direction from those of charcoal and wood, with deltaic samples in the lowest abundance (about 19%) and shallow marine samples having the highest proportion (around 29%). The abundance of fungal tissues varied among different samples, with the delta samples in the highest percentage, the tidal flat samples in the lowest abundance, and the neritic samples having a moderate amount. AOM contents for the tidal flat and neritic were more variable, they were generally relatively low in the delta samples. The AOM content ranges from 0% to 45.4% in the tidal flat samples and from 0% to 19.2% in the neritic samples.

Figure  5.  The pie charts of the average values of the percentage content of each component in the SOM for samples of the delta (a), tidal flat (b), and neritic (c). SOM, sedimentary organic matter.

DownLoad: Full-Size Img PowerPoint

4.3 DA model and results

The Percentages of SOM compositions in the studied samples (Supplementary Table S2) were used as candidate parameters for the DA model. The discriminant analysis takes into account both inter-group differences and intra-group differences in the sample point groups when they are classified. Three parameters, i.e., cuticles (%), exinites (%), and fungal issues (%) (Table 1) were automatically filtered on the principle of maximizing the inter-group variability and minimizing the intra-group variability of the sample point groups. The significance thresholds of the F value were set to 0.05 for entering and 0.1 for removing parameters respectively. Initially, learning samples were used to calibrate a discriminant model, and then all learning samples were designated as “unclassified samples” and were classified experimentally using the developed model. The correct assignment of the test sample served as evidence of the discriminant model’s predictability and dependability. According to the original validation procedure, the proper discrimination percentage is 92.2% (59/64, see Supplementary Table S3). Leave-one-out cross-validation provides an unbiased criterion for the DA model (Zhou et al., 2021), and the leave-one-out cross-validation correct rate is 89.1% (57/64). The high correct discriminant rate implies that these three parameters may successfully differentiate between delta and tidal flat samples.

Table  1.  Classification function and canonical discriminant function coefficients

Plynofacies parameter	Classification function coefficient		Fc coefficient	Fs coefficient
	Class 1	Class 2
Cuticles/%	0.861	1.644	0.363	0.230
Exinites/%	0.686	1.041	0.165	0.321
Fungal issues/%	0.812	−2.228	−1.408	−0.586
Constants	−8.312	−12.090	−2.589	−
Note: The posterior probabilities are computed from the DA based on the Bayesian rule; and the discriminant function scores, Fischer’s rule. Class 1 represents the class of delta; Class 2, of tidal flat. Fc, canonical discriminant function; Fs, standardized canonical discriminant function.

 | Show Table

DownLoad: CSV

Using the discriminant model for the three parameters, the scores of the canonical discriminant functions (Fc) based on Fischer’s rule and the posterior probability (P₁ for Class 1, i.e., delta; P₂ for Class 2, i.e., tidal flat) based on the Bayesian rule were calculated for each sample (Table 2 and Supplementary Table S2). A high posterior probability P_i (i=1, 2) of a neritic sample determined from the DA using the Bayesian rule suggests that this sample is very similar to delta (high P₁) or tidal flat (high P₂). According to posterior probability, 24 neritic samples were classified into Class 2 and only one was classified into Class 1 (Table 2). The score boundary of the canonical discriminant function of Fischer DA’s must be rectified or calibrated with the posterior or prior probability of the Bayesian DA. The revised principle is specified by Zhou et al., 2021. The modified border yields the same outcomes as the Bayesian DA (Supplementary Table S2).

Table  2.  Results of discriminant analysis of the neritic shale samples of Well T-1

No.	Well depth/m	Member	Fc score	Regrouped class (Fischer’s)	Posterior probability (Pi)		Regrouped class (Bayesian)
					Class 1 (P1)	Class 2 (P2)
1	4 230	E2p3	−0.43	2	0.29	0.71	2
2	4 250	E2p3	−0.64	2	0.40	0.60	2
3	4 270	E2p3	1.75	2	0.00	1.00	2
4	4 290	E2p3	−0.47	2	0.31	0.69	2
5	4 310	E2p3	1.82	2	0.00	1.00	2
6	4 330	E2p3	−1.42	1	0.78	0.22	1
7	4 350	E2p3	0.58	2	0.05	0.95	2
8	4 370	E2p3	−0.42	2	0.29	0.71	2
9	4 390	E2p4	1.82	2	0.00	1.00	2
10	4 410	E2p4	0.85	2	0.03	0.97	2
11	4 440	E2p4	1.17	2	0.01	0.99	2
12	4 460	E2p4	0.32	2	0.08	0.92	2
13	4 480	E2p4	1.38	2	0.01	0.99	2
14	4 500	E2p4	−0.33	2	0.25	0.75	2
15	4 518	E2p4	0.52	2	0.05	0.95	2
16	4 538	E2p4	2.91	2	0.00	1.00	2
17	4 557	E2p4	−0.12	2	0.17	0.83	2
18	4 576	E2p4	0.94	2	0.02	0.98	2
19	4 593	E2p4	0.90	2	0.02	0.98	2
20	4 614	E2p4	0.62	2	0.04	0.96	2
21	4 634	E2p4	2.21	2	0.00	1.00	2
22	4 652	E2p4	1.56	2	0.01	0.99	2
23	4 667	E2p4	0.32	2	0.08	0.92	2
24	4 687	E2p4	1.80	2	0.00	1.00	2
25	4 706	E2p4	2.23	2	0.00	1.00	2
Note: The posterior probabilities are computed from the DA based on the Bayesian rule; and the discriminant function scores, Fischer’s rule. Class 1 represents the class of the mudstones from the delta front; Class 2, from the tidal flat. Fc, canonical discriminant function. The correct percentage of the original validation is 92.2%; the leave-one-out cross-validation is 89.1%.

 | Show Table

DownLoad: CSV

4.4 Biomarker characteristics

In general, the Pr/Ph values in the mudstones from the delta source rocks (5.0−9.7, average 7.37) are higher than those in the tidal flat (1.71−2.60, average 2.12) and neritic source rocks (1.64%−3.33%, average 2.31%) (Fig. 6; Table 3). High Pr/Ph values demonstrate higher oxidizing conditions and a larger influx of plant materials in the delta environment.

Figure  6.  Partial total ion chromatograms (TIC) and m/z 123, 217, and 191 mass chromatograms showing the distribution of n-alkanes, diterpanes, steranes, and hopanes, for samples of delta front (a), tidal flat (b), and neritic facies (c). Pr, pristane; Ph, phytane; 1, 8β(H)-labdane; 2, 4β(H)-19-norisopimarane; 3, fichtelite; 4, pimarane; 5, isopimarane; 6, 16β(H)-phyllocladane; 7, 16α(H)-phyllocladane; 8, 13β(H)-atisane; 9, 20-normethylatisane; St, sterane; Ol, oleanane; H, hopane; Ga, gammacerane.

DownLoad: Full-Size Img PowerPoint

Table  3.  Biomarker ratios of the source rock samples in the study area

No.	Structural belt	Well	Well depth/m	Sedimentary facies	Sedimentary subfacies	Member	Pr/Ph	Pr/ nC17	Ph/ nC18	Ga/ C30 H	C27/ C29 St	16$\beta $(H)- Phyllocladane/ C30H
1	North Pinghu Slope	P-1	3 446	delta	delta front	E2p1-2	9.70	5.71	0.48	0.05	0.08	2.60
2	North Pinghu Slope	P-1	3 654	delta	delta front	E2p3	9.63	3.21	0.57	0.05	0.11	4.44
3	North Pinghu Slope	P-1	3 709	delta	delta front	E2p4	5.93	17.80	0.65	0.04	0.07	5.41
4	North Pinghu Slope	P-1	3 740	delta	delta front	E2p4	8.53	4.74	0.56	0.05	0.03	12.05
5	North Pinghu Slope	P-1	3 953	delta	delta front	E2p5	5.43	2.38	0.32	0.05	0.12	6.56
6	North Pinghu Slope	P-1	3 994	delta	delta front	E2p5	5.00	2.31	0.34	0.09	0.13	9.04
7	South Pinghu Slope	P-2	3 841	coast	tidal flat	E2p4	1.71	3.30	2.21	0.11	0.36	0.39
8	South Pinghu Slope	P-2	4 203	coast	tidal flat	E2p4	2.60	1.62	0.68	0.13	0.34	3.21
9	South Pinghu Slope	P-3	3 312	coast	tidal flat	E2p3	2.35	1.08	0.36	0.28	0.97	0.72
10	South Pinghu Slope	P-3	3 650	coast	tidal flat	E2p4	1.81	0.89	0.43	0.20	0.70	1.08
11	South Pinghu Slope	P-3	3 812	coast	tidal flat	E2p4	2.12	0.77	0.30	0.20	0.37	0.71
12	CISB	T-1	4 250	marine	neritic	E2p3	2.61	1.14	0.34	0.24	0.39	0.58
13	CISB	T-1	4 290	marine	neritic	E2p3	2.13	1.29	0.44	0.19	0.34	0.66
14	CISB	T-1	4 410	marine	neritic	E2p4	1.64	1.07	0.41	0.21	0.45	0.76
15	CISB	T-1	4 460	marine	neritic	E2p4	1.86	1.05	0.31	0.21	0.56	1.07
16	CISB	T-1	4 652	marine	neritic	E2p4	2.32	1.04	0.33	0.30	0.40	0.52
17	CISB	T-1	4 706	marine	neritic	E2p4	3.33	1.09	0.23	0.20	0.54	3.31
Note: Pr, pristane; Ph, phytane; nC17, C17 n-alkanes; nC18, C18 n-alkanes; St, regular sterane; Ga, gammacerane; H, hopanes. CISB, central inverted structural belt.

 | Show Table

DownLoad: CSV

Anomalously abundant gymnosperm-derived diterpenes including 8β(H)-labdane, 4β(H)-19-norisopimarane, fichtelite, pimarane, isopimarane, 16β(H)-phyllocladane, 16α(H)-phyllocladane, 13β(H)-atisane, and 20-normethylatisane were identified in the m/z 123 mass chromatogram (Fig. 6). Three of the largest peaks labeled as 2, 5, and 6 have been assigned to 4β(H)-19-norisopimarane, isopimarane, and 16β(H)-phyllocladane, respectively. Their identifications were mainly based on their mass spectra and relative retention time in previous studies (Noble et al., 1985; Cheng et al., 2020).

The sterane composition has a predominance of C₂₉ steranes relative to very low abundance of C₂₈ steranes, (m/z 217). Even though C₂₇ regular steranes were prevalent in neritic samples and even more prevalent in certain tidal flat samples, they were comparatively in low abundance in delta shale samples (Fig. 6).

There were no detective tricyclic terpanes in all the samples and only a minute abundance of pentacyclic triterpanes was found in some samples. 17α(H),21β(H)-C₃₀ hopane is the most prominent component in the C₂₇-C₃₃ hopane series. Gammacerance was detected in very low abundance in the delta samples but it has a higher abundance in tidal flat and neritic samples. There is a very low abundance of 18α(H)-oleanane in the samples. Thus, the oleanane index 18α(H)-oleanane/C₃₀ hopane was not used in this study.

5. Discussion

5.1 Origins of terrestrial organic matter in neritic mudstones

The properties of the particulate organic matter in the neritic sediments are determined by their origins and sedimentary environments. Compositions of the SOM in mudstones from the neritic facies (samples from Well T-1) were compared with those from the delta front facies (samples from Well P-1) and the tidal flat facies (samples from Well P-2) to trace the environments from which they come.

The P-E-A ternary plot (Tyson, 1995; Cai et al., 2020) was employed to assess the properties of SOM (Fig. 7). Both delta and tidal flat samples have high P contents (average values of 79.19% and 71.96% respectively) and a comparatively low E content (average values of 18.66% and 23% respectively), indicating that residues of higher plants were mainly piled up in situ. The neritic samples contain a slightly lower P value (average value of 65.77) and higher E value (average value of 28.57%) than the other two, indicating that the organic matter may have experienced a brief movement before deposition. As algae and other aquatic organisms are difficult to be preserved in the oxidizing environment, high AOM content generally dares suggestive of good preservation conditions of organic matter. The absence of AOM in the majority of the deltaic samples indicates the oxidizing condition. AOM contents of tidal flat samples are variable in a large range from 0% to 45%, due to the complicated depositional environment of tidal flat. In the supratidal zone, hydrodynamic forces are relatively mild, and the residual seawater or freshwater influx from rivers or rainfall in low-lying places creates marshes that are favorable for algae enrichment. The hydrodynamic force of intertidal to the subtidal area is strong, which is difficult to sustain algae. The concentration of AOM in the neritic environment is also low, and it appears to be related to the shallow depth of seawater and insufficient reduction, resulting in poor preservation circumstances of the organic matter. However, the information in the ternary diagram is limited, which is not enough to judge the sedimentary environment of the neritic organic matter.

Figure  7.  Ternary diagram of the P-E-A for samples of delta front (a), tidal flat (b), and neritic (c). P, charcoal+woods+undegraded woods; E, exinites+cuticles+spores; A, amorphous organic matter.

DownLoad: Full-Size Img PowerPoint

DA was used to distinguish the deltaic sample from the tidal flat samples and to determine the source of neritic organic matter based on the variation in sedimentary organic matter composition. Three parameters, i.e., cuticles (%), exinites (%), and fungal issues (%), with the strongest classification ability were automatically assessed for classification, and a discrimination model was constructed (Table 1). A three-dimensional scatter diagram for these three significant parameters chosen by stepwise DA show how they separate sample groups in 3-D space (Fig. 8a). The scatter plot of posterior probabilities (P₁, for Class 1) versus Fc scores clearly illustrates the difference between delta and tidal flat samples (Fig. 8b). On this graph, neritic samples were categorized and the majority of them were classified into Class 2, implying that they are associated with tidal flat coal-bearing mudstones.

Figure  8.  Three-dimensional scatter diagram of the three most significant parameters (i.e., exinites, curticles, and fungal tissues) showing the delta front, tidal flat, and neritic samples groups in a three-dimensional space (a), and cross-plot between the scores of the canonical discriminant function (Fc) based on Fisher’s discriminant rule and the posterior probabilities (P₁ for Class 1, i.e., delta) calculated from the DA based on the Bayesian discriminant rule (b). The boundaries between the two classes are shown by the black solid (P₁=0.5) and dotted lines (Fc=−0.44), and the corrected boundary is indicated by the red solid line (Fc=−0.88).

DownLoad: Full-Size Img PowerPoint

5.2 Similarity of organic matter in neritic and tidal flat analyzed by biomarker parameters

Biomarker parameters were employed to confirm the similarity between neritic organic matter and tidal flat organic matter (Fig. 9). For example, the ratio of pristine (Pr) to phytane (Ph) has been utilized to determine the origin and depositional environment of the organic matter (Didyk et al., 1978). Generally, high Pr/Ph ratios (>3) indicate a dominant terrigenous input in an anoxic environment, while low Pr/Ph ratios (<0.8) suggest a dominant algae input in an anoxic sedimentary environment. The cross-plot of phytane/nC₁₈ versus pristane/nC₁₇ (Fig. 9a) shows that the majority of mudstones were deposited in oxic settings, with Type III organic matter. Additionally, the Pr/Ph ratio suggests that the deltaic environment exhibits a large amount of terrigenous organic matter input in a stronger oxidizing condition (Fig. 9b), which is consistent with the results of SOM analysis.

Figure  9.  Plots of biomarker ratios of source rock samples in the study area. Pr, pristine; Ph, phytane; nC₁₇, C₁₇ n-alkanes; nC₁₈, C₁₈ n-alkanes; Ga, gammacerane; H, hopane.

DownLoad: Full-Size Img PowerPoint

Gammacerane is generated in the reduction of tetrahymanol during diagenesis and signals the stability of water column stratification in marine and lacustrine environments (Haven et al., 1989; Damsté et al., 1995). Both tidal flat and neritic samples exhibit a greater abundance of gammacerane than delta samples, suggesting that the tidal flat and neritic waters are stratified due to their higher salinity and poor hydrodynamics. The stable water column stratification is typically correlated with the preservation condition of organic materials.

C₂₇, C₂₈, and C₂₉ steranes can be used as a proxy for organic matter inputs, with a high abundance of C₂₉ indicating great input of higher plants in the organic matter (Volkman, 1986). Scatter plots of gammacerane/C₃₀ hopane versus C₂₇/C₂₉ regular sterane indicate a positive correlation between salinity and inputs of lower aquatic organisms, suggesting that tidal flat and neritic sedimentary environments likely have higher input of planktons. This is likely due to good preservation conditions of organic matter in tidal flat and neritic sedimentary environments.

Diterpanes are abundant in all the studied samples, indicating the presence of gymnosperm-dominated higher plant materials in the organic matter (Bojesen-Koefoed et al., 1996; Otto et al., 2005). In addition, the species and relative abundance of diterpenoids were remarkably similar in each of the three depositional habitats, indicating that the plant sources were comparable. Nonetheless, the total amount of diterpenoids varies considerably. There are two possible explanations for the increased relative abundance of diterpenoids in delta samples: more input from higher plants and a more oxidizing environment unfavorable for algae and bacteria preservation. Hopanes are generated from hopanoids, which are membrane lipids of prokaryotic organisms or diagenetic derivatives of these lipids (Peters et al., 2005). The relative contribution of higher plants and bacteria was characterized by the highest content of the diterpenoids, 16β(H)-phyllocladane compared to the highest content of the patchouli series, C₃₀ hopane. The plot of 16β(H)-phyllocladane/C₃₀ hopane versus C₂₇/C₂₉ regular sterane (Fig. 9d) demonstrates that the deltaic environment receives a significant input of higher plants, whereas the tidal flat and neritic environments have received more algal and bacteria input.

In summary, the influx of freshwater from the delta destroys the reducing sedimentary environment. In contrast, water bodies on tidal flat marshes are tranquil, with a high salinity influenced by seawater. The presence of still water is beneficial for the enrichment and preservation of organic matter. Biomarker characteristics in the neritic mudstone are extremely comparable to those in the tidal flat mudstones but they differ in the delta samples. This indicates that the neritic organic matter mainly originated from terrigenous organic matter on the tidal flat.

5.3 Supply of terrestrial organic matter from tidal flat to neritic environment

Based on the characteristics of SOM and biomarkers, neritic organic matter and tidal flat organic matter exhibit strong similarities. The primary source of neritic terrestrial organic debris is likely the coal-bearing sediments on tidal flats. Due to that the coal-bearing sediments were primarily formed in marshes on supratidal flats where tidal currents rarely reach, a subject that need more debate is that how the organic matter in the supratidal marshes was transferred to the ocean.

Tempestites have been discovered in the Tiantai area in the Xihu Depression (He, 2020). The tearing shape of the grayish-purple mud gravels in the tempestites shows that ancient storm eddies were capable of destroying the supratidal mud layer and dumping its debris into the sea. Microfossil evidence indicates that the Xihu Depression was located in the tropical to sub-tropical climate zone during the Eocene period (Jiang, 2011), when there were prevalent tropical storms. The storm surges damaged supratidal marshes and flushed away the plant debris and the in-site organic-rich muds.

During deposition of the Pinghu Formation, the Xihu Depression has experienced an entire process from sea erosion to sea recession. The sea level rose from E₂p₅ to E₂p₃ and dropped from E₂p₃ to E₂p_1-2, including multiple fluctuations during the process. The lithology assemblages of the tidal flat sedimentary facies have recorded the fluctuations of sea levels (Fig. 10a). Therefore, the tidal flat environment occasionally shifted because of the sea level changes. Consequently, there was frequent interbedding of coals, mudstones, and sandstone layers that were deposited. Sandstones were most commonly found in subtidal and tidal channels, whereas coalbeds and mudstones were commonly found in supratidal and intertidal zones. This provides tangible evidence that hydrodynamic power abruptly increased. As the sea level rose, the supratidal marshes were destroyed and organic-rich sediments were washed away by tidal currents and entered the ocean through tidal channels (Fig. 10b).

Figure  10.  Lithology assemblage of tidal flat sedimentary facies of Well P-3 showing contacts between coal beds and sandstones (a), and interpretation of the process of organic debris being carried to the marine via the tidal channel (b).

DownLoad: Full-Size Img PowerPoint

Terrigenous clastic organic matter can be transported by vigorous water flow along with fine sand. It was believed that the organic sources of oil and gas in the deep-water strata in the Kutei Basin were made up of leaves and other plant detritus that accumulated in turbidite sands (Saller et al., 2006). Similar cases can be found in the Congo turbiditic system, where terrigenous clastic materials and silt were carried to the deep sea combined (Baudin et al., 2017). In the Xihu depression, charcoal chips were found throughout the sandstones. As a result, particularly high TOC values were found in the tidal sand ridge samples in the Pinghu Formation in marine facies (Fig. 3). The increased tidal currents carry debris from higher plants and supratidal mud into the sea along with fine sands.

6. Conclusions

(1) The neritic strata in the Pinghu Formation of the Xihu Depression received a significant amount of terrestrial higher plant deposits. Carbonaceous debris was discovered in marine sandstone samples, with abnormally high TOC values. This carbonaceous matter was thought to be transported from the coal-bearing coastal formation and then subsequently deposited in the marine environment. The sedimentary organic debris composition of the inner shallow marine samples resembles that of the tidal flat and deltaic samples, both of which are dominated by P and E and have a low AOM abundance.

(2) The result of a discriminant model based on the plynofacies analysis of sedimentary organic matter concluded that the organic matter in neritic mudstones is more comparable to the organic matter in tidal flat mudstones than in delta mudstones. This was determined by comparing the two types of organic matter. According to this concept, most of the organic matter in neritic source rocks came from tidal sources.

(3) The results of the discrimination were supported by the fact that organic-related biomarkers revealed a significant degree of similarity between neritic samples and tidal flat samples. Tidal flats provide better preservation conditions than deltas with higher water salinity, as shown by relatively low Pr/Ph and high Ga/C₃₀H. This may be the explanation for the larger enrichment of organic matter in the tidal flats.

(4) Tidal currents have the potential to move enormous amounts of organic material. When a storm occurred or the sea level rose, the supratidal marshes were severely destroyed. The organic material was subsequently carried out to the sea by tidal water and then deposited in the shallow marine environment.