Depositional Processes and Paleoenvironmental Evolution of the Middle Eocene Lacustrine Shale in Beibu Gulf Basin, South China

Abstract

This study focuses on the middle Eocene lacustrine shales of the Lower Member 2 of the Liushagang Formation (L–LS2) in the Weixi’nan Depression of the Beibu Gulf Basin. Employing an integrated approach that combines core observation, thin-section analysis, Scanning Electron Microscopy (SEM), X-ray Diffraction (XRD), and geochemical proxies, we systematically characterize the lithofacies, sedimentary processes, and paleoenvironmental evolution. Six distinct lithofacies were identified: clay-rich mudstone, calcium-bearing mudstone, clay-rich siltstone, siliceous siltstone, ankerite-bearing sandstone, and siliceous sandstone. Based on depositional processes and structural features, these are grouped into three lithofacies assemblages: interbedded lithofacies assemblage, laminated lithofacies assemblage, and matrix lithofacies assemblage. Vertical facies distribution shows that the interbedded lithofacies assemblage dominates the lower L–LS2, reflecting active faulting, volcanism, a low lake level, prevalent gravity flows, and episodic oxidative conditions. The laminated lithofacies assemblage dominates the middle section and results from the combined influence of chemical and mechanical deposition, indicating fluctuating climate conditions that affected water depth, salinity, and redox dynamics. The upper section is characterized by matrix lithofacies assemblage, representing a stable, deep water, anoxic environment with low energy suspension settling. We propose a depositional model in which tectonics and climate jointly control lacustrine shale deposition. During the middle Eocene, intensified tectonic activity expanded accommodation space and increased clastic input, while climate fluctuations influenced chemical weathering, nutrient supply, and salinity. Together, these factors drove lake deepening and variability, affecting sedimentary energy and redox conditions. This study not only clarifies the sedimentary evolution of L–LS2 but also provides a critical geological framework for lacustrine shale oil exploration.

1. Introduction

The U.S. shale oil revolution has dramatically transformed the global energy landscape, highlighting the vast potential of unconventional hydrocarbon resources [1,2,3]. In China, the exploration of lacustrine basin shale oil, with its abundance and significant resource potential, has emerged as a key research focus [4,5,6]. However, the pronounced lithofacies heterogeneity, diverse interbeds, and complex depositional mechanisms of lacustrine shales present significant challenges for reservoir prediction and development, compared to marine shales [7,8,9,10].

Lithofacies, as products of distinct depositional settings, are crucial for interpreting organic matter enrichment, sedimentary processes, water chemistry variations, and high-quality reservoir formation [3,11,12,13,14]. Furthermore, lithofacies assemblages—ordered combinations of lithofacies within a depositional system—play a key role in controlling the distribution of shale oil “sweet spots” and the heterogeneity of the reservoir [15,16,17]. Therefore, detailed characterization of lithofacies assemblages is essential for understanding shale oil enrichment patterns.

During the Middle Eocene, climatic conditions in East Asia greatly enhanced lacustrine primary productivity and organic matter accumulation, creating favorable conditions for the formation of organic-rich shales [18,19]. The L–LS2 in the Weixi’nan Depression was deposited between approximately 48.5 and 46.7 Ma [20], coinciding with the transitional phase from the Early Eocene Climatic Optimum (EECO; ~53–49 Ma) to a subsequent cooling period [21,22]. This interval of frequent climate fluctuations likely had a profound impact on the depositional environment and biological productivity of the lake basin. Lacustrine depositional systems are highly sensitive to climate change, often reflected in enhanced terrigenous input, organic matter accumulation, and rhythmic sedimentation [23,24]. At the same time, the Early Middle Eocene was marked by intense tectonic activity, with the regional tectonic framework shaped by interactions between the Eurasian, Indo–Australian, and Pacific plates. Intense fault-related subsidence not only expanded accommodation space but also increased coarse clastic input, establishing the material basis for a coupled model of organic-rich deposition and enhanced preservation conducive to shale oil formation [25,26,27].

To date, detailed studies on the depositional processes and environments of these tectonically influenced, climatically transitional lacustrine shales from the Early Middle Eocene in the Beibu Gulf Basin remain limited, hindering effective shale oil development [28,29]. This study integrates core observations, thin-section analysis, XRD, and geochemical data to (1) establish lithofacies classification criteria for the L–LS2, construct a lithofacies assemblage framework, and reveal its vertical evolution; (2) analyze the depositional processes of lacustrine shales using lithofacies assemblages as fundamental units, and identify their depositional environments through geochemical proxies (e.g., Cr/Co, Corg/P, Sr/Cu); (3) propose a depositional model driven by the combined effects of tectonic activity and climate fluctuations, providing a theoretical basis for the efficient development of shale oil in continental lacustrine basins.

2. Geological Setting

The Beibu Gulf Basin is located on the northwestern continental shelf of the South China Sea. It is a Cenozoic intracontinental rift basin, structurally influenced by the combined interaction of the Pacific, Eurasian, and Indo–Australian plates. The basin evolution can be divided into two major stages: a rifting stage during the Eocene–Oligocene and a passive continental margin stage from the Miocene to the present [30]. In the early Eocene (~50 Ma), subduction of the Pacific Plate triggered mantle upwelling, leading to peak rifting activity [23]. This process generated NE–NEE-trending fault systems (e.g., Fault No. 1), which controlled the deposition of the thick Liushagang Formation [25,31]. The second extensional episode during the Eocene was the most intense, with an extension rate of 8.09% and a subsidence rate of up to 520 m/Ma. This episode was crucial for the widespread development of mid- to deep-lacustrine source rocks in the Member 2 of the Liushagang Formation (LS2) [26,27].

The Weixi’nan Depression lies in the northern Beibu Gulf Basin, trending NW–SE with an area of 3454 km². It is bounded by the Weixi’nan Fault to the north and bordered by the Qixi Uplift and the Weixi’nan Low Uplift to the south (Figure 1A). The depression is subdivided into three sags: A, B, and C [32]. Sags A and B serve as the main depocenters for the LS2. Sag A is currently the primary exploration target, while an oil-bearing zone has recently been discovered in Sag B [33], both showing great exploration potential. The structural pattern of the depression is controlled by NE–NEE-trending faults (Figure 1A). Fault No. 1, with a maximum displacement of over 1.69 km, defines the boundary and controls the development of Sag A. Fault No. 2 trends NEE, with a displacement of ~0.33 km, influencing sedimentary differentiation between Sags B and C [31].

During the Early to Middle Eocene, intensified collision between the Indian and Eurasian plates caused the regional extensional stress field to rotate clockwise to an NNW–SSE orientation [36,37]. Under this tectonic regime, the NE–NEE-trending (e.g., Fault No. 1) and NEE-trending (e.g., Fault No. 2) faults in the Weixi’nan Depression experienced reactivation [38]. During the Middle Eocene, the area underwent intense extensional tectonism (Figure 1B), with faulting reaching its Paleogene peak and average slip rates reaching 350 m/Ma. Rapid tectonic subsidence combined with undercompensated sedimentation significantly deepened the lake, creating a semi-deep to deep lacustrine depositional environment [25,26].

The Liushagang Formation is subdivided from bottom to top into LS3, LS2, and LS1. The LS2 member, deposited during the peak of lake basin expansion, represents a semi-deep to deep lacustrine setting, and is primarily composed of black shale, oil shale, and mudstone interbedded with thin siltstone layers (Figure 1B). The lower submember of LS2 (L–LS2) is a key target for shale oil exploration due to its high organic content (TOC > 6%), favorable kerogen type, and substantial thickness [39,40]. Vertically, the oil shale interval is typically 40–80 m thick, locally exceeding 100 m. From bottom to top, it comprises interbedded mudstone and sandstone/siltstone in the lower section, mudstone and calcareous mudstone with minor siltstone in the middle, and thick mudstone in the upper section. Laterally, the shale is mainly distributed in the hanging wall of Fault No. 1 and the peripheral zone of Fault No. 2, covering an area of over 1000 km² [41]. On seismic profiles, this interval is characterized by low-frequency, continuous, and strong reflections, making it a prominent regional marker bed [35].

The deposition of the L–LS2 (~48.5–46.7 Ma) coincided with the transition from the Early Eocene Climatic Optimum (EECO) into a global cooling phase [20,21,22] (Figure 1B). The age of L–LS2 is constrained by previous high-resolution cyclostratigraphic studies of the Liushagang Formation, which were based on gamma-ray (GR) logging data and involved the identification of Milankovitch-scale orbital cycles through spectral and evolutionary Fourier analyses as well as sedimentation-rate estimation [18,20]. These cyclostratigraphic results provide a robust chronological framework for our paleoenvironmental analysis.

Although the climate was beginning to transition toward an icehouse state, conditions remained warm and humid, characteristic of a subtropical to tropical monsoonal environment [24]. Palynological data show a decline in herbaceous plants and an increase in conifers and evergreen broadleaf trees, indicating a semi-humid to semi-arid mid-subtropical climate, with mean annual temperatures of 13.3–19.8 °C and annual precipitation of 797.5–1113.2 mm [24,26]. The combination of warm and humid climate, undercompensated subsidence, and anoxic conditions in semi-deep to deep lacustrine environments jointly facilitates efficient organic matter accumulation and preservation [18,42].

3. Methods

This study focuses on core samples obtained from Well A, located in Sag A of the Weixi’nan Depression. The well penetrates the L–LS2, a key target for shale oil exploration. In total, 23.6 m of core were recovered from depths of 3005.5–3029.1 m, with samples collected at approximately 1 m intervals.

A total of 109 core samples were analyzed using petrographic thin sections, Field Emission Scanning Electron Microscopy (FE–SEM), and porosity–permeability measurements. Petrographic analysis included 77 standard thin sections and 32 epoxy-impregnated casting sections to examine mineralogy, grain size, texture, and biogenic features, using a Leica DM2000 microscope (Leica Microsystems GmbH, Wetzlar, Germany) under 50× and 100× magnifications. FE–SEM analyses were performed on 20 representative samples, prepared by polishing and argon-ion milling. Mineral characteristics, microtextures, and porosity features were examined using a Hitachi SU8010 FE–SEM (Hitachi High-Tech Corporation, Tokyo, Japan).

Mineralogical compositions were determined by X-ray diffraction (XRD) at the Zhanjiang Branch of CNOOC Ltd. ImageJ (version 1.52i) software was used to measure quartz grains, bioclasts, and pyrite framboids across multiple fields of view, and the average values were reported. Porosity and permeability tests were performed externally at the Craton Research Service Platform.

To investigate sedimentary environmental changes in the L–LS2, geochemical proxies from Wells A and B were utilized. Data for Well A were provided by the Zhanjiang Branch of CNOOC Ltd., whereas Well B data were obtained from Cao et al. (2020) [43].

Statistical analyses, including principal component analysis (PCA) and analysis of variance (ANOVA), were applied to evaluate inter-lithofacies differences and to minimize overinterpretation of correlated variables. All statistical analyses were conducted using the SPSSAU online data analysis platform.

4. Results

The L–LS2 of the Weixi’nan Depression exhibits a thickness ranging from 40 to 80 m, with the Sag A region potentially exceeding 100 m in thickness. This interval primarily consists of mudstone and a small amount of siltstone and sandstone. Mudstone displays typical high GR characteristics in well logs, indicating a high clay mineral content. The GR response of siltstone and sandstone is significantly lower than that of the adjacent high GR mudstone. Based on mineralogy, texture, structure, and biological features, six lithofacies can be identified: (1) clay-rich mudstone, (2) calcium-bearing mudstone, (3) clay-rich siltstone, (4) siliceous siltstone, (5) ankerite-bearing sandstone, and (6) siliceous sandstone.

4.1. Mineral Compositions and Organic Matter Characteristics

Mineral composition and organic matter characteristics are key factors in lithofacies classification. Thin-section observations and XRD analysis indicate that the rocks in the L–LS2 of the Weixi’nan Depression are primarily composed of quartz, clay minerals, and carbonate minerals. Ternary diagram analysis shows clustered mineral compositions (Figure 2). Total organic carbon (TOC) analysis reveals that clay-rich mudstone and calcium-bearing mudstone contain significantly higher organic matter than other lithofacies, while siliceous siltstone, ankerite-bearing sandstone, and siliceous sandstone are characterized by low TOC content (

Table 1. Mineral contents, quartz diameters, structures, TOC, and kerogen types of different lithofacies.

Lithology	Lithofacies	Quartz, Feldspar, Pyrite (%)	Carbonate (%)	Clay (%)	Quartz Diameters (μm)	Structure	TOC (%)	Organic Matter Types
Shale	clay-rich mudstone	39–53/48.75	0–6/1.83	46–59/49.42	10–80	inconspicuous laminar structure	3.41–7.41/4.76	I–II1
	calcium-bearing mudstone	30–35/31.71	22–33/27.14	37–46/41.14	10–50	inconspicuous laminar structure	3.64–7.1/5.11	I
Siltstone	clay-rich siltstone	61–72/67.89	0–3/2	26–36/29.44	50–200	laminar structure	1.16–3.79/1.96	II2–III
	siliceous siltstone	68–73/70	3–8/6	~23	50–300	laminar structure	~0.8%	II2–III
Sandstone	ankerite-bearing sandstone	55–70/62.29	14–38/23.86	7–16/13.57	100–500	massive structure	~0.5%	II2–III
	siliceous sandstone	78–84/81.33	0–1/0.67	15–22/17.67	150–1500	graded bedding	~0.5%	II2–III

).

Clay-rich mudstone is characterized by a quartz content ranging from 33% to 41% (average 37.25%), a clay mineral content between 46% and 59% (average 49.42%), a low carbonate content (less than 5%), and a notable pyrite content averaging 10.08%. This mudstone contains a substantial amount of organic matter, with TOC ranging from 3.41% to 7.41%, averaging 4.67%. The kerogen type is classified as Type I–II₁, and its organic matter maturity is relatively low (Table 1).

Calcium-bearing mudstone features a quartz content of 23.86%, an average clay mineral content of 41.14%, an average calcite content of 23.43%, and a pyrite content averaging 6.14%. This mudstone exhibits TOC values similar to clay-rich mudstone, ranging from 3.64% to 7.1%, with an average of 5.11%. However, the kerogen type is primarily Type I (Table 1).

Clay-rich siltstone has a quartz content ranging from 59% to 70% (average 65.44%) and a clay mineral content between 26% and 36% (average 29.44%). It has a lower TOC content, ranging from 1.16% to 3.79% (average 1.96%), with kerogen classified as Type II₂–III (Table 1).

Siliceous siltstone has a quartz content between 67% and 72% (average 69%), an average clay mineral content of 23%, and an average ankerite content of 4%. It has a low TOC content of about 0.8%, with kerogen categorized as Type II₂–III (Table 1).

Ankerite-bearing sandstone shows a quartz content ranging from 54% to 70% (average 61.29%), a low clay mineral content between 7% and 16% (average 13.57%), and a significant ankerite content ranging from 12% to 38% (average 22.71%), with pyrite content not exceeding 1%. This sandstone has a low TOC content of about 0.5%, with kerogen classified as Type II₂–III, indicating a terrestrial organic source (Table 1).

Siliceous sandstone has a quartz content greater than 70%, reaching up to 82%, with a clay matrix averaging 17.67% and carbonate minerals present in very low amounts, not exceeding 1%. It has low organic content, with TOC around 0.5%, and kerogen classified as Type II₂–III, suggesting an external organic matter input (Table 1).

4.2. Lithofacies

4.2.1. Clay-Rich Mudstone

Clay-rich mudstone is dark gray to black in color, commonly containing bitumen nodules. It has relatively low hardness and appears homogeneous in hand specimens (Figure 3A). Thin-section observations reveal well developed lamination, primarily controlled by variations in organic matter content (Figure 3B). The rock is predominantly composed of mud-sized clay minerals with minor silt-sized quartz grains (Figure 3C). The grains are poorly sorted, exhibit sub-angular to sub-rounded shapes, and range from 10 to 80 μm in diameter. Clay minerals fill the interstitial spaces between quartz grains, forming the matrix.

Carbonate minerals are present in minor amounts, with locally developed euhedral ankerite crystals (Figure 3C). Pyrite is relatively abundant, occurring predominantly as granular or framboidal forms, and often forms nodular or stratified accumulations (Figure 3A,D). In addition, algal cyst structures are observed [11,44], which are commonly infilled with quartz, pyrite, and bitumen (Figure 3C,E). This lithofacies was typically deposited in a low-energy, stagnant water environment under strongly reducing conditions.

4.2.2. Calcium-Bearing Mudstone

Calcium-bearing mudstone appears as a homogeneous dark gray rock in hand specimens, with locally developed alternating light and dark laminations caused by mineralogical variations (Figure 4A). It is primarily composed of fine-grained clay, calcite, and silt-sized quartz (Figure 4B). The carbonate minerals mainly consist of bioclastic fragments, detrital calcite, and calcareous mud, with minor occurrences of ankerite (Figure 4C).

Thin-section analysis reveals that the clay minerals are intermingled with organic matter and distributed within the matrix. Bioclastic components, including ostracod shells (Figure 4D) that are 100 to 500 μm in size, are also notable, paralleling bedding planes and forming distinct layers with smooth, concave–convex contacts (Figure 4D). SEM analysis reveals detailed pore structures associated with framboidal pyrite. Algal fragments are also present, typically arranged horizontally (Figure 4B,C,E,F). Localized bioturbation features are observed as reddish-brown lenticular structures, primarily filled with calcite (Figure 4E,G,H). This facies was deposited under weakly reducing to slightly oxic conditions, suggesting elevated lake salinity with intermittent oxidation events during sedimentation [11].

4.2.3. Clay-Rich Siltstone

Clay-rich siltstone appears gray in core samples and shows soft sediment deformation, such as deformed silty lenses (Figure 5A). Features like pillow structures and load-induced deformations suggest settlement and downward intrusion of liquefied density currents (Figure 5B). This lithofacies contains dark gray clayey clasts and light gray silty clusters, with gradual transitions at the base and erosion by overlying siliceous siltstone at the top.

Microscopic examination shows quartz grains ranging from 50 μm to 200 μm, poorly sorted with moderate to poor rounding (Figure 5C). Clay minerals are layered, filling interstitial spaces between terrigenous clasts (Figure 5D,E). Additionally, pyrite is present in granular or layered distributions, often accompanied by minor amounts of muscovite (Figure 5F).

4.2.4. Siliceous Siltstone

Siliceous siltstone appears gray to gray-white and exhibits prominent parallel and ripple laminations (Figure 6A–C). Deformation and load-induced features, such as flame structures, are also present (Figure 6D). It has high quartz content, with grain sizes ranging from 50 μm to 300 μm, classified as silty–fine sand. It has poor sorting and moderate rounding, indicating lower compositional and structural maturity (Figure 6E,F). Quartz grains occasionally show dissolution features along their edges.

4.2.5. Ankerite-Bearing Sandstone

Ankerite-bearing sandstone has a massive structure with indistinct graded bedding. It appears gray–white to light gray in hand specimens (Figure 7A,B). The quartz content ranges from 54% to 70%, with fine to medium sand-sized particles (100 μm to 500 μm), which are angular to sub-angular (Figure 7C). The clay mineral content is low, ranging from 7% to 16%, and the ankerite content is relatively high, between 12% and 38% (Figure 7D,E). Minor amounts of siderite and pyrite are also present, with pyrite content not exceeding 1%.

4.2.6. Siliceous Sandstone

Siliceous sandstone is light gray and exhibits a massive structure (Figure 8A). It shows a sharp contact with underlying ankerite-bearing sandstone or clay-rich siltstone, while the upper contact with clay-rich mudstone is gradational (Figure 8B). Multiple cycles of normal grading are observed, with a distinct and sharp basal boundary (Figure 8A,B).

Thin-section analysis shows that the facies is dominated by quartz, comprising over 70% and up to 82% of the composition. The quartz grains range from 150 to 1500 μm in diameter, corresponding to the fine to coarse sand size range (Figure 8C,D). The matrix is primarily composed of clay with a low overall content (Figure 8E,F). Compared to the ankerite-bearing sandstone, this facies has coarser grains and a significantly lower carbonate content (<1%) (Figure 8F).

4.3. Lithofacies Assemblages and Their Vertical Distribution

4.3.1. Lithofacies Assemblages

Given the pronounced heterogeneity and rapid vertical transitions typical of lacustrine successions [45], this study employs a process-oriented classification of lithofacies assemblages to elucidate depositional dynamics. Six lithofacies identified in the L–LS2 are grouped into three lithofacies assemblages: interbedded lithofacies assemblage, laminated lithofacies assemblage, and matrix lithofacies assemblage (Figure 9). The grouping is based on three explicitly defined criteria: (1) sedimentary structures and fabric (e.g., grading, sharp contacts, lamination, bioturbation); (2) grain-size relationships and textural trends; and (3) mineralogical composition as shown in the clay–carbonate–siliceous ternary diagram (Figure 2). These three lithofacies assemblages represent genetic end-members reflecting distinct depositional processes—gravity-flow dominance, stratified suspension-chemical deposition, and stable suspension fallout—rather than purely descriptive categories.

Interbedded lithofacies assemblage is characterized by alternating layers of clay-rich mudstone and (silt) sandstone, including clay-rich siltstone, siliceous siltstone, ankerite-bearing sandstone, and siliceous sandstone, all showing sharp and well-defined contacts (Figure 9). Normal grading within some sandstone layers (Figure 7A, Figure 8A,B and Figure 9A) and soft-sediment deformation structures in siltstone units (Figure 5A,B and Figure 6B) indicate deposition by density currents under waning flow energy [46,47,48]. Mudstone clasts at the base of certain sandstone beds (Figure 9A) further suggest erosion of underlying strata by high-energy gravity flows [49,50,51]. Based on these textural and structural features, together with their compositional distribution on the ternary diagram (Figure 2), the interbedded lithofacies assemblage represents a gravity-flow-dominated end-member.

Laminated lithofacies assemblage mainly comprises interbedded calcium-bearing and clay-rich mudstones, with individual layers typically 1–2 m thick (Figure 9B). The calcium-bearing mudstone exhibits fine, parallel lamination and a scarcity of siliceous clasts, suggesting rhythmic suspension settling combined with authigenic precipitation of fine-grained calcite (Figure 4). Coarser bioclastic fragments and detrital calcite are interpreted as autochthonous, based on their alignment parallel to bedding and the preservation of primary structures, indicating in situ deposition rather than long-distance transport. Alternations in calcite and clay mineral content (Figure 10) reflect lake-level fluctuations and variations in water-column stratification [52]. These compositional and structural characteristics justify grouping the two lithofacies into a single laminated (suspension-chemical) lithofacies assemblage.

Figure 10. Composite graphic column of core in the L–LS2 of Well A (see locations in Figure 11). Displaying vertical trends of mineral composition and geochemical proxies in the sampled shales, with notable contrasts observed between the three lithofacies assemblages. ILA = Interbedded lithofacies assemblage; LLA = Laminated lithofacies assemblage; MLA = Matrix lithofacies assemblage.

Figure 10. Composite graphic column of core in the L–LS2 of Well A (see locations in Figure 11). Displaying vertical trends of mineral composition and geochemical proxies in the sampled shales, with notable contrasts observed between the three lithofacies assemblages. ILA = Interbedded lithofacies assemblage; LLA = Laminated lithofacies assemblage; MLA = Matrix lithofacies assemblage.

Figure 11. Vertical evolution of lithofacies assemblages in the L–LS2. Lithofacies identification was based on lithology, GR, DT, and predicted mineral composition from elemental logging. Analysis of lithofacies stacking patterns further reveals a vertical development sequence, evolving from interbedded lithofacies assemblages at the base, through laminated lithofacies assemblages with minor matrix types in the middle, to matrix lithofacies assemblages at the top. ILA = Interbedded lithofacies assemblage; LLA = Laminated lithofacies assemblage; MLA = Matrix lithofacies assemblage.

Figure 11. Vertical evolution of lithofacies assemblages in the L–LS2. Lithofacies identification was based on lithology, GR, DT, and predicted mineral composition from elemental logging. Analysis of lithofacies stacking patterns further reveals a vertical development sequence, evolving from interbedded lithofacies assemblages at the base, through laminated lithofacies assemblages with minor matrix types in the middle, to matrix lithofacies assemblages at the top. ILA = Interbedded lithofacies assemblage; LLA = Laminated lithofacies assemblage; MLA = Matrix lithofacies assemblage.

Matrix lithofacies assemblage consists of thick, continuous intervals of clay-rich mudstone characterized by fine-grained, smooth, and homogeneous lamination (Figure 9C). The mudstone contains abundant pyrite and organic matter (Figure 3), indicating long-term accumulation under low-energy, suspension-dominated, and anoxic conditions. The indistinct lamination and absence of bioturbation suggest deposition in a persistently anoxic and stable deep-lake environment (Figure 3). Based on its fabric, mineralogy, and ternary compositional characteristics (Figure 2), the matrix lithofacies assemblage represents the stable suspension–fallout end-member typical of prolonged deep-water sedimentation.

4.3.2. Vertical Distribution of Lithofacies Assemblages

The lithofacies of L–LS2 have been characterized through an integrated analysis of lithology, GR logs, sonic transit time (DT) logs, and predicted mineral composition derived from elemental logging data. The vertical distribution patterns of lithofacies assemblages were determined (Figure 11).

The interbedded lithofacies assemblage is primarily developed in the lower part of the L–LS2, reflecting active tectonism and substantial clastic input. The laminated lithofacies assemblage dominates the middle section, indicating periodic fluctuations in lacustrine depositional conditions. The matrix lithofacies assemblage is concentrated in the upper part of the L–LS2, representing continuous deposition under deep-water, low-energy conditions (Figure 11). Overall, the vertical profile of the L–LS2 shows a clear upward transition from interbedded lithofacies assemblage to laminated lithofacies assemblage to matrix lithofacies assemblage. This trend reflects a shift from high-energy, clastic-dominated conditions to a more stable, deep-water, low-energy depositional environment.

4.4. Geochemical Proxies

4.4.1. Provenance

The Al₂O₃/TiO₂ ratio is a key indicator of source rock composition [53]. In this study, the interbedded lithofacies assemblage exhibits the lowest Al₂O₃/TiO₂ ratios (14.78–24.70), followed by the matrix lithofacies assemblage (20.35–26.07), while the laminated lithofacies assemblage exhibits the highest values (20.70–26.64,

Table 2. Geochemical proxies for lithofacies assemblages in the L–LS2 from Wells A and B. The geochemical data for Well B are quoted from Cao et al. (2020) [43].

Sample	Well	Depth (m) /Strata	Lithofacies Assemblage	Provenance		Paleoproductivity			Paleoclimate	Redox Conditions	Paleosalinity	Paleowater Depth	TOC (%)
				Cr/Co	Al2O3/TiO2	P (ppm)	Cu (ppm)	Mo (ppm)	Sr/Cu	Corg/P (molar)	Sr/Ba	Zr/Rb
1	A	3005.71	Laminated	5.90	23.01	636.00	38.40	2.29	4.04	287.47	0.20	0.74	7.09
2	A	3006.8	Laminated	6.87	23.41	429.00	29.90	1.77	15.92	328.80	0.18	0.68	5.47
3	A	3007.3	Laminated	6.18	23.84	1434.00	24.00	2.10	15.92	171.55	0.17	0.68	9.54
4	A	3008.28	Laminated	6.59	23.09	413.00	40.30	1.78	4.07	416.46	0.08	0.74	6.67
5	A	3008.9	Laminated	6.03	24.10	424.00	34.80	1.92	4.31	495.06	0.25	0.73	8.14
6	A	3010.2	Laminated	6.77	26.64	556.00	30.10	1.71	8.57	254.62	0.59	0.73	5.49
7	A	3011.2	Laminated	5.85	20.70	1278.00	43.10	1.29	4.41	108.96	0.10	0.73	5.40
8	A	3011.8	Laminated	6.25	25.60	424.00	30.60	1.46	13.92	298.62	0.25	0.58	4.91
9	A	3013.3	Laminated	6.63	25.31	637.00	34.20	1.28	5.56	219.41	0.15	0.75	5.42
10	A	3014.6	Laminated	5.76	22.58	768.00	50.00	1.46	3.52	217.24	0.28	0.75	6.47
11	A	3015.9	Matrix	5.83	22.30	578.00	45.80	0.92	3.78	221.73	0.25	0.68	4.97
12	A	3017	Matrix	6.38	22.63	491.00	43.40	1.31	4.19	288.85	0.08	0.74	5.50
13	A	3017.6	Matrix	6.48	21.45	531.00	47.60	1.09	3.76	234.56	0.10	0.96	4.83
14	A	3018.3	Matrix	6.77	22.14	616.00	60.00	1.58	5.35	229.82	0.13	0.83	5.49
15	A	3018.9	Matrix	6.69	20.35	669.00	44.40	1.24	5.16	151.87	0.29	0.83	3.94
16	A	3019.5	Matrix	6.42	22.99	691.00	50.40	1.14	3.69	195.17	0.07	0.95	5.23
17	A	3020	Matrix	7.57	21.70	508.00	65.00	1.19	4.12	265.48	0.13	1.19	5.23
18	A	3021.3	Matrix	8.00	21.84	3534.00	51.70	1.07	4.53	27.58	0.14	1.06	3.78
19	A	3022.59	Laminated	6.93	22.87	1028.00	42.90	1.69	7.81	140.22	0.42	1.02	5.59
20	A	3023.2	Laminated	6.14	24.49	1638.00	32.80	1.87	16.07	70.37	0.42	0.85	4.47
21	A	3023.8	Laminated	5.20	21.31	1599.00	51.10	1.38	3.03	112.73	0.13	0.97	6.99
22	A	3025	Laminated	6.48	23.25	1517.00	51.40	1.43	3.70	92.81	0.13	0.98	5.46
23	A	3026.3	Laminated	6.68	22.03	1770.00	31.20	1.40	24.42	66.14	0.72	1.37	4.54
24	A	3026.93	Interbedded	3.38	20.64	124.00	4.25	0.17	7.46	101.90	0.08	2.58	0.49
25	A	3027.64	Interbedded	5.65	16.38	321.00	13.90	0.23	4.73	304.46	0.29	2.50	3.79
26	A	3028.17	Interbedded	4.58	17.66	116.00	5.29	0.18	18.62	113.37	0.50	2.06	0.51
27	A	3029	Interbedded	4.17	14.78	374.00	21.20	0.47	3.18	93.77	0.22	2.15	1.36
28	B	Upper L–LS2	Matrix	7.71	26.07	603.00	43.33	5.88	1.67	230.96	0.06	0.05	6.32
29	B	Upper L–LS2	Matrix	7.26	25.60	751.00	42.73	6.60	2.03	145.84	0.05	0.07	6.28
30	B	Middle L–LS2	Laminated	8.35	24.83	563.00	40.00	4.56	2.53	270.97	0.11	0.07	5.64
31	B	Middle L–LS2	Laminated	8.44	25.84	555.00	39.09	4.44	2.58	216.10	0.08	0.06	4.96
32	B	Lower L–LS2	Interbedded	6.58	24.70	969.00	34.55	3.60	4.64	258.54	0.09	0.06	4.26
33	B	Lower L–LS2	Interbedded	6.63	23.92	594.00	34.55	3.12	6.05	231.17	0.10	0.06	3.99
34	B	Lower L–LS2	Interbedded	6.84	24.15	594.00	39.09	3.48	3.32	113.40	0.10	0.05	2.66

). Chromium (Cr) and cobalt (Co) are effective provenance indicators. Their Cr/Co ratios are generally higher in ultramafic and mafic rocks than in felsic rocks [43,54,55,56]. The Cr/Co values in this study range from low in the interbedded lithofacies assemblage (3.38–6.84), to moderate in the laminated lithofacies assemblage (5.20–8.44), to the highest in the matrix lithofacies assemblage (5.83–8.00, Table 2).

4.4.2. Paleoclimate and Water Depth

The Sr/Cu ratio is widely used to infer paleoclimate conditions [57,58,59,60,61]. The laminated lithofacies assemblage exhibits the highest and most variable Sr/Cu ratios (2.53–24.42), followed by the interbedded lithofacies assemblage (3.18–18.62), which also shows considerable fluctuation. The matrix lithofacies assemblage displays the lowest and most stable values (1.67–5.35, Table 2).

The Zr/Rb ratio is a sensitive proxy for hydrodynamic conditions and water depth [62,63,64,65]. In this study, it shows a decreasing trend from bottom to top (Figure 10). The interbedded lithofacies assemblage has the highest ratios (0.05–2.58). The laminated lithofacies assemblage shows lower but more variable values (0.06–1.37), while the matrix lithofacies assemblage shows low and relatively stable values (0.05–1.19, Table 2), suggesting deposition in a deep-water environment, possibly exceeding 20 m in depth [42].

4.4.3. Paleoproductivity and Organic Matter Preservation

Total organic carbon (TOC) content ranges from 0.49% to 9.54%. The interbedded lithofacies assemblage shows the lowest TOC (0.49–4.26%), the laminated lithofacies assemblage is the highest (4.47–9.54%), and the matrix lithofacies assemblage falls in between (3.78–6.32%). Elements such as phosphorus (P), copper (Cu), and molybdenum (Mo) serve as reliable proxies for estimating primary productivity [66,67,68]. The interbedded lithofacies assemblage has the lowest contents of P (116–969 ppm), Cu (4.25–39.09 ppm), and Mo (0.17–3.60 ppm). In contrast, the laminated lithofacies assemblage shows the highest P (413–1770 ppm) and Mo (1.28–4.56 ppm) contents, and moderate Cu (24.00–51.40 ppm). The matrix lithofacies assemblage has intermediate P (491–3534 ppm) and Mo (0.92–6.60 ppm) contents, but the highest Cu levels (42.73–65.00 ppm, Table 2).

The ratio of organic carbon to phosphorus (Corg/P) reflects the redox conditions of the depositional environment [69]. In this study, Corg/P (molar) values range from 66.14 to 495.06. The interbedded lithofacies assemblage shows the lowest ratios (93.77–304.46), while the laminated lithofacies assemblage shows the widest variability (66.14–495.06). The matrix lithofacies assemblage generally exhibits high Corg/P ratios (145.84–288.85, Table 2).

4.4.4. Paleosalinity

Strontium (Sr) and barium (Ba) are sensitive indicators of paleosalinity [70]. Sr/Ba ratio analysis reveals that the interbedded lithofacies assemblage has low values (0.08–0.50), while the laminated lithofacies assemblage displays relatively higher and more variable ratios (0.08–0.72). The matrix lithofacies assemblage shows the lowest and most stable Sr/Ba values (0.05–0.29, Table 2).

4.5. Reservoir Characteristics

Clay-rich mudstone develops limited organic matter pores and intercrystalline pores within pyrite (Figure 12A,B). This lithofacies is predominantly compact, with porosity ranging from 2.38% to 5.06% (average 3.23%) and mean permeability of 0.05 × 10⁻³ μm², indicating poor reservoir potential. Calcium-bearing mudstone likewise exhibits few organic pores but well-developed pyrite intercrystalline pores (Figure 12C). Porosity ranges from 1.70% to 3.05% (average 2.34%), and permeability ranges from 0.01 × 10⁻³ μm² to 0.03 × 10⁻³ μm² (average 0.02 × 10⁻³ μm²), suggesting poor reservoir potential. Clay-rich siltstone has few pores, with porosity values between 6.89% and 8.58% (average 7.64%). Permeability ranges from 0.02 × 10⁻³ μm² to 0.03 × 10⁻³ μm² (average 0.025 × 10⁻³ μm²), indicating limited reservoir quality. In siliceous siltstone, sparse intergranular pores predominate (Figure 12D,E). Porosity ranges from 7.13% to 9.98% (average 8.72%), while permeability ranges from 0.01 × 10⁻³ μm² to 0.08 × 10⁻³ μm² (average 0.07 × 10⁻³ μm²), reflecting average reservoir potential. Ankerite-bearing sandstone is dominated by intergranular pores often occluded by carbonate cements (Figure 12F,G). Porosity ranges from 4.69% to 12.52% (average 8.64%), and permeability ranges from 0.05 × 10⁻³ μm² to 1.64 × 10⁻³ μm² (average 0.41 × 10⁻³ μm²), suggesting moderate reservoir potential. Siliceous sandstone has more developed porosity compared to ankerite-bearing sandstone (Figure 12H,I), with porosity ranging from 13.15% to 16.66% (average 14.84%). Permeability ranges from 0.36 × 10⁻³ μm² to 2.75 × 10⁻³ μm² (average 1.48 × 10⁻³ μm²), indicating favorable reservoir potential.

5. Discussion

5.1. Depositional Processes of Lithofacies Assemblages

The L–LS2 records deposition in a relatively deep lacustrine setting characterized by a stratified water column and oxygen-deficient bottom waters. Sedimentation likely occurred below the storm wave base, with intermittent fluctuations in bottom-water conditions and redox states influencing facies development.

5.1.1. Interbedded Lithofacies Assemblage

This lithofacies assemblage records recurrent turbidity-current activity. Mudstones represent suspension fallout, whereas unstructured and graded sandstones (or siltstones) reflect gravity-flow deposition on the lake floor, most likely triggered by episodic river-flood events [48]. High-density currents transported and deposited siliceous sandstones and ankerite-bearing sandstones, while low-density currents formed siliceous siltstones and clay-rich siltstones. Normal grading and sharp basal contacts indicate waning flow energy, and the occurrence of mud clasts and stacked graded packages suggests repeated high-energy events [49,50,51].

The ankerite-bearing sandstone displays massive structures and disordered clastic particles, indicating rapid sedimentation, with ankerite cement primarily filling intergranular spaces (Figure 12F). In deep-water gravity-flow sandstones, adjacent mudstones may have acted as an external carbonate source for cementation [71,72,73]. Although δ¹³C data for ankerite are unavailable, its fine, uniform texture and pore-filling occurrence support an authigenic origin (Figure 7D,E). Such diagenetic cementation likely reduced reservoir quality relative to siliceous sandstone. Rhythmic siliceous siltstones and clay-rich siltstones containing soft-sediment deformation structures record the terminal stages of low-density turbidity currents [74,75,76].

5.1.2. Laminated Lithofacies Assemblage

The laminated lithofacies assemblage developed under variable bottom-water conditions within a stratified lake. Laminations and mineral alternations (Figure 10) record periodic hydrological and climatic oscillations [52], while calcium-bearing mudstones formed by suspension settling coupled with authigenic fine-grained calcite precipitation, reflecting relatively low terrigenous input and periods of elevated salinity.

Vertical trends show that quartz, pyrite, and clay minerals covary and vary inversely with carbonate content (Figure 10), indicating alternating episodes of detrital influx and authigenic calcite precipitation. Transitions between clay-rich and carbonate-rich mudstones correspond to fluctuations in salinity and terrigenous supply, whereas variations in kerogen type—from Type I in calcium-bearing mudstone to Type I–II₁ in clay-rich mudstone—reflect shifts in organic matter sources [3].

Short-lived bioturbation and the presence of benthic fossils record brief oxygenation events (Figure 4D), with bioturbation inhibiting pyrite formation [77,78]. Calcium-bearing mudstone contains fine-grained calcite formed by authigenic precipitation, whereas the occurrence of biogenic carbonate fragments, such as ostracod shells, indicates an additional biogenic contribution.

5.1.3. Matrix Lithofacies Assemblage

The matrix lithofacies assemblage represents long-term suspension fallout in a deep, stagnant lake. Thick clay-rich mudstones with indistinct lamination, abundant pyrite, and high TOC (Figure 3B and Figure 12B) accumulated under persistently anoxic conditions. The absence of benthic fauna and bioturbation confirms exclusion of macrobenthos from bottom waters. Planktonic algal debris, abundant clay, and distal quartz imply deposition in a stratified, nutrient-rich lacustrine system far from clastic input. This lithofacies assemblage thus represents the most stable deep-water setting, conducive to organic matter accumulation and preservation.

5.2. Paleoenvironmental Analysis

To quantitatively extract paleoenvironmental signals and evaluate geochemical distinctions among lithofacies assemblages, principal component analysis (PCA) was applied to the Z-score standardized geochemical dataset. The first two components (PC1 and PC2) explain 89.67% of the total variance (PC1: 75.40%; PC2: 14.23%) (Figure 13A), providing a statistically robust basis for subsequent environmental interpretation [79].

PC1 exhibits strong positive loadings of Al₂O₃, TiO₂, Rb, Cr, Zr, and Ni, reflecting dominance of terrigenous clastic input and provenance control. Minor loadings of P, Cu, and Mo imply a secondary influence from redox or productivity-related processes. PC2, dominated by Sr, Ba, and Mo, reflects variations in water chemistry, including salinity fluctuations and redox dynamics associated with transient water-column stratification and organic matter cycling (Figure 13B).

In PC1–PC2 space (Figure 13A), the lithofacies assemblages form distinct clusters: interbedded lithofacies assemblages occupy lower PC1 and PC2 values; laminated lithofacies assemblages plot at intermediate PC1 and higher PC2; matrix lithofacies assemblages show higher PC1 and moderate PC2. These patterns indicate systematic differences in clastic input, water chemistry, and redox conditions.

Boxplots (Figure 13C–E) and ANOVA results confirm statistically significant differences for Al₂O₃/TiO₂, Cr/Co, Cu, and Zr/Rb (p < 0.05), supporting their reliability as environmental proxies. In contrast, Sr/Cu (p = 0.172) and Corg/P (p = 0.599) do not significantly discriminate among lithofacies assemblages, and interpretations based on them should be considered qualitative (Supplementary Tables S3–S8).

5.2.1. Provenance

PC1 primarily reflects terrigenous influx and is interpreted together with provenance-sensitive ratios such as Al₂O₃/TiO₂ and Cr/Co. Al₂O₃/TiO₂ ratios are widely used to infer source lithology: 3–8 for mafic, 8–21 for intermediate, and 21–70 for felsic rocks [53]. Cr/Co, being relatively immobile during weathering, provides additional constraints [43,54,55,56].

Interbedded lithofacies assemblages, with intermediate Al₂O₃/TiO₂ ratios (8–21) and low Cr/Co values (Table 2, Figure 13C), indicate a dominant intermediate-to-mafic source, consistent with episodic volcaniclastic input during extensional faulting [80,81]. Laminated lithofacies assemblages and matrix lithofacies assemblages display felsic-dominated Al₂O₃/TiO₂ and higher Cr/Co, suggesting a stable felsic provenance and more intense chemical weathering under sustained lake deepening (Table 2, Figure 13C).

Although interbedded lithofacies assemblages formed under episodic, high-energy gravity flows, their lower PC1 scores likely reflect shorter depositional durations and dilution of geochemical signals by coarse detritus. In contrast, the matrix lithofacies assemblage, deposited in low-energy deep-water settings, records prolonged fine-grained sedimentation and higher cumulative terrigenous signatures on PC1 (Figure 13A).

5.2.2. Paleoclimate

Sr/Cu ratios are commonly used as paleoclimate indicators, with values >5 suggesting arid conditions and 1–5 indicating warm-humid conditions [57,58,59,60]. Although inter-assemblage differences are not statistically significant (p = 0.172), the overall stratigraphic trend remains informative. Laminated lithofacies assemblages exhibit higher and more variable Sr/Cu values, implying episodic aridity or enhanced evaporation, whereas interbedded lithofacies assemblages and matrix lithofacies assemblages show lower and relatively stable values, consistent with persistently humid conditions (Figure 13D).

These patterns are broadly consistent with regional palynological evidence, which shows an upward increase in evergreen broad-leaved pollen, a slight decline in mesophytic taxa, and a modest rise in hygrophytic forms—collectively indicating a mid-subtropical warm climate with semi-humid to semi-arid conditions and short-term fluctuations [24].

5.2.3. Paleoproductivity

TOC reflects organic matter accumulation, while P, Mo, and Cu serve as paleoproductivity indicators [67,68,69,82,83,84].

Interbedded lithofacies assemblages exhibit low TOC, P, Mo, and Cu, suggesting limited productivity combined with strong dilution by clastic input. Laminated lithofacies assemblages show elevated TOC, P, and Mo but moderate Cu, reflecting enhanced productivity and intermittent preservation under fluctuating bottom-water redox conditions. Matrix lithofacies assemblages yield high TOC, moderate P and Mo, and relatively high Cu, indicative of sustained productivity and efficient preservation in stratified deep-water settings (Table 2).

Correlation analysis shows strong positive TOC–Mo relationships, weaker correlation with P, and negative correlation with Cu (Figure 13F–H). These relationships suggest that organic carbon enrichment was primarily preservation-controlled, although productivity also contributed. Further integration of Rock-Eval pyrolysis, biomarker, and Fe-speciation data would help to quantitatively constrain the relative roles of productivity and preservation.

5.2.4. Paleoredox Conditions

PC2 captures variation in redox-sensitive and water-chemistry-related elements (Mo, Ba, Sr), serving as an integrated indicator of redox state and stratification intensity (Figure 13B). Laminated lithofacies assemblages exhibit higher and more variable PC2 values than interbedded lithofacies assemblages or matrix lithofacies assemblages (Figure 13A), indicating fluctuating redox conditions linked to transient stratification and periodic bottom-water oxygenation.

Vertically, the transition from interbedded to matrix lithofacies assemblages corresponds to declining depositional energy and increasing bottom-water anoxia. This trend collectively supports progressive deepening and enhanced redox stability, as evidenced by the upward decrease in turbidite frequency, reduced bioturbation, and absence of shallow-water indicators (Figure 9A–C).

Corg/P ratios indicate redox state (higher = more reducing) [69,85,86], whereas Zr/Rb reflects hydrodynamic energy. Low Zr/Rb indicates deeper, low-energy settings, while high values suggest shallower, high-energy conditions [63,64,65,87]. Although Corg/P alone shows no significant inter-assemblage difference (p = 0.599), its co-variation with TOC and Zr/Rb delineates redox gradients across facies. Zr/Rb negatively correlates with both Corg/P and TOC, while Corg/P and TOC are positively correlated (Figure 13I,J).

Interbedded lithofacies assemblages show high Zr/Rb and low Corg/P, consistent with shallow, oxic-suboxic, high-energy conditions. Laminated lithofacies assemblages display lower Zr/Rb and higher, variable Corg/P, reflecting alternating redox conditions. Matrix lithofacies assemblages exhibit low Zr/Rb and high Corg/P, indicative of deep, low-energy, anoxic settings favorable for organic matter accumulation (Figure 13E,I,J).

Available TOC and HI data provide a further basis for assessing the relative contributions of productivity and preservation. Clay-rich mudstones occupy the high HI–high TOC region, indicating both high productivity and favorable preservation; calcium-bearing mudstones show high HI but slightly lower TOC, reflecting consistently high productivity with preservation differences across subenvironments; ankerite-bearing sandstones fall into medium-low HI–low TOC, suggesting either limited productivity or poor preservation under oxidizing, high-energy conditions; siliceous siltstones display low to medium HI with variable TOC, implying TOC variability is mainly controlled by preservation; clay-rich siltstones and siliceous sandstones occupy the low HI–low TOC region, indicating limited productivity and preservation (Figure 13L).

5.2.5. Paleosalinity

In low-salinity waters, Sr and Ba occur as soluble bicarbonates. As salinity rises, Ba precipitates as BaSO₄, reducing its concentration relative to Sr [88]. Consequently, elevated Sr/Ba ratios indicate increasing water depth and salinity [89,90].

Interbedded lithofacies assemblages exhibit moderate Sr/Ba, reflecting shallow, low-salinity to weakly brackish waters under humid conditions. Laminated lithofacies assemblages show the highest and most variable Sr/Ba, indicating fluctuating salinity in mid-to-deep-lake settings, likely driven by short-term climatic and hydrological oscillations. Matrix lithofacies assemblages display low and stable Sr/Ba, consistent with freshwater deep-lake deposition under relatively stable hydrological conditions (Table 2). The positive correlation between Sr/Ba and Sr/Cu reinforces the role of climate in modulating water salinity and lithofacies development (Figure 13K).

5.3. Depositional Model of the L–LS2

The tectonic evolution of the Beibu Gulf Basin was driven by interactions among the Pacific, Eurasian, and Indo–Australian plates, resulting in an NE–NEE-trending fault system within the Weixi’nan Depression. This fault system intensified markedly during the second rifting climax (L–LS2), with extension rates of ~8% and peak subsidence rates up to 520 m/Ma [26,30]. Sustained tectonic subsidence and progressive lake deepening provided ample accommodation space for sediment accumulation. The L–LS2 corresponds to the initial phase of the post-EECO cooling transition (~53–49 Ma) [20,21,22]. Although this was a global cooling transition, the study period remained generally warm and humid, characterized by frequent short-term climatic fluctuations. This climatic setting, combined with ongoing tectonic subsidence, promoted vertical lithofacies differentiation through the early, middle, and late stages (Figure 14). Based on these observations, a conceptual depositional model is proposed to illustrate the coupled effects of tectonics and climate on lithofacies evolution.

During the early stage, as the basin initiated this phase of strong extension [26], it was characterized by interbedded lithofacies assemblage. Although fault-controlled subsidence had not yet reached full compensation, the absence of wave-generated or shallow-water structures (e.g., hummocky cross-stratification, oscillatory ripples, desiccation cracks) indicates deposition in an overall deep-lake setting (Figure 9A). Fault activity and frequent volcanism [23,25,31] supplied coarse detritus transported by high-density turbidity currents, forming alternations of sandstone and clay-rich mudstone layers (Figure 9 and Figure 11). Gravity-flow structures such as sedimentary deformation (Figure 5A,B), load structures (Figure 6A), flame structures (Figure 6D), normal grading (Figure 7A,B), and basal scour (Figure 8A,B) record rapid deposition under high-energy conditions [74,75,76]. Geochemical indicators (high Zr/Rb, high Sr/Cu, and low Sr/Ba) suggest slightly shallower and more oxic conditions relative to later stages, developed under a warm–humid climate with limited seasonal evaporation. Low Mo concentrations and TOC values (<4.5%) reflect nutrient limitation and subdued algal productivity, whereas moderate Corg/P ratios indicate intermittent oxygenation and partial degradation of organic matter (Figure 10 and Figure 14A).

In the middle stage, the lithofacies assemblage is dominated by laminated types, with a small amount of matrix lithofacies assemblage also present. Sustained high subsidence rates led to rapid basin deepening, producing a low-energy, deep-water depositional environment. Vertical variations (Figure 9B and Figure 10), including decreasing Zr/Rb ratios and declining turbidite frequency, indicate a progressive reduction in hydrodynamic energy and continued deepening. Riverine input delivered abundant sediments and nutrients under a persistently warm–humid yet increasingly variable climate. Pronounced fluctuations in Sr/Cu ratios (Figure 10) indicate alternating humid and semi-arid intervals that enhanced chemical weathering and nutrient fluxes (P, Cu, Mo) [42,43], stimulating high organic productivity (TOC > 5%). Climatic oscillations also induced salinity shifts (highly variable Sr/Ba), promoting migration of the redox boundary and leading to the interbedding of calcium-bearing and clay-rich mudstones (Figure 10, Figure 11 and Figure 14B). The presence of oriented ostracod assemblages, localized bioturbation (Figure 4D–H), and Sr/Ba variability collectively suggest episodic oxic intervals within an otherwise stratified and reducing water column. Overall, this stage marks the peak of organic productivity and facies diversification, jointly controlled by high subsidence and climate variability.

In the late stage, a matrix lithofacies assemblage developed. The climate remained warm and humid, accompanied by continued tectonic activity and progressive lake deepening, as reflected by low Zr/Rb values. The upsection disappearance of turbidites (Figure 9C and Figure 10) records a further decline in depositional energy, consistent with ongoing basin deepening likely caused by the combined effects of continued tectonic subsidence and rising lake level. Low-energy suspension settling dominated, producing thick, laterally continuous clay-rich mudstones (Figure 3 and Figure 11). Although subsidence remained pronounced, rapid deepening created an underfilled basin, where sediment supply lagged behind accommodation generation, limiting terrestrial input and promoting fine-grained organic matter accumulation [26]. Strongly reducing conditions (high Corg/P) and persistent anoxia suppressed bioturbation [11], facilitating efficient organic preservation, as evidenced by TOC ≈ 5% and abundant pyrite nodules and algal cysts (Figure 3). Increased felsic input (high Al₂O₃/TiO₂) and low Sr/Ba ratios indicate persistent freshwater conditions, further enhancing organic carbon burial [42]. Thus, lithofacies evolution during this stage was primarily governed by water-depth fluctuations and redox stability, reflecting the combined influence of tectonic subsidence and climatic stabilization (Figure 14C).

5.4. Implications for Shale Oil Exploration

Based on the heterogeneity of lithofacies assemblages, paleoenvironmental controls, and depositional dynamics of the L–LS2, key implications for shale oil occurrence and exploration can be summarized. Source and reservoir quality are primarily governed by the interplay among depositional texture, redox state, and subsequent diagenetic modification, which together control pore structure and hydrocarbon storage potential.

5.4.1. Implications for the Source Rock

The interbedded lithofacies assemblage is characterized by high-energy gravity flows and abundant coarse-grained clastic input, which results in generally low TOC values (<4.5%). These conditions indicate that organic matter preservation was limited under shallow-water and oxidizing conditions, yielding relatively poor hydrocarbon generation potential. In contrast, the laminated lithofacies assemblage developed during the middle stage features alternating suspension and chemical sedimentation driven by climate fluctuations. TOC values commonly exceed 5% but display significant variability, reflecting both high primary productivity and partial loss of organic matter due to intermittent oxidation. The matrix lithofacies assemblage formed under deep-water, low-energy, and anoxic conditions exhibits TOC values of approximately 5%, suggesting a high-quality source rock. Although continuous fine-grained deposition may adversely affect reservoir properties, it strongly favors long-term organic matter preservation (Table 1 and Table 2).

5.4.2. Implications for the Reservoir

The interbedded lithofacies assemblage, although characterized by relatively low organic content, contains sandstone interlayers (siliceous sandstone, ankerite-bearing sandstone, and siliceous siltstone) that are relatively well preserved and retain intergranular pores (Figure 12D,F,H). These layers exhibit relatively high porosity and permeability (Figure 15), indicating favorable vertical juxtaposition for shale-oil accumulation and enabling local “source–reservoir” coupling. The laminated lithofacies assemblage, composed of alternating clay-rich mudstones and calcium-bearing mudstones with minor siltstone beds (Figure 11), may locally develop enhanced reservoir zones due to lake-level or redox fluctuations that generated permeability contrasts. In contrast, the matrix lithofacies assemblage is dominated by organic-matter pores and intercrystalline pyrite pores (Figure 12A,B), which constitute the principal storage space for shale oil [1,91,92]. However, continuous fine-grained sedimentation resulted in generally low porosity (2–5%) and permeability (~0.05 × 10⁻³ μm²), posing challenges to efficient hydrocarbon extraction. Overall, these lithofacies-related variations highlight the primary influence of depositional texture on reservoir quality.

Although depositional textures exerted the primary control on the initial pore architecture, the reservoir quality of the L–LS2 was further modified by diagenetic processes. Mechanical compaction markedly reduced primary intergranular pores in clay-rich mudstones and calcium-bearing mudstones, whereas authigenic ankerite cementation occluded much of the remaining pore space in ankerite-bearing sandstones, thereby locally degrading reservoir quality. Despite moderate TOC contents, the relatively low and vertically uniform thermal maturity (Ro ≈ 0.8) [41], owing to limited burial-depth variations within the thin L–LS2 (~40–80 m thick), hindered the formation of organic pores and limited secondary porosity development. As a result, the overall storage capacity is dominated by inorganic (clay- and pyrite-related) pores that are themselves prone to compaction. These diagenetic features are consistent with previous observations in fine-grained shale systems [93,94]. Nevertheless, the pronounced lithofacies-dependent variations in porosity and permeability indicate that depositional texture exerted the dominant control, with diagenetic modification acting as a secondary overprint.

5.5. Study Limitations and Future Directions

This study reconstructs the depositional and paleoenvironmental evolution of the L–LS2 based on a single core (Well A). Although this approach provides excellent vertical resolution, the lack of multi-well correlation and seismic constraints limits the assessment of lateral facies heterogeneity, making it difficult to distinguish between regional lake-level rise and local subsidence. In the absence of backstripping-derived subsidence rates and quantitative isotopic or palynological climate data, the proposed tectonic-climatic deepening model should therefore be regarded as a preliminary interpretive framework rather than a quantitative reconstruction.

Geochemical proxies such as Sr/Cu, Zr/Rb, Al₂O₃/TiO₂, and Corg/P provide qualitative insights but are affected by multiple environmental and diagenetic factors. PCA and ANOVA results indicate that Al₂O₃/TiO₂, Cr/Co, and Zr/Rb differ significantly among lithofacies assemblages, whereas Sr/Cu and Corg/P do not, and should thus be interpreted with caution. Sr/Cu-based paleoclimate inferences remain speculative, and redox interpretations derived from Corg/P are tentative due to the absence of direct proxies (e.g., Fe–speciation, Mo/U). The relative contributions of primary productivity versus preservation to TOC enrichment also remain uncertain, requiring additional geochemical, biomarker, and sedimentological indicators for better constraint.

Although this study provides a preliminary qualitative analysis of diagenetic processes, the quantitative impacts of cementation, clay mineral transformation, and dissolution on pore networks and reservoir quality remain poorly constrained. Future investigations combining SEM/EDS observations, isotopic analyses, and multi-well datasets could further clarify the timing, extent, and impact of diagenetic modification.

Future research should integrate multi-well sedimentological and geochemical datasets, high-resolution seismic stratigraphy, isotopic and biomarker analyses, and detailed diagenetic evaluation. Incorporating quantitative subsidence modeling and independent climate proxies will enable a more rigorous, basin-scale assessment of the tectonic–climatic model and provide a clearer understanding of the L–LS2 succession and its implications for shale oil exploration.

6. Conclusions

(1) The L–LS2 in the Weixi’nan Depression preserves a complete record of lacustrine shale deposition during the Early Middle Eocene, under the combined influence of active tectonics and climatic transition. Based on core observations, thin-section analysis, and mineralogical data, six lithofacies were identified and grouped into three lithofacies assemblages: interbedded lithofacies assemblage, laminated lithofacies assemblage, and matrix lithofacies assemblage. Each lithofacies assemblage reflects distinct depositional processes and paleoenvironmental settings.

(2) The L–LS2 was deposited during a period of lacustrine expansion. In the early stage, the interbedded lithofacies assemblage dominates, formed through alternating high- and low-density turbidity flows and suspended sediments, resulting in interbedded siltstone/sandstone and clay-rich mudstone. This reflects a relatively shallower, high-energy depositional setting with frequent tectonic and volcanic activity, limited nutrient supply, and poor organic matter preservation. In the middle stage, deposition shifted to a laminated lithofacies assemblage, indicating alternating chemical and mechanical sedimentation driven by relative water-depth changes and stratification of the water column. Enhanced nutrient influx boosted primary productivity, while increased climatic variability caused water column stratification and salinity changes, leading to frequent redox boundary shifts and intermittent oxygenation. These changes, driven by the interaction of tectonics and climate, are critical in controlling redox conditions and organic matter preservation by modulating the position of the oxic–anoxic interface. In the late stage, the matrix lithofacies assemblage became dominant, representing continuous suspension settling of fine particles in a deep-water, low-energy, and anoxic environment. Under strongly reducing conditions, organic matter was effectively concentrated and well preserved.

(3) This depositional evolution clearly highlights the dual control of tectonics and climate on the sedimentary environment of lacustrine shales. Tectonic subsidence governed the basin architecture and overall water depth, while climate fluctuations influenced weathering intensity, nutrient supply, and salinity, thereby affecting sedimentary processes and biological productivity. Minor water-depth changes, driven by the interaction of tectonics and climate, are critical in controlling redox conditions and organic matter preservation. The ankerite in the ankerite-bearing sandstone represents post-sedimentary authigenic carbonate, which fills intergranular pores, reducing reservoir quality and potentially influencing hydrocarbon storage and migration. From an unconventional hydrocarbon exploration perspective, the L–LS2 constitutes a high-quality source rock. In the early-stage interbedded lithofacies assemblage, siltstone and sandstone layers with favorable reservoir properties and close proximity to source rocks offer potential exploration targets. The laminated lithofacies assemblage of the middle stage may form localized reservoirs. Although the matrix lithofacies assemblage of the late stage is a high-quality source rock, its reservoir quality is generally poor.