Rare-Earth Element Geochemistry for the Characterization of Sedimentary Environment and Provenance: A Case Study of the Eocene Liushagang Formation, Weixi’nan Sag, Beibuwan Basin, China

Abstract

This study investigates the rare-earth element (REE) geochemistry of twenty-nine clastic rock samples from the Paleogene Liushagang Formation in the Weixi’nan Sag. The primary objectives were to quantitatively evaluate the depositional paleoenvironment, determine the provenance lithology, and constrain the tectonic setting of the source area. Results reveal distinct chondrite-normalized REE distribution patterns characterized by light REE (LREE) enrichment, relatively flat heavy REE (HREE) segments, and pronounced negative Eu anomalies. The cerium anomaly index (Ce_anom, normalized to the North American Shale Composite) ranges from −0.06 to 0.00, implying broadly suboxic to anoxic-reducing conditions in the water column during deposition. The chondrite-normalized (La/Yb)_N ratio, utilized as a proxy for relative depositional residence time, decreases stratigraphically from member 3 to member 1, reflecting a transition to shorter residence times and higher relative sedimentation rates. Laterally, (La/Yb)_N increases toward the basin center, accurately recording progressively lower sedimentation rates basinward. Provenance analysis indicates that the sediments were predominantly derived from felsic igneous rocks of the upper continental crust. Spatially, the northern steep-slope belt reflects a uniform source, whereas the southern gentle-slope belt and the Weixi’nan low-uplift periphery record multisource mixed inputs. Finally, tectonic discrimination reveals an “active continental margin” affinity. This geochemical signature represents the inherited tectonic environment of the Mesozoic parent rocks in the surrounding source uplifts, rather than the Cenozoic extensional rift setting of the Weixi’nan Sag itself.

1. Introduction

Rare-earth elements (REEs) are robust tracers of sedimentary provenance, tectonic setting, and paleoenvironmental conditions owing to their coherent geochemical behavior and low mobility during weathering, transport, and diagenesis [1,2,3,4]. Fine-grained sedimentary rocks, such as mudstones, are particularly reliable archives of these primary signatures because they are less affected by sorting and burial diagenesis compared to coarser clastics [3,5]. Various REE parameters, including total REE abundance (ΣREE), ΣLREE/ΣHREE ratios, chondrite-normalized patterns, and specific anomalies (e.g., δEu, δCe, and (La/Yb)_N), are widely utilized to constrain source-rock lithology, evaluate paleoredox conditions, and estimate relative sedimentation rates [6,7,8,9].

The Paleogene Liushagang Formation in the Weixi’nan Sag is a prolific hydrocarbon-bearing sequence and a primary focus of exploration in the Beibuwan Basin. While previous studies have established the structural framework, sequence stratigraphy, and sand-body distribution [10,11,12,13,14,15,16,17], the basin-scale tectono-sedimentary setting and provenance characteristics remain poorly constrained. Crucially, the systematic application of REE geochemistry to decode the specific depositional environments and source-area properties of this formation has not yet been reported, which hinders a comprehensive understanding of the basin’s “source-to-sink” system.

To address this gap, this study presents a quantitative REE geochemical analysis of mudstones from the Liushagang Formation in the Weixi’nan Sag. By integrating REE distribution patterns and diagnostic geochemical indices, we aim to (1) determine the paleoenvironmental conditions and relative sedimentation rates; (2) discriminate the provenance lithology; and (3) clarify the tectonic setting of the source area. These findings will provide essential geological constraints for understanding the evolutionary history of the Weixi’nan Sag.

2. Geological Setting

The Weixi’nan Sag lies within the northern depression of the Beibuwan Basin. Its northwestern margin is bounded by the Weixi’nan fault, the southeastern margin adjoins the Qixi uplift, and its southwestern side connects to the Haizhong Sag, separated by the Weixi’nan low uplift. The exploration area of the study region is about 2.3 × 10³ km². Three approximately NE–SW-trending normal faults (No. 1 Fault to No. 3 Fault in Figure 1) are developed within the sag; their activity governs the overall structural framework and gives rise to a characteristic half-graben geometry, with a steep northwestern slope and a gentle southeastern slope.

The Weixi’nan Sag exhibits a two-tier structural architecture (i.e., rift beneath and a sag above), with Cenozoic evolution partitioned into a rifting stage and a post-rifting stage [18,19]. The rifting stage comprises an initial rift phase, a peak rift phase, and a waning phase, during which the Changliu, Liushagang, and Weizhou Formations were deposited successively. The post-rifting stage is characterized by regional subsidence, during which the Xiayang, Jiaowei, Dengloujiao, and Wanglougang Formations accumulated (Figure 2). During deposition of the Liushagang Formation, the basin was in the second Eocene rifting episode and underwent expansion, peak rifting, and subsequent waning [20,21]. From base to top, the Liushagang Formation is divided into member 3 (El₃), member 2 (El₂), and member 1 (El₁). El₃ corresponds to the initial expansion of the lacustrine basin and is dominated by fan-delta deposits in shore-shallow-lake to middle-deep-lake settings. El₂ represents peak rifting and comprises very thick middle-deep-lake mudstone and oil shale; it forms the principal source-rock interval of the Weixi’nan Sag. During El₁ time, the lake progressively shrank, and deposition was characterized by shore-shallow-lake mudstone interbedded with fan-delta sandstone.

3. Materials and Methods

For the present study, twenty-nine samples were collected from seventeen wells in the Weixi’nan Sag to allow us to characterize the geochemical properties of the Liushagang Formation (Table S1 in the Supplementary Material). The sampled lithologies consist of mudstone, silty mudstone, and muddy siltstone. Due to the limited availability of continuous cores, these samples were strategically distributed across multiple wells to maximize regional representativeness. This strategy ensures comprehensive coverage across diverse structural belts (i.e., the northern steep-slope, southern gentle-slope, and low-uplift periphery) and stratigraphic members (El₁–El₃), thereby minimizing sampling bias and providing a reliable, basin-wide three-dimensional perspective of the geochemical variations.

All analyses were conducted at the State Key Laboratory of Geological Processes and Mineral Resources, China University of Geosciences (Wuhan). Prior to analysis, all samples were crushed and ground to 200 mesh and stored in sealed bags. The digestion and measurement procedures were as follows: samples were oven-dried at 105 °C for 12 h; 50 ± 1 mg of powdered material was weighed into Teflon crucibles, moistened with 1–2 drops of high-purity water, and then 1.5 mL high-purity HNO₃ and 1.5 mL high-purity HF were added sequentially. The Teflon crucibles were placed in steel jackets, sealed, and heated at 190 ± 5 °C for >48 h. After cooling, lids were removed and the solutions were evaporated to dryness on a hotplate at 140 °C; 1 mL HNO₃ was added and again evaporated to dryness (ensuring no residual liquid on the crucible walls). Next, 3 mL of 30% HNO₃ was added, the crucibles were resealed in steel jackets and heated at 190 ± 5 °C for >12 h. The resulting solutions were transferred to polyethylene bottles and diluted with 2% HNO₃ to a final mass of 100 g (overall dilution factor 2000), then analyzed by ICP–MS. Certified reference materials (AGV-2, BHVO-2, BCR-2, GSP-2, and RGM-2) and procedural blanks showed good linearity for REEs. Relative analytical uncertainty was generally <3% and only rarely exceeded 10%, indicating reliable results. Analytical data are reported in Table S1 of the Supplementary Material.

4. Results

4.1. REE Concentrations and Characteristics

Analytical results (Table S2 in the Supplementary Material) show that the total REE abundances (∑REE) of the Liushagang samples from the Weixi’nan Sag range from 52.74 to 328.54 μg·g⁻¹, with a mean of 200.51 μg·g⁻¹. By subarea, the northern steep-slope belt yielded 52.74~328.54 μg·g⁻¹ (mean 202.80 μg·g⁻¹); the southern gentle-slope belt yielded 96.87~266.13 μg·g⁻¹ (mean 193.88 μg·g⁻¹); and samples from the periphery of the Weixi’nan low uplift yielded 108.49~312.04 μg·g⁻¹ (mean 200.15 μg·g⁻¹). Overall, the total REE content of the Liushagang samples is clearly higher than the average value for the continental crust (146.4 μg·g⁻¹) and slightly higher than for the North American Shale Composite (NASC: 173.21 μg·g⁻¹) [22,23].

The ratios ∑LREE/∑HREE and (La/Yb)_N quantify the degree of REE fractionation and, to some extent, inform on provenance [24]. For a given lithology, larger values indicate stronger LREE–HREE fractionation, i.e., relative enrichment of LREEs and depletion of HREEs. Here, ∑LREE/∑HREE is the ratio of total LREE to total HREE concentrations. For the Liushagang samples from the Weixi’nan Sag, ∑LREE/∑HREE ranges from 3.95 to 10.13 (mean 7.58). By subarea, the northern steep-slope belt yielded 3.95~10.13 (mean 7.65); the southern gentle-slope belt, 5.75~8.46 (mean 7.64); and the periphery of the Weixi’nan low uplift, 4.57~9.38 (mean 7.45). Overall, inter-belt differences are small, and values are comparable to the North American Shale Composite (NASC) average (7.50; from Haskin et al. [22]), indicating relative LREE enrichment and HREE depletion. (La/Yb)_N denotes the La/Yb ratio normalized to NASC. For the Liushagang Formation, (La/Yb)_N spans 0.61~1.68 (mean 1.14). By subarea, the northern steep-slope belt shows 0.61~1.68 (mean 1.15); the southern gentle-slope belt, 1.07~1.26 (mean 1.17); and the periphery of the Weixi’nan low uplift, 0.07~1.50 (mean 1.10). On the whole, (La/Yb)_N values are slightly higher than the NASC benchmark (1.00), consistent with modest LREE enrichment [22].

(La/Sm)_N and (Gd/Yb)_N are key parameters that characterize intra-group fractionation among LREEs and HREEs, respectively; larger ratios indicate more pronounced fractionation among LREEs (or HREEs) during transfer from the source area to the depositional basin [25,26]. For the Liushagang Formation in the Weixi’nan Sag, (La/Sm)_N ranges from 0.76 to 1.21 (mean 1.03), indicating minimal fractionation within the LREEs. (Gd/Yb)_N ranges from 0.77 to 1.70 (mean 1.11), likewise implying weak fractionation among the HREEs.

Post-Archean Australian Shale (PAAS) and the average upper continental crust (UCC) are characterized by LREE enrichment, relatively uniform HREE content, and a negative Eu anomaly, and are commonly used to represent upper-crustal REE signatures [27]. For the Liushagang Formation in the Weixi’nan Sag, δEu ranges from 0.56 to 1.18 with a mean of 0.79. By subarea, the northern steep-slope belt recorded 0.70~1.18 (mean 0.82); the southern gentle-slope belt, 0.71~0.86 (mean 0.82); and the periphery of the Weixi’nan low uplift, 0.56~0.84 (mean 0.72). Overall, aside from a few samples with δEu slightly >1.00 or ≈1.00, most values are clearly below the North American Shale Composite benchmark (δEu = 1.00) and slightly higher than PAAS and UCC (δEu ≈ 0.72 and 0.66, respectively), indicating an overall negative Eu anomaly across the study area.

Diagenetic overprinting can remobilize and redistribute REEs, thereby modifying primary geochemical signatures [28]. Under intense diagenesis, REE concentrations may become enriched and ΣREE can increase markedly; this, coupled with the effects on the mobility of Ce and Eu, may attenuate the Ce anomaly—reducing or even reversing a negative anomaly—and/or producing a pronounced negative Eu anomaly. Correlation analysis among REE parameters (e.g., δCe, δEu, and ΣREE) provides a means to assess the extent to which diagenesis has altered the primary signal. Weak or absent correlations indicate limited diagenetic influence and, hence, infers that REE-based paleoenvironmental inferences remain robust. As shown in Figure 3, Liushagang samples from the Weixi’nan Sag exhibit no obvious correlations between δCe and δEu or between δCe and ΣREE, implying only minor diagenetic modification.

4.2. REE Distribution Patterns

REE distribution patterns provide an intuitive view of REE geochemical characteristics [29]. Using the chondrite normalization values of Taylor et al. [27], Liushagang samples from the Weixi’nan Sag were normalized and plotted separately for the northern steep-slope belt, the southern gentle-slope belt, and the periphery of the Weixi’nan low uplift. The patterns exhibit three consistent features: (1) The three subareas show broadly similar patterns: curves have a steep LREE segment (right-inclined) and a relatively flat HREE segment, indicating pronounced LREE enrichment and HREE depletion. (2) A distinct “V-shaped” feature at Eu is present, with Eu displaying an overall negative anomaly. (3) Curves are quasi-parallel across subareas, implying broadly synchronous variation in REE abundances and suggesting internally consistent provenance characteristics within the study area (Figure 4).

5. Discussion

Rare-earth elements (REEs), owing to their distinctive geochemical properties and systematic fractionation under varying redox conditions, have become indispensable geochemical tracers in paleo-depositional environment analysis.

5.1. Depositional Environment Analysis

Cerium (Ce) is the only REE that undergoes valence change under surficial redox conditions (Ce³⁺ ↔ Ce⁴⁺). In oxidizing waters, soluble Ce³⁺ is readily oxidized to the sparingly soluble Ce⁴⁺ and scavenged onto Mn/Fe oxyhydroxide particles, removing Ce from the solution and generating a deficit relative to neighboring REEs in the water column and in chemically precipitated sediments (negative Ce anomaly, Ce/Ce* < 1). Conversely, in reducing waters, Ce occurs mainly as soluble Ce³⁺, is less efficiently removed, and may even be locally enriched, yielding a positive anomaly (Ce/Ce* > 1). Accordingly, the Ce anomaly is a sensitive indicator of the paleo-redox state [30,31,32].

The cerium anomaly index (Ceanom) quantifies the magnitude of the anomaly using North American Shale Composite (NASC)–normalized values and is calculated as (from Elderfield and Greaves [33]):

$${Ce}_{anom}=\mathrm{l}\mathrm{g}\left({\displaystyle \frac{3{Ce}_N}{2{La}_N+{Nd}_N}}\right)$$

(1)

Using NASC as the reference, Ce_anom < −0.10 denotes Ce depletion and indicates oxic conditions, whereas Ce_anom > −0.10 denotes Ce enrichment and indicates anoxic-reducing conditions [32]. For the study area, NASC-normalized values yield Ce_anom = −0.06 to 0.00 (mean −0.03;

Table 1. Statistical summary of key rare-earth element (REE) parameters for the Liushagang Formation samples across different structural belts in the Weixi’nan Sag.

Structural Belt	No. of Samples (n)	ΣREE (μg·g−1)	ΣLREE/ ΣHREE	δEu	δCe	(La/Yb)N
Northern steep-slope belt	14	157.68–328.54 (222.87)	3.95–10.13 (7.74)	0.70–0.87 (0.80)	0.91–0.96 (0.93)	0.61–1.68 (1.17)
Southern gentle-slope belt	5	96.87–266.13 (193.88)	5.75–8.46 (7.64)	0.71–0.86 (0.82)	0.87–0.94 (0.91)	1.07–1.26 (1.17)
Weixi’nan low-uplift periphery	10	108.49–312.04 (200.15)	4.57–9.38 (7.45)	0.56–0.84 (0.72)	0.90–0.96 (0.92)	0.70–1.50 (1.11)
Overall (Total)	29	96.87–328.54 (210.04)	3.95–10.13 (7.62)	0.56–0.87 0.78	0.87–0.96 (0.92)	0.61–1.68 (1.15)

ΣREE = ΣLREE + ΣHREE, ΣLREE = La + Ce + Pr + Nd + Sm + Eu, ΣHREE = Gd + Tb + Dy + Ho + Er + Tm + Yb + Lu, δEu = Eu_N/(Sm_N × Gd_N)^1/2, and δCe = Ce_N/(La_N × Pr_N)^1/2. _N represents values normalized to the North American Shale Composite (NASC). (La/Yb)_N values are normalized to chondrite.

), indicating slight Ce enrichment, which may reflect a suboxic to weakly reducing environment, or fluctuating redox conditions during deposition of the Liushagang Formation in the Weixi’nan Sag. Furthermore, in bulk-rock clastic sediments, the primary authigenic Ce anomaly from the water column can be partially masked by the uniform REE patterns of terrigenous detrital inputs [34]. Therefore, in the absence of independent redox-sensitive trace metal proxies (such as Mo, U, and V enrichments) or Fe speciation data, an interpretation of the precise redox state based solely on Ce anomalies carries some uncertainty. Based on the current REE parameters, we conservatively interpret the depositional environment of the Liushagang Formation as a suboxic to weakly reducing setting rather than a strictly anoxic one.

5.2. Relative Depositional Residence Time and Sedimentation Rate

As depositional conditions fluctuate, REEs undergo varying degrees of fractionation, particularly between light and heavy REEs (LREEs and HREEs) [3,8,26]. In lacustrine systems, clay-sized detritus and other suspended particles serve as primary carriers for both REEs and organic matter [35]. The extent of REE fractionation is closely coupled to the settling dynamics of these particles. Consequently, the chondrite-normalized (La/Yb)_N ratio, which defines the slope of the REE distribution pattern, acts as a reliable indicator of the residence time of suspended particles in the water column [7,33]. When the residence time is short (i.e., rapid deposition), the interaction between REEs, ambient pore waters, and clay minerals is restricted. This limited interaction prevents extensive fractionation, resulting in (La/Yb)_N values closer to 1 and relatively flat REE patterns. Conversely, prolonged suspension allows for extensive adsorption onto clays and complexation with organic matter, driving strong fractionation characterized by LREE enrichment and HREE depletion. Under such slow settling conditions, the (La/Yb)_N ratio increases significantly above 1, producing a steeper, right-inclined distribution pattern.

Although the (La/Yb)_N ratio has been widely utilized to evaluate sedimentation rates [21], it is critical to emphasize that this parameter is not exclusively controlled by depositional speed. The fractionation between LREEs and HREEs can also be substantially modified by competing factors, including provenance composition, grain-size sorting, clay mineralogy, and sediment recycling. Therefore, rather than representing absolute sedimentation rates, the (La/Yb)_N values in this study are more accurately interpreted as reflections of relative depositional residence times and hydrodynamic energy conditions. Given that the provenance of the Liushagang Formation remained relatively stable throughout the depositional period (as established in Section 5.3), the observed variations in (La/Yb)_N primarily record fluctuations in the water-column residence time prior to final burial. In this context, a lower (La/Yb)_N value indicates a shorter residence time, corresponding to a higher relative sedimentation rate or higher-energy hydrodynamic conditions, which restricts the prolonged fractionation of REEs.

Spatial and stratigraphic analyses of the twenty-nine samples reveal distinct (La/Yb)_N trends across the Weixi’nan Sag. In both the northern steep-slope and southern gentle-slope belts, (La/Yb)_N values generally decrease upward from member El₃ to El₁. During El₃ deposition, (La/Yb)_N values reached up to 1.68 (mean: 1.23), whereas during El₁, they ranged from 0.61 to 1.50 (mean: 1.09). This upward decrease in (La/Yb)_N signifies a reduction in residence time, coinciding with the stratigraphic transition from El₃ fan-delta systems to El₁ gravity-flow deposits, as the latter are inherently characterized by much more rapid, episodic accumulation [13]. Interestingly, an opposite temporal trend is observed around the periphery of the Weixi’nan low uplift, where El₃ values are generally lower than those of El₁. Laterally across the basin, (La/Yb)_N values exhibit a consistent increasing trend from the basin margins toward the depocenter. This spatial distribution perfectly captures the progressive basinward deceleration of sediment transport and the corresponding decline in relative sedimentation rates.

5.3. Provenance Characteristics

REEs are highly immobile under normal depositional and diagenetic conditions, making their distribution patterns (e.g., ∑LREE/∑HREE ratios and δEu) robust geochemical fingerprints of sediment provenance [1,2,4,36].

Chondrite-normalized REE distribution patterns of the Liushagang Formation samples (Figure 4) exhibit LREE enrichment, flat HREE segments, and negative Eu anomalies, closely mirroring the upper continental crust (UCC) [1,36]. High ∑LREE/∑HREE ratios (mean 7.62 and range 3.95–10.13) and pronounced negative δEu values (mean 0.78 and range 0.56–0.87) further point to a chiefly felsic source, as plagioclase fractionation typically produces Eu depletion in evolved magmas [1,8]. While relying solely on δEu carries uncertainty, our geochemical inferences are highly consistent with the regional geology. Borehole data confirm that the primary sediment source areas for the Weixi’nan Sag (e.g., the Weixi’nan low uplift and Qixi uplift) consist extensively of Mesozoic granitic, intermediate-acidic volcanic, and metamorphic rocks. Thus, the negative δEu signature effectively corroborates the felsic nature of this regional “source-to-sink” supply.

Spatially, samples from the northern steep-slope belt display highly parallel REE curves with tightly clustered LREE content, indicating a stable, homogeneous provenance. Conversely, the southern gentle-slope belt and the Weixi’nan low-uplift periphery exhibit dispersed LREE content and lower inter-sample similarity, suggesting mixed-source inputs.

The La/Yb–∑REE diagram (Figure 5) supports this spatial variation. Northern steep-slope data points concentrate mainly in the alkaline basalt–granite–sedimentary and granite–sedimentary fields, implying a uniform source dominated by igneous rocks. In contrast, samples from the southern gentle-slope and low-uplift periphery are scattered across multiple fields (including sedimentary- and calcareous-mudstone), confirming multisource input. This mixed-source signal at the low-uplift periphery aligns perfectly with its stratigraphic context. Most samples here belong to El₁ gravity-flow deposits. During the earlier El₂ deposition, intense activity on the No. 3 Fault and lake-level rise submerged the low uplift; by El₁ time, it was re-exposed to erosion. Consequently, these El₁ deposits record a mixture of autochthonous materials eroded from the uplift itself and allochthonous sediments transported from adjacent regions.

5.4. Tectonic Setting of the Provenance Area

Because REEs are relatively stable during weathering, erosion, transport, and deposition, they effectively preserve the geochemical fingerprints of their source rocks. Provenance formed in different tectonic settings—such as oceanic island arcs, continental island arcs, active continental margins, and passive continental margins—displays distinctive REE compositions, total abundances, and chondrite-normalized distribution patterns. By comparing sedimentary REE patterns with these archetypal signatures, one can infer the tectonic position of the basin and the attributes of its source area [38].

REE parameters of greywacke from contrasting tectonic settings provide a comparative yardstick for diagnosing provenance. Because mudstone typically has ΣREE values ~20% higher than coeval greywacke within the same tectonic regime due to the concentration of REEs in the clay fraction [4], the REE parameters for mudstone samples were recalculated by dividing ΣREE by 1.2. This normalization allows for an unbiased cross-comparison with the greywacke-based reference fields in the tectonic discrimination diagrams (

Table 2. REE data for greywacke from different tectonic settings (from Bhatia [4]).

Tectonic Setting	Provenance Type	REE Content			La/Yb	ΣLREE/ ΣHREE	δEu
		La	Ce	ΣREE
Oceanic Island Arc *	Undissected Magmatic Arc	8 ± 1.7	19 ± 3.7	58 ± 10	4.2 ± 1.3	3.8 ± 0.9	1.04 ± 0.11
Continental Island Arc *	Dissected Magmatic Arc	27 ± 4.5	59 ± 8.8	146 ± 20	11 ± 3.6	7.7 ± 1.7	0.79 ± 0.13
Active Continental Margin *	Uplifted Basement	37.00	78.00	186.00	12.60	9.10	0.60
Passive Continental Margin *	Cratonic Interior Tectonic Highlands	39.00	85.00	210.00	15.90	8.50	0.56
North steep-slope #		38.18	77.39	185.73	12.05	8.23	0.67
South gentle-slope #		32.98	66.79	161.57	12.08	8.26	0.68
Weixi’nan low-uplift periphery #		34.61	69.29	166.79	11.49	7.97	0.60

* Data from Bhatia [4]. δEu represents chondrite-normalized values. # Data are averages.

; Figure 6). Pattern comparison shows that REE curves from the study area resemble those of active continental margins, indicating that the provenance of the Liushagang Formation in the Weixi’nan Sag is most consistent with an active-margin tectonic setting.

Ce anomalies are commonly linked to basin-scale tectonic settings. Prior studies indicate that near spreading ridges Ce shows a pronounced negative anomaly (δCe ≈ 0.29); in open-ocean basins δCe indicates a moderate negative anomaly (≈0.55); whereas along continental margins the negative anomaly diminishes or even becomes positive, with δCe typically ranging from ~0.90 to 1.30 [39]. For the Liushagang Formation in the Weixi’nan Sag, δCe spans 0.87~0.96 with a mean of 0.92, consistent with the continental-margin affinity of the provenance.

Integrating La, Ce, La/Yb, ∑REE, ∑LREE/∑HREE, and δEu, we infer that the northern steep-slope belt reflects an active continental-margin provenance, whereas the southern gentle-slope belt and the periphery of the Weixi’nan low uplift record the mixed signatures of a continental island arc and an active continental margin—consistent with a heterogeneous (mixed-lithology) source.

When interpreting these geochemical signals, it is crucial to explicitly distinguish between the original tectonic setting of the source area (provenance) and the tectonic regime of the depositional basin itself. Geologically, the Beibuwan Basin is a Cenozoic extensional rift basin, specifically characterized as a dextral pull-apart basin controlled by pre-existing NE-trending large-scale strike-slip faults in the South China Block. During the deposition of the Paleogene Liushagang Formation, the basin experienced intense rifting. Controlled by fault activities, the surrounding uplifts, such as the Weixi’nan low uplift and the Qixi uplift, were rapidly uplifted and exposed, serving as the primary provenance areas within a typical “source-to-sink” system.

However, the REE tectonic discrimination diagrams (e.g., [4]) primarily indicate the tectonic setting in which the parent rocks were originally formed, rather than the subsequent basinal environment. Regional geological evidence shows that the basement and these surrounding uplifts are extensively composed of pre-Cenozoic metamorphic and magmatic rocks (mainly Mesozoic granitic and intermediate-acidic volcanic rocks). These ancient parent rocks were generated under an active continental margin setting associated with the subduction of the Paleo-Pacific plate beneath the South China Block during the Mesozoic. Therefore, the “active continental margin” geochemical signature observed in the Liushagang Formation clastic rocks is an inherited chemical memory from the eroded Mesozoic basement rocks in the provenance area, which perfectly reconciles the REE signals with the Cenozoic extensional rifting history of the Weixi’nan Sag.

6. Conclusions

Based on the geochemical analysis of rare-earth elements (REEs) in the samples of the Paleogene Liushagang Formation in the Weixi’nan Sag, the following key conclusions are drawn:

(1)

The clastic rocks exhibit pronounced light REE (LREE) enrichment, relatively flat heavy REE (HREE) segments, and distinct negative Eu anomalies. These geochemical signatures closely mirror the upper continental crust (UCC), indicating that the sediments were predominantly derived from felsic to intermediate-acidic igneous parent rocks.

(2)

The cerium anomalies (Ce_anom) are generally normal to slightly negative, indicating that the depositional water column during the Liushagang Formation period was primarily a suboxic to anoxic-reducing environment.

(3)

The chondrite-normalized (La/Yb)_N ratio effectively tracks the relative depositional residence time. Vertically, (La/Yb)_N values decrease from member El₃ upward to El₁, reflecting a transition to shorter residence times and higher relative sedimentation rates (driven by El₁ gravity flows). Laterally, (La/Yb)_N values increase from the basin margins toward the center, accurately recording progressively lower sedimentation rates basinward.

(4)

Spatial mapping of REE parameters reveals distinct sediment routing. The northern steep-slope belt reflects a uniform, stable provenance dominated by igneous rocks. Conversely, the southern gentle-slope belt and the Weixi’nan low-uplift periphery record multisource mixed inputs, heavily influenced by localized fault activities and topographic variations.

(5)

While REE discrimination diagrams point to an “active continental margin” affinity, this represents an inherited geochemical memory. It reflects the original tectonic environment of the Mesozoic parent rocks in the surrounding source areas (e.g., the Weixi’nan low uplift), rather than the Cenozoic extensional rift setting of the Weixi’nan Sag depositional basin itself.