Authigenic clay mineral constraints on spatiotemporal evolution of restricted, evaporitic conditions during deposition of the Ediacaran Doushantuo Formation

The carbonate-rich shelf facies of the Ediacaran Doushantuo Formation (South China) have produced a broad array of microfossil and isotope proxy records which underpin much of our understanding of environmental change and biospheric evolution during this key time period. Recent reports of locally abundant authigenic, Mg-rich saponite clay in the Yangtze Gorges Area hint at potentially widespread evaporitic conditions in a lagoonal setting. Despite the implications for interpretation of proxy records and the environmental setting of early metazoans, the spatiotemporal extent of restricted, evaporitic conditions across the Yangtze Block remains largely unexplored. Here we use mineralogical and petrographic techniques to document the spatial and strat-igraphic distribution of authigenic clay minerals across seven sites, representing a transect from shallow proximal shelf to deep basinal settings. Our results demonstrate the widespread and persistent occurrence of authigenic saponite in proximal shelf settings, whereas talc is identified at more distal shelf sites, consistent with Mg-clay authigenesis in an evaporitic lagoon with spatially variable detrital aluminosilicate input. Slope and basinal sites contain no Mg-clays and are instead characterized by abundant authigenic illite, consistent with an open marine setting. Stratigraphically, illite was found in the basal cap carbonate and uppermost Doushantuo Formation, with Mg-clays only present in the interval in-between, suggesting gradual development of an extensive carbonate rimmed lagoon over the Yangtze shelf that restricted seawater exchange, followed by stronger marine influence due to erosion of the carbonate rim and subsequent marine transgression. We conclude that restricted, evaporitic conditions were much more expansive than previously assumed, potentially providing favorable conditions for the Doushantuo Biota. Further, the widespread presence of authigenic clays across the shelf supports the case for enhanced reverse weathering before the advent of biosilicification.


Introduction
The Ediacaran Doushantuo Formation (DST, ca.635-551 Ma) of South China is an important archive of Earth System and biospheric evolution in the lead up to the Phanerozoic, having produced an array of key fossil and geochemical proxy records (Jiang et al., 2007;Knoll et al., 2006;Sahoo et al., 2012;Xiao et al., 2014).Paleogeographic reconstructions show that the DST was deposited widely across the Yangtze Block, from shallow shelf to deep basinal settings, but with substantial differences in lithology, geochemistry and paleontology across the region (Jiang et al., 2011(Jiang et al., , 2007;;Li et al., 2010;Liu et al., 2013;Wang et al., 2020;Xiao et al., 2014).In general, the shallow-water shelf settings are dominated by carbonates or interbedded shaly limestone/dolostone and calcareous shale, while the deep-water basinal settings show more shale intervals.
Geochemical data show major differences between shelf versus slope-basin settings.For example, slope-basin settings show consistently more negative carbonate δ 13 C (Cui et al., 2017;Jiang et al., 2007;Lu et al., 2013;Wang et al., 2020;Zhu et al., 2013) and pyrite δ 34 S signatures (Gao et al., 2021).Ferruginous-dominated conditions in slope-basin settings are distinct to the more commonly sulfidic conditions on the shelf (Li et al., 2010).Enrichment in redox sensitive elements (including Mo, Re, U, and V) also show distinct patterns (Bristow et al., 2009).The distinct characteristics of shelf versus basin locations have been widely discussed (Gao et al., 2021;Jiang et al., 2007;Wang et al., 2020) and suggest that proxy records from the DST record local conditions and thus may not always be representative of global seawater signatures (Jiang et al., 2011).Interestingly, eukaryotic fossils (e.g.acanthomorphic acritarchs) in the DST are mostly observed in shelf settings or at the shelf margin to slope transition, but more rarely found in slope to basin settings (Bowyer et al., 2017;Jiang et al., 2011;Xiao et al., 2014).This may be due to beneficial environmental conditions developing in shallow water settings, whereas conditions in the open Neoproterozoic ocean were more challenging due to widespread anoxia (Knauth, 2005) and nutrient deficiency (Cox et al., 2018;Reinhard et al., 2017).Alternatively, this distribution may reflect a taphonomic, preservational bias because of local enrichments of silica or phosphorous that are facilitated in shallow settings (Muscente et al., 2015;Slagter et al., 2021;Tarhan et al., 2016).
The current palaeogeographic model for the DST, based largely on sedimentological and stratigraphic evidence, infers deposition under open marine conditions during formation of the cap carbonate, followed by the gradual build-up of an outer shelf carbonate rim on a topographic high inherited from the Nantuo glaciation or possibly formed through syndepositional faulting.Shelf facies were subsequently deposited in a rimmed shelf lagoon setting, also described as an intrashelf basin (Jiang et al., 2011;Zhu et al., 2022).While this model is broadly accepted, key questions with major implications for the interpretation of geochemical proxy records and assessing environmental controls on biotic evolution remain, including: 1) the extent to which the lagoon was restricted versus open to exchange with the ocean, 2) the timing and cause of lagoon formation and demise, and 3) the geographic extent of the lagoon (Jiang et al., 2011).Recent work shows that the nature and distribution of authigenic clays can offer valuable additional constraints on the timing and extent of the lagoonal conditions in this region (Han et al., 2022a).
Abundant authigenic saponite as well as its burial diagenetic products, corrensite and chlorite, were found in member 2 (M2) of the DST in the Yangtze Gorges Area (YGA), which has been interpreted to record deposition in a restricted, evaporitic environment distinctly different from an open marine setting (Bristow et al., 2009;Han et al., 2022a).Saponite, a trioctahedral Mg-rich clay mineral (Mg-clay), can form by 1) hydrothermal alteration of (ultra)mafic igneous rocks, volcanic ash, siliceous dolostone or other clay minerals (Cuadros et al., 2013;Dill et al., 2011;Voigt et al., 2020), 2) pedogenic weathering of (ultra)mafic silicates under humid conditions (Lessovaia and Polekhovsky, 2009), and 3) authigenic precipitation from supersaturated waters at elevated pH (Akbulut and Kadir, 2003;Furquim et al., 2008;Milesi et al., 2019).Authigenic saponite formation requires high Si(OH) 4 and Mg 2+ concentrations as well as alkaline conditions, so that saponite authigenesis is most commonly associated with alkaline lake settings (Bristow et al., 2012(Bristow et al., , 2009) ) and is not known to occur in Phanerozoic open marine, non-hydrothermal settings (Han et al., 2022a).However, clay synthesis experiments and thermodynamic modelling illustrate that modest evaporation of already Si-rich Neoproterozoic seawater is likely to produce conditions exceeding the critical supersaturation required to precipitate authigenic Mg-clays including saponite, talc and sepiolite (Han et al., 2022a;Strauss and Tosca, 2020;Tosca et al., 2011), or their metastable precursors (e.g.kerolite) (Noack et al., 1989).Importantly, Mg-clays are absent in open marine, deep basinal settings during deposition of the DST and have also not been documented in other unambiguously open marine settings globally, indicating that a restricted, evaporitic lagoon setting with limited connection to the open ocean was required for authigenic Mg-clay formation.The unique occurrence of authigenic marine Mg-clays in restricted, evaporative settings during the Neoproterozoic (Han et al., 2022a;Strauss and Tosca, 2020;Tosca et al., 2011) therefore offers a novel proxy with which to probe the spatial and temporal evolution of the Yangtze lagoon.To date, Mg-clays (saponite) in the DST have only been documented in the YGA.
To further explore the spatio-temporal evolution of environmental conditions across the Yangtze Block, we use mineralogical and petrographic techniques to investigate the identity and origin of clays in representative sections spanning inner shelf to deep basin locations (Fig. 1), covering each of the main paleogeographic settings identified in previous studies (Jiang et al., 2011;Wang et al., 2020).We identify an assemblage of authigenic Mg-clays including saponite, talc, sepiolite and palygorksite, the first time these clays (except saponite) have been reported in the DST.We discuss the likely origins of these clays and use their spatial and stratigraphic distribution, together with independent geochemical constraints, to assess the duration and spatial extent of restricted, evaporitic conditions over the Yangtze shelf.Finally, we consider likely implications for the evolution and fossilization potential of early metazoan organisms.

Materials and Methods
Ninety-one samples collected from seven different sites were characterized in this study, forming a shelf to basin transect spanning the major depositional settings of the DST according to Jiang et al. (2011) (Fig. 1).The Sishang section (Lu et al., 2013) here interpreted as an inner shelf setting based on the presence of shallow water carbonate facies and largely positive carbon isotope values characteristic of shelf sections of the DST (Jiang et al., 2011); Tianjiayuanzi (Lu et al., 2012), Sixi (Bristow et al., 2009) and Qinglinkou sections (An et al., 2014) represent shelf lagoon settings; the Zhongling section (Cui et al., 2015) represents a shelf margin setting; the Fengtan section (Lu et al., 2013) represents a slope setting; and the Yuanjia section (Sahoo et al., 2012) represents a deep basin setting (Supplementary Table 1).
The DST is commonly subdivided into four lithostratigraphic members: Member 1 (M1) comprises a cap carbonate unit that sharply overlies the glacial diamictite of the Cryogenian Nantuo Formation; Member 2 (M2) comprises interbedded shale and dolostone with chert and phosphatic nodules; Member 3 (M3) is dominated by dolostone and bedded chert; and Member 4 (M4) appears as a black shale unit that is overlain by the upper Ediacaran Dengying Formation (Fig. 1).However, substantial regional variability means that sections which can be clearly separated into the distinct four lithological members are mostly located in the YGA (Jiang et al., 2011;McFadden et al., 2008).Correlation of the sections in this study therefore relies on a combination of lithostratigraphy and chemostratigraphy.The deposition of cap carbonate marks the beginning of the DST and the black shale interval of the M4 marks the end.The cap carbonate was observed in all sampled sections except the Sishang section, while the black shale marker beds were observed in the three sections in the YGA and the Fengtan section (Fig. 1).Two global negative carbon isotope excursions (EN1 and EN3, Jiang et al. (2007)) further constrain the interval between the cap carbonate (EN1) and upper M3 (EN3, Shuram excursion).Both excursions are present in out of 7 studied sections (Fig. 1), but EN1 and EN3 have not been recognized in the Sishang and Yuanjia sections, respectively.More clay-rich intervals were prioritized for this study and stratigraphic positions of the samples studied here are shown in Fig. 1c.Clay mineralogy for samples of M2 at Sixi and Qinglinkou sections has previously been reported (Han et al., 2022a(Han et al., , 2022b)), and is supplemented here with additional samples from M1 and M3.The M4 black shale was not sampled in the YGA sections for this study, and the upper M4-equivalent black shale interval is absent in sections at other locations.
X-ray diffraction (XRD) analyses of micronized bulk samples and oriented clay films were used to identify and quantify all minerals present, while illite polytype analyses (Grathoff, 1996) show illite origins for comparison to petrographic interpretations.Polished resin mounts of selected samples were analyzed using scanning electron microscopeenergy dispersive spectroscopy (SEM-EDS) mineral mapping (Thermofischer Maps Mineralogy Version 3.1) to identify clay origins.Mineral identification is achieved by matching individual EDS spectra to reference spectra collected on mineral standards, using an automated deconvolution algorithm for mixed-phase spectra and a list of mineral candidates ('mineral recipe') informed by sample lithology and independent XRD-based characterization.This approach produces quantitative mineralogical data with accuracy and precision comparable to quantitative XRD analysis, as well as high-resolution backscatter electron images with false-colour mineral map overlays which allow the origin of individual sedimentary components to be identified (Han et al., 2022b).We use diagnostic petrographic criteria, listed in Supplementary Table 2, to differentiate clays of three different origins: 1) authigenic clays, precipitating from pore fluids during early diagenesis; 2) clays of burial diagenetic origin; and 3) detrital clays, originating from weathering of the adjacent landmasses.The carbonate fraction of carbonate-rich samples from the Tianjiayuanzi (shelf lagoon), Zhongling (shelf margin) and Sishang (inner shelf) sections was analyzed for trace elements by ICP-MS (see extended methods in Supplementary Information).Jiang et al., (2011)) with locations of the investigated sections (red points).1: Sishang section.2: Yangtze Gorges Area, with Sixi, Qinglinkou and Tianjiayuanzi sections.3: Zhongling section.4: Fengtan section.5: Yuanjia section.c: Lithological logs and sample positions.Lithological logs are based on new field observations (for Sixi and Qinglinkou sections) and published studies (An et al., 2014;Bristow et al., 2009;Cui et al., 2015;Lu et al., 2013Lu et al., , 2012;;Sahoo et al., 2012).The numbers behind section names represent locations shown in b.The orange and green shading cross sections represent negative carbon isotope excursions in cap carbonate (EN1) and the upper DST (EN3), respectively, used here for chemostratigraphic correlation of the sections.NT: Nantuo Formation; DY: Dengying Formation; LCP: Liuchapo Formation.
S. Han et al.

Inner shelf setting
The inner shelf setting is represented by Sishang section which is located at the northern margin of the Yangtze Block (Fig. 1).It consists mainly of limestone and dolomitic limestone in which calcite is the predominant carbonate mineral, unlike the dolomite-dominated sections at other locations (Supplementary Table 3).
In terms of clay mineralogy, corrensite (i.e.regularly interstratified chlorite-saponite) was detected by XRD in lower interval sample SS107.1, confirming the presence of Mg-clay in this section (up to 5.3 wt %; Supplementary Fig. 1d).SEM-EDS mineral mapping identifies two samples which contain mixed layer chlorite-saponite with fine-grained, intergrown, cornflake-like texture (Fig. 2e-h).These Mg-clays are concentrated in large domains (> 100 μm in long axis), indicating a precompaction origin, and are associated with chemical sediments including authigenic carbonate and chert (Supplementary Table 2).

Shelf lagoon setting (Yangtze Gorges Area)
Sixi (SX), Qinglinkou (Qlk) and Tianjiayuanzi (TJ) sections represent the shelf lagoon setting of the YGA.The dominant clay mineral assemblage in all three sections changes from illite in M1 to saponite and chlorite in M2 before returning to illite in M3.Minor muscovite (< 3 wt %) is also observed (Fig. 3, Supplementary Table 4 and 5, Supplementary Fig. 1e-h).
Samples from M1 (the cap carbonate) are cherty dolostone (Sample JTA1.7), cherty limestone (Sample TJA2.8) or dolomitic limestone (Sample SX1).Clays within M1 are only observed in the Tianjiayuanzi section, which contains two classes of illite, both associated with authigenic quartz and dolomite, i.e. illite with curvy, flaky texture (Fig. 3g) and more fine-grained fluffy texture (Fig. 3h).We interpret illite in M1 to be mostly of authigenic origin based on a) close association with chemical sediments (i.e., chert and carbonate minerals), b) occurrence in domains that are largely free of detrital grains (Han et al., 2022a) and c) illite polytype analysis which shows the dominance of 1M d illite (Supplementary Table 8), suggestive of an authigenic origin, potentially via the burial alteration of an authigenic smectite precursor (Grathoff, 1996).
Two silica rich samples were found in the transition zone between M1 and M2 in the Qinglinkou section (Supplementary Table 4, Fig. 3e).One of these is a chert sample (Qlk2), while the other sample (Qlk3) comprises alternating laminae of fine-grained quartz and illite, with minor feldspar (Fig. 3e, f).Larger, platy illite flakes (Fig. 3f, red arrow) are of similar size to feldspar and quartz grains, and some grains are deformed by compaction at the contacts with brittle minerals, consistent with a detrital origin.Fibrous, pore-filling illite was also observed (Fig. 3f, yellow arrow).Fibrous illite is interpreted to be of authigenic origin, likely precipitating from pore water during early diagenesis, but may also be the product of feldspar alteration during later burial diagenesis.Polytype analysis shows that Sample Qlk 3 contains 25% detrital and 75% authigenic or diagenetic illite (Supplementary Table 8).
Above the basal transitional zone, the main part of M2 in the YGA is interbedded dolostone and dolomitic shale, associated with chert and phosphatic nodules.Here the clay mineralogy of M2 becomes saponitedominated, as also reported in previous studies (Bristow et al., 2009;Han et al., 2022a).Saponite constitutes up to 31 wt% of bulk samples (Supplementary Table 4).This saponite is interpreted to be of authigenic origin, because (1) it is concentrated in large pore filling domains or bedding-parallel laminae, both showing evidence of compaction, (2) the fragile grains and aggregates cannot withstand transportation, and are mostly associated with carbonate minerals rather than detrital clasts, ruling out a detrital origin; and (3) it is closely associated or intergrown with authigenic dolomite, suggesting a chemical precipitation origin.(Fig. 3c, d, Supplementary Fig. 2) (Han et al., 2022a).
M3 is comprised mainly of dolostone and largely clay-free.XRD analysis detected illite in 2 out of 8 samples in the upper part of the Tianjiayuanzi section (Supplementary Table 5) and only one sample (SXA0.2) out of 8 analyzed by SEM-EDS mineral mapping contains minor illite as pore fill between dolomicrite crystals (Fig. 3a).The illite is fine-grained with a fibrous texture (Fig. 3b, yellow arrow), indicating an authigenic (burial diagenetic) origin.Long flake illite (Fig. 3b, red arrow) and other detrital components such as quartz and feldspar silt are very rare, suggesting negligible detrital inputs.
In summary, the shelf lagoon setting shows an evolving clay mineral assemblage which changes from illite-dominated in M1 and the basal M2 to saponite-dominated in most of M2, returning to illite-dominated in M3 (Fig. 4).Except for the abundant sheet-like illite in lower-most M2 which is probably of detrital origin, other abundant clay phases show intergrown, curvy to wavy, fibrous to flaky fragile morphologies typical of in situ clay mineral phases (i.e.authigenic or diagenetic origins), and large clay mineral domains commonly display compactioninduced deformation, consistent with an earliest diagenetic, authigenic origin.
The single sample in M1 (ZL255.7,Fig. 1) is a quartz and calcite cemented dolostone, containing minor feldspar and illite that is present mainly within quartz-cemented layers (Fig. 5h; Supplementary Fig. 3).Illite is the dominant clay, filling the intercrystal pore space of dolomite, displaying fluffy texture and associated with quartz and calcite cements (Fig. 5h).Illite shows compaction deformation in contact with adjacent brittle minerals, demonstrating a pre-compaction, authigenic origin (Fig. 5i white arrow).Most illite, especially in the center of the illite domains is randomly oriented (Fig. 5i yellow arrow), as expected for authigenic clays shielded from compaction by adjacent brittle silt to sand size particles (Schneider et al., 2011).Alternatively, some of this clay may be the burial diagenetic replacement product of detrital feldspar and muscovite, as minor feldspar grains were found in the same area with similar shape and size (Fig. 5h).
The basal 30 m of M2 (transition from M1) is dominated by black shale (Fig. 1).Sample ZL254 is closest to the cap carbonate and contains mainly quartz, feldspar and illite (Supplementary Table 6), with minor dolomite intraclasts (Fig. 5f).Two classes of illite are distinguished.Illite of fluffy texture (Fig. 5g yellow arrow) is too fragile to survive transportation, ruling out a detrital origin, and is most likely an early authigenic phase or, alternatively, a burial diagenetic alteration phase of feldspar.Flaky illite grains (Fig. 5g red arrow) are larger, show sharp grain boundaries and cleavage, and are associated with feldspar and quartz grains, suggesting a detrital origin.Black shale samples ZL244 and ZL224 are mainly quartz cemented and almost free of carbonate (Supplementary Table 6), similar to samples from the stratigraphically ~equivalent lowermost M2 in the YGA (Fig. 3e, Supplementary Table 4).The polytype analysis of the three shale samples identify 90% to 100% 2M 1 illite, suggesting a predominantly detrital origin (Supplementary Table 8).We suggest that the authigenic, fluffy illite observed in Sample ZL254 is not sufficiently abundant to be detected by the XRD-based polytype analysis.Chlorite grains in these samples show similar flaky morphologies to detrital illite, and are associated with feldspar and quartz grains, indicative of a detrital origin.
Other than the 3 shale samples in the transition interval from M1, M2 mainly comprises carbonate rocks or shaly carbonate rocks, with talc as the dominant clay mineral (Fig. 5d), constituting up to 13.1 wt% of the bulk rock.Talc occurs as pore-fill (Fig. 5d) or in concretions ~ 2 mm in size (Fig. 6), with a fine-grained, randomly intergrown flaky texture (Fig. 5e), and is consistently associated with carbonate and quartz cements (Fig. 5d).The concretions mainly consist of intergrown talc and quartz with calcite wrapping them (Fig. 6b, c).The calcite laminae split around the concretions, showing compaction-related differential deformation.Together, these features argue for an authigenic origin for talc.
In addition, minor sepiolite and palykorskite are detected by XRD and SEM-EDS mineral mapping in the two samples with the highest talc content (Sample ZL114 and ZL121; Fig. 5d, Fig. 6c), and minor glauconite is also detected by XRD (Supplementary Table 6).Three samples towards the top of M2 contain large flakes of biotite and vermiculite (Fig. 5c) with less authigenic Mg-clay content and show an increase in feldspar content; this contrasts with the low siliciclastic content in the talc-rich intervals of M2 (Supplementary Table 6).The biotite is likely a detrital phase, derived from physical breakdown of mafic rocks, with the vermiculite formed by burial diagenetic alteration of biotite (Chamley, 1989).

Slope-basin setting
The slope setting is represented by Fengtan section and the basin setting is represented by Yuanjia section, which are located ~250 km and ~400 km south of the YGA, respectively.The main lithologies in the Fengtan section are dolostone and dolomitic shale (Fig. 7a-f), while samples from the Yuanjia section are mostly shale and free of carbonate minerals (Supplementary Table 7).The dominant clay mineral in both sections is illite.SEM-EDS mineral mapping results show that illite in the Fengtan section occurs as three distinct morphotypes.The basal sample FTA5.7 is dominated by large, flaky illite grains of detrital origin, as inferred from their sharp crystal boundaries, internal cleavage, compaction deformation and association with detrital quartz (Fig. 7e, f).Sample FTA22.9 contains mainly fibrous illite that is of authigenic origin (Fig. 7c, d), based on morphological criteria, a pore filling occurrence and close association with fine-grained dolomite crystals.The uppermost sample (FTB53) shows silt size dolomicrite floating within a clay rich matrix comprised of illite as well as minor quartz and feldspar (Fig. 7a, b).Illite shows fine-grained, platy to poorly defined outlines, consistent with an authigenic origin, as also supported by compaction deformation around dolomite crystals, (Fig. 7a, yellow arrows).
XRD-based polytype analyses of samples in Yuanjia section in the lower DST also confirm that it contains both detrital and authigenic illite, where proportions of detrital illite is no less than 50% (Supplementary Table 8).

Mineralogical trends across the different settings
Mineralogical and petrographic evidence identifies authigenic Mgclays including saponite, talc, sepiolite and palygorskite in shelf settings.Although based on a comparatively low number of samples, slope and basin settings do not appear to contain Mg-clays, consistent with previous work (Bristow et al., 2009), and are instead dominated by a mixture of detrital and authigenic illite (Fig. 8).In the shelf settings, there is a consistent stratigraphic progression in clay mineralogy from illite dominated phases during the deposition of cap carbonate and a short silica-rich interval in the lower M2 to a Mg-clay dominated phase in the middle of the DST, and then mineralogy returns to an illite-dominated phase in the upper DST (Fig. 8).However, the type of authigenic Mg-clay varies according to the setting.In the more proximal shelf lagoon and inner shelf settings the main types of Mg-clays are the 'aluminous' clays saponite and mixed layer chlorite-saponite.Distal shelf margin settings, in contrast, are dominated by 'non-aluminous' authigenic talc, with only traces of sepiolite, palygorskite and glauconite.Our results are consistent with previous studies reporting saponite and mixed layer chlorite-saponite in the YGA (Bristow et al., 2009;Derkowski et al., 2013;Han et al., 2022a), but show a much wider geographic and stratigraphic distribution of the Mg-clays, as well as more variable mineralogy in the DST, including talc, sepiolite and palygorskite.

(caption on next page)
S. Han et al.

Sr concentration
Sr concentrations in carbonate leaches range from 31-967 ppm with an average of 309 ppm (SD = 274) in the Sishang section, 9-294 ppm with an average of 65 ppm (SD = 83) in the Tianjiayuanzi section, and 30-1038 ppm with an average of 378 ppm (SD = 317) in the Zhongling section (Supplementary Data).The intervals with higher Sr contents generally overlap with the intervals rich in Mg-clays (Fig. 9).
We calculated the calcite / dolomite ratios (C:D) from the quantitative XRD results to test if the variation in the Sr concentration is correlated to the type of carbonate minerals present (Fig. 9, Supplementary Fig. 4).Within the Mg-clay rich interval, the Sr concentration is positively correlated with C:D, for example in the Tianjiayuanzi section, Zhongling section, and lower Sishang section.However, when comparing intervals with and without Mg-clays, the Sr concentration and C:D are decoupled.For example, in the Sishang section, the Mg-clayfree limestone samples with high C:D in the upper Sishang section show a lower Sr concentration than those Mg-clay bearing samples in the lower Sishang section (Fig. 9, Supplementary Fig. 4).

Spatial variation of clay mineralogy and depositional environment
Mineralogical and petrographic observations confirm the authigenic origin of Mg-clays in the DST (Fig. 2-7).Authigenic Mg-clays are known to form under specific environmental conditions, requiring elevated pH, H 4 SiO 4 and Mg 2+ concentrations compared to open-marine conditions , representative example of authigenic, pore filling talc in calcareous samples.Talc is commonly associated with quartz cements.e: ZL147, talc exhibits fine-grained, flaky, randomly intergrown texture.f: ZL254, shale at the bottom of M2 in the DST.Illite is the main clay mineral, with dolomite intraclasts, quartz and feldspar.g: ZL254, minor authigenic or diagenetic illite is fibrous (yellow arrow), while detrital illite occurs as short flakes with well preserved cleavage (red arrow).h: ZL255.8,cap carbonate with well-crystalized dolomite rhombs, calcite and quartz cements, as well as small amount of feldspar and illite.i: ZL255.8,illite is mostly fibrous and fluffy (yellow arrow), which could be authigenic pore fill material or possibly diagentic alteration/replacement of feldspar.A small amount of illite was deformed by compaction at the contact with brittle minerals (white arrow), indicating a pre-compaction origin for the pore fill illite.prevailing during the Ediacaran or indeed throughout the Phanerozoic (Bristow et al., 2009;Han et al., 2022a;Pozo and Calvo, 2018;Strauss and Tosca, 2020;Tosca et al., 2011).Consequently, most authigenic Mg-clays in the geological record have been documented in alkaline or saline lake deposits (Bristow et al., 2012;Deocampo, 2015;Furquim et al., 2008;Hindshaw et al., 2020;Milesi et al., 2019) or, more rarely due to the relatively low concentration of H 4 SiO 4 in seawater from the Cambrian onwards (Conley et al., 2017), from restricted evaporitic marine settings (Han et al., 2022b;Hover, 1999;Noack et al., 1989;Tosca et al., 2011).Previous studies have reported saponite in the YGA, which has been interpreted as evidence for deposition in a restricted, evaporitic lagoon (Han et al., 2022a) or an alkaline lake (Bristow et al., 2009), distinctly different from an open marine environment.
We find that the distribution of Mg-clay extends well outside the YGA with Mg-clays present in all shelf settings studied here, including Sishang section, which is ~350 km distant from the YGA.Authigenic Mgclays are only absent in sections from slope and basin settings, where clay fractions are characterized by abundant illite (both detrital and authigenic, with the 'authigenic' illite potentially forming via an authigenic smectite precursor).The marked change in clay mineralogy from shelf to basin argues for major differences in water chemistry between these regions, consistent with existing paleogeographic (Jiang et al., 2011;Wang et al., 2020) and clay-based environmental reconstructions (Han et al., 2022a), implying the presence of a restricted, evaporitic lagoon complex bounded by a carbonate rim on the outer shelf margin (Fig. 8).Such a lagoon environment could have facilitated favorable high pH and evaporative concentration of the key ingredients for Mg-clay formation, in particular H 4 SiO 4 and Mg 2+ ions, to levels not possible in the open ocean (Han et al., 2022a).
The transition from saponite to talc-dominated assemblages from the inner shelf lagoon towards the outer lagoon/shelf sites is likely due to changing availability of dissolved Al 3+ .Higher concentrations of dissolved Al 3+ favor the formation of saponite, an Al-bearing Mg-clay, and inhibits the formation of talc and sepiolite (Deocampo, 2015;Friedman, 1965).The reverse is also true, meaning that the dominance of talc and trace occurrence of sepiolite in the outer lagoon and shelf margin settings can be attributed to lower dissolved Al 3+ compared to proximal inner lagoon settings.Early diagenetic dissolution of detrital aluminosilicates was likely the main source of pore water Al 3+ , consistent with petrographic observations that the sections where saponite is present contain more detrital material such as feldspar and quartz (Fig. 2c, Han et al., 2022a), likely sourced from weathering and erosion of the adjacent emergent parts of the Yangtze Block.The talc-bearing Zhongling section, situated on the carbonate rimmed outer shelf margin, likely received limited detrital inputs due to trapping of detritus in the central lagoon, consistent with petrographic observations (Fig. 4c, Fig. 5).The talc can be altered from sepiolite or disordered talc (i.e.kerolite) during early diagenesis, as they are metastable relative to talc (Noack et al., 1989;Tosca and Masterson, 2014) and microbial sulfate reduction and organic matter decay in early diagenesis can lead to elevated pH    (Cui et al., 2017;Lu et al., 2013Lu et al., , 2012Lu et al., , 2011)).required for talc formation (Han et al., 2022a;Tosca et al., 2011).
While the spatial variability of Mg-clay types illustrates contrasting detrital input across the Yangtze Block, the widespread distribution of Mg-clays suggests that the entirety of the carbonate-rimmed shelf experienced restricted conditions, so that evaporitic conditions covered a much greater expanse than the localized 'intrashelf basins' (e.g.YGA) interpreted as alkaline lakes in previous work (Bristow et al., 2009).Occurrence of minor palygorskite and glauconite are also compatible with such environments.Palygorskite is usually found in (semi)arid environments, which requires alkaline conditions and high Mg, Si and Al activities in solution, and glauconite frequently forms in semi-confined micro-environments, such as in fecal pellets, preferentially on the shelf during early diagenesis, with supply of i) K and Mg from (modified and evaporitic) seawater and ii) Si, Fe and Al from dissolution of the host sediment.

Clay mineral constraints on the temporal evolution of the Yangtze lagoon
Systematic, broadly correlative (via lithostratigraphy and carbon isotope chemostratigraphy; see Fig. 1 and Fig. 9) stratigraphic patterns in lithology and clay mineral type across the studied shelfal sections allow us to distinguish five evolutionary phases in the development of the lagoon complex (Fig. 8), namely 1) cap carbonate deposition characterized by abundant authigenic illite in M1; 2) siliceous rocks with detrital illite and abundant chert at the transition between M1 and M2; 3) shaly carbonate rocks spanning the majority of M2, deposited after formation of a carbonate rim and containing authigenic Mg-clay and elevated Sr in the evaporitic shelf settings but not in the open marine slope-basin settings; 4) carbonate rocks in the lower M3 with minor authigenic illite; and 5) carbonate rocks in the uppermost DST corresponding to the global Shuram/Wonoka carbon isotope excursion In Phase 1, cap carbonate deposition records a postglacial warming period in the aftermath of the Marinoan glaciation (Hoffman et al., 1998;Penman and Rooney, 2019;Shields, 2005).The lithologically distinctive, globally recognized cap carbonate and its characteristic negative carbon isotope excursion have been widely used for stratigraphic correlation.U-Pb zircon dating of volcanic ash beds in South China (Condon, 2005) suggest that the cap carbonate was deposited between ca.635.2 ± 0.6 Ma to ca. 632.5 ± 0.5 Ma.The global distribution of this cap carbonate is consistent with an open marine depositional environment, with elevated chemical weathering delivering silicic acid and dissolved inorganic carbon species to seawater in response to super-greenhouse conditions (Penman and Rooney, 2019).Consistent with this, and the reverse weathering hypothesis suggesting that the silica-rich Precambrian seawater favored marine clay authigenesis (Isson and Planavsky, 2018), the clay fraction in cap carbonate samples is dominated by authigenic illite closely associated with chert (Fig. 3g, h, Fig. 8).Illite readily forms via burial alteration of dioctahedral smectite (illitisation) (Aubineau et al., 2019), the latter being a common authigenic phase in both ancient and recent marine settings (Chamley, 1989).We therefore infer that 'authigenic' illite formed via an authigenic smectite precursor and preserved its characteristic authigenic morphology.A smectite precursor pathway for authigenic illite is consistent with XRD evidence for abundant 1M d illite (Rafiei et al., 2020).
Phase 2 and 3 correspond to M2 of the DST, which was deposited from 632.5 ± 0.5 Ma to 579.3 ± 0.8 Ma, based on tuff zircon U-Pb geochronology (Condon, 2005) and the astronomical time scale of Sui et al. (2019).Phase 2 shows a distinct silica-rich and carbonate-free lithology, which was only found in a short interval at the boundary between M1 and M2 (Fig. 3,5).Zhongling section at the shelf margin shows the most prolonged Phase 2, in which the carbon isotope values transition from negative to positive values in Phase 3 (Fig. 9).This phase could represent a period of elevated silica burial in the post-glacial period (Penman and Rooney, 2019) combined with a pulse of detrital input, with the latter potentially occurring in response to increased sedimentation related to reworking of glacial detritus as well as erosion linked to post-glacial rebound.
Phase 3 represents the main part of Member 2. In this phase, the development of distinct differences in clay mineralogy between shelf versus slope-basin settings indicates the formation of a restricted lagoon, consistent with the depositional model of Jiang et al. (2011) which invokes the formation of a carbonate-rimmed shelf.However, Sr isotope ratios as well as REE patterns with persistently marine characteristics (Chen et al., 2022;Zhao et al., 2016) demonstrate that some degree of exchange with the open ocean continued through this phase.The mid latitude position of the Yangtze Block during this time (Merdith et al., 2021) likely contributed to the development of mildly evaporitic conditions in the restricted lagoon, which could have favored locally elevated H 4 SiO 4 , Mg 2+ and pH required for the precipitation of authigenic Mg-clays (Han et al., 2022a).In contrast, the open ocean settings (i.e.Fengtan and Yuanjia section) maintained illite as the main clay mineral during this time, consistent with our interpretations.Petrographic results from the Fengtan section show that part of this illite is of authigenic origin and XRD-based polytype analysis from the Yuanjia section confirmed the presence of 1M and 1M d illite which is likely derived from the dioctahedral authigenic smectite, as discussed above.These results are consistent with an open marine environment at Fengtan and Yuanjia in which reverse weathering is facilitated (Fig. 8).
Additional support for the development of evaporitic conditions in the shelf lagoon complex comes from carbonate-bound Sr data (Fig. 9).Our results show that the interval of high Sr concentration largely overlaps with the interval with highest Mg-clay abundance.More detailed Sr concentration data reported from the YGA show similar trends to our samples, with Sr concentration generally higher in M2, and decreasing in lower M3 (Sawaki et al., 2010).We suggest that elevated Sr concentrations are the result of evaporative concentration, or that the evaporitic conditions facilitated aragonite precipitation and the metastable Sr-rich aragonite altered to calcite during early diagenesis, retaining comparatively elevated Sr (Sun et al., 2015).The deposition of aragonite is also supported by recent cathodoluminescence microscope observation in the Zhongling section (Cui et al., 2022).
Phase 4 corresponds to the lower M3 of DST, which was deposited from approximately 579.3 ± 0.8 Ma to 571.1 ± 0.8 Ma (Sui et al., 2019).The lithology is dominated by carbonate rocks with minor illite clay at the top of M3, indicating the end of the lagoon environment as connections to the open ocean increased (Fig. 8).Erosional surfaces in the middle and upper DST in shelf margin sections (e.g.Yangjiaping section, (Jiang et al., 2011)) indicate partial destruction of the carbonate rim by erosion.The erosion of the carbonate rim may have resulted from sea-level drawdown during the 'Gaskiers' glaciation (between 580.9 ± 0.4 and 579.9 ± 0.4 Ma, (Pu et al., 2016)), the timing of which places it close to the boundary age of M2 and M3 of the DST (the onset of the Phase 4) (Sui et al., 2019).Furthermore, pulses of detrital input of different types are also consistent with sea-level drawdown when enhanced erosion of different provenances provided abundant detrital materials to the depositional basin (Fig. 8).
Phase 5 covers the uppermost DST which includes the global Shuram/Wonoka carbon isotope excursion that begins in upper M3.This phase represents the period postdating the Gaskiers glaciation and lasted from about 571.1± 0.8 Ma to 551.1 ± 0.8 Ma.In this phase, a transgression reconnected the shelf and basin settings, restoring the open shelf environment (Fig. 8).Decreased carbonate-bound Sr concentrations in this interval are consistent with the above interpretation.

Implications for early metazoan life and climate regulation
Fossil-bearing localities in the DST are largely restricted to the shelf settings, despite some exceptions documented in deeper, slope to basin settings (Bowyer et al., 2017;Jiang et al., 2011;Liu et al., 2013), and acanthomorphic acritarchs appear much earlier in South China than in other locations (e.g.South Australia, Jiang et al., 2011).Both environmental and taphonomic controls have been proposed to explain these observations.One suggestion, based mainly on local datasets obtained from sections in the YGA, is that an expansive evaporitic shelf with limited exchange with the open ocean could have contributed to the accumulation and retention of nutrient elements such as P sourced from continental weathering and fluvial inputs to the Yangtze shelf, contributing to early metazoan life evolution (Han et al., 2022a).Alternatively, a restricted lagoonal setting featuring elevated silica and phosphorous concentration may have been conducive to fossil preservation, leading to the preferential preservation of acanthomorphic acritarchs in this environment (Slagter et al., 2021;Tarhan et al., 2016).This study supports both hypotheses by demonstrating that restricted lagoonal conditions were widespread during deposition of the DST, likely affecting most of the shelf, providing the conditions conducive for both nutrient accumulation and enhanced preservation.
Finally, reverse weathering (i.e.marine clay authigenesis) involves the removal of cations from seawater and release of CO 2 (Mackenzie and Kump, 1995).Recent modelling studies have proposed that the silica-rich Precambrian seawater could have enhanced reverse weathering, which potentially played an important role in stabilizing atmospheric pCO 2 and maintaining a mostly equable climate during a period of time when solar luminosity was low (Isson and Planavsky, 2018).In this study, we document abundant authigenic clays throughout the DST in both restricted lagoon and open marine settings, providing empirical support for model predictions.

Conclusion
The presence and nature of authigenic clay minerals allows us to delineate the presence and duration of restricted, evaporitic conditions across the Yangtze Block.The consistent occurrence of authigenic Mgclays in the shelf settings across the Yangtze Block during the deposition of Doushantuo Member 2 indicates a much more widespread restricted, evaporitic lagoon environment than previously thought.The temporal changes in clay mineralogy record an open shelf setting during the deposition of the cap carbonate and the lowermost Member 2, a lagoonal setting with restricted and evaporitic conditions in the main part of Member 2 and the opening of the lagoon back to an open shelf setting during Member 3. The spatiotemporal evolution of the lagoon inferred from clay mineralogy is supported by stratigraphic trends in carbonate associated Sr and the distribution of fossils across shelf to basin settings.Our findings imply that caution is required when interpreting proxy records derived from shallow, shelfal sections of the DST as these may record local conditions rather than a 'global ocean' signal.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Fig. 3 .
Fig. 3. Representative SEM-EDS mineral mapping images of sections in Yangtze Gorges Area.a, c, e, g and h are SEM-EDS mineral mapping images; b, d and f are zoomed BSE images.The panels are arranged in stratigraphic order.The white arrows identify the main clay minerals in each sample.Ilt: illite, Sap: saponite, Chl: chlorite.a: SXA0.2, dolomicrite with illite pore fill.b: SXA0.2, illite of fibrous texture could be of either authigenic or burial diagenetic origins (yellow arrows), while long flake illite (red arrows) is likely of detrital origin.c: SX22, abundant saponite associated with dolomite.The sample is matrix supported with few detrital quartz and feldspar grains.d: SX22, saponite is present as fine-grained, curvy flakes concentrated in laminae ~40 × 5 μm in size (yellow arrows).The laminae are deformed by compaction (red arrows) and individual particals in laminae domains are oriented parallel to bedding, identifying a pre-compaction, early diagenetic origin for saponite.e: Qlk3, laminated, carbonate-free shale in the transition zone between M1 and M2, with alternating laminae of fine-grained quartz and illite.Illite and mixed layer illite-smectite are the main clay minerals.XRD analysis of this sample shows that the smectite in the mixed-layer I/S is dioctahedral montmorillonite.f: Qlk3, illite shows thin, sheet-like (red arrow with dash line delinating the illite grain; detrital) or fluffy, fibrous (yellow arrow with dash line delinating illite domain; authigenic) morphology, and occurs together with fine-grained quartz and feldspar.g: TJA1.7, curvy, flaky authigenic illite closely associated with chert and carbonate and minor platy chlorite flakes on the edge of quartz lenses.h: TJA2.8 Authigenic illite concretion in cap carbonate, enclosed by chert and calcite cements.

Fig. 4 .
Fig. 4. Sedimentary column and mineralogy of the DST in Yangtze Gorges Area.The colored areas in mineralogy columns are based on quantitative XRD data.Arrows denote samples for which SEM-EDS mineral mapping-based petrographic observations were made; yellow arrows represent sample positions where authigenic Mg-clays were observed; black arrows indicate sample positions where only other clays (e.g.illite) were observed; white arrows represent samples where clay fractions are lower than 5 wt%.

Fig. 5 .
Fig. 5. Representative SEM-EDS mineral mapping images of Zhongling section.a, b, c, d, f and h are SEM-EDS mineral mapping images.The panels are arranged in stratigraphic order.Bt: Biotite, Vrm: Vermiculite, Tlc: talc, Ilt: illite*(also includes minor I/S mixed layer).a: ZL5, phosphatic dolostone with minor detrital biotite flakes, upper DST.b: ZL23, partially calcite-replaced oolitic dolostone.c: ZL81.5, dolostone with large flakes of vermiculite and biotite.d: ZL147, representative example of authigenic, pore filling talc in calcareous samples.Talc is commonly associated with quartz cements.e: ZL147, talc exhibits fine-grained, flaky, randomly intergrown texture.f: ZL254, shale at the bottom of M2 in the DST.Illite is the main clay mineral, with dolomite intraclasts, quartz and feldspar.g: ZL254, minor authigenic or diagenetic illite is fibrous (yellow arrow), while detrital illite occurs as short flakes with well preserved cleavage (red arrow).h: ZL255.8,cap carbonate with well-crystalized dolomite rhombs, calcite and quartz cements, as well as small amount of feldspar and illite.i: ZL255.8,illite is mostly fibrous and fluffy (yellow arrow), which could be authigenic pore fill material or possibly diagentic alteration/replacement of feldspar.A small amount of illite was deformed by compaction at the contact with brittle minerals (white arrow), indicating a pre-compaction origin for the pore fill illite.

Fig. 6 .
Fig. 6.Talc concretion in ZL114, M2 of the DST at Zhongling.Mineral color overlay is as per Fig. 5. ZL114 is a shaly dolostone with talc as the main clay mineral.Talc is consistently associated with chert.Tlc: talc, Sep: sepiolite.a: backscatter overview image.The talc concretion (darker grey) in the middle has a long axis of ~ 2 mm length and is oriented parallel to bedding.The yellow rectangular area is the selected region of interest for SEM-EDS mineral mapping, covering the upper right quadrant of the talc concretion as well as overlying shaly dolostone.b: SEM-EDS mineral map.Talc (gold color) is intergrown with authigenic quartz to form a concretion splitting calcite lamina and showing differential compaction (dashed yellow lines).Talc pore-fills are pervasive in the surrounding shaly dolostone as shown in Fig. 5d.c: zoomed mineral mapping image of the red box in b.Minor sepiolite is intergrown with the talc in the concretion.d: zoomed BSE image of the white square area in c.Talc displays a typical authigenic morphology, occuring as tightly intergrown flake bundles, with distinct orientation in adjacent bundles.

Fig. 7 .
Fig. 7. Representative SEM-EDS mineral mapping images in slope settings.a, c, e are SEM-EDS mineral mapping images; b, d, f are zoomed BSE images.The panels are arranged in stratigraphic order.Ilt: illite.a: FTB53, a shaly dolostone from the uppermost DST.Note the matrix supported texture, with fine-grained illite, quartz, and feldspar grains occupying space between dolomicrite.Yellow arrows show deformed illite matrix around the dolomite crystals.b: FTB53, small flakes of illite (likely authigenic).c: FTA22.9, pore-filling illite in dolostone.d: FTA22.9, illite is fibrous or fluffy and concentrated in irrgular intercrystal domains.e: FTA5.7, dolomitic shale at the base of the DST, with illite as the dominant clay mineral.f: FTA5.7, long flake detrital illite with larger grain size (yellow dashed line).

Fig. 8 .
Fig. 8. Evolution of the Yangtze shelf lagoon based on spatiotemporal trends in clay mineralogy across the DST.Blue: illite-dominated mineralogy; Green: saponite-dominated mineralogy; Yellow: talc-dominated mineralogy.M1-M4: Member 1 to Member 4 of the DST.The blue horizontal dotted lines represent sea level.The black dotted lines in the Phase 4 represent potential exposed and eroded carbonate rims, which are supported by field observations(Jiang et al., 2011).

Fig. 9 .
Fig. 9. Mineralogy, δ 13 C values, Sr concentration and calcite / dolomite ratio curves of three sections from shelf lagoon (Tianjiayuanzi), shelf margin (Zhongling), and inner shelf (Sishang) settings and δ 13 C values of Fengtan section from slope setting.The light purple columns illustrate corresponding phases discussed in section 4.2.Mineralogy is from bulk XRD data.Yellow arrows represent samples in which authigenic Mg-rich clay minerals were observed by SEM-EDS mineral mapping.Black arrows represent illite-dominated samples.Green arrows represent samples with vermiculite, which was only found in the upper part of the Zhongling section.White arrows represent samples where clay fractions are lower than 5 wt% referring to mineral mapping results.δ 13 C data refer to(Cui et al., 2017;Lu et al., 2013Lu et al., , 2012Lu et al., , 2011)).