1. INTRODUCTION
The marine δ13C curve is used to understand the evolution of the carbon cycle and for chemostratigraphic correlation through geological time (Cramer & Jarvis, 2020). Long-term changes in seawater δ13C are often considered to reflect the net production, flux, burial, and oxidation of isotopically light organic matter (e.g., Veizer & Hoefs, 1976). However, local pools of DIC with distinct isotopic composition and sediment-buffered versus fluid-buffered diagenetic regimes may result in significant deviation of the carbon isotopic composition preserved by carbonate cements and sediments (δ13Ccarb) from global average seawater δ13CDIC (Ahm & Husson, 2022; Cui et al., 2020; Geyman & Maloof, 2019; Hoffman & Lamothe, 2019; Melim et al., 2002). Notwithstanding these local effects, global composite δ13Ccarb records that are calibrated using available radiometric constraints in multiple regions throughout the late Neoproterozoic to early Phanerozoic appear to suggest that numerous δ13Ccarb excursions recorded within individual successions may be globally synchronous (e.g., Bowyer et al., 2022; Halverson et al., 2005; Maloof et al., 2010; Nelson et al., 2023; Yang et al., 2021). These excursions may therefore be driven by globally synchronous processes including, but not limited to, changes in the relative volumes of siliciclastic vs carbonate sedimentation, nutrient delivery, or eustatic sea level (Ahm et al., 2021).
The δ13C record is assembled using carbonate rocks or bioclasts, but clastic rocks (sandstones, siltstones, shales, or conglomerates) are commonly excluded because their authigenic carbonate cements are assumed to have precipitated later under burial diagenetic conditions, rather than from seawater. However, carbonate cements within some Neoproterozoic clastic rocks have also recently been shown to record trends in δ13C that are radiometrically calibrated to be synchronous with global excursions in the δ13C record derived from coeval carbonate rocks, albeit with some variable offsets (Canfield et al., 2020). This raises the possibility that δ13C data derived from shale-hosted carbonate cements (herein termed δ13Ccarb-sh to aid distinction from δ13Ccarb of coeval carbonates), in some depositional and diagenetic settings, may be useful to establish stratigraphic correlations and to infer carbon cycle behavior in otherwise poorly constrained clastic successions. A crucial pre-requisite for the application of δ13Ccarb-sh for chemostratigraphy is to understand the mechanisms for, and extent of, δ13Ccarb-sh deviation from regional δ13Ccarb records.
The isotopic composition of a carbonate mineral is dependent upon the composition of the solution from which it precipitates, and the effects of post-depositional diagenetic alteration (e.g., Ahm et al., 2018; Swart, 2015). The oxygen isotopic composition (δ18O) of inorganic carbonates is largely dependent upon the δ18O composition of the precipitating fluid, the temperature of precipitation, and the resulting carbonate mineralogy (e.g., Epstein et al., 1953; Epstein & Mayeda, 1953; Tarutani et al., 1969; Urey, 1947). Seawater is generally enriched in 18O relative to freshwater, and carbonate precipitates that are altered within the meteoric mixing zone therefore commonly show positive covariation between δ13C and δ18O that reflects these two end-member solution compositions (Allan & Matthews, 1982; Swart, 2015). Lower values of δ18O are also associated with higher burial temperatures (Urey, 1947), and co-variation between δ13C and δ18O may therefore also occur as a consequence of mixing between low temperature, fluid-buffered carbonates and high temperature, sediment-buffered carbonates (Ahm et al., 2018).
Here we present new δ13Ccarb-sh data from the fossiliferous terminal Ediacaran Nama Group, Namibia (≥550.5 Ma to <539.6 Ma). The succession comprises mixed carbonates and clastics with well-established intra- and inter-basinal correlations and abundant dated ash beds, but the regional composite δ13Ccarb curve is discontinuous (Bowyer et al., 2022; Germs, 1983; Germs & Gresse, 1991; Linnemann et al., 2019; Nelson et al., 2022; Saylor et al., 1998; Wood et al., 2015). A proposed chemostratigraphic marker for the Ediacaran-Cambrian boundary corresponds to the stratigraphic position of a large magnitude negative δ13Ccarb excursion, termed the ‘1n/BACE’ (min. δ13Ccarb = ~-10‰), relative to key fossil occurrences, including the first appearance datum (FAD) of the ichnospecies Treptichnus pedum (Brasier et al., 1994; Zhu et al., 2006). Whilst recent radiometric ages from Laurentia constrain recovery from the 1n/BACE at ca. ≤533 Ma (Nelson et al., 2023), numerous uncertainties remain in the precise age of the 1n/BACE onset (reviewed in Bowyer et al., 2022), and the sequence of regional and global biotic first appearances across the 1n/BACE (e.g., Bowyer, Zhuravlev, et al., 2023; Topper et al., 2022). The 1n/BACE has been recorded in multiple successions globally (e.g., Bowyer, Uahengo, et al., 2023; Hodgin et al., 2021; Kouchinsky et al., 2007; Maloof et al., 2010; E. F. Smith, Nelson, et al., 2016; Topper et al., 2022; Zhu et al., 2019), but is notably absent from the Nama Group (Nelson et al., 2022; Saylor et al., 1998; Wood et al., 2015). This may suggest that the onset of this excursion is younger than ca. 538 Ma, and therefore postdates carbonate sedimentation in the Nama Group succession (Bowyer et al., 2022; Bowyer, Zhuravlev, et al., 2023; Nelson et al., 2022, 2023) (but see discussion).
We assess the potential for siliciclastic rocks to record δ13C values that approximate seawater composition throughout deposition of the Nama Group. First, measurements of δ13Ccarb-sh are compared with the magnitudes and trends in δ13Ccarb recorded by carbonate interbeds and laterally correlative carbonate-clastic successions. We then interrogate stratigraphic intervals that show covariation between δ13C and δ18O, using published carbonate data and new clastic data, to identify possible alteration of δ13C from seawater composition associated with meteoric diagenesis. Covariation between δ13Ccarb-sh and the concentrations of calcium carbonate (CaCO3) and total organic carbon (TOC) are also evaluated in order to explore the potential for differences in bulk shale composition to result in deviation of δ13Ccarb-sh from δ13Ccarb. Lastly, we explore the chemostratigraphic alignment of new δ13Ccarb-sh data within alternative lithostratigraphic correlations for the Urusis Formation of the Nama Group (≤543 to ≤538.6 Ma) and consider the utility of the δ13Ccarb-sh data to infill gaps in the composite regional δ13C curve across the critical ≥550.5 to <538 Ma interval. We compare the composite δ13C record for the upper Nama Group with regional composite δ13C records from other approximately-contemporaneous successions deposited across the terminal Ediacaran to lowermost Cambrian interval, in order to evaluate the inter-regional consistency of magnitudes and trends in δ13C and the implications for calibrated global biostratigraphy.
2. GEOLOGICAL SETTING OF THE NAMA GROUP
The Nama Group in southern Namibia (≥550.5 to <538 Ma) is a mixed carbonate and siliciclastic foreland basin succession deposited in supratidal to outer ramp settings in the Zaris and Witputs sub-basins (fig. 1A; Germs, 1983; Gresse & Germs, 1993). Sediments within the Zaris and Witputs sub-basins were deposited during cratonic convergence along the Damara and Gariep orogenies, to the north and southwest, respectively (Germs, 1983; Germs & Gresse, 1991; Gresse & Germs, 1993). Strata within both sub-basins have been correlated using litho- and chemostratigraphy, and correlation of stratal stacking patterns have informed sequence stratigraphy (figs. 1 and 2; e.g., Bowyer, Uahengo, et al., 2023; Germs, 1983; Nelson et al., 2022; Saylor et al., 1995, 1998; Wood et al., 2015). Intervals of the Nama Group succession have also been accurately age-calibrated via U-Pb zircon geochronology of interbedded tuff deposits (fig. 2; Bowring et al., 2007; Grotzinger et al., 1995; Linnemann et al., 2019; Nelson et al., 2022). The age of the base of the Nama Group remains uncertain but is estimated to be ≥550.5 Ma (Bowyer et al., 2022; Saylor et al., 1998), and the youngest dated tuff deposit in the Witputs Sub-basin, immediately overlying the sub-Nomtsas Formation unconformity, yields a U-Pb age of 538.58 ± 0.19 Ma (Linnemann et al., 2019).
The succession is divided into the lower Nama Group (Kuibis Subgroup, ca. ≥550.5 to <547 Ma), and upper Nama Group (Schwarzrand Subgroup, ca. ≥546 to <538 Ma; fig. 1). A mixed carbonate-siliciclastic succession deposited in the Vioolsdrif Sub-basin of northwest South Africa is radiometrically constrained to be temporally equivalent to, at minimum, the Schwarzrand Subgroup of the Witputs Sub-basin (Nelson et al., 2022). The Witputs and Vioolsdrif sub-basins may have been separated by a tectonic forebulge (the Koedoelaagte Arch; Germs & Gresse, 1991).
The Zaris and Witputs sub-basins deepened to the (present-day) north and southwest, respectively, during deposition of the Kuibis Subgroup, with increasing distance from an intervening paleobathymetric high (the Osis Arch), and with distance from the Kalahari Craton to the present east (fig. 1B; Germs, 1983). However, gradual infill of the Zaris Sub-basin shifted the orientation of facies belts to northwest-southeast across both sub-basins during deposition of the Schwarzrand Subgroup (fig. 1C; Germs, 1983; Saylor, 2003; Saylor et al., 1995). Facies belts range from clastic-dominated braided-fluvial to muddy tidal, to inner, mid- and finally outer ramp carbonate-dominated facies, which deepened, on average, to the west or south-west during deposition of the Schwarzrand Subgroup (Germs, 1983; Saylor, 2003). The Kalahari craton was the main source of clastic sediment into the Nama sub-basins during deposition of the Kuibis Subgroup (fig. 1B; Germs, 1983), but during deposition of the Schwarzrand Subgroup, the Zaris Sub-basin received additional detrital material directly from the Damara Belt to the north (fig. 1C; Blanco et al., 2009, 2011; Germs, 1983). In the Nama Group successions, early Transgressive Systems Tracts are generally dominated by siliciclastic rocks whereas various carbonate facies distinguish late Transgressive to Highstand Systems Tracts (Saylor et al., 1995, 1998).
3. METHODS
δ13Ccarb-sh and δ18Ocarb-sh data were obtained from 107 shale/mudstone samples (defined as <16 wt% CaCO3) collected from eight outcrop sections distributed across both the Zaris and Witputs sub-basins, which together cover ≥12 Myr from the Kanies Member of the Dabis Formation (≥550.5 Ma) to the Nomtsas Formation (≤538.6 Ma; fig. 2, table S1). Samples were taken from three settings: (1) three sections of the lower Kuibis Subgroup that contain intervals composed of decimeter to meter-scale interbedded carbonates and clastics (≤1 m) deposited on the inner to mid-outer ramp (n = 21, sections 6, 9 and 10; figs. 1B, 2, 3A,B); (2) three extended shale packages (12–90 m) within carbonate-clastic successions, one from a shallow-inner ramp section of the Kuibis Subgroup (n = 3, section 8; figs. 1B and 2), and two from the Schwarzrand Subgroup deposited on the mid- to outer ramp (n = 21, sections 3 and 4; figs. 1C, 2, 3C–E); (3) two composite sampling transects through clastic-only successions of the uppermost Kuibis Subgroup and entire Schwarzrand Subgroup (n = 62, sections 11 and 12; figs. 1B and C, 2, 3F,G) that record an overall shallowing-upward succession from mid-ramp to inner ramp and lagoon. Samples from these three settings are classified into three groups: Group 1: m-scale shale interbeds within carbonate-clastic successions; Group 2: extended shale packages within carbonate-clastic successions, and Group 3: clastic-only successions.
Analyses of δ13Ccarb-sh followed the method of Canfield et al. (2020). Powdered bulk clastic rock samples in sealed vials were reacted with 100% orthophosphoric acid at 75 °C and left for 24 h using an Elementar iso FLOW system. Any resulting CO2 gas produced was then extracted from the vial and analyzed for its carbon and oxygen isotopic ratios using an Elementar PRECISION stable isotope ratio mass spectrometer. The standard deviation (n = 48) of a powdered coral laboratory standard (COR1D, δ13C = -0.649‰, δ18O = -4.924‰) run as a sample on the same days as the study samples, was ± 0.074‰ for δ13Ccarb and ± 0.111‰ for δ18Ocarb. All isotopic values are normalized relative to the Vienna Pee Dee Belemnite (VPDB) standard. Total wt% CaCO3 in shales was estimated using a linear regression model between peak height area from the isotopic measurements and the analyzed mass of powdered coral laboratory standard (COR1D, assumed 100% CaCO3). The uncertainty associated with this regression model is <1% (see Supplementary Information). A subset of shale samples of varying CaCO3 concentration from different sections were also analyzed by colorimeter, and the relative concentrations of CaCO3 were consistent between both methods (table S2). We also consider published TOC concentration data from the same samples analyzed herein for δ13Ccarb-sh (Bowyer et al., 2020).
δ13Ccarb-sh data (n = 107) were compared to available δ13Ccarb from 38 sections (n = 1611) of the Nama Group succession in Namibia and northwest South Africa. In order to test the degree to which these data may reflect early diagenetic resetting from seawater δ13C composition associated with mixing between two end-member compositions, we assessed the co-variation of δ13C and δ18O. In order to test whether values of δ13Ccarb-sh are affected by the concentrations of CaCO3 or TOC within each sample, we tested the co-variation of δ13Ccarb-sh and CaCO3, and δ13Ccarb-sh and TOC. In each case, the Shapiro-Wilk test was used to evaluate whether these data are normally distributed (tables S3 and S4). The strength and significance of correlations were tested using either Pearson’s correlation coefficient (r), or the non-parametric equivalent test of Spearman’s rank correlation coefficient (ρ). Spearman’s rank correlation coefficient was used when assumptions about the normality of the distribution of the variables, constant residual variability, and linearity were not fulfilled. In each case, significant correlations are indicated when p ≤0.05.
Lastly, we explore the alternative regional δ13C chemostratigraphies that result from different possible lithostratigraphic correlations between sections of the Urusis Formation within the Witputs Sub-basin and between the Witputs and Vioolsdrif sub-basins. All new and existing data are calibrated with available radiometric dates and placed within the resulting composite Nama Group δ13C chemostratigraphic age frameworks (updated from Bowyer et al., 2022 and Bowyer, Zhuravlev, et al., 2023).
4. RESULTS
4.1. Group 1: Meter-scale shale interbeds within carbonate-clastic successions
The δ13Ccarb-sh data for m-scale shale interbeds within carbonate-clastic successions range from -6.63‰ to +2.50‰ (mean = -1.62‰, sd = 2.38‰, n = 21, table S1). Through the Dabis Formation and lower Zaris Formation (≥550.5 to ca. 547 Ma), δ13Ccarb-sh values from Arasab, Omkyk and Brak (sections 6, 9, 10, fig. 2) closely follow the negative-to-positive trend recorded by contemporaneous δ13Ccarb (fig. 2).
δ13Ccarb-sh values do not show a statistically significant correlation with δ18Ocarb-sh (r = 0.42, p = 0.06, R2 = 0.17, fig. 4A). These interbedded shales are characterized by CaCO3 concentrations in the range 0.02–15.51 wt% (mean = 2.42 wt%, fig. 4B), TOC concentrations in the range 0.04–0.11 wt% (mean = 0.07 wt%, fig. 4C), and CaCO3/TOC in the range 0.4–85.0 (mean = 11.83, fig. 4D). There is no significant correlation between δ13Ccarb-sh and CaCO3 content (ρ = 0.07, p = 0.77, fig. 4B), δ13Ccarb-sh and TOC content (ρ = -0.49, p = 0.15, fig. 4C), or δ13Ccarb-sh and CaCO3/TOC (ρ = -0.32, p = 0.37, fig. 4D).
4.2. Group 2: Extended shale packages within carbonate-clastic successions
Values of δ13Ccarb-sh from a 12.5 m-thick shale package of the Dabis Formation (Kuibis Subgroup, ca. 550.5 Ma) at Zwartmodder (section 8, fig. 2) range from -7.18‰ to -2.38‰ (n = 3). This interval records siliciclastic deposition during initial transgression across the Kalahari basement and has been litho- and chemostratigraphically correlated with transgressive limestone, dolostone and shale of the basal Omkyk Member at Omkyk (section 9), and the dolostone-dominated Dabis Formation at Brak (sections 10, fig. 2; Wood et al., 2015). δ13Ccarb-sh data at Zwartmodder overlap in magnitude and trend with both δ13Ccarb-sh and δ13Ccarb at Omkyk and Brak throughout this interval (figs. 2 and 4E).
δ13Ccarb-sh data from extended shale packages of the Urusis Formation (≤543 to ≤538.6 Ma), comprising the Swartpunt (section 3) and Swartkloofberg (section 4) sections, show dominantly negative values that range from -8.52‰ to +0.17‰ (mean = -4.01‰, sd = 2.69, n = 21, figs. 2 and 4E). δ13Ccarb-sh values at section 3 (-8.52‰ to +0.17‰, mean = -3.16‰, sd = 2.24, n = 12) appear scattered and depleted relative to δ13Ccarb above and below the siliciclastic interval (0.33–1.84‰, mean = 1.32‰, n = 36). However, δ13Ccarb-sh values at section 4 (-8.12‰ to -0.40‰, mean = -5.14‰, sd = 2.95, n = 9) broadly display both a falling limb from δ13Ccarb values recorded by underlying pinnacle reef carbonate (0.63–1.72‰, mean = 1.24, n = 4), followed by a nadir and a rising limb, which together appear to track the overall shape of a negative δ13Ccarb-sh excursion (fig. 2).
Group 2 shales do not show a statistically significant correlation between δ13Ccarb-sh and δ18Ocarb-sh (r = 0.37, p = 0.08, R2 = 0.13; fig. 4E). These shale packages have CaCO3 concentrations in the range 0.01–4.43 wt% (mean = 1.05 wt%, fig. 4F), TOC concentrations in the range 0.05–0.19 wt% (mean = 0.08 wt%, fig. 4G), and CaCO3/TOC in the range 0.1–54.8 (mean = 18.49, fig. 4H). There is no statistically significant correlation between δ13Ccarb-sh and CaCO3 content (ρ = 0.31, p = 0.14, fig. 4F), δ13Ccarb-sh and TOC content (ρ = -0.34, p = 0.28, fig. 4G), or δ13Ccarb-sh and CaCO3/TOC (r = 0.23, p = 0.47, R2 = 0.05, fig. 4H).
4.3. Group 3: Clastic-only successions
δ13Ccarb-sh data from one clastic-only succession of the Urikos Member and Nudaus Formation of the Zaris Sub-basin (section 11, fig. 2), and a second composite clastic-only succession of the Urikos Member, and Nudaus, Urusis and Nomtsas formations of the Zaris Sub-basin (section 12, fig. 2), range from -15.45‰ to -1.00‰ (mean = -7.72‰, sd = 3.31, n = 62). These data are significantly depleted relative to δ13C values of contemporaneous carbonates and shales from carbonate-clastic successions of the Zaris and Witputs sub-basins (figs. 2, 4I), and show a strong and statistically significant positive correlation between δ13Ccarb-sh and δ18Ocarb-sh (ρ = 0.70, p < 0.01, fig. 4I). This positive correlation also appears to be associated with the temporal distribution of samples, whereby values of δ13Ccarb-sh and δ18Ocarb-sh increase with relative age (figs. 2 and 4I).
With the exception of some of the oldest samples, the CaCO3 content of Group 3 shales is lower than shales from groups 1 and 2, ranging from 0.00–3.16 wt% (mean = 0.09 wt%, fig. 4J). Group 3 shales have TOC concentrations in the range 0.05–0.10 wt% (mean = 0.08 wt%, fig. 4K) and CaCO3/TOC in the range 0.1–0.4 (mean = 0.11, fig. 4L). There is a significant positive correlation between δ13Ccarb-sh and CaCO3 content (ρ = 0.57, p < 0.01, fig. 4J), but no significant correlation observed between δ13Ccarb-sh and TOC content (r = 0.56, p = 0.09, R2 = 0.31 fig. 4K), or δ13Ccarb-sh and CaCO3/TOC (ρ = -0.43, p = 0.29, fig. 4L).
4.4. Qualitative observations of δ13C and δ18O through the Nama Group
Carbonates that were deposited following initial transgressive onlap of the Nama basement record negative values of δ13Ccarb that correspond to the ‘basal Nama excursion’ (BANE; Bowyer et al., 2022; Maloney et al., 2020; Saylor et al., 1998; O. Smith, 1999; Wood et al., 2015). Following the BANE, δ13Ccarb values increase through the lower Omkyk Member (Zaris Sub-basin) and Kliphoek and Mooifontein members (Witputs Sub-basin), to reach peak values (~5‰) associated with the Omkyk excursion (OME, fig. 2; Bowyer et al., 2022; Saylor et al., 1998; O. Smith, 1999; Wood et al., 2015). Following the OME interval, δ13Ccarb values show a gradual decrease towards a minor negative excursion (Saylor et al., 1998). This interval has been tentatively correlated with a negative δ13Ccarb excursion recorded in the A0 Member of the Ara Group, Oman, based on a preliminary global chemostratigraphic correlation anchored by available radiometric data from tuff interbeds in the underlying Hoogland Member of the Nama Group and A0 Member of the Ara Group (Bowring et al., 2007). This negative excursion has therefore been termed the ‘A0’ excursion, for ease of reference (but see discussions of uncertainty in Bowyer et al., 2022 and O. Smith, 1999).
In the Schwarzrand Subgroup of the Witputs Sub-basin, carbonates of the lower Huns Member record recovery from a positive δ13Ccarb excursion (max δ13Ccarb = 4.24‰, section 1, fig. 2). This interval stratigraphically overlies the Nasep Member, wherein a tuff bed at a neighbouring section to the north of section 1 has been dated to 542.65 ± 0.15 Ma (Nelson et al., 2022). The δ13Ccarb peak recorded in the lower Huns Member at section 1 has been tentatively correlated with radiometrically-constrained positive δ13Ccarb values in the Tamengo Formation of Brazil (Boggiani et al., 2010; Parry et al., 2017) and the A3 Member of the Ara Group, Oman (Bowring et al., 2007). This interval of positive δ13Ccarb has therefore been termed the ‘A3’ excursion (Bowyer et al., 2022) but this is dependent upon the absolute age of the base of the Huns Member (see Discussion).
A compilation of all Nama Group carbonate data shows no correlation between δ13Ccarb and δ18Ocarb (ρ = -0.09, p <0.01, fig. 4M). However, individual sections and discrete intervals of the Nama Group record are characterized by significant correlation between δ13Ccarb and δ18Ocarb (fig. 2). As such, we consider associated temporal changes in δ13C and δ18O, and differences in recorded values between carbonates and shales, through the Nama Group succession.
In the Kuibis Subgroup, Group 1 and Group 2 shales show generally reciprocal trends in δ13C and δ18O relative to carbonates within individual sections, and between contemporaneous sections, recording the BANE (fig. 2). Carbonates in the lowermost 10–20 m of this interval at sections 5, 8, 9 and 10, record increasing δ18Ocarb values that broadly covary with δ13Ccarb (fig. 2). Throughout this interval, δ13Ccarb-sh values of Group 1 shales at sections 9 and 10 and Group 2 shales at section 8, and δ18Ocarb-sh values of Group 1 shales at section 10 and Group 2 shales at section 8, covary in magnitude and trend with δ13Ccarb and δ18Ocarb, respectively. Group 1 shales in the lowermost 10 m of section 9 record values of δ18Ocarb-sh that are more positive than interbedded δ18Ocarb (fig. 2). By contrast, in the Mara Member of the Witputs Sub-basin, carbonates at sections 6 and 7, and Group 1 shales at section 6, record scattered negative δ13C values that correspond to the BANE, but show no positive correlation between δ13C and δ18O (fig. 2). Indeed, trends in δ13Ccarb and δ18Ocarb appear to be negatively correlated throughout this interval at section 7 (fig. 2).
Throughout the OME interval, there is no significant correlation between δ13Ccarb and δ18Ocarb in any section. Following the OME, δ13Ccarb values show a gradual decrease towards a minor negative excursion in the Urikos Member (or upper Hoogland Member) at one section to the northwest of section 9 (Zebra River, not shown on fig. 2), which is associated with positive correlation between δ13Ccarb and δ18Ocarb (Saylor et al., 1998).
Following the A3 excursion, carbonates of the Urusis Formation in the Witputs Sub-basin (sections 1–4, fig. 2) show limited variability in δ13Ccarb (-0.80‰ to 2.40‰, mean = 1.34‰, sd = 0.48) and no observable correlation between δ13Ccarb and δ18Ocarb (fig. 2). Values of δ18Ocarb-sh in Group 2 shales at sections 3 and 4 (-9.96‰ to -3.08‰, mean = -6.93‰, sd = 1.84) overlap with, or are more positive than, δ18Ocarb of approximately contemporaneous carbonates throughout this interval of the Neint Nababeep Plateau composite section (-9.39‰ to -4.87‰, mean = -6.99‰, sd = 1.15, figs. 2 and 4E).
In the Schwarzrand Subgroup of the Zaris Sub-basin, Group 3 shales at sections 11 and 12 are characterized by decreasing δ13Ccarb-sh and δ18Ocarb-sh up through the stratigraphic succession (fig. 2), and corresponding positive covariation between δ13Ccarb-sh and δ18Ocarb-sh (fig. 4I). These values are significantly depleted relative to δ13Ccarb and δ18Ocarb in samples of the contemporaneous mixed carbonate-clastic succession of the Witputs Sub-basin (sections 1–4, figs. 2, 4I).
5. DISCUSSION
5.1. δ13C and δ18O covariation and potential meteoric influence
Sedimentary rocks of the lower Kuibis Subgroup record diachronous deposition across basement rocks of the Kalahari craton in both the Zaris and Witputs sub-basins (Germs, 1974, 1983). As such, the thickest measured sections of the lower Nama Group that accumulated in the deeper parts of each sub-basin (e.g., sections 7 and 10) where accommodation space was greatest, also contain the oldest units. This is supported not only by lithostratigraphic and sequence stratigraphic correlation, but also by observations of δ13Ccarb chemostratigraphy and associated preliminary biostratigraphic considerations in both sub-basins (fig. 2; Bowyer, Uahengo, et al., 2023; Maloney et al., 2020; Saylor et al., 1995, 1998; O. Smith, 1999; Wood et al., 2015).
The oldest transgressive carbonates and shale-hosted carbonate cements at all studied sections record recovery from a negative δ13C excursion (fig. 2). This interval is also characterized by a general shift in dominant lithology from dolostone to limestone, where the δ13C trend is recorded in both dolostones and limestones within and between individual sections. In the lowermost 10–20 m of sections 5, 8, 9 and 10, this δ13Ccarb recovery is accompanied by significant δ13Ccarb-δ18Ocarb covariation, with a corresponding increase in δ18Ocarb (fig. 2). By contrast, basal transgressive deposits of the Dabis Formation in the Witputs Sub-basin at sections 6 and 7 do not show any clear positive correlation between δ13Ccarb and δ18Ocarb (fig. 2). Carbonates of the Mara Member in section 6 are frequently characterized by evaporitic fabrics, and show scattered values of δ13Ccarb and δ18Ocarb, whilst carbonates at section 7 show a general negative correlation between δ13Ccarb and δ18Ocarb through the lower ~50–75 m (fig. 2; Wood et al., 2015).
Carbonates of the lower Nama Group were most commonly deposited upon granitic basement rock, where the δ13C of freshwater input was likely depleted due to the dominant influence of organic carbon respiration and lack of carbonate weathering that would otherwise contribute elevated δ13Ccarb (Khadka et al., 2014; Rodriguez Blanco et al., 2020). Positive δ13Ccarb-δ18Ocarb covariation and low initial δ13Ccarb recorded in the lowermost 10–20 m of transgressive carbonate deposits in sections 5, 8, 9 and 10 may therefore reflect a greater degree of mixing between marine and freshwater associated with meteoric diagenesis. However, the overall trend from negative to positive δ13Ccarb recorded across the boundary between the Kanies and lower Omkyk members at section 10, and within the Kliphoek Member at sections 6 and 7, is accompanied by no synchronous positive shift in δ18Ocarb (fig. 2). Therefore, the degree to which meteoric diagenesis has altered δ13Ccarb from the composition of seawater DIC remains uncertain.
The recovery from a negative δ13Ccarb excursion recorded in dolostone of the Dengying Formation, South China, is radiometrically constrained to be <550.14 ± 0.63 Ma (Yang et al., 2021), and may correlate with the trend in δ13Ccarb recorded by carbonates of the lower Nama Group (Bowring et al., 2007; Saylor et al., 1998; Yang et al., 2021). However, δ13Ccarb values of the basal Nama Group are notably depleted relative to all other global δ13Ccarb data that postdate the Shuram δ13Ccarb excursion (ca. 575–565 Ma; Rooney et al., 2020; Yang et al., 2021) but pre-date the 1n/BACE. Therefore, if the BANE is distinct from the Shuram excursion, as suggested by current global chemostratigraphic age models (Bowyer et al., 2022; Bowyer, Uahengo, et al., 2023; Rooney et al., 2020; Yang et al., 2021), then the magnitude of the BANE may reflect regional amplification of negative δ13Ccarb values associated with local meteoric diagenesis.
Positive correlation between δ13Ccarb and δ18Ocarb is also recorded in carbonate interbeds of the Urikos Member (or upper Hoogland Member) of the Zebra River section (to the northwest of section 9), coincident with the inferred A0 negative δ13Ccarb excursion (Bowyer et al., 2022; Saylor et al., 1998). Lateral differences in δ13Ccarb recorded between sections in the vicinity of Zebra River, and meteoric dissolution of ooids observed within the upper Hoogland Member, have previously been suggested as possible evidence to support deviation of δ13Ccarb from the composition of seawater δ13CDIC in this interval associated with diagenetic alteration (O. Smith, 1999). Alternatively, the observed differences in trend and magnitude of δ13Ccarb between these sections may record lateral differences in section completeness associated with increasing paleodepth and accommodation space from southeast to northwest within the Zaris Sub-basin (for more information, see supplementary correlation chart of Bowyer, Uahengo, et al., 2023).
Group 3 shale samples in sections 11 and 12 also show positive covariation between δ13Ccarb-sh and δ18Ocarb-sh, and decreasing values of both δ13Ccarb-sh and δ18Ocarb-sh during progressive infill and shallowing of the Zaris Sub-basin (figs. 2 and 4I). At this time, paleocurrent data indicate that clastic input to the Zaris Sub-basin was sourced from the present north and east (fig. 1C; Germs, 1983). Group 3 shales were deposited distant from the locus of carbonate sedimentation (Germs, 1983; Gresse & Germs, 1993; Saylor et al., 1995). Due to outcrop availability, samples were taken along road transects, and record increasing proximity to the source of clastic input with decreasing age (fig. 1C). As such, an increase in the contribution of isotopically light riverine freshwater to δ13Ccarb-sh with decreasing age appears to be the most parsimonious explanation for the significant deviation of δ13Ccarb-sh in Group 3 samples, relative to contemporaneous δ13Ccarb recorded throughout the carbonate-clastic succession of the Schwarzrand Subgroup in the Witputs Sub-basin (figs. 2, 4I).
5.2. Potential significance of δ13Ccarb-sh data from carbonate-clastic successions
The isotopic composition of marine carbonates is commonly considered to approximate the isotopic composition of DIC in seawater, provided that precipitated carbonate minerals have not undergone significant subsequent diagenetic alteration. In the Kuibis Subgroup, whole rock carbonates and carbonate cements within Group 1 and Group 2 shales show δ13C values that are consistent in magnitude and trend (fig. 2). We may infer this to reflect a consistent source for the measured carbonate, potentially seawater DIC.
Nama Group reef carbonates from the Omkyk Member show multiple phases of syndepositional through to late burial cements (Wood et al., 2018), but the timing and type of carbonate cements found in Nama Group clastics is not known, and indeed these may potentially have formed at any time during diagenesis, and from any diagenetic fluid. δ13C and δ18O values will only show a correlation in scenarios where there is mixing between different end-members of alteration or by mixing of two different diagenetic fluids. For example, in meteorically-altered carbonates the most altered end-member will have low δ13C and δ18O, while the least altered end-member will show high δ13C and δ18O values relative to contemporaneous seawater DIC, and a similar trend can also be created by mixing of two different diagenetic fluids such as seawater and meteoric water. Values of δ13Ccarb-sh from Group 1 and 2 shales in carbonate-clastic successions show no significant correlation with δ18Ocarb-sh, which may suggest that they formed from a single diagenetic fluid and during a single diagenetic stage (fig. 4A, E). By contrast, Group 3 shales in clastic-only successions show a statistically significant positive correlation between δ13Ccarb-sh and δ18Ocarb-sh, and more negative and variable δ13Ccarb-sh values than contemporaneous δ13Ccarb (fig. 4I), and so may therefore have formed by either a mixture of variably diagenetically-altered end-members, or via the mixing of two different diagenetic fluids.
5.3. δ13Ccarb-sh, organic carbon and CaCO3 content
Bulk δ13C values derived from impure carbonates can deviate from the composition of seawater DIC due to the incorporation of organic matter and subsequent formation of authigenic carbonate during diagenesis. This deviation is most likely to occur in sediments with low carbonate content relative to organic carbon (Saltzman & Thomas, 2012). Scattered negative δ13Ccarb values recorded from samples of the Neint Nababeep Plateau composite section have also been associated with stratigraphic proximity to siliciclastic-rich intervals, and they have thus not been considered useful for basin-wide δ13C correlation (Nelson et al., 2022).
Group 1 and 2 shales were deposited in settings where CO32- concentrations are inferred to have been high, consistent with contemporaneous, laterally extensive carbonate platform development. In such settings, authigenic carbonate may readily precipitate from seawater. Clastic-hosted carbonate cements from samples of the mixed carbonate-clastic Isaac Formation of the Ediacaran Windermere Supergroup of Laurentia have similarly been shown to record δ13C values that approximate contemporaneous δ13Ccarb (Canfield et al., 2020; Cochrane et al., 2019). We infer that seawater was the primary DIC source for precipitation of carbonate cements in Group 1 shales given elevated concentrations of CaCO3 (0.02–11.96 wt%, mean = 1.77 wt%) compared to Group 3 shales from clastic-only successions (0.00–3.16 wt%, mean = 0.22 wt%), and the consistency between δ13Ccarb-sh in Group 1 shales and δ13Ccarb from associated, interbedded carbonates (figs. 2 and 4A). Group 2 shales have an intermediate range of CaCO3 concentrations (0.00–4.43 wt%) relative to shales from groups 1 and 3, and a mean CaCO3 concentration (1.05 wt%) that is elevated relative to Group 3 shales. The median CaCO3 concentration of Group 2 shales (0.89 wt%) is also significantly elevated relative to Group 1 (0.18 wt%) and Group 3 (0.01 wt%) shales. Despite these observations, Group 2 shales of the Schwarzrand Subgroup record δ13Ccarb-sh values that are depleted relative to carbonates above and below the sampled shale package in section 3 and below the sampled shale package at section 4 (fig. 2). If seawater was the primary DIC source for carbonate cement precipitation in Group 2 shales, then the apparent deviation of δ13Ccarb-sh from approximately contemporaneous δ13Ccarb in sections 3 and 4 might be a consequence of contamination by additional 12C associated with elevated TOC, and this is explored below.
The combined shale data from groups 1–3 show a significant positive correlation between δ13Ccarb-sh and CaCO3 (ρ = 0.69, p < 0.01, fig. 4N), and a moderate but statistically significant negative correlation between δ13Ccarb-sh and TOC (ρ = -0.43, p = 0.01, fig. 4O). These correlations appear to support the inference that samples of lower purity (characterized by low CaCO3/TOC, fig. 4P) will tend to record more negative δ13Ccarb-sh. However, given the low TOC concentrations of shales from the Nama Group (<0.20 wt%), the power of these statistical correlations are impacted to a degree by analytical uncertainties associated with TOC measurements (precision of better than ±0.04 wt%; Bowyer et al., 2020). Future studies that aim to further test this hypothesis would therefore benefit from sample sets with a larger range in TOC concentrations.
A carbonate to organic carbon concentration ratio of 7:1 has been suggested as a potential threshold below which the incorporation of 12C from organic matter may result in deviation of bulk δ13C from the composition of seawater DIC (Saltzman & Thomas, 2012). Indeed, very low CaCO3/TOC ratios (0.06–0.36) in Group 3 shales may be partly responsible for the clearly depleted δ13Ccarb-sh relative to contemporaneous δ13Ccarb (fig. 4I). However, values of δ13Ccarb-sh in Group 1 shales clearly track the magnitude and trend of δ13Ccarb recorded by carbonate interbeds (fig. 2). These, after removing one outlier (CaCO3/TOC = 84.99), have a mean CaCO3/TOC of 3.70 (fig. 4D). The average purity of Group 2 shales (mean CaCO3/TOC = 18.49, n = 12) is greater than Group 1 shales, even when this outlier value is included (mean = 11.83, n = 10, figs. 4D, H, P). It is therefore difficult to disregard values δ13Ccarb-sh derived from Group 2 shales, which are more negative than average δ13Ccarb of coeval carbonates in the Urusis Formation, on the basis of sample purity alone. We therefore explore the lithostratigraphic correlation of the Schwarzrand Subgroup in the vicinity of sections 3 and 4, in order to investigate the stratigraphic position of negative δ13Ccarb-sh data from Group 2 shales relative to regional δ13Ccarb chemostratigraphy.
5.4. Implications of alternative lithostratigraphic correlations for regional chemostratigraphy
The Urusis Formation in the Witputs Sub-basin is a mixed carbonate-siliciclastic succession deposited over a ca. 4-million-year period from ≤543 Ma to ≥538.5 Ma (Linnemann et al., 2019; Nelson et al., 2022; Saylor et al., 1998). Extensive studies of litho-, chemo-, bio- and sequence stratigraphy, alongside radiometric dating of the Urusis Formation succession, have focused on sections that outcrop on farms Nord Witputz 22, Swartpunt 74 and Swartkloofberg 95 (fig. 5; Darroch et al., 2015; Grotzinger et al., 1995; Jensen et al., 2000; Linnemann et al., 2019; Saylor et al., 1998; Saylor & Grotzinger, 1996; Wood et al., 2015). Confident lithostratigraphic correlation of strata in this region is complicated by laterally discontinuous outcrop associated with eroded topography, and the recognition of regional thrust faulting and folding associated with the Gariep Orogeny (fig. 5A; Saylor & Grotzinger, 1996). Here, we present four possible lithostratigraphic correlations for sections 1–4 (fig. 5B), which result in a range of possible chemostratigraphic correlations for the δ13Ccarb-sh data in section 4 relative to the radiometrically constrained Nama Group composite δ13Ccarb age framework. Each of these four lithostratigraphic correlations are discussed below.
Correlation 1 corresponds to the original lithostratigraphic correlation of Saylor and Grotzinger (1996) wherein the extended shale package at section 4 is correlated with the Feldschuhhorn Member (fig. 5B). This was justified by Saylor and Grotzinger (1996) on the basis of correlating an underlying unit of pink thrombolitic and stromatolitic lime mudstone at section 4 to a lithologically similar unit in the upper Huns Member at sections further to the east. According to this correlation, available radiometric data would constrain the negative δ13Ccarb-sh data recorded at section 4 to between 542.65 ± 0.15 Ma and 540.095 ± 0.099 Ma (Linnemann et al., 2019; Nelson et al., 2022), however the precise age and duration of the Feldschuhhorn Member within this interval remains uncertain. This is further complicated by uncertainties in the reliability of available radiometric ages and lithostratigraphic correlation between sections 2 and 3
Recent re-dating of an ash bed within section 2 has been preliminarily interpreted to suggests that this section may have been laterally equivalent to, rather than stratigraphically beneath, section 3 (Messori et al., 2021). Sections 2 and 3 show consistency in lithostratigraphic architecture and number of ash beds (fig. 5B). However, Nelson et al. (2022) suggest that the age of the lowermost ash bed in section 3 (dated at 540.095 Ma, Linnemann et al., 2019) may instead represent a maximum depositional age. If future structural reassessments allow for sections 2 and 3 to represent laterally equivalent depositional successions, then section 2 may have been deposited in a shallower depositional environment than section 3 based on the available sedimentological details. However, out-of-sequence ages of ash beds in the vicinity of Farm Swartpunt associated with possible zircon reworking in the source magma chamber(s) cannot presently be ruled out. The lithostratigraphic and radiometric correlation between sections 2 and 3 demands future clarification, as it has significant implications for the duration of Huns Member deposition in the Witputs Sub-basin, lithostratigraphic subdivision, and lateral differences in sedimentation rate. Differences in sedimentation rate between correlative sections also have significant implications for interpretations of associated geochemical proxy data (Nelson et al., 2022). Integrated stratigraphic, geochemical, and radiometric dating may resolve these ongoing uncertainties.
Correlation 2 implies that the extended shale package at section 4 is equivalent to the transgressive systems tract of medium scale sequence E17 of Saylor (2003), which immediately underlies section 3. Given the uncertainty in lithostratigraphic correlation between sections 2 and 3, this would result in an age range for the section 4 δ13Ccarb-sh data equivalent to, or slightly younger than, correlation 1, but still ≥540.095 ± 0.099 Ma, if this is the accepted depositional age of the lowermost ash bed in section 3 (fig. 5B, but see Nelson et al., 2022). Alternatively, correlation 3 implies that the two extended shale packages in sections 3 and 4 are coeval, but were deposited at significantly different depositional rates, and that both are <538.99 ± 0.21 Ma (Linnemann et al., 2019).
Lastly, correlation 4 re-positions section 4 stratigraphically above section 3, and implies that the negative δ13Ccarb-sh data in section 4 are younger than the negative δ13Ccarb-sh data in section 3, <538.99 ± 0.21 Ma (fig. 5B; Linnemann et al., 2019). Despite the recognition of thrust faulting in this area, this correlation also appears to be the most parsimonious when considering the consistent regional dip to the north/northwest for sections to the east of the large NW–SE trending fold and fault system to the west of section 4 (figs. 5A and 6). Correlation 4 would simply assume a less significant offset associated with the thrust fault that separates sections 3 and 4 (fig. 6D). Linnemann et al. (2019) note that tuff bed 6 may be reworked, and indeed if this reworking is from the underlying Spitskop Member (now removed by erosion), then the onset of the δ13Ccarb-sh excursion at section 4 may be younger than 538.6 Ma. A possible further alternative stratigraphic correlation (not shown in fig. 5) would imply that the entire Feldschuhhorn Member at section 4 corresponds to ongoing deep water shale deposition equivalent to the entirety of the Feldschuhhorn and Spitskop members in shallower sections 1–3.
Correlation 4 may find further support in recent high resolution litho-, bio- and δ13Ccarb chemostratigraphic assessment, and radiometric dating of correlative strata on the Neint Nababeep Plateau that were deposited in the Vioolsdrif Sub-basin (Nelson et al., 2022). Figure 7 shows two possible litho- and chemostratigraphic correlations between all relevant sections of the Urusis and Nomtsas formations in the Witputs and Vioolsdrif sub-basins. In this figure, the lithostratigraphic correlation of Witputs Sub-basin sections follows correlation 4. The two lithostratigraphic correlations between the Witputs and Vioolsdrif sub-basins have different implications for rates of sediment accumulation in the Witputs sub-basin (fig. 8).
Nelson et al. (2022) interpret the age of ash bed 1 at section 3 as a maximum depositional age and correlate the lower carbonate-dominated unit of the Neint Nababeep Plateau to the Huns Member (fig. 7). However, if ash beds 1–5 at section 3 approximate the ages of deposition, as originally proposed (Linnemann et al., 2019), then the implication is that the entire carbonate-dominated Huns-Spitskop members of the composite Neint Nababeep Plateau succession instead correspond to an expanded lateral equivalent of the Spitskop Member in the Witputs Sub-basin (correlation 5 in figs. 7C, 8). Whilst possible, this would imply that, in the Neint Nababeep Plateau succession, time-equivalent deposits to the Huns Member are dominantly siliciclastic (correlation 5 in figs. 7C, 8). In this correlation, the lithostratigraphic architecture, δ13Ccarb chemostratigraphy, and radiometric ages of the Neint Nababeep Plateau succession also appear consistent with lateral correlation to the Spitskop Member and lower Nomtsas Formation in sections 1–4, to the north, which were themselves deposited more slowly (correlation 5, figs. 7B, C, 8). Unfortunately, the largely invariant δ13Ccarb data throughout this interval make this correlation difficult to verify chemostratigraphically. Adopting correlation 4 for sections of the Witputs Sub-basin and either the original correlation of Nelson et al. (2022) or correlation 5 between the Witputs and Vioolsdrif sub-basins, also allows for basin-wide transgression and the contemporaneous development of pinnacle reefs of the lower Nomtsas Formation at section 4 and the Neint Nababeep Plateau (fig. 7C). This would, by extension, imply that the fragmented ash bed at section 4 (538.58 ± 0.19 Ma, Linnemann et al., 2019) is redeposited from the uppermost Spitskop Member. The veracity of correlation 5 requires future verification of the lithostratigraphic and radiometric correlation between sections 2 and 3 (e.g., Messori et al., 2021; Nelson et al., 2022), particularly by integration of data from core with outcrop.
5.5. Calibrating δ13Ccarb-sh data within the current terminal Ediacaran age framework
Nelson et al. (2022) record relatively stable, positive δ13Ccarb values throughout the Neint Nababeep Plateau section, with scattered negative values that may be laterally correlative with negative δ13Ccarb-sh values at section 3, documented herein (figs. 8, 9). This may also imply that the magnitude of negative δ13Ccarb-sh in Group 2 shales at section 3 is exaggerated relative to contemporaneous δ13Ccarb, possibly associated with sample impurity (Nelson et al., 2022). However, given the uncertainty in the lithostratigraphic correlation of section 4, it remains possible that δ13Ccarb-sh data recorded by Group 2 shales at section 4 reflect trends in seawater δ13CDIC.
A composite chemostratigraphic curve for the Nama Group has been constructed via visual alignment of δ13Ccarb data within the well-established litho- and chemostratigraphic framework of the Kuibis and Schwarzrand subgroups, and temporally constrained by all available radiometric ages within and between sections (fig. 9, full details of the methodological approach and associated uncertainties are provided in Bowyer et al., 2022; Bowyer, Uahengo, et al., 2023; Bowyer, Zhuravlev, et al., 2023). Within this framework, three chemostratigraphic alignments for section 4 result from lithostratigraphic correlations 1 to 4 of the Urusis Formation in the Witputs Sub-basin, and correlation 5 between the Witputs and Vioolsdrif sub-basins (fig. 9). In lithostratigraphic correlations 1 and 2, radiometric ages of 542.65 Ma and 540.099 Ma that bracket deposition of (at maximum) the upper Nasep, Huns and Feldschuhhorn members (Linnemann et al., 2019; Nelson et al., 2022) would correlate the δ13Ccarb-sh data at section 4 with radiometrically-constrained negative δ13Ccarb data in the A4 Member of the Ara Group, Oman (figs. 9A–C; Bowring et al., 2007). However, it now appears most likely that the radiometric age from the ash bed in the basal A4 Member does not constrain the onset age of this excursion (see Nelson et al., 2023), and that the putative ‘A4 excursion’ does not exist in this temporal window. By contrast, lithostratigraphic correlation 3 results in direct correlation between δ13Ccarb-sh data at sections 3 and 4 (fig. 9D). In both correlations, δ13Ccarb-sh values are significantly depleted relative to δ13Ccarb and cannot be considered informative of any trend in seawater δ13CDIC.
The 1n/BACE has not been recorded in the Witputs or Vioolsdrif sub-basins (fig. 7) and associated data therefore likely support a 1n/BACE onset after 538.04 ± 0.14 Ma (Nelson et al., 2022, 2023), consistent with age models C to F of Bowyer et al. (2022) and Bowyer, Zhuravlev et al. (2023) and age model K of Bowyer, Uahengo, et al. (2023). If δ13Ccarb-sh values of Group 2 shales at section 4 approximate the trend of seawater δ13CDIC, then lithostratigraphic correlation 4 suggests that the apparent negative δ13Ccarb-sh excursion recorded at section 4 may correlate with an unnamed excursion that predates the 1n/BACE, and is <538.04 Ma (fig. 9E; see Bowyer, Uahengo, et al., 2023; Nelson et al., 2022, 2023). These data may correspond in time with the negative δ13Ccarb excursion recorded in the lower Zuun-Arts Formation of Mongolia (e.g., E. F. Smith, Macdonald, et al., 2016; Topper et al., 2022) but the global significance of this excursion currently remains uncertain due to a dearth of radiometrically calibrated δ13Ccarb data ca. 538–535 Ma. However, similar to correlations 1-3, correlation 4 also implies that the δ13Ccarb-sh data in section 4 are depleted relative to contemporaneous δ13Ccarb. We note that lithostratigraphic correlation 4, whereby pinnacle reefs developed contemporaneously at section 4 and the Neint Nabebeep Plateau, remains possible even if δ13Ccarb-sh at section 4 are not considered useful for chemostratigraphy.
5.6. Potential biostratigraphic considerations and future target intervals
A preliminary global biostratigraphy has been constructed for the Ediacaran-Cambrian transition, which calibrates first and last appearances of key fossils directly within the global δ13Ccarb chemostratigraphic age framework (fig. 9C–F; Bowyer, Uahengo, et al., 2023). This age framework is undergoing continuous calibration with the publication of new data and lithostratigraphic considerations (e.g., Bowyer, Zhuravlev, et al., 2023; Nelson et al., 2023; Topper et al., 2022), and may be a useful predictive tool for targeting uncertain intervals of the stratigraphic record for geochemical and paleontological sampling. For example, the output of each age framework allows visualization of the series of biotic first and last appearances across the 1n/BACE interval and may help to inform targeted sampling of stratigraphic intervals in order to clarify temporal and spatial distributions of critical transitional biota.
The Bayesian age-depth model of Nelson et al. (2022) constrains ages for the last appearances of erniettomorphs and Cloudina in the Neint Nababeep Plateau composite section (fig. 9F) and also permits the calculation of uncertainties for each age. In contrast to Bayesian age-depth models, uncertainties in the precise ages of biotic first and last appearances are difficult to constrain by visual δ13Ccarb alignment alone, especially when considering local effects, including diagenesis, on regional δ13Ccarb variability. However, it is possible to make broad observations concerning the chemostratigraphic position of biotic first and last appearances relative to large magnitude δ13Ccarb excursions such as the 1n/BACE, if long-term trends in δ13Ccarb within each succession reflect changes in global seawater δ13CDIC.
The lowest occurrence of the cloudinid Zuunia chimidtsereni in the Zuun-Arts Formation of the Zavkhan Terrane, Mongolia pre-dates negative δ13Ccarb values associated with the regional expression of the 1n/BACE (Topper et al., 2022). Visual δ13Ccarb alignment of the Zavkhan and Laurentian successions of Nevada and California, with associated uncertainties, may permit a degree of temporal overlap between the maximum first appearance of Z. chimidtsereni and a terminal Ediacaran cloudinid assemblage, including (but not limited to) Saarina and Costatubus (fig. 9F; Bowyer, Uahengo, et al., 2023; Nelson et al., 2023; Selly et al., 2020; E. F. Smith, Nelson, et al., 2016; Topper et al., 2022). A conservative maximum age for the first appearance of morphologically simple anabaritids in Siberia is set by fossils assigned to Cambrotubulus within the Turkut Formation of the Olenek Uplift, which stratigraphically underlie the onset of a negative δ13Ccarb excursion interpreted to correlate with the 1n/BACE (fig. 9F; Bowyer, Zhuravlev, et al., 2023; Pelechaty et al., 1996; Rogov et al., 2015). Therefore, the present global age framework appears to suggest that the terminal Ediacaran cloudinid assemblage, and morphologically simplest anabaritids, post-date deposition of the lowermost Nomtsas Formation of the Nama Group (fig. 9).
Given that the maximum age for the onset of the 1n/BACE remains uncertain, and notwithstanding possible issues associated with endemism of terminal Ediacaran to lower Cambrian skeletal fossils, this age framework suggests that tubular fossils that predate the 1n/BACE in Laurentia (e.g., Selly et al., 2020) and Mongolia (e.g., Topper et al., 2022), may yet be identified within the youngest interbedded carbonate-siliciclastic units of the Nomtsas Formation that predate the regional erosional unconformity. According to lithostratigraphic correlation 4, herein, strata of the Nomtsas Formation in section 4 and the Neint Nababeep Plateau composite section, warrant continued, focused paleontological study
6. CONCLUSIONS
We present new δ13Ccarb-sh and δ18Ocarb-sh data from carbonate cements within 107 shale samples of the terminal Ediacaran Nama Group, Namibia. These data are compared with the published δ13Ccarb and δ18Ocarb record derived from interbedded and laterally coeval carbonates throughout the Nama Group succession. Our preliminary results suggest that δ13Ccarb-sh recorded by samples deposited within mixed carbonate-clastic settings can approach δ13Ccarb if CaCO3/TOC is sufficiently elevated. By contrast, shale samples with low CaCO3/TOC deposited in clastic-only settings record values of δ13Ccarb-sh that are significantly depleted relative to δ13Ccarb in coeval carbonate-clastic successions. Intervals of the Nama Group succession characterized by both δ13C-δ18O covariation and depleted δ13C values appear to be restricted to those stratigraphic units and shallower facies that were more susceptible to mixing with meteoric fluids during early diagenesis. Recognition of these regional diagenetic effects permits a more accurate and detailed assessment of regional and global δ13Ccarb chemostratigraphy.
Despite these observations, the utility of δ13Ccarb-sh to infill gaps in the existing δ13Ccarb record remains unclear, even when using shale samples with high CaCO3/TOC that were deposited in carbonate-clastic settings. For example, in the Nama Group, shale samples of the upper Schwarzrand Subgroup of the Witputs Sub-basin with elevated CaCO3/TOC ratios record values of δ13Ccarb-sh that appear depleted relative to δ13Ccarb from underlying and overlying carbonate rocks. Robust comparison of trends in δ13Ccarb-sh and δ13Ccarb between sections in this interval is also impaired by the recognition of several possible lithostratigraphic correlations for key sections of the upper Schwarzrand Subgroup. Perhaps the most parsimonious lithostratigraphic alignment for this interval appears to support the synchronous development of carbonate pinnacle reefs in the Witputs and Vioolsdrif sub-basins of the Nama Group, associated with a major flooding (transgressive) event. This alignment may also result in temporal overlap of new δ13Ccarb-sh data in one section of the Witputs Sub-basin with a negative δ13Ccarb excursion that predates the 1n/BACE, and warrants future targeted paleontological, geochemical and stratigraphic study.
ACKNOWLEDGMENTS
We acknowledge funding from UKRI grant NE/T008458/1 and Beca de Doctorado en el Extranjero, Becas Chile from ANID. We thank C. Chilcott, G. Sim, H. Mocke, C. Hoffmann, A. and S. Horn, U. Schulze Neuhoff, C. Husselman, B. Romer and L. Gessert. We thank two anonymous reviewers, Chris Jones, Associate Editor Tim Lyons for thorough and thoughtful comments that helped to significantly improve this manuscript. For the purpose of open access, the authors have applied a Creative Commons Attribution (CC BY) license to any Author Accepted Manuscript version arising.
SUPPLEMENTARY INFORMATION
https://doi.org/10.17632/x2dnn8xcvy.1
Editor: C. Page Chamberlain, Mark Brandon, Associate Editor: Timothy W. Lyons