Evidence for local carbon‐cycle perturbations superimposed on the Toarcian carbon isotope excursion

A Jurassic negative carbon isotope excursion (CIE), co‐evolved with Toarcian Oceanic Anoxic Event (OAE) at ~183 Ma, is suggested to be linked to a global carbon‐cycle perturbation and is well documented for Toarcian terrestrial fossil woods and marine sediments around the globe. A theoretically coupled δ13Ccarb‐δ13Corg pattern due to such dubbed global carbon‐cycle event from the negative CIE in Dotternhausen Toarcian stratigraphic profile (southwest Germany) is unexpectedly disturbed by two‐step δ13Ccarb‐δ13Corg decoupling in which the last step, upper in the stratigraphic order, is of higher magnitude. However, the trigger(s) for these sudden decoupling disturbances are still poorly constrained. Here, connecting new carbon and oxygen isotope data with documentary lipid biomarkers shows that the global carbon cycle during the Toarcian OAE was disturbed by enhanced green sulfur bacteria (GSB) metabolisms and early diagenesis at local scales. The first step δ13Ccarb‐δ13Corg decoupling was induced in the initial stage of the GSB bloom. The second step of much larger δ13Ccarb‐δ13Corg decoupling arising from a GSB prosperity was, however, exaggerated by early diagenesis through the respiration of sulfate‐reducing bacteria (SRB). Paleo‐geographically distinct localities of the Tethys region show contrasting decoupled δ13Ccarb‐δ13Corg patterns, which implies that the second‐order carbon‐cycle perturbations have pervasively and independently impacted the global carbon event during the Toarcian OAE.

Nevertheless, considering the confirmed global perturbation of the carbon cycle during the Toarcian OAE and the subsequent equilibration between the dissolved inorganic carbon (DIC) and dissolved organic carbon (DOC) reservoirs, a co-variation between δ 13 C carb and δ 13 C org would be expected (Knoll, Hayes, Kaufman, Swett, & Lambert, 1986). However, a number of studies have observed unpaired curves between δ 13 C carb and δ 13 C org records in Toarcian OAE stratigraphic profiles (Fu et al., 2016;Han, Hu, Kemp, & Li, 2018;Hermoso et al., 2012;Röhl, Schmid-Röhl, Oschmann, Frimmel, & Schwark, 2001), indicating that second-order carbon cycles might have played a significant role during the Toarcian OAE. In general, unmatched δ 13 C carb and δ 13 C org records can be diagnostic for post-depositional diagenetic alteration, hydrocarbon contamination, DOC contribution, terrestrial contamination Maloof et al., 2010), or metabolisms by Chlorobiaceae, such as green sulfur bacteria (GSB), which are anoxygenic phototrophs that are able to oxidize hydrogen sulfide (H 2 S) to sulfate (SO 4 2-) at low light availability (Riccardi, Kump, Arthur, & D'Hondt, 2007). However, how and to what extent second-order carbon cycling is responsible for the observed Toarcian OAE δ 13 C carb -δ 13 C org decoupling remain poorly constrained. Therefore, little attention has been attributed to carbon-cycle perturbations at a local scale. As such, a more comprehensive understanding of the features of the Toarcian negative CIEs might be, in part, hindered when correlating the Toarcian OAE chemostratigraphy worldwide.
This study re-investigates the nature of the Toarcian CIE for both the DIC and DOC reservoirs at the Toarcian OAE sedimentary section of Dotternhausen, Germany, which is characterized by well-preserved organic-rich sediments (Röhl, Schmid-Röhl, Oschmann, Frimmel, & Schwark, 2001). Thus, these sediments provide an especially valuable archive for tracking coeval carbon-cycle processes and evolution of the contemporaneous biosphere . To do so, new outcrop and drillcore samples from the Dotternhausen section were sampled at a high-stratigraphic resolution for integrated analyses of bulk δ 13 C carb , δ 13 C org , δ 18 O carb , total organic carbon on a carbonate-free base (TOC cf ), carbonate concentration, and major-and trace-elemental contents.
Additionally, we compiled and compared the previously published δ 13 C carb -δ 13 C org records from geographically different Toarcian OAE localities (Bilong Co section in eastern Tethys, Fu et al., 2016; Sancerre section in northwestern Tethys, Hermoso et al., 2012; and Nianduo section in southeastern Tethys, Han, Hu, Kemp, & Li, 2018), in order to investigate whether second-order carbon-cycle processes may have superimposed the Toarcian OAE δ 13 C carb and δ 13 C org records on a global scale.

| G EOLOG I C AL BACKG ROUND
During the Jurassic, most of present-day Europe was located on a broad and shallow continental shelf that deepened toward the southeast Tethys Ocean (Jenkyns, 2010) (Figure 1a). The coeval shallow shelf was marked by several sub-basins within which the water-mass circulation was hydrographically restricted (McArthur et al., 2008;Röhl, Schmid-Röhl, Oschmann, Frimmel, & Schwark, 2001). Moreover, global warming linked to the massive carbon emissions (Cohen, Coe, Harding, & Schwark, 2004;Hesselbo et al., 2000;Kemp, Coe, Cohen, & Schwark, 2005;Ruebsam, Münzberger, & Schwark, 2014;Ruebsam, Mayer, & Schwark, 2019) as well as tectonic activities (e.g., breakup of the Pangaea Supercontinent) may jointly have been responsible for a major marine transgression and deoxygenation during the early Jurassic (Toarcian), at least in the vastly extending shallow marine areas developed by the opening Tethys Ocean. These prominent changes in paleoceanic conditions and paleoclimate were likely responsible for the widespread deposition of organic-rich sediments during that time (e.g., Baudin, Herbin, & Vandenbroucke, 1990).
The Toarcian OAE Dotternhausen sedimentary section (48°13′32.60″N, 8°46′29.76″E), geographically situated in southwestern Germany, represents a mid-paleo-latitude basin, at the southern margins of European epicontinental seaways, which bridged the northwest Tethys Ocean and the Boreal Sea during the Early Jurassic (Ziegler, 1988) (Figure 1b,c). This section exhibits the well-studied, and laterally widespread, Lower Toarcian Posidonia Shales, which are world-famous for their exceptionally well-preserved fossils and their high organic matter content (up to 16 wt.% TOC) (Bour, Mattioli, & Pittet, 2007;Fantasia, Föllmi, Adatte, Spangenberg, & Montero-Serrano, 2018;Montero-Serrano et al., 2015;Ruebsam, Münzberger, & Schwark, 2014;Song, Littke, & Weniger, 2017). The sedimentary rocks are exceptionally well preserved and thus provide an ideal stratigraphic sequence to investigate the Toarcian OAE and compare it with the worldwide negative CIE records (Röhl, Schmid-Röhl, Oschmann, Frimmel, & Schwark, 2001). The Dotternhausen sedimentary section has been lithoand bio-stratigraphically well characterized, mainly including three ammonite zones (tenuicostatum, falciferum, and bifrons) and some subzones (elegantulum, exaratum, elegans, and falciferum; Figure 2), though the transition between the tenuicostatum and falciferum zones remains not precisely constrained due to a lack of standard fossils and/or to the existing stratigraphic gap in the Pliensbachian-Toarcian boundary interval (Kuhn & Etter, 1994). Furthermore, the whole section has been investigated and subdivided into three contrasting lithofacies units in ascending order: bioturbated mudstones (1,180-980 cm), laminated oil shales (980-550 cm), and bituminous mudstones (550-0 cm), respectively, in terms of the amounts of organic matter and the types of fabric (cf. Röhl, Schmid-Röhl, Oschmann, Frimmel, & Schwark, 2001). This study only shows the sedimentary strata from 1,180 to 560 cm (Figure 2), which include the Toarcian negative CIE. The laminated oil shale unit is largely limited to the uppermost part of tenuicostatum zone and the lower and middle parts of the falciferum zone (Figure 2), is marked by a complete lack of benthic fauna, and exhibits a distinct micro-lamination.

| MATERIAL S AND ME THODS
Samples of this study include gray mudstones, black shales, and diagenetic limestones taken from the stratigraphic section of the Dotternhausen profile ( Figure 2)  Survey; USGS) prepared as above and using the major element values reported in Govindaraju & Roelandts (1989). Secondary quality control standards were also prepared from the international reference materials (QS-1, OU-6, SCo-1, and AGV-2), and the precision range of major elements relative to the recommended values from GeoReM. Data are reported as elemental concentrations expressed in weight percent (wt.%), and generally, uncertainties for major elements are better than 1 wt.% (1σ).

| Trace elements analyses
200 mg of dried powder materials was weighted and ashed in a furnace at 500°C for 15 hr to oxidize organic matter. These ashed sample powders were re-weighed immediately upon cooling to determine the loss of weight during the ashing process. ~25 mg of ashed powders was weighed and digested in a 4 ml (4:1) HF-HNO 3 volumetric mixture on hot plates at 120°C for 2 days. Following digestion, the dissolved sample mixtures were evaporated, and the residues were reacted with 2 ml 6 m HCl at 120 C° for 24 hr to dissolve potential fluorides. Upon drying the samples, they were taken up in 1 ml 14.5 m HNO 3 , heated to 120°C for 1 hr and dried again to volatilize excess fluorine and chlorine. This last step was repeated to ensure full conversion of the samples to nitric form. The samples were then taken up in 1 ml 5 m HNO 3 and gravimetrically diluted to ~30 ml 2% HNO 3 stock solutions with a nominal dilution factor of ~1,000. Powders of rock reference materials used for calibration and quality control were also digested using this procedure. For analyses, all stock solutions were then further diluted with 2% HNO 3 to a nominal, gravimetric dilution factor of ~10,000. This 2% HNO 3 , as an internal standard during ICP-MS analysis, contains a mixed spike of 6 Li (~3 ng/g), In (~1 ng/g), Re (~1 ng/g), and Bi (~1 ng/g). All samples were measured using a Thermo Fisher Scientific iCap-Qc ICP-MS coupled to an ESI SC-2 DX autosampler with an ESI Fast uptake system equipped with a 4 ml sample loop. All sample liquids were introduced from the loop using the iCap-Q peristaltic pump (at 30-35 rpm) and aspirated with a PFA nebulizer into a Peltier-cooled cyclonic spray chamber. The nebulizer and cool gas flow rates were typically ~1 and 14 L/min, respectively, and the interface was configured with Ni sampler cone, with a Cu core and Ni skimmer cone with a high-matrix insert.
The analytical procedure was analogous to that described in previous studies (Albut et al., 2018;Babechuk, Widdowson, Murphy, & Kamber, 2015;Kamber, Webb, & Gallagher, 2014). Oxide/hydroxide interference rates of Ba on Eu, Nd on the MREE to HREE (Gd, Tb, Dy, Er), Zr, on Ag, and the isobaric overlap of 160 Dy on 160 Gd were quantified ahead of each experiment. Remaining interference "rates" were determined according to a previous quantification scaled to the daily Nd oxide on Gd production rate (Aries et al., 2007;Ulrich, Kamber, Woodhead, & Spencer, 2010). A daily measurement of the 6 Li/ 7 Li ratio in unspiked USGS standard AGV-2 was also applied to determine a correction factor for the contribution of natural 6

| TOC cf and carbonate content analyses
To avoid biases related to carbonate (CaCO 3 ) dilution, this study reports carbon-free total organic carbon contents (TOC cf  For the carbonate content measurement, the treatment procedures for the studied samples are described below for δ 13 C carb (Section 3.3). An almost perfect linear correlation (R 2 = .9917) between samples weight (for samples between 10 and 150 mg mg CaCO 3 ) and the 2nd peak area on the Finnigan MAT 252 gas source mass spectrometer (for further details see Spötl & Vennemann, 2003) allows for the calculation of the carbonate content in each specific sample. These calculated carbonate contents were then divided by total sample weights and multiplied with 100 to obtain the carbonate contents of the bulk samples in %. in an oxidation tube and at 650°C in a reduction tube, before they were cooled in a watertrap and transferred through a GC gas column into the mass spectrometer. Sample organic carbon was measured relative to an internal acetanilide standard which is calibrated against an in-house (e.g., Laaser marble) and international reference material (USGS24, δ 13 C org = −16.00‰). δ 13 C org results have an external reproducibility of ±0.1‰ (1σ) for shale and ±0.2‰ (1σ) for limestones.

| Element geochemistry
The concentrations of major element oxides and trace element data of the measured samples from Dotternhausen are reported in Table 1.

| TOC cf , carbonate and isotopic data
The TOC cf contents and bulk stable isotope (carbon and oxygen) data are given in Table 2 and are plotted in Figures 6 and 7.
Carbonate concentrations are reported in  Figure 6). Unterer Stein shows high carbonate contents with an average of 92 wt%, the rest of this studied samples exhibit a variation between 3 to 88 wt.%. According to the distribution of organic-rich sediments in the Dotternhausen section, this study places the onset of the Toarcian OAE to a depth of 980 cm in our profile. Four data points at depths between 1,012 cm and 986 cm reveal constant δ 13 C carb and δ 13 C org values prior to the Toarcian OAE ( Figure 7). The negative δ 13 C carb and δ 13 C org excursions rang-

| Negligible effects of terrigenous-sourced organic carbon, thermal maturity, and hydrothermal fluids on δ 13 C org across the Toarcian CIE
Due to the contrasting δ 13 C signatures between terrestrial and marine organic matters, terrestrial-sourced organic carbon can lead to a significant influence on δ 13 C org record of marine sediments depos- can considerably contribute to lower this index (Dembicki, 2009;de Kock et al., 2017), the Dotternhausen black shale and limestone beds exhibit relatively invariable HI values of ~580 mg HC/g TOC and high TOC cf contents of ~13 wt.% ( Figure 6). This demonstrates that the organic matter was well preserved and of predominantly marine origin-here, bolstered by the oil shale character of the studied section (Röhl, Schmid-Röhl, Oschmann, Frimmel, & Schwark, 2001).   and references therein). Yttrium is less effectively scavenged from seawater than the trivalent REE Ho, leading to high residual Y/Ho ratios in the oceans. High-temperature hydrothermal fluids are also characterized by chondritic Y/Ho ratios, but exhibit higher Eu/Sm ratios than oceanic or continental crust due to the higher solubility of reduced divalent Eu compared to trivalent Sm. The Dotternhausen black shales have Eu/Sm ratios of 0.21 and 0.23, which are within the range of the modern shale average, but an order of magnitude lower than typical high-temperature hydrothermal fluids at midocean ridges (Figure 4b). This observation demonstrates negligible alteration by hydrothermal activity on the studied black shale facies.
In summary, the δ 13 C org records of Dotternhausen Toarcian CIE samples were not significantly influenced by terrigenous-derived organic matters, contaminated by recycling of old sediments, and altered by hydrothermal activity.

| Marine redox condition record in the Dotternhausen succession during the Toarcian CIE
During the Toarcian CIE high rates of organic carbon production oc-

F I G U R E 7
Stratigraphic profiles of δ 13 C carb , δ 13 C org , Δ 13 C (Δ 13 C = δ 13 C carb − δ 13 C org ) and δ 18 O in the Dotternhausen section. Stratigraphic distributions of aryl isoprenoid (GSB biomarker) abundance are from . The red dashed arcs denote the theoretic Toarcian CIE curves without suffering local-scale carbon-cycle perturbations. The blue lines represent the 5-point moving average of the isotopic data. Int.: interval. White rectangles particularly represent the isotopic data from limestone beds 2004; . These were commonly tied to high marine productivity rates (e.g., Jenkyns, 2010;Röhl, Schmid-Röhl, Oschmann, Frimmel, & Schwark, 2001) caused by excessive nutrient input from enhanced continental weathering (e.g., Cohen, Coe, Harding, & Schwark, 2004;Percival et al., 2016). It was proposed that the generated organic matter subsequently suffered aerobic degradation during which the seawater oxygen inventory was massively depleted (Röhl, Schmid-Röhl, Oschmann, Frimmel, & Schwark, 2001). Furthermore, at a high sea-level stand during the Toarcian CIE period (Ruebsam, Mayer, & Schwark, 2019), the strong watercolumn stratification and hydrological restriction have prevented efficient ventilation between bottom waters and the outside open ocean, which eventually promoted the formation of anoxic benthic seawaters in the northwestern Tethys shelf-sea settings (e.g., Guillaume and Yannick, 2012;Hermoso et al., 2013;McArthur et al., 2008;Ruvalcaba Baroni et al., 2018;Röhl, Schmid-Röhl, Oschmann, Frimmel, & Schwark, 2001). The euxinic depositional condition for the Dotternhausen black shale facies was determined by the existence of aryl isoprenoids ( Figure 5), biomarkers for GSB (cf. Schwark , which implies relatively high aqueous hydrogen sulfide (H 2 S) concentration in the seawater. Such a seawater redox structure is further supported by the strong authigenic enrichment of redox-sensitive elements, that is, Mo, U, and V ( Figure 5). For the Dotternhausen sedimentary profile, a gradual increase in aryl isoprenoid concentration is exhibited in its lower part, followed by a decreasing trend in its upper part ( Figure 5). The GSB, which are anoxygenic autotrophs, are able to oxidize H 2 S to sulfate (SO 4 2-) (e.g., Riccardi, Kump, Arthur, & D'Hondt, 2007). A bloom of GSB in the seawater will thus enhance the assimilative consumption of H 2 S (Hurse, Kappler, & Keller, 2008), which can, to a certain extent, lower its concentration in the bottom anoxic/euxinic water mass, and thus deepen the seawater sulfidic oxygen minimum zone (chemocline).
Alternatively, molecular oxygen generated by oxygenic photosynthesizers can also oxidize H CIEs (Hesselbo et al., 2000;Kemp, Coe, Cohen, & Schwark, 2005;Kuypers, Pancost, & Damste, 1999) and Neoproterozoic Snowball Earth events (Och & Shields-Zhou, 2012;Sahoo et al., 2012) were commonly associated with changes in coeval δ 13 C carb and δ 13 C org records. Generally, a common approach used to distinguish whether a carbon cycle is global-or local scale is to evaluate the nature of the concurrently preserved δ 13 C carb and δ 13 C org signatures in sediments (Knoll, Hayes, Kaufman, Swett, & Lambert, 1986). Unpaired δ 13 C carbδ 13 C org is often suggested to have been impacted by second-order local-scale carbon-cycle perturbations (e.g., Jiang et al., 2012;Meyer et al., 2013); paired δ 13 C carb -δ 13 C org is usually attributed to changes in the global carbon cycle (e.g., Ader et al., 2009;Li et al., 2018). It is widely accepted that the global injection of isotopically light carbon (CH 4 and/or CO 2 ) into the coeval atmosphere-ocean system during the early Toarcian led to widespread negative CIEs recorded in northwestern Tethys (e.g., Hesselbo et al., 2000, Hesselbo, Jenkyns, Duarte, & Oliveira, 2007Svensen et al., 2007), eastern Tethys (Fu et al., 2016), southwestern Tethys (Ruebsam et al., 2020), southeastern Tethys (Newton et al., 2011;Han, Hu, Kemp, & Li, 2018), northwestern Panthalassa (Izumi, Kemp, Itamiya, & Inui, 2018), and northeastern Panthalassa (Caruthers, Gröcke, & Smith, 2011). If the widespread Toarcian CIE was thus a global alteration of the carbon cycle, it is expected to be accompanied by a constant Δ 13 C carb-org between DIC and DOC reservoirs. In this study, negative CIEs are observed in both δ 13 C carb and δ 13 C org records from the Dotternhausen section (Figure 7) of the northwestern Tethys (Figure 1). Pronounced  (Figures 7 and 10a), which most likely reflect global-scale carbon-cycle perturbations during the Toarcian. However, a prominent two-step δ 13 C carb -δ 13 C org decoupling signature, along with two distinct magnitudes of decreased Δ 13 C carb-org values, is observed around the climax of the CIE (interval 2a and 2b; Figures 7 and 10a), which points to a local-scale disturbance of the carbon cycle. The first step in interval 2a is of smaller magnitude compared to the much larger second step in interval 2b (Figures 7 and 10a).

| Possible mechanism(s) for the local-scale
δ 13 C carb -δ 13 C org decoupling

Interval 2a
The negative excursion in Δ 13 C carb-org is more influenced by the variation in the δ 13 C carb than δ 13 C org values (Figure 7). However, carbonates in the black shales of interval 2a represent a combination of biogenic calcites related to calcareous phytoplankton (Frimmel, Oschmann, & Schwark, 2004) and disseminated authigenic grains.
A rough estimate of contribution from organic matter remineralization (C org ) and seawater inorganic carbon (C sw ) in carbonate can be obtained with the following equation (cf. Heimann et al., 2010;Konhauser et al., 2017):  This implies that carbon components of these carbonates are contributed by ~80% of DIC and only ~20% of remineralized organic carbon ( Figure 10b). The remineralization of organic matter occurs through anaerobic or microaerophilic microbial processes that require electron acceptors (e.g., Heimann et al., 2010;Konhauser et al., 2017 and references therein), which here are sulfate and possibly Mn-oxyhydroxides (see Section 5.2). These processes produce carbonates with lighter δ 13 C carb values compared to the ambient DIC composition.
It can be noticed that the shift toward negative δ 13 C carb values in interval 2a is accompanied by a very minor shift of δ 13 C org toward heavier values (Figure 7). The metabolisms of anoxygenic photoautotrophs (GSB) yield 13 C-enriched biomass through primary productivity (Takahashi, Kaiho, Oba, & Kakegawa, 2010;van Breugel, Baas, Schouten, Mattioli, & Sinninghe Damsté, 2006) and can explain the small increase in δ 13 C org values observed in interval 2a. However, since this δ 13 C org increase is minimal, it does not have a major impact on the overall Δ 13 C carb-org negative excursion-though this interval represents the early stage of the GSB bloom. In view of this, carbonates in the interval 2a precipitated in equilibrium with seawater isotope composition, but were further aided by a small contribution of isotopically lighter carbon generated through microbial-induced organic carbon remineralization (Figure 8a). This small contribution of light carbon isotopes likely explains the smaller magnitude of the Δ 13 C carb-org negative excursion recorded in this interval.

Interval 2b
The negative Δ 13 C carb-org excursion in this interval is accompanied by a shift to heavier δ 13 C org and lighter δ 13 C carb values (Figure 7).
Assuming that diagenesis of organic-rich sediments is widely described to have a minimal effect on δ 13 C org values (e.g., Jiang et al., 2012;Watanabe et al., 1997), it cannot thus explain the positive shift in δ 13 C org values recorded in interval 2b (see Section 5.1). This δ 13 C org record rather reflects a pristine isotope signal imparted by the dominant productive marine biota. As stated above (see interval 2a), the biomass produced by GSB is 13 C-enriched. Across interval 2b the height of GSB activity is obviously indicated by the maximum concentration of aryl isoprenoid biomarkers, which generated more isotopically heavy biomass and the associated δ 13 C org positive excursion ( Figure 7). The δ 13 C carb -δ 13 C org decoupling in interval 2b can thus partly be explained by the climax of GSB metabolism.
Importantly, interval 2b includes changes in lithofacies from carbonate-bearing laminated black shales in its lowermost part to the diagenetic Unterer Stein carbonate bed upward and then back to carbonate-bearing laminated black shales (Figure 7). Therefore, the light carbon isotope composition of the Unterer Stein may be dominantly attributed to diagenetic carbonate precipitation through microbial sulfate reduction (MSR) with a minor contribution from dissimilatory Mn-oxyhydroxides reduction (see Section 5.2). Such diagenetic processes produce carbonates with light δ 13 C carb values, which may enhance the negative Δ 13 C carb-org excursion in this interval (cf. Heimann et al., 2010;Konhauser et al., 2017). However, the δ 13 C carb versus δ 18 O diagram indicates that the limestone bed samples fall between C org : C sw = 1:1 and C org : C sw = 1:2 lines (Figure 10b) Figure 8b).
In summary, the coeval decreases in δ 13 C carb and increases in δ 13 C org values, which contributed to the two-step Δ 13 C carb-org decoupling in interval 2, reflect a combined effect of enhanced activity of GSB and sulfate-reducing bacteria (SRB) (Figure 8). The GSB bloom caused high H 2 S consumption in the euxinic bottom water and likely expanded the chemocline near to the water-sediment interface. This allowed higher sulfate flux to the anoxic sediment pore waters and ultimately caused carbonate precipitation aided by higher rates of SRB-induced organic carbon remineralization. In contrast, the following coeval increase in δ 13 C carb and decrease in δ 13 C org values, which ended this two-step Δ 13 C carb-org decoupling in the upper part of interval 2b, were mainly caused by a decline of GSB activity as a consequence of exhausted H 2 S availability in the water column. However, interval 2b represents the peak of the GSB bloom and explains the more pronounced Δ 13 C carb-org decoupling in this interval compared to the lower 2a. Interestingly, this GSB bloom in interval 2 might have consumed enough H 2 S in the Toarcian OAE water column to re-establish environmental conditions favorable for an increased expansion of aerobic metabolisms, F I G U R E 9 Crossplots of δ 13 C carb -δ 13 C org in different intervals of the Dotternhausen carbon isotopic profiles that is, oxygenic photosynthesis, and ultimately trigger the progressive recovery to the end of the Toarcian CIE observed in the overlying interval 3 (Figure 7).

| Implication for ecosystem structure on localscale carbon-cycle perturbation
A previous study has shown that co-varying δ 13 C carb and δ 13 C org imply that organic carbon in marine sediments is mainly derived from primary (photosynthetic) production without substantial post-depositional alteration (cf. Jiang et al., 2012). Here, it is shown that an abrupt carbon-cycle perturbation by intense metabolisms of GSB has the capacity of decoupling the primary carbon isotopic signals. In Earth's history, the dramatic rise in GSB activity is not only limited to Toarcian OAE epicontinental seas (Pancost et al., 2004;Röhl, Schmid-Röhl, Oschmann, Frimmel, & Schwark, 2001;Saelen, Tyson, Telnaes, & Talbot, 2000;van Breugel, Baas, Schouten, Mattioli, & Sinninghe Damsté, 2006;Xu et al., 2018), but also significant in a number of ancient marine realms during the end-Ordovician, end-Devonian, end-Permian and end-Triassic mass-extinction events, which were, at least in part, forced by an increase in H 2 S in the seawater (e.g., Joachimski et al., 2001;Riccardi, Kump, Arthur, & D'Hondt, 2007;Richoz et al., 2012). Decoupled δ 13 C carb -δ 13 C org signals across these mass extinctions broadly coincide with enhanced GSB activities (van de Schootbrugge & Gollner; 2013, and references therein), which implies the potential role of GSB in perturbing coeval carbon cycling. Thus, to investigate the possible mechanisms for decoupling δ 13 C carb -δ 13 C org in sedimentary rocks, carbon-cycle turnover arising from a sharp proliferation of GSB may play a major role and should not be neglected.

| Tethys-wide but locally variable second-order perturbations of the Toarcian carbon cycle
Theoretically, coupled δ 13 C carb -δ 13 C org signatures are expected in all existing Toarcian OAE sedimentary sections due to the global scale of this event (described in Section 5.3.1). Opposed to this expectation, however, decoupled δ 13 C carb -δ 13 C org signals (yellow areas in Figure 11a and yellow squares in Figure 11b Baroni et al., 2018) that prevented water mass exchange between the two basins.
However, further detailed work beyond the scope of this study is needed to understand the δ 13 C carb -δ 13 C org decoupling patterns observed in the various Toarcian OAE sections.

| CON CLUS ION
The black shale facies in the Dotternhausen section was deposited in a redox stratified marine environment and was not significantly impacted by terrestrial-sourced organic carbon, thermal maturity, F I G U R E 11 δ 13 C carb -δ 13 C org decoupling signals (yellow areas in panel (a) and yellow squares in panel (b)) from geographically distinct localities (Bilong Co in eastern Tethys, Fu et al., 2016;Sancerre in northwestern Tethys, Hermoso et al., 2012;Nianduo in southeastern Tethys, Han et al., 2018) are identified across Toarcian CIE shown in panel A. Blue and orange colors in both panel A and B standing for the observed coupled δ 13 C carb -δ 13 C org signals across Toarcian CIE are comparable to the marked interval 1 and 3 in the studied Dotternhausen section. Fm., formation; Plien., Pliensbachian; Serp., Serpentinum; Spin., Spinatum and hydrothermal fluids. Two prominent decoupled δ 13 C carb -δ 13 C org signals superimpose the Toarcian CIE in the Dotternhausen section, southwest Germany. The stratigraphically first δ 13 C carb -δ 13 C org decoupling (shown in black shales in interval 2a) was of a small magnitude and most likely caused by a carbon-cycle disturbance from the onset of GSB metabolic activity. The second and in magnitude much larger decoupling in interval 2b resulted from a GSB bloom, but was further exaggerated by early diagenetic effects. Thereby, dissimilatory respiration of SRB led to the formation of 12 C-enriched authigenic calcites within the sediment throughout organic matter remineralization. This process amplified the δ 13 C carb -δ 13 C org decoupling between the diagenetic carbonate (Unterer Stein) and the remaining organic matter. The δ 13 C carb and δ 13 C org profiles from geographically distinct Toarcian OAE localities (e.g., Bilong Co, Nianduo, and Sancerre) in the Tethys region show strong variability in their δ 13 C carb -δ 13 C org decoupling signals, which indicates the substantial impact of local carbon-cycle perturbations on the global carbon cycle during the Toarcian CIE.

ACK N OWLED G M ENTS
We are grateful for the help and technical support from Elmar Reitter and Bernd Steinhilber in the laboratory. We thank Holcim GmbH Dotternhausen for access to their quarry and for providing drillcore materials for this study. Y.W. thanks the China Scholarship Council (CSC) for his financial support. We are grateful to the two anonymous reviewers and to the Editor N. Planavsky for their comments and suggestions that highly improved the original manuscript. We would also like to thank the Editor K. Konhauser for handling this manuscript.
Open access funding enabled and organized by Projekt DEAL.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest and no compelling financial interests.