Restricted Oxygen‐Deficient Basins on the Northern European Epicontinental Shelf Across the Toarcian Carbon Isotope Excursion Interval

The worldwide recognition of the Toarcian carbon isotope excursion (T‐CIE) in organic‐rich sedimentary rocks has been linked to an oceanic anoxic event (OAE) which implies the world's deep oceans were anoxic ∼183 Ma. The majority of independent redox observations used to build this argument were mainly obtained from T‐CIE organic‐rich sediments deposited on northern European epicontinental shelf. However, increasing evidence has shown that this shelf had limited connection with the open ocean, making it unsuitable for reconstructing the T‐CIE ocean redox structure. To unveil such controversy, we present integrated δ15Nbulk and δ15Nker from Dotternhausen profile, Germany, combined with literature data from other T‐CIE profiles. Both δ15Nbulk and δ15Nker values are predominantly between +0.3 and +2.5‰. These positive near‐zero δ15N values imply enhanced N2 fixation by cyanobacteria using molybdenum (Mo)‐based nitrogenase to compensate bioavailable N loss following quantitative denitrification and/or anammox in a strongly redox‐stratified marine setting. Such N isotope composition contradicts the typical sedimentary δ15N values (>3‰) induced by partial water‐column denitrification and/or anammox in modern‐ocean oxygen minimum zones. We rather propose the existence of local oxygen‐deficient basins on northern European epicontinental shelf where dissolved N underwent extensive denitrification and/or anammox causing bioavailable N deficiency. Mo‐based diazotrophy thus played a critical role in discriminating N isotope compositions among multiple hydrographically restricted T‐CIE marginal basins. Restricted oxygen‐depleted environments on the northern European epicontinental shelf unlikely represent the open‐ocean redox landscape. The existence of the global OAE thus needs comprehensive redox investigations on Tethys and/or Panthalassa deep‐sea T‐CIE successions to validate.

In this study, the biogeochemical cycle of nitrogen (N), a redox-sensitive and bio-limiting element, is used in the attempt to unravel the controversy of the Toarcian oceanic redox structure. Before the T-CIE and inferred co-evolved OAE, the oxygenated ocean is expected to have built up a large and steady nitrate (NO 3 − ) reservoir. When entering the T-CIE, if sedimentary deposition occurs in open marine environments, the extensive seawater deoxygenation (OAE) would partially consume NO 3 − through water-column denitrification, the most prominent oceanic NO 3 − removal process (Sigman & Fripiat, 2019;Stüeken et al., 2016).
Considering that partial denitrification and/or anammox preferentially releases 14 N as the form of N 2 /N 2 O to the atmosphere, resulting in higher δ 15 N values of the residual N pool, organic matter (OM) deposited in open marine environment records such 15 N-rich seawater signal through assimilation and imparts δ 15 N values ≥3‰ in sediments (e.g., Garvin et al., 2009;Gilleaudeau et al., 2020;Stüeken et al., 2016;Tesdal et al., 2013). This is because a large dissolved marine N reservoir is required to record the isotopic fingerprints of its redox-driven cycle in co-deposited sediments (Stüeken et al., 2016). In contrast, if hydrographic water-mass restriction represents the main depositional settings of T-CIE sedimentary rocks, quantitative denitrification and/or anammox is most likely to occur as a result of strongly redox-stratified water column Wang et al., 2020). Biomass deposited under such conditions is expected to preserve the δ 15 N value of the main input source that is, N 2 fixation (Stüeken et al., 2016). Collectively, whether the northern European T-CIE epicontinental shelf successions were deposited in marine environments well connected to the open ocean or in very restricted environments can thus be reliably validated through evaluating the contemporaneous sedimentary N isotope records. This study reports high resolution of whole rock and separated kerogen δ 15 N (δ 15 N bulk and δ 15 N ker ) data from the well-recognized T-CIE stratigraphic profile of Dotternhausen, southwestern (SW) Germany, to address the evolution of marine N biogeochemical cycling and ultimately shed new light on the expansion of anoxia in the coeval ocean.

Geological Background and the Studied Sedimentary Section
Early Jurassic experienced a warm period induced by a global temperature rise, reaching ∼6ºC-10ºC higher than the present day (Chandler et al., 1992;Dera et al., 2011;Rosales et al., 2018;Svensen et al., 2007). This global warming event was suggested to be linked to the massively released CO 2 owing to the Karoo-Ferrar large igneous province emplacement (Pálfy & Smith, 2000;Svensen et al., 2007) and/or CH 4 released from Gondwanan organic-rich shales or methane clathrates (e.g., Hesselbo et al., 2000;Kemp et al., 2005;McElwain et al., 2005). Across the greenhouse early Jurassic (Toarcian) interval, the hydrological cycle was intensified, followed by the formation of humid climate conditions and the enhanced chemical weathering rates (Brazier et al., 2015;Cohen et al., 2004;Fantasia et al., 2018;Hermoso & Pellenard, 2014;Montero-Serrano et al., 2015;Them et al., 2017). In addition, the majority of the present-day Europe was situated on the northern European epicontinental shelf which deepened toward the southeastern part of the contemporaneous Tethys Ocean (Jenkyns, 2010) (Figure 1a). However, the northern European epicontinental shelf during the Early Toarcian was characterized by multiple isolated basins/sub-basins in which the frequency of the deep-water renewal was to some extent restricted (e. g., McArthur, 2019;McArthur et al., 2008;Röhl et al., 2001). Moreover, the warming event during the early Toarcian tied to the massive emission of the greenhouse gases (Hesselbo et al., 2000;Kemp et al., 2005;Ruebsam et al., 2019Ruebsam et al., , 2020 together with the coeval large-scale tectonic activities for example, the breakup of Pangaea supercontinent, likely jointly led to seawater transgression and deoxygenation in the northwestern Tethyan shallow shelf marine areas. The pervasive deposition of the organic-rich sedimentary rocks may thus be related to the pronounced change in the paleoceanographic and paleoclimatologic conditions during the early Toarcian period (e.g., Baudin et al., 1990;Jenkyns, 2010).
The studied Dotternhausen sedimentary section (48°13′32.6″N, 8°46′29.8″E) is geographically located in the present-day SW Germany, and was deposited in the northern European epicontinental shelf seaway throughout the Early Toarcian interval (Ziegler, 1988) (Figures 1b,and 1c). This sedimentary section belongs to the well-recognized lower Toarcian Posidonia Shales that are enriched in high abundance of well-preserved OM and are widely distributed over Europe (Bour et al., 2007;Fantasia et al., 2018;Montero-Serrano et al., 2015;Song et al., 2017 Wang et al., 2020). This Dotternhausen section has been also extensively studied and been well characterized in terms of lithostratigraphy and biostratigraphy (Röhl et al., 2001;Röhl & Schmid-Röhl, 2005). From the bottom up, this section was divided into three ammonite zones tenuicostatum, falciferum and bifrons, and was further subdivided into several subzones including elegantulum, exaratum, elegans, falciferum, commune, etc (cf. Röhl et al., 2001). In addition, three lithologic units for the Dotternhausen section have been defined in the light of the distinct abundances of the OM and siliciclastic component and different fabric types (Röhl et al., 2001;van Acken et al., 2019;Wang et al., 2020). They are bioturbated light-gray (organic-lean) shales from the depth of 1,180-980 cm, laminated fine-grained organic-rich shales and marlstones from 980-550 cm, and bituminous light-gray (organic-lean) shales from 550 to 0 cm, respectively (Röhl et al., 2001). The bioturbated light-gray shale unit starting with the limestone bed Spinatum Bank is abundant in benthic fauna and contains two intercalated thin organic-rich black shale beds ( Figure 2). The fine-grained organic-rich sedimentary rocks deposited during the T-CIE are depleted in benthic fauna, and marked by high total organic carbon (TOC) concentration. These organic-rich sedimentary rocks were not metamorphosed and only experienced diagenesis in the oil window which is reflected by their persistently high hydrogen index(HI) values (580 mg hydrocarbon (HC)/g TOC on average) and their oil shale character (cf. Frimmel et al., 2004;Röhl et al., 2001;Wang et al., 2020   organic-rich rock facies, Unterer Schiefer and Mittlerer Schiefer Members are partitioned by the Unterer Stein limestone bed. A further description of the Dotternhausen sedimentary section with aspect to its paleontological and sedimentary characteristics can be found in previous studies (cf . Röhl et al., 2001;Röhl & Schmid-Röhl, 2005).

Sampling
In this study, the Dotternhausen sedimentary section, SW Germany, from 1,050 to 610 cm in stratigraphic depth containing the T-CIE was investigated. This stratigraphic interval starts with the upper semicaletum subzone and ends with the Oberer Stein in the upper elegans subzone (Figure 2). In detail, it comprises a lower 0.8 m-thick bioturbated light-gray (organic-lean) shale unit preceding the T-CIE and an upper 3.6 m-thick interbedded organic-rich shale and marlstone unit deposited during the T-CIE. High-resolution sampling at a small spacing (3.5 cm on average) was performed on the organic-lean shales and find-grained organic-rich sedimentary rocks from the two units in this study. A total of 31 samples were collected and used for further geochemical analyses.

Analytical Methods
To avoid potential influence from weathered materials and any handling contamination, we cut the collected sedimentary rock samples into 2 cm-thick slabs with a water-cooled saw. The slabs were cleaned and then rinsed by distilled water and analytical grade and glass distilled ethanol and dichloromethane, respectively. Following preparing the cleaned sample slabs, they were dried at 50°C for one day, and then fragmented and powdered using an agate mill. The obtained sample powders were subsequently analyzed for δ 15 N bulk and δ 15 N ker , and δ 13 C for TOC and extracted kerogen (δ 13 C org and δ 13 C ker ) at the laboratory of the Institute of Earth Surface Dynamics, University of Lausanne. The analytical procedure is based on the previous description provided by Spangenberg and Frimmel (2001), Spangenberg and Herlec (2006), and Spangenberg and Macko (1998).
The measurement of all C and N data was carried out by combined elemental analyzer (Carlo Erba 1,108) and isotope ratio mass spectrometry (Delta V Plus; Thermo Fisher Scientific) (EA-IRMS) under a continuous helium flow condition through a ConFlo III open split interface (cf. Spangenberg et al., , 2014. For TOC, total nitrogen (TN), δ 13 C org , and δ 15 N bulk analyses, powdered whole rock materials were treated by a drop-wise addition of 10% HCl to remove all inorganic carbon. For the analyses of N ker , δ 13 C ker , and δ 15 N ker , powdered whole rock samples were gradually treated to extract the kerogen fraction that cannot be dissolved in organic solvents, non-oxidizing acids and bases. On the basis of the kerogen extraction procedure described by Durand and Nicaise (1980), this study used Soxhlet extraction method in which mixed methanol and dichloromethane were added to remove bitumen (a soluble fraction in OM), 6 M HCl was added to remove carbonate, sulfide, sulphate, and (hydro)oxide, and a mixed acid composed of 40% HF and 6 M HCl was added to remove siliciclastic fraction. These stepwise treatments were all carried out at the temperature of 65ºC-70°C, and were accelerated through using a PTFE-coated magnetic stirrer. After removing the supernatant, the residues were thoroughly washed and purified using deionized water and ultrapure water produced by Direct-Q UV3 Millipore®, respectively, and then dried at the temperature of 40°C. Following drying the solid residues, mineralogical analysis performed by a Thermo Scientific ARL X-TRA Diffractometer showed that all the silicate fraction in the whole rock samples has been removed completely.
The measured stable C and N isotope compositions are reported using delta (δ) notation in per mil relative to their standards which are Vienna Pee Dee Belemnite and atmospheric nitrogen (Air-N 2 ), respectively. They are thus defined as: δ 13 C = (( 13 C/ 12 C) sample /( 13 C/ 12 C) VPDB − 1) × 1,000 and δ 15 N = (( 15 N/ 14 N) sample / ( 15 N/ 14 N) Air-N2 − 1) × 1,000. In this study almost all the C and N isotope data were performed in duplicate, and the analytical reproducibility of the repeated sample measurements for the C and N isotope values conducted by the EA-IRMS was, respectively, better than ± 0.1% and ± 0.3%. The TOC abundance for the whole-rock samples, and the abundance of the TN bulk for whole-rock samples and of N ker for kerogen extracts were obtained from the peak areas of the major isotopes through the use of the calibrations for the δ 13 C org , δ 15 N bulk , and δ 15 N ker values, respectively. The analytical repeatability was better than ± 0.2 wt.% for the contents of TOC, TN bulk and N ker .

Evaluation of Primary Nitrogen Isotopic Signals
Continental run-off of OM from land plants may provide an additional source of N and may influence the isotope composition of marine sediments. In this study, the Dotternhausen sedimentary section from 1,050 to 980 cm is composed of organic-lean shales deposited prior to the T-CIE (Figure 2). These shale samples contain a considerable component of land plant-derived OM supported by low HI value (<400 mg HC/g TOC) and high oxygen index values (up to 100 mg CO 2 /g TOC) (cf. Frimmel et al., 2004;Wang et al., 2020). In contrast, from 986 to 600 cm of this section, the OM in T-CIE fine-grained organic-rich sedimentary rocks ( Figure 2) was primarily of marine origin Wang et al., 2020) and thus likely reflect N biogeochemical processes in marine environments during the T-CIE.
The δ 15 N of sedimentary rocks is a robust proxy to record the primary biomass signal of living organisms if it is not substantially imprinted by allochthonous inorganic N, diagenesis, metamorphism or metasomatism (e.g., Ader et al., 2006;Stüeken et al., 2016). Nitrogen in the sedimentary rocks is primarily present in organic-and clay-bound forms (mainly as ammonium (NH 4 + )  (Table 1) imply the negligible alteration on the primary δ 15 N by inorganic N. The positive correlation between TOC and TN bulk (R 2 = 0.43; p = 0.00; Figure 3a) observed for the studied T-CIE samples implies that most N was derived from marine organic matter. Furthermore, the lack of correlation between TN bulk and detrital indicator aluminum (Al) concentration (Figure 3b) indicates the N contribution from detrital input does not compromise the robustness of the primary N isotope signal.  Table 1 Carbon and Nitrogen Data for the Dotternhausen Samples diagenesis that is accompanied by an isotopic fractionation of <1‰ (Ader et al., 2006;Altabet et al., 1999;Stüeken et al., 2016). In this study, early diagenetic influence on the studied δ 15 N bulk values can be discounted because fine-grained Dotternhausen sedimentary rocks across the T-CIE interval are enriched in TOC, prone to anoxic diagenesis, and were deposited in a strongly redox-stratified basin with oxic surface and euxinic bottom water masses (Röhl et al., 2001;Wang et al., 2020) which further prevented oxic diagenesis. This inference is supported by the absence of negative correlations of δ 15 N bulk versus TOC and δ 13 C org versus TOC (Figures 3c,and 3d) that are often expected following early diagenetic alteration processes (e.g., Wang et al., 2018).
Thermal devolatilization of OM due to burial diagenesis and metamorphism tends to preferential free organic-bound 14 N, leading to a δ 15 N bulk increase by 1‰-2‰ for greenschist facies, 3‰-4‰ for lower amphibolite facies, and up to 6‰-12‰ for upper amphibolite facies (e.g., Ader et al., 2006). However, it is constrained that the OM of the Dotternhausen T-CIE samples was well preserved and did not apparently experience thermal volatilization (Wang et al., 2020), which suggests a negligible rise in the studied δ 15 N bulk values. On the other hand, the preferential release of 14 N due to thermal volatilization is typically accompanied by positive covariations of δ 15 N bulk versus (C/N) bulk and δ 15 N bulk versus δ 13 C org , and negative WANG ET AL. 10.1029/2020PA004207 7 of 15 In addition, δ 15 N ker can be used to calibrate and evaluate the preservation extent of the primarily deposited N isotope signal imparted by the originally sinking biomass (e.g., Ader et al., 2016;Kipp et al., 2018;Stüeken et al., 2016Stüeken et al., , 2017. A majority of the studied samples show an overall minimal offset between δ 15 N bulk and δ 15 N ker (Δ 15 N bulk-ker = δ 15 N bulk -δ 15 N ker , 0.2 ± 0.8‰; average ± 1 standard deviation), excluding three samples (DH-836, DH-825 and DH-729) with high δ 15 N bulk values (up to +5.7‰), where the offset is up to +4.1‰ (Figure 4; Table 1). The primary N isotope compositions of these three samples were likely affected by locally small-scale post-depositional modifications (Ader et al., 2016;Kipp et al., 2018;Stüeken et al., 2017) and will thus be excluded for further discussion. By contrast, the minimal δ 15 N bulk -δ 15 N ker offset displayed for the rest of the Dotternhausen samples implies that the primary N isotopic signals are well preserved.

Marine N Biogeochemical Cycling During the Dotternhausen Toarcian Carbon Isotope Excursion
Well preserved primary N isotope compositions of sedimentary OM can serve as an effective proxy to reconstruct the input (e.g., diazotrophy which is an atmospheric N 2 fixation metabolism utilizing metal-based nitrogenase enzymes) and output (e.g., denitrification and/or anammox) fluxes of bioavailable N in marine environments. Diazotrophy relying on molybdenum (Mo)-based nitrogenase is the most representative form of N 2 fixation in natural environments (Stüeken et al., 2016;Tyrrell, 1999). In marine settings with aqueous Mo limitation, N 2 fixation can alternatively use vanadium (V)-or iron (Fe)-based nitrogenase.
In the Dotternhausen sedimentary section, the initiation of the T-CIE corresponds to the onset of the prominent enrichment of Mo, V and uranium (U) which are redox-sensitive elements (Figure 4). These elements are more soluble and less particle-reactive under oxygen-rich seawater states, which tends not to express authigenic enrichment in the oxic sediments; on the other hand, these elements are less soluble and more particle-reactive under oxygen-depleted seawater conditions, which can lead to pronounced authigenic accumulation in sediments (Algeo & Maynard, 2004;Tribovillard et al., 2006). The obvious authigenic enrichment of the Mo, V and U in this study thus implies the development of bottom-water euxinic conditions during the deposition of the T-CIE Dotternhausen section. Furthermore, there is a strong correlation between the height of the T-CIE curve and aryl isoprenoid, green sulfur bacteria (GSB) biomarker abundance (see Figure 4; Wang et al., 2020) indicating that the peak of the T-CIE coincides with the highest level of euxinic conditions in the water column. Evidence for this anoxia was traditionally described to reflect the Toarcian OAE (e.g., Pearce et al., 2008;Thibault et al., 2018).
Both δ 13 C for TOC and separated kerogen show similar negative excursion, reaching minimum values of −33.1‰ and −33.4‰, respectively ( Figure 4; Table 1). The δ 15 N values predominantly range between +0.3 and + 2.5‰ (average ± 1 standard deviation; δ 15 N bulk = +2.0 ± 0.6‰ and δ 15 N ker = +1.8 ± 0.5‰) (Figure 4; Table 1). Across the T-CIE, the increasing bottom-water euxinia facilitated the emergence of water-column denitrification and/or anammox at the redoxcline ( Figure 5). However, the reliable δ 15 N values between +0.3 and + 2.5‰ found in the uppermost tenuicostatum and lower falciferum zones (Figure 4) do not reflect the isotope effect of water-column partial denitrification and/or anammox (traditionally > + 3‰; cf. Garvin et al., 2009;Gilleaudeau et al., 2020;Stüeken et al., 2016). By contrast, such a N isotope signal is in accordance with Mo-based diazotrophy likely under N-limited conditions. The severe scarcity of bioavailable N in the coeval depositional environment was driven by quantitative denitrification and/or anammox which was accompanied by the lack of an isotopic effect. The other possibility is that the concentration of dissolved N may have been high in anoxic bottom waters similar to the situation in the modern Black Sea (Ader et al., 2016;Fulton et al., 2012;Stüeken et al., 2016). However, the lack of N exchange between bottom and surface waters may have limited the availability of this dissolved deep N pool to organisms in surface waters and likely caused N scarcity. Cyanobacteria may thus have enhanced the rate of N 2 fixation to compensate this bioavailable N famine, here marked by the lack of any large isotopic variability. A small WANG ET AL. 10.1029/2020PA004207 9 of 15 δ 15 N depletion to ∼0‰ is found in the middle of the exaratum subzone (Figure 4). This depletion correlates with the highest level of bottom-water euxinia Wang et al., 2020) and may be associated with enhanced Mo-based diazotrophic activity through which more atmospheric N 2 would have been fixed (∼0‰) ( Figure 5). Alternatively, this small negative δ 15 N excursion is also coincident with highest abundance of aryl isoprenoid biomarkers, leading to elevated contribution of 14 N-enriched GSB biomass (Fogel & Cifuentes, 1993; (Figures 4 and 5). Nonetheless, N famine occurred in surface waters as a result of N loss to the atmosphere (quantitative denitrification and/or anammox) and/ or lack of mixing between bottom and surface waters during the deposition of the Dotternhausen T-CIE interval. In view of this, atmospheric N 2 was substantially fixed through Mo-based diazotrophs in order to keep pace with this seawater N scarcity.

Can the Northern European Epicontinental Shelf Settings Help to Validate an Evolved Oceanic Anoxic Event During the Toarcian Carbon Isotope Excursion?
T-CIE sedimentary profiles in the northern European epicontinental shelf are commonly marked by high abundance in TOC (e.g., Jenkyns, 2010). In addition to hydrographically restricted conditions (e.g., modern Black Sea), organic-rich sediments are also distributed in modern upwelling (unrestricted)-induced OMZ settings where partial denitrification and/or anammox is at play (Lam & Kuypers, 2011). By contrast, the studied T-CIE fine-grained organic-rich sedimentary rocks were deposited in a characteristic marine environment with enhanced water-column quantitative denitrification and/or anammox (see above). This finding convincingly manifests that the Dotternhausen fine-grained organic-rich rock deposition happened in a hydrographically restricted regime with minimal exchanges between bottom and surface waters during the T-CIE. Such hydrographic water-mass restriction heavily impeded bioavailable N replenishment from the bottom waters or open ocean and only accommodated an extremely limited NO 3 − inventory in surface waters. The restricted amount of NO 3 − was quantitatively consumed when the seawater became extensively euxinic during the T-CIE   Stüeken et al., 2016). However, this possibility is markedly inconsistent with the T-CIE δ 15 N values (+0.3‰ -+2.5‰) of the Dotternhausen section ( Figure 4). It further indicates that the ecosystem assimilating this newly fixed N was too small to outbalance the δ 15 N values of the preserved biomass as a consequence of a restricted setting. An analog exhibiting such a restricted marine environment is the Black Sea across the Holocene during which the prominently prevalent biological N 2 fixation was accompanied by biomass δ 15 N values of ∼0‰ -+2‰ stimulated by the deficit in bioavailable N (Fulton et al., 2012;Tyrrell, 1999) owing to (near-) complete consumption of NO 3 − . This scenario has also been proposed for ancient Archean and mid-Proterozoic marine environments (Ossa Ossa et al., 2019;Stüeken et al., 2016).
Additionally, whole rock δ 15 N data compiled from multiple shallow shelf T-CIE sedimentary successions including Rietheim (Switzerland), Réka Valley (Hungary), Yorkshire (UK), Winterborne Kingston (UK) and Mochras (UK) in the northern European epicontinental shelf range between −4‰ and +2‰ ( Figure 6). However, the pelagic T-CIE profile from Dogna in Italy exhibits a slightly larger variation in δ 15 N bulk values between −3‰ and +4‰ ( Figure 6). Environmental redox conditions during the deposition of the pelagic T-CIE Dogna sedimentary succession were proposed to have been unstable (Dickson et al., 2017). Nevertheless, δ 15 N values up to +4‰ observed in this profile may also be tied to NO 3 − assimilation or partial denitrification and/or anammox in OMZs elsewhere in a likely oxygenated Tethyan Sea (see review from Ader et al., 2016). It is thus worth noting that the δ 15 N records of other T-CIE profiles located in the shallow shelf settings fall in the Mo-based diazotrophic range. This N 2 fixation isotope record contradicts the possibility of existing water-column partial denitrification and/or anammox of marginal marine basins connected to the open ocean (δ 15 N values traditionally > +3‰; e.g., Ader et al., 2016;Garvin et al., 2009;Gilleaudeau et al., 2020;Stüeken et al., 2016), but is in line with redox-controlled N biogeochemical cycle in hydrographically restricted environments.
To confirm the scenario for restricted marine settings, organic-rich sediment Mo/TOC mass ratio was used to constrain the degree of hydrographically restricted water-mass conditions of T-CIE sections deposited in the northern European epicontinental shelf. This proxy relies on the observation that the slope of the Mo versus TOC regression line is low at ∼4.5 μg⋅g −1 ⋅wt.% −1 (Figure 7a) in modern Black Sea organic-rich sediments deposited under severe hydrographic restriction with limited water-mass renewal    Mo drawdown to the sediment along with limited resupply. When hydrographic restriction condition decreased, the increase in open ocean water mass renewal and Mo replenishment led to high slope values in organic-rich marine sediments. Slopes in Mo versus TOC space for T-CIE sedimentary successions are very variable with values between 0.4 and 1.1 μg⋅g −1 ⋅wt.% −1 , implying that for example the depositional environment of the Yorkshire succession was more restricted than that of Dotternhausen ( Figure 7a). Furthermore, both slope values are lower than ∼4.5 μg⋅g −1 ⋅wt.% −1 (Figure 7a), supporting that their depositional settings were even more restricted than the modern Black Sea (Figure 7a). Given the drawdown of dissolved Mo is very fast when adequate amounts of aqueous hydrogen sulphide are present (Algeo & Maynard, 2004;Reinhard et al., 2013), the persistently high Mo enrichment throughout the entire Dotternhausen T-CIE section ( Figure 4) points to limited but permanent Mo feeding from a larger oxic Mo reservoir that is, open ocean   (Dickson et al., 2017) further support that the water column, at least at the site of deposition, was not anoxic and sulfidic (euxinic) during the T-CIE. This may explain why the Dogna sedimentary succession did not record the N isotopic effect of water-column denitrification and/or anammox at the site of the deposition-though the water-column of deposition was neither euxinic nor sulfidic oxygen-depleted OMZ.
In addition, hydrographic conditions can also be evaluated using a cross-plot of sedimentary cadmium (Cd)/ Mo versus manganese×cobalt (Mn*Co) since distinct hydrographic regimes can be effectively constrained by modern high-TOC marine sediment Cd/Mo and Co×Mn values (Figure 7b; cf. Sweere et al., 2016Sweere et al., , 2020. High Cd uptake in organic material relative to Mo often promotes high Cd/Mo ratios (>0.1) in organic-rich sediments in upwelling (unrestricted) regions of the modern ocean (Conway & John, 2015). In restricted settings, oxygen deficit is in favor of Cd and Mo deposition but relatively weak marine productivity limits organic-driven output channel of Cd, resulting in low sedimentary Cd/Mo ratios (<0.1). The contents of Co and Mn in rivers and/or surface seawaters are significantly enriched compared to open-ocean sub-surface waters which feed upwelling systems, and thus the supply of both elements is typically higher in restricted basins (Sweere et al., 2016 and references therein). The different behaviors of these elements eventually promote high Co×Mn values (>0.4 μg⋅g −1 × wt.%) in organic-rich sediments that are deposited under restricted settings, whereas low Co×Mn values (<0.4 μg⋅g −1 × wt.%) in the counterpart from upwelling (unrestricted) settings. An empirical comparison between elemental data from the representative T-CIE sedimentary successions with the modern marine sediment records is further indicative of hydrographically restricted water-mass conditions for the northern European epicontinental shelf seas ( Figure 7b).
Collectively, combined N isotope and elemental concentration data of sedimentary successions deposited during the T-CIE add an important insight in the evolution of the ocean redox structure during the early Jurassic. These data support that the deoxygenated marine conditions commonly identified in several T-CIE localities from the northern European epicontinental shelf do not reflect the redox structure of the coeval Tethyan ocean, but rather indicate the development of oxygen-deficient basins. However, testing whether the global deep-ocean anoxia existed during the T-CIE requires robust redox evaluation (e.g., by trace element geochemistry and Fe speciation) of deep open ocean sedimentary successions.

Conclusion
Combined δ 15 N bulk and δ 15 N ker data obtained from the well-recognized T-CIE sedimentary section of Dotternhausen, SW Germany, are used to investigate the N biogeochemical cycle and its isotopic effect during the Early Toarcian marine deoxygenation. The δ 15 N bulk and δ 15 N ker values range predominantly between +0.3‰ and +2.5‰, implying enhanced N 2 fixation by cyanobacteria using Mo-based nitrogenase enzyme to keep pace with bioavailable N loss as a result of quantitative denitrification and/or anammox in a strongly redox-stratified marine environment lacking exchanges between bottom and surface water masses. This inference is further confirmed by the sedimentary N isotope records ranging between −4‰ and +2‰ from other northern European epicontinental shallow-shelf sections. The N 2 -fixing signal exhibited in these T-CIE sections contradicts the typical sedimentary δ 15 N values (>3‰) induced by partial water-column denitrification and/or anammox in marginal marine settings well connected to the open ocean. Instead, this study proposes the development of oxygen-deficient basins on the northern European epicontinental shelf.
In such hydrographically restricted marine environments, dissolved N pool suffered extensive denitrification and/or anammox which led to severe bioavailable N scarcity and thus stimulated Mo-based diazotrophic activity. These oxygen-depleted restricted settings situated on the northern European epicontinental shelf cannot thus be considered for evaluating the redox landscape of the coeval open ocean during the T-CIE. A robust redox evaluation of sedimentary successions deposited in both Tethyan and Panthalassan oceans is thus still missing for validating the putative global deep ocean anoxia across the T-CIE interval.

Data Availability Statement
Geochemical data for this research are available in the Zenodo repository at https://zenodo.org/record/4778565#.YKe1LqgzaUl.