Cenomanian-Turonian sea-level transgression and OAE2 deposition in the Western Narmada Basin, India

a Department of Geosciences, Princeton University, Princeton, NJ 08544, USA b Department of Geology, Mohanlal Sukhadia University, Udaipur, Rajasthan, India c Majestic Shore, 208 Choolaimedu High Road, Chennai 600094, India. d Anagha, 4th Cross Street, CB Nagar, Tumkur 572102, India. e Institute of Earth Surface Dynamics (IDYST), University of Lausanne, Lausanne 1015, Switzerland f Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, CA 91125, USA g Institute of Earth Sciences, University of Lausanne, Lausanne 1015, Switzerland


Introduction
The Narmada Basin in India is an intracratonic rift basin of intense interest because of its tectonic history and northward journey after the breakup of Gondwana in the Early Cretaceous~130 Ma (Tandon, 2000;Chatterjee et al., 2013;Kumari et al., 2020). Madagascar separated from India during the Turonian to Santonian transition and from the Seychelles during the Campanian to early Maastrichtian. Highlights of this journey are recorded in the Bagh Group exposed in disconnected outcrops along the edges of the Deccan basalt, and overlie Precambrian rocks from Madhya Pradesh to Gujarat (Tandon, 2000;Tripathi, 2006;Tripathi, 2005;Kumari et al., 2020) (Fig. 1). Sediment deposition in the Narmada Basin occurred in a tectonic graben that was largely controlled by uplift, block faulting, volcanic activity, erosion and sea level transgressions (Biswas, 1987;Kumar et al., 1999;Tandon, 2000;Tripathi, 2006;Bansal et al., 2020;review in Kumari et al., 2020). Clues to this history can be found in the Late Cretaceous sediments roughly parallel to the present course of the Narmada River. In both eastern and western parts of the Narmada Basin, the Nimar Sandstone overlies the Archean crystalline basement rocks in a westerly sloping basin (Tripathi, 2006). However, the overlying sedimentary sequences vary.
In the western Narmada Basin of Gujarat, the Turonian-Coniacian limestone units of the Bagh Group from Madhya Pradesh are not present, likely because of tectonic activity, uplift, erosion and nondeposition (Tripathi, 2006). Instead, the estuarine facies of the Nimar Sandstone underlie in ascending order, a micritic sandstone, oyster beds and limestone with oysters, revealing the first major sea-level transgression into the western Narmada Basin (Tandon, 2000;Tripathi, 2006). The depositional age was determined as Late Cenomanian to Early Turonian based on ostracods and calcareous algae (Kundal and Sanganwar, 1998;Chaudhary et al., 2017a).
The focus of this study is the first sea-level transgression into the western Narmada Basin in Gujarat during the late Cenomanian that originated central India's Narmada Seaway, which reached about 800 km across India by the end of the Maastrichtian . We explore whether this sea-level transgression could be linked to the global transgression that occurred during the late Cenomanianearly Turonian Oceanic Anoxic Event 2 (OAE2) based on two outcrops in Gujarat near the villages of Bilthana and Bhundmariya (Fig. 1A, B). We focus on four objectives: (1) improve biostratigraphy and age control; (2) determine the paleoenvironment and onset of the sea-level transgression; (3) identify the link between the sea-level transgression and the OAE2; and (4) test the potential link between volcanism, LIPs, and Hg toxicity during this time.
Our investigations focus on biostratigraphy of planktic foraminifera and ostracods for age control, the nature of sediment deposition and characteristic fossil life for paleoenvironmental reconstructions, carbon isotopes for the identification of the OAE2, oxygen isotopes for paleoclimate and fresh water influx reconstructions, and mercury concentrations in sediments for assessing the potential influence (toxicity, warming) of large volcanic eruptions.

Materials and methods
We analyzed Bhundmariya and Bilthana localities from Gujarat in the western Narmada Basin. The Bilthana section is located near the village of Bilthana, along the Men River (N 21°57′; E 73°39′) (Fig. 1A, C). The Bhundmariya section, is located near the village of Bhundmariya, also along the Men River (N 21°59′; E 73°59′) (Fig. 1B, C). In both sections, the Nimar Sandstone followed by the Micritic Sandstone mark the base of the Bagh Group and disconformably overlie crystalline Archaean rocks. Above this interval, oyster biostromes transition into sandy Limestones with oysters at Bilthana and marly Limestone with oysters at Bhundmariya (Figs. 2,3).
Samples were initially collected for ostracod biostratigraphy at 30, 60 and 100 cm intervals at Bilthana and analyzed for foraminifera, oxygen and carbon stable isotopes, and mercury concentrations. The Bhundmariya outcrop was collected at closer sample intervals (10-15 cm) to confirm the stratigraphic record and identify missing intervals.
For foraminifera biostratigraphy, samples were crushed into peasized fragments, placed in beakers and covered with a 3% hydrogen peroxide solution to oxidize organic carbon, and let stand for 24-48 h. After disaggregation of sediments, the sample residue was processed through >38 μm, >63 μm and > 150 μm screens to catch foraminifers in dwarfed, standard and larger size fractions, respectively. Cleaned residues were dried in the oven and analyzed for microfossils. Most lithologies are strongly micritized and foraminifera difficult to free from such sediments. Therefore, thin sections were also prepared for planktic foraminiferal analyses. Preservation is poor due to recrystallized foraminiferal shells in micritized sandstones and limestones. Species identifications are based on processed samples and thin sections. Identification of recrystallized specimens is mainly restricted to larger species with recognizable number of chambers, shape and keel characteristics.
The concentration of mercury in the sediment was analyzed at the Geosciences Department of Princeton University (USA) using a Zeeman R-915F (Lumex, St. Petersburg, Russia), a high-frequency atomic absorption spectrometer specifically designed for Hg determination with a detection limit of 0.3-3 ppb. Measurements are based on the direct thermal evaporation of Hg from solid samples and do not require chemical pre-treatment of samples, thus avoiding potential contamination during sample preparation. Analyses were conducted on two aliquots. Accuracy of the measurements was confirmed by the analysis of certified reference materials NIST 1646a with Hg calibrated at 30.2 ppb. Fig. 2. Panorama view of the Bilthana outcrop (Gujarat), which consist of the Nimar Sandstone and overlying Micritic Sandstone at the base, followed by oyster beds that transition into sandy Limestones with oysters at Bilthana. Inserts show close-up views of bedded micritic sandstone and oyster beds.
G. Keller, M.L. Nagori, M. Chaudhary et al. Gondwana Research 94 (2021) 73-86 Precision based on relative standard deviation of repeated sample measurements is 5-10%. Total organic carbon (TOC) contents were determined at the Institute of Earth Sciences, University of Lausanne (Switzerland) based on RockEval™ 6 and quantified by flame-ionization and infrared detection. The IFP 160000 Rock-Eval standard was used for calibration. Analytical precision is 0.05 wt%. Results show TOC consistently <0.02 wt% likely due to diagenetic alteration of organic matter as evident by tan orange and ochre outcrop colors.
Carbon and oxygen stable isotope analyses were performed on bulk rock sediments at the Institute of Earth Surface Dynamics (ISTE) of the University of Lausanne (Switzerland) using a Thermo Fisher GasBench II preparation device interfaced with a Thermo Fisher Delta Plus XL continuous flow isotope ratio mass spectrometer. The carbon and oxygen stable isotope ratios are reported in the delta (δ) notation as the permil (‰) deviation relative to the Vienna Pee Dee belemnite standard (VPDB). The reproducibility was better than 0.05‰ for δ 13 C and 0.1‰ for δ 18 O.

Age and biostratigraphy
The Nimar and Micritic Sandstones are the oldest sediments overlying the Archean crystalline rocks in this area and the age was previously attributed to the Cenomanian based on ostracods and calcareous algae (Kundal and Sanganwar, 1998;. Planktic foraminifera are rare to few, small, long-ranging and stress tolerant species but not age diagnostic (Muricohedbergella, Planoheterohelix, Heterohelix). Benthic foraminifera are generally long-ranging and not useful age indicators either (Haplophragmoides, Textularia, Spiroplectammina, Trochammina, Dentalina, Vaginulina) (Rajshekhar, 1987(Rajshekhar, , 1995Rajshekhar and Atpalkar, 1995). A summary of previous biostratigraphic studies based on a large variety of fossils (e.g., ammonites, inoceramids, ostracods, echinoids, benthic and planktic foraminifera, calcareous algae) in the Bagh Group of the western Narmada Basin is shown in Table 1 (references therein).
Planktic foraminifera at Bilthana and Bhundmariya localities are common to abundant in the oyster beds and Limestone with oysters, but benthic foraminifera are rare or absent. A single low salinity tolerant planktic foraminifer, Muricohedbergella delrioensis, thrived in the oyster biostrome and oyster-rich sandy Limestone at Bilthana. Also present are rare Whiteinella archaeocretacea (Plate 1), the index species that marks the latest Cenomanian to early Turonian W. archaeocretacea zone, which coincides with the OAE2 (Leckie, 1987;Leckie et al., 1998;Keller et al., 2001Keller et al., , 2008Keller and Pardo, 2004;Gertsch et al., 2008Gertsch et al., , 2010Heimhofer et al., 2018). Benthic foraminifera include the brackish water tolerant Anomalinoides newmanae and diverse arenaceous species (Fig. 4).
At Bhundmariya, more diverse planktic foraminiferal assemblages survived in the marly Limestone with oysters. In the basal 1.5 m of the section, M. delrioensis dominated with few Muricohedbergella hoelzli and Muricohedbergella simplex species (Figs. 3,5). Diversity increased with the appearance of larger species, including W. archaeocretacea, Whiteinella baltica, Whiteinella brittonensis, Dicarinella imbricata and Dicarinella hagni, providing good biostratigraphic age control during the late Cenomanian-early Turonian OAE2.

Late Cenomanian-early Turonian OAE2
The late Cenomanian to early Turonian is an extremely well-studied interval because the global OAE2 occurred during this transition. Agedefining biozones characteristic of OAE2 are commonly based on ammonites, inoceramids, and deep-sea planktic foraminifera. We illustrate these biozonations along with the defining OAE2 δ 13 C excursion based on the Global Stratotype Section and Point (GSSP) in Pueblo, Colorado (USA), where sediment deposition occurred in the middle neritic environment of the USA Western Interior Seaway (Leckie, 1987; Leckie Table 1 Summary of fossil contents and age diagnostic species (bold) of ostracods, ammonites and planktic foraminifera in late Cenomanian sediments. This is the first application of shallow, nearshore planktic foraminiferal assemblages to identify the OAE2 in the western India interior seaway of the Narmada Valley, although such applications have been successfully done in Egypt and Morocco Gertsch et al., 2008Gertsch et al., , 2010El-Sabbagh et al., 2011 Rajshekhar 1987Rajshekhar , 1995Nayak, 1987;Leckie et al., 1987Leckie et al., , 1998 6). For age comparison with shallower environments, we plot the zonal schemes and OAE2 δ 13 C excursions of the subtidal to estuarine Wadi El Ghaib section in the eastern Sinai of Egypt (Gertsch et al., 2008) and the Gujarat sections of India ( Fig. 6).
In the middle neritic environment of Pueblo and the estuarine to subtidal environment of the Sinai, ammonites have been commonly used for regional age control as shown for comparison with planktic foraminifera ( Fig. 6). In the western Narmada Basin of Gujarat, only the monospecific nektobenthic ammonite Placenticeras mintoi was reported in the Nimar Sandstone (Gangopadhyay and Bardhan, 2007;Jaitley and Ajane, 2013). However, by Turonian age, the sea-level transgression reached middle neritic depths in the eastern Narmada Basin of Madhya Pradesh supporting diverse ammonites in the Nodular Limestones (Gangopadhyay and Bardhan, 2007;Kumar et al., 2018).
Planktic foraminifera are the most commonly used age-defining biozones characteristic of OAE2. The upper Rotalipora cushmani zone marks the gradual onset to the first peak of the OAE2 δ 13 C excursion, which coincides with the extinction of R. cushmani, the last survivor of this genus (Fig. 6). The W. archaeocretacea zone defines the interval from the R. cushmani extinction to the first appearance of Helvetoglobotruncana helvetica in the early Turonian. This interval spans most of the OAE2 δ 13 C excursion.
The W. archaeocretacea zone is subdivided into three subzones Keller and Pardo, 2004), two of which span the Plenus Cold Event and maximum δ 13 C excursion (Fig. 6) (Keller and Pardo, 2004;O'Connor et al., 2020). Subzone Globigerinelloides bentonensis defines the interval between the R. cushmani extinction at the base and the G. bentonensis extinction at the top, but also coincides with the first appearances (FAs) of D. imbricata at the base and D. hagni at the top. Subzone D. hagni defines the interval from the FA of this species to the Planoheterohelix shift, an interval dominated by low oxygen tolerant biserial species that continues into the early Turonian and ends with the return to normal δ 13 C values (Leckie, 1987;Leckie et al., 1998;Keller and Pardo, 2004;. Presence of the global OAE2 δ 13 C excursion in the open ocean and shallow near-shore sequences is an excellent correlation tool, provided sedimentation is relatively complete. The δ 13 C excursion began with a gradual rise to the maximum excursion coupled with the short Plenus cold event, also identified as the Plenus Carbon Isotope Excursion (CIE) (O'Connor et al., 2020). This event merged into a broad fluctuating δ 13 C plateau, followed by the gradual decrease to normal δ 13 C values in the early Turonian (Gertsch et al., 2008(Gertsch et al., , 2010Jarvis et al., 2011;O'Connor et al., 2020).
In the subtidal to estuarine environment of the Wadi El Ghaib in the Sinai, the OAE2 δ 13 C excursion is very similar to Pueblo, Colorado (Fig. 6). However, at Bilthana and Bhundmariya in the western Narmada Basin, the δ 13 C excursion peak is not well expressed due to incomplete sedimentation, erosion, non-deposition and partly low sample resolution. At Bilthana, sediment deposition occurred in a brackish estuarine environment dominated by the nearly monospecific assemblage of M. delrioensis but the OAE2 δ 13 C excursion peak is present in the oyster biostrome (Figs. 4,6).
In contrast, the Bhundmariya section, just~25 km from Bilthana, lacks oyster beds or oyster biostromes at the base of the Limestone with oysters and the OAE2 δ 13 C excursion peak is not observed (Figs. 5,6). This suggests non-deposition or erosion (hiatus) of this interval, as also implied by the significantly deeper and more diverse, larger planktic foraminiferal assemblages, including late Cenomanian index species D. imbricata and D. hagni, beginning just 2 m above the Plate 1. Common planktic foraminifera from the oyster beds and Limestone with oysters at Bilthana and Bhundmariya. Scale bar = 100 μm. Fig. 1-9 Muricohedbergella delrioensis, common to abundant small species tolerant of brackish water environments. Figs. 10-12. Whiteinella archaeocretacea, index species for late Cenomanian age and OAE2. Figs. 13-15. Thin section images of Whiteinella archaeocretacea, Bhundmariya. Similar Muricohedbergella-dominated assemblages have been documented in near-shore environments of Egypt, Morocco and USA (Pueblo, Colorado, GSSP) Gertsch et al., 2008;El-Sabbagh et al., 2011). Micritic Sandstone (Fig. 5). These index species are absent at Bilthana's shallower estuarine environment dominated by the salinity tolerant M. delrioensis. Nevertheless, the mere presence of these index species, which evolved during OAE2 after the δ 13 C excursion peak reflects the OAE2 plateau correlative with Pueblo, Colorado, and the Wadi El Ghaib Sinai sections (Fig. 6). We conclude, the biostratigraphic correlation between the western Narmada Basin with sections at Pueblo, Colorado and the Sinai based on planktic foraminifera index species provides good age control and reveals the presence of OAE2 in all three localities.

Paleoenvironment
Few planktic foraminiferal species survive in high-stress, shallow neritic environments, such as the western Narmada Basin during the late Cenomanian to early Turonian. Survivor species are generally tolerant of variations in oxygen, salinity, temperature, nutrients and toxicity (e.g., mercury from large volcanic eruptions) (Keller, 2003;Keller et al., 2007, Keller et al., 2020. These species tend to be small, unornamented with simple test morphologies (biserial, triserial, trochospiral) and frequently dwarfed in high-stress environments (Leckie et al., 1998;Keller and Abramovich, 2009). Depending on the degree of biotic stress, foraminiferal assemblages vary from few species in shallow neritic environments to a single survivor in brackish estuarine conditions. During the late Cenomanian OAE2, one to six survivor species was the norm in shallow neritic environments of the Tethys from Egypt to Morocco and the USA Western Interior Seaway (Leckie, 1987;Leckie et al., 1998;Keller and Pardo, 2004;Gertsch et al., 2008Gertsch et al., , 2010El-Sabbagh et al., 2011).
In the Nimar and Micritic Sandstones of the Gujarat sections, few small planktic foraminifera from Planoheterohelix and Heterohelix genera were reported along with arenaceous benthic foraminifers (Haplophragmoides, Textularia, Spiroplectammina, Trochammina, Dentalina, Vaginulina) (Rajshekhar, 1987(Rajshekhar, , 1995Rajshekhar and Atpalkar, 1995). These benthic species indicate deposition in an estuarine environment, consistent with the onset of the first sea-level transgression into the Narmada Basin. However, heterohelicid species (e.g., Heterohelix globulosa, Planoheterohelix moremani) generally thrived in the oxygen minimum zone of deeper neritic environments (Leckie, 1987;Leckie et al., 1998;Keller and Pardo, 2004) and may have washed into this estuarine environment, as suggested by their absence in the rest of the sequence.
At Bilthana, the oyster biostromes above the Nimar and Micritic Sandstones, are characteristic of high-energy, shallow and faunally restricted environments with low salinity and mesotrophic nutrient levels (Pufahl and James, 2006). Deposition generally occurred in shallow subtidal (<20 m) and estuarine environments (Pufahl and James, 2006). With the deepening environment, the oyster biostrome morphed into a sandy Limestone with oysters and a nearly monospecific planktic foraminiferal assemblage thrived consisting of the low salinity tolerant M. delrioensis (Fig. 4, Plate 1). This species is tolerant of brackish waters, similar to M. planispira (Keller and Pardo, 2004). Benthic foraminifera are rare, including the low salinity tolerant brackish water Anomalinoides newmanae, and arenaceous species. Oysters, echinoids and ostracods are common. Towards the top of the sequence, the sharply decreased abundances of M. delrioensis, oysters, ostracods and echinoids may be related to deepening of the subtidal environment as the sea-level transgression progressed. These faunal changes occurred during the late Cenomanian W. archaeocretacea zone, as indicated by rare occurrences of this index species and also two ostracod index species, Rostrocytheridea divergence and Rostrocytheridea jaisalmerensis (Fig. 4). The overall high-stress environment and near absence of benthic foraminifera may also be amplified by relatively high mercury concentrations (8-9 ppb) and a one peak eruption event of 30 ppb. The high background Hg values suggest increased Hg runoff from land during the humid conditions resulting in toxicity for marine life .
Bhundmariya reveals the biotic trend during the late Cenomanian W. archaeocretacea zone in the marly Limestone with oysters where the Muricohedbergella assemblage was followed by Whiteinella species but near absence of benthic foraminifera. The temporary disappearance of this assemblage (samples 9, 10; Fig. 5), coupled with negative δ 18 O (−3.4‰) and δ 13 C (−2.4‰) excursions and extreme δ 18 O data Fig. 6. Late Cenomanian biostratigraphy of planktic foraminifera and ammonites, and δ 13 C excursion characteristic of the OAE2 at the GSSP at Pueblo, Colorado (USA) (Keller and Pardo, 2004;, compared and correlated with subtidal-estuarine sections from the Sinai (Egypt; Gertsch et al., 2008) and Gujarat (India). Note the onset of the global δ 13 C excursion is present in Gujarat with the overall OAE2 δ 13 C trend preserved in an environment alternating between brackish and normal marine salinities.
fluctuations (−5‰ to −8.5‰), suggests fresh water influx and eutrophication. Above this barren interval, planktic foraminifera and ostracods are more diverse and species are larger and more complex (Whiteinella and Dicarinella), which suggests migration of open marine planktic foraminifera into the deepening Narmada Seaway and thriving in more normal salinity and oxygen conditions and gradually decreasing δ 13 C values. Near the top of the section, another fresh water event resulted in the negative (−4.5 to −9‰) δ 18 O excursion indicative of fresh water influx and eliminating all but a few fossils. Mercury concentrations doubled, suggesting increased volcanic eruptions and/or mercury runoff from land during the humid conditions resulting in toxicity for marine life .

Carbon and oxygen stable isotopes
We analyze carbon and oxygen stable isotopes of the Bhundmariya and Bilthana sections from Gujarat (Figs. 4,5) to evaluate nutrient and salinity trends in shallow neritic environments during the OAE2. We compare the Gujarat δ 13 C record with similar shallow marine sequences in the Tethys Ocean from the Sinai, Egypt, to Morocco and Pueblo, Colorado, USA Western Interior Seaway (Fig. 6). The OAE2 has a 2‰ to 3‰ positive δ 13 C excursion, prolonged plateau and gradual decrease across the Cenomanian-Turonian boundary in shallow marine environments Gertsch et al., 2008Gertsch et al., , 2010El-Sabbagh et al., 2011;Jarvis et al., 2011;Bomou et al., 2013;Wendler, 2013;Kuhnt et al., 2017). Shallow neritic environments reveal this characteristic δ 13 C excursion dependent on the continuity of the sedimentation record (Gertsch et al., 2008(Gertsch et al., , 2010O'Connor et al., 2020). We illustrate the OAE2 δ 13 C excursion for Pueblo, Colorado, and the Sinai Wadi El Ghaib sections compared with the western Narmada Basin of India (Fig. 6).
The carbonate δ 13 C values are generally not affected by diagenesis (Schrag et al., 1995), except for organic-rich sediments due to the diagenetic alteration of organic carbon (Pratt, 1984;Marshall, 1992). Cross-plots of δ 13 C and δ 18 O of the Sinai and Morocco sections revealed no significant correlation, thus indicating no major diagenetic alteration (Gertsch et al., 2008(Gertsch et al., , 2010. In contrast, the Bilthana and Bhundmariya δ 13 C records show intervals of significantly negative δ 13 C values, which can only be explained by 13 C-depleted dissolved inorganic carbon (DIC) by contribution of oxidized organic carbon. The oxidation of organic matter is evident by the low TOC values and the tan-orange and ochre colors of the sediments. This early diagenetic alteration of more abundant organic matter in well-oxygenated shallow to middle neritic environments may have produced 13 C depleted CO 2 that contributed to the DIC involved during neoformation/recrystallization of the carbonate phases.
Bilthana and Bhundmariya sections mark overlapping parts of the OAE2 δ 13 C excursion with the major difference in the variable sedimentation/erosion patterns and the lower sample resolution (Fig. 6). The overall more positive δ 13 C ratios at Bilthana can be explained by higher terrestrial influx nearer to shore, which is also evident in the sandy limestone. More negative δ 13 C ratios at Bhundmariya reflect early diagenetic alteration of abundant organic matter in well-oxygenated shallow to middle neritic environments (discussed above). These two δ 13 C records for Gujarat are successfully compared with the biostratigraphy of the Tethys Seaway (Morocco, Egypt) and the Western Interior Seaway (Pueblo, Colorado, USA) (Fig. 6).
In the Micritic Sandstone at Bilthana, Gujarat, the negative δ 18 O ratios (−8‰ to −6.8‰) indicate fresh water influx, correlative with the gradual 2‰ positive δ 13 C excursion and sea-level transgression associated with the OAE2 (Fig. 4). No foraminifera or invertebrates were observed in this interval. The second fresh water pulse (δ 18 O -7‰, samples 8-9) correlates with the OAE2 δ 13 C plateau. At Bhundmariya, the correlative fresh water influx (δ 18 O -9.5‰, samples 10-12) also marks the OAE2 δ 13 C plateau (Fig. 5). In both localities, the fresh water influx resulted in absent or strongly reduced marine life, as also observed in the fresh water pulse near the top of the Bhundmariya section.
The main source of Hg to the environment is from volcanic emissions, atmospheric transport over 6 months to one year and fallout into terrestrial and marine environments. Secondary sources include natural coal combustion and scavenging by organic matter and clay in sediments. Hg in organic and clay-rich sediments thus contain a component of sequestered Hg that is normalized when TOC is >0.2 wt% (Thibodeau et al., 2016;Percival et al., 2018;review in Grasby et al., 2019;Keller et al., 2020). In Bilthana and Bhundmariya, TOC is consistently <0.05 wt% and < 0.08 wt%, respectively, which is likely due to diagenesis of organic matter as evident by tan orange and ochre outcrop colors. However, there is no indication of organic-rich clay layers in these predominantly oyster-rich limestones.
Our preliminary analyses from the late Cenomanian at Bilthana reveal Hg concentrations are stable and average 8-9 ppb (detection limit <3 ppb), except for a significant peak of 30 ppb in the sandy Limestone with oysters (sample 8, Fig. 4). The peak concentration suggests a volcanic eruption, which may or may not be of LIP origin. The consistently high values (8-9 ppb) suggest a significant contribution from terrestrial runoff. At Bhundmariya, Hg values are generally low (<3 ppb), likely due to greater distance from terrestrial sources, as also suggested by marly Limestone compared with sandy Limestone at Bilthana. However, the increased concentration to 6 ppb at the top could reflect increased terrestrial runoff and/or volcanic eruptions.

Discussion
This study began with a single objective: a routine biostratigraphic age analysis based on planktic foraminifera of the Bilthana section in Gujarat. It turned out to be a high-stress nearly monospecific assemblage in an estuarine to subtidal environment of late Cenomanian age. We confirmed this observation by collecting and analyzing a second locality, Bhundmariya located~25 km to the east (Fig. 1C). Results showed deposition at Bhundmariya occurred in slightly deeper subtidal waters with higher species diversity and good index species (W. archaeocretacea, D. imbricata, D. hagni) for the late Cenomanian correlative with the OAE2.
With this new information, we added objective-2: evaluate the onset of the sea-level transgression based on sediments and fossil assemblages. At Bilthana and Bhundmariya, the Nimar and Micritic Sandstones at the base of the Bagh Group overlie Precambrian Archean rocks from Gujarat in the west to Madhya Pradesh in the east. In the western Narmada Basin, few benthic and planktic brackish water foraminifera are present along with ostracods and calcareous algae of late Cenomanian age (Table 1), thus demonstrating that the first sea-level transgression began in late Cenomanian sands and reached peak in oyster beds or biostromes at Bilthana. As the transgression progressed, oyster biostromes drowned giving way to sandy Limestone with oysters ( Fig. 4).
At Bhundmariya, oyster beds are absent but similar low diversity assemblages (2-3 species) overlie the sandstones in the first 1.5 m of the marly Limestone with oysters. Upsection, planktic foraminifera diversified rapidly with larger species and index species characteristic of the OAE2 (W. archaeocretacea, D. imbricata, D. hagni), suggesting rapid deepening to shallow neritic depths (Fig. 5). Thus, we concluded that the sea-level transgression into the Narmada Basin began in the late Cenomanian at the time of the OAE2.
This conclusion led us to objective-3: testing the link between the sea-level transgression and the OAE2 based on carbon isotopes, and environmental changes based on oxygen isotopes. In Bilthana and Bhundmariya, the δ 18 O records, reveal intervals of very negative values indicative of fresh water runoff from land areas, which affected all species intolerant of brackish water environments (Figs. 4, 5).
At Bilthana, the δ 13 C excursion began in the sandstones above the Archean rocks and reached the excursion peak in the oyster beds (Fig. 4), decreased by 2‰ and rebounded. At Bhundmariya, the δ 13 C excursion was recorded in the marly Limestone with oysters, with results similar to the characteristic prolonged plateau with gradual decrease and fluctuating values after the OAE2 δ 13 C excursion peak. We concluded that the outcrops in both sections contain partial OAE2 records.
We compared these records with the GSSP at Pueblo, Colorado, USA, and the Wadi El Ghaib section in the Sinai, Egypt, which have the globally characteristic OAE2 δ 13 C excursions (Fig. 6). This comparison is also based on planktic foraminifera biostratigraphy. The δ 13 C record from Bilthana reveals presence of the δ 13 C excursion gradual increase and peak, plus the partial presence of the subsequent plateau. At Bhundmariya, both the plateau and gradual decrease are present. The δ 13 C excursion peak is missing, supported by the faunal assemblages indicating erosion or non-deposition. The comparison of the OAE2 δ 13 C excursion in the Gujarat sections with Pueblo and Wadi El Ghaib is based on biostratigraphy and indicates this excursion is partially present and can be recognized even though the sequences are incomplete.
Last, we tested objective-4: the potential link between volcanism, based on mercury fallout, and mercury toxicity during this time. Mercury concentrations at Bilthana are relatively stable averaging 8-9 ppb, three times the detection limit, which likely indicates terrestrially enriched Hg runoff. There is one significant peak of 30 ppb, indicative of a larger volcanic eruption possibly from one of the LIPs associated with the OAE2 (Caribbean Plateau, Ontong-Java Plateau, Madagascar Province (Ernst, 2014). At Bhundmariya, Hg values are generally at or below the detection limit of 3 ppb and reach 6 ppb at the top. However, the record is incomplete and inconclusive. Toxicity based on Hg cannot be evaluated in these shallow, estuarine environments, where the major limiting factor for planktic and benthic foraminifera is fresh water influx from land that significantly reduced abundance and diversity.
In summary, given the data and its limitations, we conclude the Narmada Seaway began in Gujarat with the sea-level transgression during the late Cenomanian and progressed eastward towards Madhya Pradesh during the Turonian through Coniacian and Santonian as indicated by current fossil data and biostratigraphy (Fig. 7A). Our preliminary analyses of Turonian-Coniacian-Santonian sequences suggest that by this time the Narmada Seaway deepened to a middle neritic environment. Consequently, the Narmada Seaway must have existed by the end of the Maastrichtian, possibly as a shallower seaway as cooling persisted from the Campanian into the late Maastrichtian. By early Danian, the presence of early Danian planktic foraminifera in shallow intertrappeans at Jhilmili suggests the end of the maximum extent of this seaway (Fig. 7B)  .
A recent review of the India seaways based on tectonic history and some fossil groups concluded there was inconclusive evidence the Narmada Seaway existed by the end of the Cretaceous (Kumari et al., 2020). Instead, they proposed the Godavari rift basin as the only inland seaway at this time. We concur that the Godavari Seaway must have existed during the late Cretaceous and could have connected with the Narmada Seaway and formed the Trans-India Seaway (Fig. 7A).
A recent study also reported organic-rich sediments of the late Cenomanian to early Turonian from subsurface cores in the Cauvery Basin of south India, correlative with the global transgression and the OAE2 (Nagendra and Reddy, 2017). In this area, sedimentation occurred in a middle-neritic environment marked by high concentrations of dark colored pyritized agglutinated foraminifera and dwarfed calcareous benthic foraminifera (e.g., Lenticulina, Gavelinella), which indicate high-stress dysoxic to anoxic environments (Govindan, 1993). Organic-rich shale deposition continued into the early Turonian correlative with the transgression and expansion of oxygen-depleted waters into shallower neritic environments (Nagendra and Reddy, 2017). The Cauvery River basin thus also recorded the Cenomanian-Turonian transgression, just as the Godavari and Krishna River basins in the Rajahmundry area and the Narmada River in the north. India's entire south-eastern seaboard must have been inundated by the Cenomanian-Turonian transgression (Fig. 7B).
We suggest that the Godavari and Narmada Seaways likely coexisted. The Narmada Seaway across central India and the Godavari Seaway from Rajahmundry north to Nagpur may have connected at times with the Narmada Seaway via the Narmada-Tapti rift zone and formed a Trans-India Seaway (Fig. 7B, C).
The global sea-level rise during the late Cenomanian to early Turonian thus inundated low-lying areas worldwide and created shallow carbonate seas and vast inland seaways, such as the Trans-Sahara Seaway (Meister et al., 1992;Pascal et al., 1993) and the US Western Inland Seaway spanning from the Gulf of Mexico to the polar area ( Fig. 7C) (e.g., Haq et al., 1987;Hallam, 1992;Haq, 2014). In southern India's Cauvery Basin, sea-level transgressions have been identified for nearly all of the Cretaceous Oceanic Anoxic Events (OAEs) (review in Nagendra and Reddy, 2017).

Global context of OAE2
Placing the Cenomanian-Turonian OAE2, the δ 13 C excursion, climate change, sea-level transgression, evolution and extinctions into the global context helps understand this complex time interval (Fig. 8). The Cretaceous was a time of rapid greenhouse warming and cooling episodes, rising sea-levels and repeated platform drowning, accompanied by major carbon isotope excursions, episodic OAEs, and increased volcanic activity from LIPs. The largest of the anoxic events is OAE2, which occurred during the late Cenomanian to early Turoniañ 94-90 Ma at a time of LIP eruptions in the Caribbean, Ontong Java and Madagascar (Fig. 8) (Ernst, 2014;Scaife et al., 2017).
By the late Cenomanian to early Turonian, climate warmed and reached peak sea-surface (SS) temperatures of 28°C during OAE2 (Huber et al., 1995;Weissert and Erba, 2004;Jarvis et al., 2006, Jarvis et al., 2011Keller, 2008), interrupted by short-term cooling (Plenus marl) (O'Connor et al., 2020). From the early Turonian through the Santonian SS temperatures remained high~25°C interrupted by cooling episodes to~20°C (Fig. 8). Warmer SS temperatures for OAE2 have been estimated based on TEX 86 from the Cassis section in southern France where temperatures reached highs of 38°C during the δ 13 C excursion and lows of 32°C during the Plenus cool event (Heimhofer et al., 2018).
During the OAE2, the sea-level reached 250-300 m above the present level, inundating carbonate platforms and continental shelf areas and forming vast inland seas, including the USA Western Interior Seaway spanning from the Gulf of Mexico to the Arctic (Gale et al., 2008), the Trans-Sahara Seaway of North Africa and now also the Trans-India Seaway combining the Narmada and Godavari Seaways (Fig. 7C). A major CIE accompanied this sea-level transgression in open marine, shallow carbonate platform environments and inland seaways. This δ 13 C excursion defines the OAE2 that occurred in four distinct phases: (1) the gradual increase, (2) the excursion peak associated with the Plenus Cold Event, (3) the plateau, and (4) the gradual decrease to normal values in the early Turonian. These δ 13 C excursion characteristics are globally identifiable from estuarine-subtidal to deep-sea environments (Fig. 6). They are the δ 13 C fingerprint of the OAE2 and an  (Jarvis et al., 2011;Wendler, 2013;O'Connor et al., 2020).
Despite these major global environmental changes, there was no mass extinction in marine life and no significantly increased extinctions in genera. Diversity in invertebrate genera increased gradually through the Cretaceous OAEs, reached a plateau between the Cenomanian to Santonian and peaked in the Campanian and Maastrichtian prior to the KPB mass extinction (Fig. 8). Extinctions remained in background values (<10%), but slightly accelerated between the Aptian and Cenomanian OAEs and increased during the Maastrichtian prior to the KPB mass extinction (Sepkoski Jr., 1996;Sepkoski Jr., 1997). However, the species record shows relatively few extinctions in planktic foraminifera or ammonites (Keller and Pardo, 2004;Monnet, 2009), but rapid evolution and/or adaptation of survivor species to the changing high-stress environments associated with OAE2 and submarine LIP eruptions (e.g., salinity, low oxygen, high productivity, temperature changes) (Scaife et al., 2017). In contrast, all five mass extinctions in Earth's history are associated with continental flood basalts LIPs, which globally resulted in rapid climate changes, distribution of toxins via the atmosphere causing adverse effects, and terrestrial acidity and toxicity .

Conclusions
(1) We report the first evidence of the late Cenomanian-early Turonian OAE2 in the western Narmada Basin associated with the largest sea-level transgression of the Cretaceous that initiated the western Narmada Seaway, which by the end of the Turonian reached from Gujarat through Madhya Pradesh. At times this seaway may have connected to the Godavari Seaway via the Narmada-Tpati rift forming a Trans-India Seaway to Rajahmundry.
(2) The characteristic OAE2 gradual onset of the δ 13 C excursion began with the transgression and peaked in the oyster beds or biostromes at Bilthana in a subtidal to estuarine environment.
The δ 13 C plateau followed in the overlying Limestone with oysters.
(3) The onset of the transgression is recognized in the Nimar Sandstone and Micritic Sandstone by an influx of few benthic and planktic foraminifera and ostracods indicative of the late Cenomanian. (4) The OAE2 δ 13 C excursion at Gujarat is dated based on planktic foraminifera biozones and correlated with the GSSP at Pueblo, Colorado, USA, and the Wadi El Ghaib section in the Sinai, Egypt. (5) Very negative δ 18 O values (−7 to −8‰) indicate episodic terrestrial fresh water influx associated with high-stress, planktic foraminiferal assemblages dominated by nearly monospecific Muricohedbergella species (M. delrioensis) tolerant of salinity fluctuations and brackish water benthic foraminifera. (6) The deepening transgression during the late Cenomanian δ 13 C plateau is marked by increased species diversity, with common Whiteinella and few Dicarinella characteristic of the latest Cenomanian W. archaeocretacea zone. Similar faunal progressions mark this global sea-level transgression from Egypt to Morocco and the USA Western Interior Seaway.   (Huber et al., 1995;Keller, 1998, 1999;Weissert and Erba, 2004), increasing planktic foraminiferal diversity through the Cretaceous and absence of accelerated extinctions by genera (Sepkoski Jr., 1996;Sepkoski Jr., 1997). Three Large Igneous Provinces (LIPs) are associated with the time of OAE2 (Courtillot and Renne, 2003;Ernst, 2014). Ages based on time scale by Gradstein et al. (2012). The late Cenomanian OAE2 began with a major sea-level transgression accompanied by~8°C climate warming and 2‰ increased productivity associated with black shale deposition. The western India interior seaway of the Narmada Valley began with this sea level transgression during the late Cenomanian (modified from Keller, 2008

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