Introduction

Paleoceanic basins are excellent depositories of deep-time biological and geological records, and geological columns from these basins can yield important information for predicting biogeochemical processes and paleoclimate in Earth history (Pearson and Palmer 2000; Zachos et al. 2001; Hong et al. 2020). In particular, proxy records relating to the production, accumulation, and preservation of sedimentary organic matter (OM) responds in a very sensitive way to past palaeolatitudinal (tectonic), climatic, and oceanographic changes. The Cretaceous to Neogene period was a time of known fluctuations in greenhouse gas concentration, warm global temperature episodes, sea-level highstands, and stagnate oceanic circulation (e.g., Arthur et al. 1985; Raymo and Ruddiman 1992; Pearson and Palmer 2000; Zachos et al. 2001; Hong et al. 2020; Yang et al. 2020). These fluctuations are documented by stratigraphic and sedimentological features (Davies et al. 1995; Lu et al. 2019), paleontological and palynological data (Kaiho et al. 1999; Gupta and Kumar 2019), petrography and geochemistry (Dypvik et al. 2011; Teng et al. 2019), and isotopic geochemistry (Bains et al. 1999; Pälike et al. 2006; Hong et al. 2020). It is widely accepted that the tectonic evolution of Asia has played a crucial role in many of these Earth systems phenomena over this critical period in the Earth’s history.

The present tectonic configuration of Asia has evolved since the breakup of the Gondwana supercontinent during the middle Mesozoic. The offshore sedimentary basins of the Indian plate record a prolonged period of isolation before the subduction of this plate with the Eurasian (Asian) plate (Molnar and Tapponnier 1975; Ali and Aitchison 2008; Chakraborty et al. 2019). Therefore, the Indian Ocean is an ideal location for gaining a better understanding of regional and global environmental and climatic changes over geological time. However, it is poorly sampled compared to the other regions. Furthermore, geochemical studies have rarely been applied to develop continuous and long-term records from Cretaceous to Neogene sediments in the Indian Ocean.

The Mannar Basin is the largest peri-cratonic paleoceanic sedimentary basin in the Sri Lankan jurisdiction of the Indian Ocean (Ratnayake et al. 2014). Geochemical studies in the Mannar Basin (Fig. 1) have much promise for helping us to understand the relationships between carbon cycling and regional and/or global paleoclimatic changes in a portion of the Indian Ocean that has often been overlooked. In this study, the author examines the long-term factors controlling organic matter burial in the Mannar Basin. Variations in organic geochemical proxies (i.e., organic matter concentration and types and depositional environments) show how sedimentary records have responded to global and regional paleoclimatic changes, such as taxonomic diversification of calcareous microorganisms during the Late Cretaceous, K-Pg mass extinction, Deccan volcanism, the tectonic movement of the Indian plate, the Eocene–Oligocene climatic transition, and Cenozoic global cooling.

Fig. 1
figure 1

Generalized map showing major geological units and offshore sedimentary basins in Sri Lanka. The inserted regional map shows the surrounding offshore sedimentary basins in the Bengal Fan and Indus Fan

Full size image

Geological background

The Mannar Basin is situated in the southern part of the Indian subcontinent between the eastern portion of south India and western coast of Sri Lanka (Fig. 1). The basin contains a relatively thick succession of sediments spanning from the Late Jurassic to the Holocene (Ratnayake et al. 2014; Ratnayake and Sampei 2015; Kularathna et al. 2020). The sediments are underlain by Precambrian high-grade metamorphic rocks of amphibolite to granulite facies (Cooray 1984). The observed geological procession has its counterparts in geological outcrops in parts of south India and Sri Lanka such as the Achankovil Terrain in south India/Wanni Complex in Sri Lanka and the Thiruvananthapuram Terrain in south India/Highland Complex in Sri Lanka (e.g. Cooray 1984; Dissanayake and Chandrajith 1999; Bandara et al. 2020). Gravitational and magnetic anomalies observed in the southernmost part of the Mannar Basin indicate the presence of a highly attenuated and intruded continental to transitional crust as a lateral propagating tip of the seafloor spreading ridge (Sreejith et al. 2008).

The separation of East and West Gondwanaland resulted in the creation of the Indian Ocean between Greater India and East Antarctica during the Middle Jurassic (McKenzie and Sclater 1971; Norton and Sclater 1979; Chakraborty et al. 2019). The identified regional magnetic and gravitational anomalies show the tectonic evolutionary history of the study area (Rotstein et al. 2001; Desa et al. 2006; Nair and Pandey 2018; Lasitha et al. 2019). A rifting of India/Sri Lanka from East Antarctica (ca. 130 Ma; Ramana et al. 2001; Gaina et al. 2007; Chakraborty et al. 2019; Kapawar and Mamilla 2020) was followed by the opening of the Mannar Basin. This basin developed as a failed rift basin according to the regional tectonic framework. The Indian plate was subjected to several volcanic eruptions due to detachments of Madagascar (~ 90 Ma), Seychelles-Laxmi Ridge (~ 70 Ma), and Seychelles (~ 65 Ma) during the northward voyage (Storey et al. 1995; Subrahmanyam and Chand 2006; Nair and Pandey 2018). The collision of the Indian and Eurasian plates occurred in the Early Eocene (~ 50 Ma) and ultimately led to the build-up of the Himalaya and Tibetan Plateau (Le Fort 1975; Molnar and Tapponnier 1975; Chung et al. 1998; Chakraborty et al. 2019). After that, oceanic basins in the Indian plates closed at ~ 25 Ma with the completion of the India-Eurasian collision (Qin et al. 2019; Kapawar and Mamilla 2020). Consequently, the Mannar Basin gradually subsided since its rifting, leading to the deposition of marine and terrestrial sedimentary OM over ca. 167 million years from the Jurassic to Recent in age.

Materials and methods

Materials

The ` (PRDS) of Sri Lanka undertook an initial stage of offshore hydrocarbon exploration in the deepwater Mannar Basin using the Chikyu drillship. The 260 drill core cuttings were obtained at 10-m intervals from the Barracuda exploration well (coordinates: 08° 20ʹ 34.46″ N, 79° 09ʹ 39.38″ E in Fig. 1). The sampling depth ranges from 2139 to 4741 m, and the present water depth of the sampling site is 1509.0 m. All cutting samples were washed extensively to remove contaminated artificial oil-based drilling mud. Manual stirring and ultrasonic vibration methods were used following the addition of dichloromethane and methanol (9:1 v/v) solvent mixture. Ratnayake and Sampei (2019) provide a detailed methodology for the cleaning of cutting samples.

Methods

CHNS elemental analysis

Total carbon (TC), total organic carbon (TOC), total nitrogen (TN), and total sulfur (TS) percentages were determined with a FISONS Carlo Erba EA 1180 elemental analyzer using combustion method, at a combustion temperature of 1000 °C (the American Society for Testing and Materials (ASTM) method D5373). Total carbon content was first determined without HCl acidification for 260 samples. Samples were ground into a fine powder before analysis. About 10 mg of each sample was placed into tin capsules, which were crimp-sealed before analysis. Total organic carbon content was determined in a separate run for the same set of samples (n = 260), using powdered samples about 10 mg in weight. Accurately weighed samples were placed in a silver film. After that, a few drops of 1 M HCl were added to remove inorganic carbon and then dried at 110 °C for 45 min. The dried samples were sealed and placed into tin capsules, which were again sealed. The BBOT [2,5-bis(5-tert-butyl-2-benzoxazolyl)thiophene] standard was placed in a tin film. Regression analysis using this standard was used for quantitative analysis. All of the elemental percentages were calculated on a dry weight basis. Carbonate carbon percentages were calculated based on the difference between TC and TOC content. Carbonate carbon values were used to calculate CaCO3 (%) in the samples.

GC–MS analysis

Sedimentary OM (bitumen, n = 27) was extracted using the Soxhlet apparatus with dichloromethane:methanol 9:1 v/v solution for 72 h. Elemental sulfur was removed using activated copper granules. The less soluble portions present after extraction (inner cutting samples) were ground into a fine powder. The inner cutting samples were again extracted using the Soxhlet apparatus, refluxing using a dichloromethane:methanol 9:1 v/v solution for 72 h. Elemental sulfur was again removed using activated copper granules. Extracted OMs were separated into aliphatic and aromatic fractions utilizing thin layer chromatography (silica gel 60 PF254 containing gypsum) using hexane as a mobile phase under room temperature. Aliphatic and aromatic hydrocarbons were identified from the rest of the geolipids using UV light and were separated after washing with hexane.

Aliphatic and aromatic hydrocarbons were subsequently analyzed by gas chromatography (GC) and a mass spectrometer (MS; Shimadzu GCMS-QP 2010) system. The GC system (Shimadzu 2010) is equipped with a 30-m fused silica capillary column (DB 5MS, 0.25 mm film thickness) and used helium as a carrier gas. The GC oven temperature was programmed to increase from 50 to 300 °C at the rate of 8 °C/min before being held at 300 °C for 30 min. The MS was scanned every 0.5 s from m/z 50 to 850, and all spectral data were automatically stored in the computer system. The ionization energy of the MS was set at 70 eV. The organic compounds were recognized by comparison of retention times and mass spectra with published data and reference standards.

δ 13 C and δ 15 N isotope analysis

Twenty-seven powdered samples were combusted at 1800 °C, with the evolved gases being carried by helium (90 mL/min), and passed through catalysts, to convert CO into CO2. Elemental copper in the reduction column reduces NO and NO2 to elemental N2. Nitrogen and carbon dioxide gasses were separated by gas chromatography (GC) column and analyzed for δ13C and δ15N using a Thermo Fisher Scientific Delta V Advantage Isotope Ratio Mass Spectrometer (IRMS) coupled with a Carlo Erba NC 2500 elemental analyzer at Illinois State Geological Survey, USA.

Ammonium sulfate standards and an internally calibrated amino acid standard were analyzed as nitrogen standards (see Appendix A for laboratory standards and 5-point calibration curve). Samples were calibrated and reported as δ15N vs. air. Atropine, IAEA-600 caffeine, USGS-40, and amino acid L-serine were analyzed as carbon standards (see Appendix A for laboratory standards and 4-point calibration curve). Samples were calibrated and reported relative to Vienna Pee Dee Belemnite (VPDB). The results show the precision of the δ15N measurements to be better than ± 0.22‰ and that of the δ13C analysis to be better than ± 0.11‰.

Stratigraphy and age model

Biostratigraphic charts of the Barracuda well were obtained from the Petroleum Resources Development Secretariat (PRDS) of Sri Lanka to tabulate the age of the sedimentary column. These unpublished data were generated using seismic data, well logs, and biostratigraphic studies of micropalaeontological, nannopalaeontological, and palynological assemblages. Kularathna et al. (2020) published the generalized stratigraphic column of the Barracuda well, providing a crucial reference for this study.

Results and discussion

Variations of carbonate deposition

Figure 2 shows the variations in carbonate accumulations through the Barracuda well record in the Mannar Basin (see Appendix B for primary data). The accumulation and preservation of carbonate particles in marine sediments is controlled by several factors, as discussed below. In this study, carbonate-rich strata were recorded from the Early Campanian to Late Maastrichtian stages (from 4270 to 4741 m, CaCO3 = 32.50% ± 9.42 in Fig. 2c). Late Cretaceous carbonate accumulation may indicate the effects of the taxonomic diversification of calcareous nannofossils and planktonic foraminifera (Boss and Wilkinson 1991; Huber and Watkins 1992; Ridgwell and Zeebe 2005), as well as the precipitation of CaCO3 under high atmospheric CO2 (ca. 2500 ppm) level (Ridgwell and Zeebe 2005; Kent and Muttoni 2008). The weathering of silicate and carbonate rocks also plays a major role in the burial of terrestrial carbonate flux under warmer and more humid climates (Berner et al. 1983; Arthur et al. 1985; Teng et al. 2019). Therefore, the reduction of carbonate deposition in the Upper Maastrichtian volcanogenic sediments can be recognized as being indicative of major environmental changes (i.e., alteration of oceanic chemistry and/or the open-ocean ecosystem) in the Indian Ocean (Fig. 2c).

Fig. 2
figure 2

Vertical distributions of a TC (%), b TOC (%), c CaCO3 (%), d TN (%), e C/N ratio, and f C/S ratio of the Barracuda well record from the Mannar Basin

Full size image

The K-Pg mass extinction event was characterized by a sharp reduction in carbonate productivity for several hundred thousand years (e.g., Caldeira and Rampino 1993; Gertsch et al. 2011). This period is globally characterized as involving the extinction of about 85% of planktonic foraminifera and calcareous nannofossils in the marine ecosystem (Zachos et al. 1989; D’Hondt et al. 1998). Therefore, the reduction of carbonate accumulation, along with TOC (Fig. 2), is likely associated with a collapse of aquatic productivity (super-stress conditions) across the K-Pg boundary in the Mannar Basin. The published records in the Indian Ocean also indicate the enhancement of carbonate dissolution due to acidification following Deccan volcanism. It has also been considered to have catalyzed biotic stress (reduction in diversity) and/or the mass extinction of planktonic foraminifera and calcareous nannofossils in the Indian Ocean during the Late Maastrichtian based on records from the Ocean Drilling Project (ODP) Leg 121, Ninetyeast Ridge Deep Sea Drilling Project (DSDP) Sites 216 and 217, Wharton Basin Site 212, and Krishna-Godavari Basin (Rea et al. 1990; Keller 2003; Keller et al. 2008; Tantawy et al. 2009).

Carbonate deposition is low in the Lower Paleocene to Upper Paleocene sediments of the Barracuda well (average CaCO3 = 11.13% ± 6.14). However, carbonate deposition drastically increased from the Late Paleocene. The development of a CaCO3 platform from the Late Paleocene (close to the Paleocene-Eocene boundary) is primarily correlated with the transition from an arid climate into a warm and humid tropical climate following the movement of the Indian plate (including Sri Lanka) northward into warmer, equatorial latitudes. According to the literature, the weathering of Himalayan rocks has been generally identified to be the main cause behind increments of carbonate flux during the Late Cenozoic in the Bay of Bengal and Indus Fan sediments (France-Lanord and Derry 1997; Métivier et al. 1999; Zachos et al. 2001; Parmar et al. 2020).

Carbonate accumulation was enhanced in the Barracuda well (average CaCO3 = 40.20% ± 12.14) since the Late Paleocene. Consequently, the uplifted Precambrian rocks in Sri Lanka and the Eastern Ghats, India (e.g., Cooray 1984; Bose et al. 2020) have been influenced by the carbonate burial in the Mannar Basin due to combined rapid physical denudation with chemical weathering within a tropical climate. This idea is supported by an observed enhancement of greenhouse conditions and monsoon-like rainfall patterns in Asia during the Eocene (Licht et al. 2014; Teng et al. 2019). In addition, based on basin modeling interpretations (Ratnayake et al. 2014), carbonate burial was likely enhanced by open marine and deepwater conditions (e.g., Adatte et al. 2002; Gertsch et al. 2011) following tectonic subsidence in the Mannar Basin from the Eocene (average subsidence rate = 16 m/Ma) to Miocene (average = 19 m/Ma).

Variations of organic matter delivery

The average TOC content in the Lower–Upper Campanian sediments (from 4540 to 4741 m) of the Barracuda well is 0.97% ± 0.23 (Fig. 2b). The Upper Campanian to Upper Maastrichtian sediments of the Barracuda sequence are rich in OM (from 4270 to 4540 m, average TOC = 1.34% ± 0.36 in Fig. 2b). Late Cretaceous organic carbon deposition has an apparent relationship with carbonate deposition (Fig. 2).

TOC values are slightly decreased in the Lower–Upper Paleocene sediments of the Barracuda well (from 3060 to 3440 m, average = 1.07% ± 0.39 in Figs. 2b). Although carbonate accumulation is considerably enhanced in the sediments from the Upper Paleocene to the Lower-Middle Oligocene, organic carbon accumulation is relatively low in this portion of Barracuda well sequence (2520–3060 m, average = 0.74% ± 0.41). The reduction of organic carbon accumulation in these strata is likely driven by limited nutrient mixing (low oceanic primary productivity) as a result of weak paleoceanic currents during the Eocene to Early Oligocene in response to low global latitudinal temperature gradients between poles and topics under the greenhouse climate (Davies et al. 1995; Licht et al. 2014). However, TOC content gradually increased from the Middle Oligocene (Fig. 2b).

The analysis of sedimentary facies revealed that OM-rich Middle Oligocene to Miocene beds consist of black carbon and laminations (Ratnayake et al. 2014), suggesting seasonal changes in the watershed. The Middle Oligocene to Miocene (ca. 380-m thick) sedimentary sequence of the Barracuda well indicates a higher amount of OM preservation (average TOC = 2.51% ± 1.20). This observation was linked with records of major sea-level regression (marked by erosional unconformity) and continental sedimentation along rift-passive margin sedimentary basins in the Indian subcontinent since the Oligocene (e.g., Adatte et al. 2002; Nair and Pandey 2018; Chakraborty et al. 2019). In this study, Middle Oligocene to Miocene OM-rich beds were deposited after the Eocene–Oligocene climatic transition ca. 34 Ma ago (Pearson and Palmer 2000; Pearson et al. 2009). It is possible that this event had a teleconnection with the formation of the Antarctic ice sheet and resulted in the drop of global sea level, as demonstrated by simulations and proxy analyses across the globe (e.g., Volk 1989; Raymo and Ruddiman 1992; Gaillardet and Galy 2008) that characterize the state of the Earth today.

Variations of organic matter type delivery

Bulk C/N ratios, biomarkers, and stable δ13C isotope values have been widely applied to paleoclimatic records in order to identify OM sources and assess changes in the availability of nutrients (e.g., Meyers and Ishiwatari 1993; Sampei and Matsumoto 2001; Ratnayake and Sampei 2015; Gupta and Kumar 2019; Li et al. 2020). C/N ratios reveal that the Lower–Upper Campanian sediments consist of a mixture of terrestrial and algal-derived OM (average = 15.13 ± 4.68 in Fig. 2e). The m/z 57 mass chromatograms of the Lower–Upper Campanian sediments in the Barracuda document a nC19 dominant bimodal pattern with maxima at the nC22/nC23 and nC31 peaks (Fig. 3a). The middle-chain length (nC23 and nC25) n-alkane homologues can be recognized as sources of bog-forming aquatic vegetation in swamps (Ficken et al. 2000; Pancost et al. 2002; Bingham et al. 2010; Ratnayake and Sampei 2015). The abundance of the nC31 homologue also indicates the occurrence of grasses (graminoids) and shrub-type herbaceous vegetation in swamps (Zhou et al. 2005; Castańeda et al. 2009; Ratnayake et al. 2017). Therefore, the Lower–Upper Campanian sediments of the Barracuda well record a higher amount of terrestrial (average C20-C26/nCall = 0.67 ± 0.08) carbon sources (Fig. 4).

Fig. 3
figure 3

Mass chromatograms (m/z = 57) showing the distribution of n-alkanes in the Barracuda well record

Full size image
Fig. 4
figure 4

Biomarker results from the Barracuda well record

Full size image

The C/N ratios of organic carbon-rich Upper Campanian to Upper Maastrichtian sediments of the Barracuda (4270–4540 m, average = 20.36 ± 7.64) well indicate an accumulation of higher plant wax dominant sediments (Fig. 2e). Biomarker studies also indicate a higher amount of middle-chain (nC21 to nC25) and long-chain (nC27 to nC31) n-alkanes from vascular plant waxes (Fig. 3b). Similarly, Ratnayake et al. (2018) identified occurrences of oil and gas prone (type II-III) and gas prone (type III) kerogen, based on Rock–Eval pyrolysis data in the Lower Campanian to Upper Maastrichtian sediments of the Mannar Basin. Therefore, the increment of TOC and vascular plants OM in this sedimentary succession suggests the deposition of continental sediments in the Mannar Basin, likely as a result of sea-level regression in nearshore environments and tectonic changes (e.g., Huber and Watkins 1992; Adatte et al. 2002; Keller 2003, 2005; Gertsch et al. 2011; Clift 2020). The available literature suggests that sea-level regression was generally associated with widespread continental erosion during the Late Maastrichtian (e.g., Abramovich et al. 2002; Adatte et al. 2002; Keller 2004). Moreover, the rapid evolution and expansion of angiosperms since the Middle Cretaceous (Hickey and Doyle 1977) may have gradually increased the burial rate of terrestrial OM in coastal swamps. This idea is supported by high values in terrestrial biomarker proxies of > nC26/nCall (average = 0.17 ± 0.04) and average-chain length (ACL, average = 26.47 ± 1.14) compared to the underling Campanian sedimentary sequence (> nC26/nCall average = 0.14 ± 0.06 and ACL average = 25.80 ± 0.73) in the Barracuda well (Fig. 4). The volcanogenic sediments of the Barracuda well also indicate the deposition of terrestrial OM (Figs. 2e and 3c).

The Paleocene sequence (3060–3440 m, average C/N = 15.75 ± 3.19 in Fig. 2e) contains mixtures of terrestrial and algal-derived OM. Similarly, the Upper Paleocene sediments of the Barracuda well show an intensification in the frequency of middle-chain length nC23 n-alkanes with higher molecular weight (> nC26) n-alkanes (Fig. 3d). The Upper Paleocene to the Lower-Middle Oligocene argillaceous marl/marlstone sediments indicate algal-derived OM with some terrestrial contributions (2520–3060 m, average = 13.78 ± 10.00 in Fig. 2e). The n-alkane distributions in this sedimentary succession also indicate an enhancement of higher plant wax (> nC20 in Fig. 3e) from terrestrial sediments (C20-C26/nCall average = 0.75 ± 0.15, > nC26/nCall average = 0.25 ± 0.15 and ACL average = 26.00 ± 1.28 in Fig. 4).

Terrestrial OM distribution increased from the Middle Oligocene to Lower Miocene portions of the sedimentary succession from the Barracuda well (2139–2520 m, average = 23.45 ± 15.02 in Fig. 2e). The n-alkane mass chromatogram indicates higher plant wax (> nC20) dominant sediments during the Early-Middle Oligocene (Fig. 3f), followed by continental erosion that has been recorded in surrounding sedimentary basins in the Indian subcontinent (e.g., Nair and Pandey 2018; Chakraborty et al. 2019; Parmar et al. 2020). In addition, the available Rock–Eval data suggested that OMs are primarily of terrigenous (type III/II) origin (Ratnayake et al. 2018). Furthermore, the Lower-Middle Miocene sediments in the Barracuda well consist of fine-grained black carbon/charcoal fragments (Ratnayake et al. 2014) with a higher amount of long-chain n-alkanes (> nC26) (Fig. 3g). This suggests the deposition of long-distance transported OM. Consequently, these changes suggest the strengthening/development of the South Asian monsoon from the Early-Middle Miocene, as has been demonstrated elsewhere (e.g., Quade et al. 1989; Dettman et al. 2001; Gupta et al. 2004; Clift 2020).

In this study, a considerable number of samples recorded abnormally low C/N ratios (< 4) due to the absorption of inorganic nitrogen (NH4+) in fine-grained sediments (Müller 1977; Sampei and Matsumoto 2001). These low C/N ratios suggest that inorganic nitrogen absorption was prominently recorded in the Middle Eocene to Lower Oligocene sediments of the Barracuda well (Fig. 2e). In addition, TN values drastically increased, without TOC increments, in the Middle Eocene to Lower Oligocene sediments of the Barracuda well (Fig. 2). These episodes signify the enrichment of inorganic nitrogen rather than nutrient accumulations over geological time. In this study, samples with abnormally low C/N values (< 4) were removed from calculations of the average and standard deviations for each stratum.

Stable carbon (δ13C) isotopic values are thus likely more reliable for reconstructing past sources of OMs under different depositional conditions in this record. Terrestrial C3 and C4 plants show an average δ13C value of ca. − 28 ‰ and ca. − 14 ‰, respectively (Bender 1971; Chikaraishi and Naraoka 2005; Lamb et al. 2006; Gupta and Kumar 2019; Hong et al. 2020). The δ13C isotopic composition of plant biomass is thus primarily a function of the photosynthetic pathway (Bender 1971), with no influence of grain size effects. Overall, based on these data, it is clear that core sediments in the Mannar Basin derive mainly from C3 plant dominant terrestrial sources (Fig. 5 and see Appendix A for primary data).

Fig. 5
figure 5

Relationship between isotopic δ13C and elemental C/N ratio shows sources of organic matters (modified after Meyers 2003 and Lamb et al. 2006) (where POM, particulate organic matter and DOC, dissolved organic carbon)

Full size image

Depositional environments

TOC to TS relationship (C/S ratio) is an indicator of depositional processes such as sulfate reduction, Fe-S cycling, paleosalinity, and early diagenesis (e.g., Berner 1982, 1984; Meyers and Ishiwatari 1993; Sampei et al. 1997a; Teng et al. 2019). C/S ratios reveal that the lowermost Campanian sediments show oxygen-poor conditions (average = 6.38 ± 4.93 in Fig. 2f). C/S ratios indicative of organic carbon-rich conditions in the Upper Campanian to Upper Maastrichtian sediments (4270–4540 m, average = 9.15 ± 3.22) suggest a comparatively less reducing environment (Fig. 2f). The relatively high C/S ratios compared to the bottom of the sedimentary succession may indicate a lower activity of sulfate-reducing bacteria in terrestrial OM-rich sediments (e.g., Berner 1984; Gong and Hollander 1997; Sampei et al. 1997a).

C/S values for the Lower–Upper Paleocene portion of the sequence, ranging from ca. 10 to less than 100, show wide and cyclic fluctuations. Furthermore, wide variations in C/S ratios of the Barracuda well record indicate weaker reducing characteristics in sand dominant turbidities (3060–3440 m, average 16.21 ± 20.14 in Fig. 2f).

C/S ratios are low from the Upper Paleocene to the Lower-Middle Oligocene portion of the record (2520–3060 m, average = 6.83 ± 12.24 in Fig. 2f). This implies an oxygen-depleted environment. Similarly, alternating dark and light color laminations in this sedimentary succession signify cyclic changes in supplies of sediments under anoxic/reducing conditions (Gong and Hollander 1997; Meyers 2003; Valdés et al. 2004; Lokho et al. 2020). This interpretation is consistent with observed monsoon rainfall patterns in Asia (Licht et al. 2014; Teng et al. 2019) and stratified and weak ocean circulation in the Indian Ocean during the Eocene (Davies et al. 1995). By contrast, a few outliers (range from 39.88 to 47.97, n = 5) increase the average C/S ratio of the Barracuda well during the Upper Paleocene to the Lower-Middle Oligocene period (Fig. 2f). This can be interpreted as an influence of strong short-term and deepwater oceanic currents under open marine conditions. C/S ratios (average = 5.98 ± 4.96) of the Middle Oligocene to Lower Miocene sediments of the Barracuda well can be explained as reflecting deposition of less reducing terrestrial OM-rich sediments within a deepwater marine setting.

Paleoenvironment and paleoclimate

Four major paleoclimatic chronozones were recognized in the Mannar Basin, based on sedimentary facies, OM quantity/type, and depositional changes (Fig. 6). The lowermost period (chronozone 1: Early Campanian) consists of organic carbon-depleted (average TOC = 0.97%) sediments. In this period, algae-dominant OM was deposited with terrestrial OM (average C/N ratio = 15.13) under oxygen-poor depositional environments (average C/S ratios of Barracuda = 6.38). The Mannar Basin was situated in the arid climate belt during the Early Campanian (Lees 2002; Scotese et al. 2011). Similarly, ca. 720 mm/year precipitation is estimated for eastern part of the Indian subcontinent at this time, according to the Fast Ocean-Atmospheric Model (Chatterjee et al. 2013).

Fig. 6
figure 6

Schematic of cross sections for the major paleoclimate chronozones identified in the Mannar Basin record

Full size image

Chronozone 2a in Fig. 6 (Late Campanian to Late Maastrichtian) is composed of organic carbon-rich sediments (average TOC = 1.34%). These terrestrial OM-rich (average C/N ratio = 20.36) sediments were deposited under less reducing conditions (average C/S ratio = 9.15). Several smaller continental blocks such as the Laxmi Ridge-Seychelles (~ 70 Ma) and Seychelles (~ 65 Ma) were separated from the Indian plate during the Late Maastrichtian (Storey et al. 1995; Subrahmanyam and Chand 2006; Calvès et al. 2011; Rao and Singh 2020). Tectonic forcing and a drop of sea level (Huber and Watkins 1992; Gombos et al. 1995; Keller 2003, 2005; Clift 2020) were followed by the deposition of terrestrial OM-rich sediments. Although the Mannar Basin was located in the arid to subtropical temperate climatic zone during the Late Campanian to Late Maastrichtian (Lees 2002; Scotese et al. 2011), a broadly warmer and wetter global climate than the present, with the exception of short-term global cooling events (Arthur et al. 1985; Adatte et al. 2002; Keller et al. 2015; Yang et al. 2020), may have influenced this process. In this period, the Indian plate became smaller and smaller in size and resided as an island continent in the southern hemisphere. Previous investigations, as well as the data recorded in this study, indicate that paleoceanic basins in the Indian plate were characterized by the deposition of thick carbonate sediments during the Late Cretaceous (e.g., Rao 2001; Chatterjee et al. 2013; Teng et al. 2019). The carbonate-rich pelagic sediments of this region are likely a product of the known evolutionary diversification of microorganisms and high partial pressure of atmospheric carbon dioxide (pCO2) under warmer climatic conditions, during the Late Campanian to Late Maastrichtian (Kent and Muttoni 2008; Keller et al. 2015). Carbonate depletion at the K-Pg boundary indicates super-stress environmental conditions, including greenhouse warming, eutrophication, and acid rain over the Indian subcontinent (e.g., O’Keefe and Ahrens 1989; Sigurdsson et al. 1992; Gertsch et al. 2011). The super-stress environmental conditions are associated with Deccan volcanism and are marked by the extinction of an abundance of foraminifera in the Indian Ocean (e.g., Keller 2003, 2005; Keller et al. 2015).

Chronozone 2b subcategory is characterized by the deposition of volcanogenic sediments due to several episodes of Deccan-Reunion basalt volcanism at the top of the Late Maastrichtian (Fig. 6). Ratnayake et al. (in preparation) examined the provenance of the Upper Maastrichtian volcanogenic sediments in the Mannar Basin using whole-rock geochemistry. These unpublished data reveal that the Upper Maastrichtian mafic sources are linked to the Deccan basalt volcanisms, and a significant difference can be observed compared to the adjacent sedimentary strata. In contrast, this sedimentary succession indicates relatively high TOC values (average = 0.84%) under a high mud sedimentation rate of 49 m/Ma (Ratnayake et al. 2014). Previous investigations showed that high sedimentation rates (> ca. 40 m/Ma) decrease TOC content due to clastic dilution (e.g., Berner 1982; Ibach 1982; Sampei et al. 1997b; Li et al. 2020). Therefore, high TOC values of the Upper Maastrichtian volcanogenic sediments also suggest warmer climatic conditions and higher fertility (primary productivity), triggered by submarine igneous events and favorable preservation of OM.

Chronozone 3a in Fig. 6 (Early-Late Paleocene) records OM-poor to moderate sediments (average TOC = 1.07%). Published work suggest that low marine productivity had gradually recovered more than 3 Ma after the K-Pg mass extinction (D’Hondt et al. 1998), under the semiarid to arid (mock aridity) climate of the Early Danian on the Indian subcontinent (e.g., Gertsch et al. 2011). In addition, this period is characterized by the deposition of algae-dominant OM with terrestrial sediments (average C/N ratio = 15.75) under oxic depositional conditions (average C/S ratio = 16.21). In contrast, the Late Paleocene was characterized by noticeably higher concentrations of CO2 (ca. 2000 ppm) and other greenhouse gases compared to the present day (Pearson and Palmer 2000; Zachos et al. 2001, 2008). For example, the Late Paleocene Thermal Maximum (LPTM) represented the warmest interval of the Cenozoic at the Paleocene/Eocene (ca. 55 Ma) boundary (Bains et al. 1999; Zachos et al. 2001, 2005; Lourens et al. 2005; Gupta and Kumar 2019). The Mannar Basin was characterized by the enhancement of carbonate-rich pelagic sediments (carbonate platform) near the Paleocene/Eocene boundary under a tropical (equatorial) humid climate.

In chronozone 3b, the Late Paleocene to the Early Oligocene, OM concentration (average TOC = 0.74%) and type (average C/N ratio = 13.78) remain almost constant (Fig. 6). However, in this sub-chronozone, the depositional environment changed to oxygen-poor marine conditions (average C/S ratio = 6.83), perhaps as a consequence of weak ocean circulation in the Indian Ocean under the greenhouse climate (Davies et al. 1995; Licht et al. 2014).

The Middle Oligocene to Miocene sedimentary succession of the Barracuda well represents chronozone 4 (Fig. 6). In this section, a higher amount of terrestrial debris (average C/N ratio = 23.45) was deposited within a marine depositional environment (average C/S ratio = 5.98). This sedimentary succession indicates the enhancement of organic burial (average TOC = 2.51%) after the Eocene–Oligocene climate transition (ca. 34 Ma), which is marked by the sharp decline of atmospheric CO2 concentration due to the growth of the Antarctic ice sheets (Raymo and Ruddiman 1992; Pearson and Palmer 2000; Pälike et al. 2006; Tripati et al. 2009). The Indian subcontinent had achieved its present-day configuration by the Miocene. The subcontinental landmass thus provided a source of insolation in the lower atmosphere during the summer, which produced a strong low-pressure system over the region. The uplifted mountains act as a barrier and resulted in the development of South Asian monsoon precipitation which is continuous to the present (e.g., Quade et al. 1989; Dettman et al. 2001; Zhisheng et al. 2001; Gupta et al. 2004; Harris 2006; Clift 2020). Therefore, OM-rich beds, lamination, back carbon/fine-grained charcoal fragments, and the domination of mass chromatograms by long-chain n-alkanes (> nC26) indicate the development of South Asian monsoon from the Middle to Late Miocene.

Conclusions

The Mannar Basin record provides the first geological and paleoclimatic proxy sequence spanning the Jurassic to the Miocene for Sri Lanka, a landmass that was at the center of tectonic change and rafting (e.g., the breakup of Gondwana), oceanic circulation evolution, and mass extinction events during this span of time. Calcareous nannofossils and planktonic foraminifera diversified during the Early Campanian to Late Maastrichtian under a warm climate. The diversity of microorganisms was then drastically reduced toward the K-Pg boundary under the influence of Deccan-Reunion volcanism in the Indian Ocean. Tectonic forcing and sea-level regression controlled the deposition of terrestrial OM and organic carbon-rich sediments during the Late Campanian to Late Maastrichtian. The development of a CaCO3 platform around the Late Paleocene-Eocene boundary was linked to the influence of a tropical climate and continuous subsidence of the basin. The Eocene epoch was characterized by weak oceanic circulation in the Indian Ocean under the greenhouse climate. Finally, the drop in sea level after the formation of the Antarctic ice sheets is associated with the deposition of terrestrial OM-rich sediments since the Middle Oligocene and the development of the Indian Ocean monsoon system which characterizes the climate of the region to this day. It is hoped that the Mannar Basin record acts as an important reference for further research into the geological and climatic evolution of the equatorial latitudes of Asia across this important time frame in Earth’s history.