Persistent increase in carbon burial in the Gulf of Mannar, during the Meghalayan Age: Influence of primary productivity and better preservation

Abstract

The oceans store a substantial fraction of carbon as calcium carbonate (CaCO₃) and organic carbon (C_org) and constitute a significant component of the global carbon cycle. The C_org and CaCO₃ flux depends on productivity and is strongly modulated by the Asian monsoon in the tropics. Anthropogenic activities are likely to influence the monsoon and thus it is imperative to understand its implications on carbon burial in the oceans. We have reconstructed multi-decadal CaCO₃ and C_org burial changes and associated processes during the last 4.9 ky, including the Meghalayan Age, from the Gulf of Mannar. The influence of monsoon on carbon burial is reconstructed from the absolute abundance of planktic foraminifera and relative abundance of Globigerina bulloides. Both C_org and CaCO₃ increased throughout the Meghalayan Age, except between 3.0–3.5 ka and the last millennium. The increase in C_org burial during the Meghalayan Age was observed throughout the eastern Arabian Sea. The concomitant decrease in the C_org to nitrogen ratio suggests increased contribution of marine organic matter. Although the upwelling was intense until 1.5 ka, the lack of a definite increasing trend suggests that the persistent increase in C_org and CaCO₃ during the early Meghalayan Age was mainly driven by higher productivity during the winter season coupled with better preservation in the sediments. Both the intervals (3.0–3.5 ka and the last millennium) of nearly constant carbon burial coincide with a steady sea-level. The low carbon burial during the last millennium is attributed to the weaker-upwelling-induced lower productivity.

1. Introduction

Carbon dioxide (CO₂) and methane (CH₄) are the two dominant gaseous forms of carbon in the atmosphere. The atmospheric CO₂ concentration depends on a multitude of processes involving the exchange of carbon between atmosphere, lithosphere, oceans and biosphere, collectively defined as the global carbon cycle (Carlson et al. 2001). The excessive use of fossil fuels (coal, petroleum) since industrialization has increased the atmospheric CO₂ concentration to levels unprecedented in the last ∼800 000 years (Lüthi et al. 2008). The atmospheric CO₂ combines with rainwater to form weak carbonic acid that dissolves rocks on the earth’s surface. The dissolution of rocks, termed as silicate weathering, releases calcium and bicarbonate ions from the rocks into the rivers and subsequently into the oceans (Misra & Froelich, 2012). The silicate weathering is one of the significant components of the global carbon cycle, on longer timescales (Brady, 1991; Raymo & Ruddiman, 1992; Wan et al. 2012).

In the oceans, calcareous organisms combine calcium ions with bicarbonate ions to form calcium carbonate (CaCO₃) (Zeebe & Wolf-Gladrow, 2001). In modern oceans, most of the CaCO₃ is precipitated as the shells of microorganism, like foraminifera, coccolithophores and corals (Ramaswamy & Nair, 1994; Schiebel, 2002). Foraminifera, single-celled organisms with hard outer shells called test, contribute a significant fraction of the marine carbonate flux (Langer, 2008). After the death of the organisms, the shells sink to the ocean floor. Foraminiferal tests are usually constructed either by secreting CaCO₃ or by cementing sand, silt and other particles, known as agglutinated tests (Kaminski & Kuhnt, 1995; Saraswat, 2015; Saalim et al. 2019). Foraminiferal shells are one of the most abundant and significant fossil remains in marine sediments. In addition to the biogenic carbonate, organic matter also removes a substantial fraction of carbon from the ocean water and buries it in the sediments, as carbon is the main component of all life forms (Ciais et al. 2013). The photosynthesizing plants, collectively termed as marine productivity or marine primary productivity, constitute a substantial part of the oceanic organic matter (Field et al. 1998). Thus, the long-term carbon burial in the oceans includes both the organic carbon (C_org) and CaCO₃ in the sediments. Therefore, oceans are vital in modulating the global carbon cycle by regulating the amount of carbon buried in the sediments (Falkowski et al. 2000).

The anthropogenic greenhouse-gas-emission-induced warming is likely to affect the marine organisms and thus their contribution to carbon cycling. However, the likely response of the marine carbon cycling to the anthropogenic greenhouse-gas-emission-induced warming is not clear. Because of the unprecedented increase in atmospheric CO₂ concentration during the last ∼150 years, it is crucial to understand its effect on marine carbon cycling. The carbon burial records from the oceans, covering the times of different atmospheric CO₂ concentrations in the past, can help understand the effects of anthropogenic contribution in global carbon cycling (Falkowski et al. 2000). The large influence of anthropogenic activities on global CO₂ supposedly began after the industrial revolution (IPCC, 2021). Therefore, long-term multi-decadal records of carbon burial, spanning the interval before significant anthropogenic activities, are required to understand the effect of human-induced perturbations on the carbon cycle.

The tropical Indian Ocean has a considerable carbon burial potential as its several regions have high primary productivity during the summer and winter monsoon (Prell & Curry, 1981; Prasanna Kumar et al. 2001; Sreeush et al. 2018). The northern Indian Ocean has the highest flux of inorganic (CaCO₃) and particulate organic carbon (Sarma et al. 2007). Many workers have documented the temporal changes in both the organic matter and CaCO₃ burial in the northern Indian Ocean and the processes affecting these changes (von Rad et al. 1999; Staubwasser & Sirocko, 2001; Reichart et al. 2002; Agnihotri et al. 2003 a, b; Naik et al. 2014; Azharuddin et al. 2017; Naik et al. 2017). The monsoon-induced productivity, terrigenous influx by the rivers, grain size and bottom water conditions strongly influence the C_org and CaCO₃ content in the modern surface sediments of the Arabian Sea (Kolla et al. 1981; Galy et al. 2007). The large spatial heterogeneity observed in the distribution of C_org and CaCO₃ in the surface sediments of the Arabian Sea (Kolla et al. 1981; Paropkari et al. 1992) is also prevalent during the geologic past. The continuous increase in CaCO₃, biogenic opal, biogenic Ba and their mass accumulation rates throughout the Holocene in the SE Arabian Sea were attributed to higher productivity due to increased upwelling during the summer monsoon (Naidu, 1991; Thamban et al. 1997; Naidu & Shankar, 1999; Bhushan et al. 2001; Pattan et al. 2003). However, the decrease in CaCO₃ abundance in the NE Arabian Sea during the Holocene was attributed to terrigenous dilution (Naidu, 1991; Azharuddin et al. 2017). Interestingly, the relatively high C_org but low CaCO₃ in the SW margin of India during the late Holocene was attributed to increased productivity and diagenesis (Kessarkar & Rao, 2007). Although, the CaCO₃ content in the NE and mid-eastern Arabian Sea was more during the interglacial period, no specific variation was observed in the organic matter content during glacial and interglacial times (Guptha et al. 2005). A rapid decrease in productivity coeval with very low dissolved oxygen in the bottom waters during the Last Glacial Maximum was reported from the SE Arabian Sea (Naik et al. 2017). A strengthened winter monsoon during the Little Ice Age was inferred from the NE Arabian Sea (Böll et al. 2014). A consistent increase in C_org throughout the Meghalayan Age was attributed to increased sedimentation and better preservation under a more reducing environment (Nagoji & Tiwari, 2017). Interestingly, an altogether different pattern with higher C_org accumulation during the cold glacials and stadials as compared to low accumulation during the warmer interglacials and interstadials in the neighbouring Bay of Bengal was attributed to intense winter-monsoon-induced increased marine primary production during the colder intervals (Weber et al. 2018; Xu et al. 2021).

Therefore, a vast spatial variation is observed in both the organic matter and CaCO₃ burial in the Arabian Sea during the past, necessitating more regional records. The majority of previous studies focused on carbon burial changes during the glacial–interglacial interval. High-resolution carbon burial studies focused on the Holocene are limited. Additionally, the sample resolution in the majority of previous studies was too coarse to understand short-term carbon burial changes in the northern Indian Ocean. Therefore, the objective of this work was to reconstruct the multi-decadal carbon burial changes from the Gulf of Mannar during the last ∼5000 years covering the Meghalayan Age (4.2 ka to recent), and to understand the factors affecting carbon burial in this region.

2. Study area

The gravity core (SSD004 GC02) collected from the upper slope of the Gulf of Mannar, off the southern tip of India (8° 37.9443′ N, 78° 44.1874′ E), during the fourth cruise of RV Sindhu Sadhana (October 2014) was used (Fig. 1). The core was retrieved from a depth of 1002 m. The study area is between the west coast of Sri Lanka and the southeast coast of India. The oceanographic processes mainly control the salinity, as except for the only perennial river of Tamil Nadu, namely the Thamirabarani River, no significant rivers drain directly into the Gulf of Mannar. A sizable seasonal change is observed in both the seawater temperature (varying from a lowest of 26.6°C during the summer monsoon season to a highest of 28.5°C during the pre-summer-monsoon season, in the top 25 m of the water column) (Locarnini et al. 2018) and salinity (varying from a lowest of 33.8 psu during the post-summer-monsoon season to the highest of 35.0 psu during the summer season, in the top 25 m of the water column) (Zweng et al. 2018). The changes are, however, restricted to the top ∼200 m of the water column, and the deeper waters are relatively stable (Locarnini et al. 2018; Zweng et al. 2018; Fig. 2). The hydrography of the area is controlled mainly by the monsoon system. The mean annual rainfall varies from 760 mm to 1270 mm (Sulochanan & Muniyandi, 2005). The study area receives copious rainfall during both the summer and winter monsoon. The summer monsoon occurs from June to September and the winter monsoon occurs from October to November. The summer monsoon brings more rainfall as compared to the winter monsoon (Gadgil & Kumar, 2006). The seasonal reversal of winds results in considerable variation in the intensity of upwelling as well as the primary productivity (Haake et al. 1993). The upwelling increases productivity in the area during May–June (Thomas et al. 2013). The Ekman Pumping induces large phytoplankton bloom increasing the chlorophyll-a content (up to 2 mg m⁻³) in the SW Bay of Bengal during the winter monsoon season (Vinayachandran & Mathew, 2003). The high-productivity water advects into the Gulf of Mannar from the Palk Bay during the winter season, increasing the productivity (Jyothibabu et al. 2021). The large influx from the rivers draining into the Gulf of Mannar also increases the nutrient availability and, in turn, productivity during the winter season (Chandramohan et al. 2001). The zooplankton biomass and chlorophyll-a concentration during the winter season is comparable with that during the summer season in the Gulf of Mannar (Jagadeesan et al. 2013). The seasonally reversing winds generate coastal currents that transport warm saltier water from the Arabian Sea into the Bay of Bengal during the summer, and cold fresher water from the Bay of Bengal into the Arabian Sea during the winter (Schott & McCreary, 2001).

Fig. 1. The core location and other cores from the eastern Arabian Sea (SO90-39 KG/SO130-275 KL, Böll et al. 2014; SO90-63 KA, Burdanowitz et al. 2019; SK240/485, Azharuddin et al. 2017; SK291 GC15, Saravanan et al. 2019; SN-6, Nagoji & Tiwari, 2017; AAS-VI/GC-05, Pattan et al. 2019; SK237 GC04, Naik et al. 2017) discussed in the paper. The filled black square is the location of core SSD004 GC02 in the Gulf of Mannar. The coloured contours are bathymetry/topography and the scale is on the right. The major bathymetric and topographic features and Thamirabarani River are also marked. The faint blue lines mark the major rivers draining in the northern Indian Ocean.

Fig. 2. Annual and seasonal water column temperature (Locarnini et al. 2018) and salinity (Zweng et al. 2018) at the core location.

The region has an extensive relict carbonate platform, with the age of the overlying sediments varying from 7.3 to 8.4 ka (Rao et al. 2003). The sediments are sandy on the continental shelf, gradually dominated by the silt and clay on the slope and further deeper regions (Hashimi et al. 1981; 1982; Singh et al. 2018). The inner shelf sediments are dominated by CaCO₃ of biogenic origin (foraminifera, molluscs, pteropods) (Hashimi et al. 1982). The C_org in surface sediments varies from 1.5 % to 6.9 % (Singh et al. 2018). The dissolved oxygen varies from 0.48 mL L⁻¹ to 3.84 mL L⁻¹, with the oxygen-deficient zone (<2 mL L⁻¹) between 152 and 1550 m (Singh et al. 2018). Based on the previous studies, sedimentation rate is comparatively high on the slope (Ray et al. 1990; Singh et al. 2017). Consequently, the core location was selected to ensure a high sedimentation rate record.

3. Materials and methodology

The core (SSD004 GC02) was 5.95 m long and the top 1.50 m section of the core was used for the study. The core was subsampled at 1 cm intervals and thus a total of 150 samples were used. The samples were processed in several stages.

3.1. Sample processing for foraminiferal studies

A small aliquot (5–10 g) of sediment was collected in a pre-weighed glass Petri dish and freeze-dried. The dried sediments were weighed and sieved using a 63 µm sieve. The material retained on the sieve (>63 µm, coarse fraction, CF) was dried, weighed and stored in clean plastic vials. For picking planktic foraminifera, one-quarter of the coarse fraction was weighed and dry-sieved using a 125 µm sieve. Both the >125 µm and <125 µm fractions were then weighed and stored in separate vials. A representative aliquot of >125 µm fraction was weighed and uniformly spread in a picking tray. A minimum of 300 completely intact planktic foraminifera tests were picked from this fraction using an Olympus SZX16 stereozoom microscope. Planktic foraminiferal abundance was normalized to 1 g dry sediment weight. The number of specimens of Globigerina bulloides was counted and its relative abundance was calculated.

3.2. Total carbon, nitrogen and inorganic carbon analysis

About 1–2 g of the freeze-dried sediment was finely powdered for the total and inorganic carbon and total nitrogen analysis. The total inorganic carbon (TIC) was analysed using a coulometer (model CM 5015 CO₂, UIC Inc. USA). The pure limestone (CaCO₃) was used as standard, with carbon percentage of 12 %. The total carbon and total nitrogen content was analysed using the Flash 2000 series CN Elemental Analyzer. For total carbon and nitrogen analysis, NC Soil Standard was used. The nitrogen content in the standard was 0.21 ± 0.01 % and the carbon was 2.29 ± 0.07 %. The C_org was estimated by subtracting inorganic carbon from the total carbon. The CaCO₃ was estimated by multiplying the inorganic carbon by 8.33 (Johnson et al. 2014).

3.3. Radiocarbon dating

The radiocarbon dating provides the age of carbon-comprising material derived from the living organisms. The chronology of the top 1.5 m section of the core was established by five accelerator mass spectrometer (AMS) radiocarbon dates (Table 1). The surface-dwelling planktic foraminifera, namely Globigerinoids ruber and Trilobatus sacculifer, were picked for radiocarbon dating. The radiocarbon dates were obtained from the Inter University Accelerator Center, Delhi, India, and the Center for Applied Isotope Studies (CAIS) at the University of Georgia. The radiocarbon ages were calibrated using the CALIB 8.2 radiocarbon calibration program (Stuiver & Reimer, 1993; Reimer et al. 2013). A reservoir correction of 77 ± 58 yr from the nearby locations (Dutta et al. 2001; Southon et al. 2002) was used to calibrate the dates.

Table 1. AMS radiocarbon age details

4. Results

4.1. Chronology

The chronology of the core was established by using the Bacon age model (Blaauw & Christen, 2011), utilizing the five AMS radiocarbon dates (Table 1; Fig. 3). Based on the sedimentation rate between the top two dated intervals (39.5 cm and 49.5 cm), the core top was assigned a modern age. The age of the bottommost section (149–150 cm) of the studied core was radiocarbon-dated to be 4600 ± 25 yr. The modelled age for the bottommost section is 4926 (−445 / +262) yr BP. The age uncertainty varies from a minimum of −46 / +100 yr towards the core top to −445 / +282 yr in the older section. Thus, the core covers the entire Meghalayan Age. The sedimentation rate varied from 13.0 cm kyr⁻¹ to 67.6 cm kyr⁻¹, with a mean sedimentation rate of 34.5 cm kyr⁻¹ (Fig. 3). The sample resolution varied from 15 years to 77 years, with an average resolution of 41 years.

Fig. 3. The chronology of core SSD004 GC02 asestablished by Bacon age model, utilizing the AMS radiocarbon ages. The core top age was interpolated to be modern, based on the sedimentation rate between the subsequent radiocarbon-dated intervals. The radiocarbon ages are plotted as grey filled points, and the age uncertainty is marked by the dotted line envelope.

4.2. Coarse fraction (>63 µm)

The coarse fraction comprises of sand-sized sediments. As the sediments were not pre-treated, the sediments also contained biogenic carbonates. A gradual increase in coarse fraction abundance is observed from the bottom of the section to a depth of 90 cm (3.45 ka). The coarse fraction was most abundant (15.5 %) at 90 cm (3.45 ka). Subsequently, it decreased abruptly, only to increase again at 72 cm (2.93 ka). A very prominent abrupt decrease in coarse fraction (4.2 %) was observed at 47 cm (2.20 ka). The coarse fraction abundance increased again at 43 cm (2.05 ka). The coarse fraction decreased gradually from 43 cm onwards, except for a minor increase in the core top sections (Fig. 4).

Fig. 4. Down-core variation in total carbon, inorganic carbon, organic carbon, C_org/N and coarse fraction (>63 µm) in core SSD004 GC02. The yellow shaded bar is the Northgrippian Age.

4.3. Calcium carbonate (CaCO₃)

A 6 % decrease in CaCO₃ is observed from the bottom of the studied section up to a depth of 120 cm (4.20 ka) (Fig. 4). Subsequently, CaCO₃ increased rapidly up to a depth of 92 cm (3.51 ka). Later, the weight percentage of CaCO₃ decreased up to a depth of 78 cm (3.10 ka). From 3.08 ka onwards, CaCO₃ increased gradually up to a depth of 28 cm (1.36 ka) to reach a peak value of 35.0 %. Two prominent peaks, centred at 72 cm and 55 cm (2.91 ka and 2.45 ka, respectively), are observed within this gradual increase. The concentration of CaCO₃ remained uniform in the top ∼25 cm section (1.17 kyr) of the core (Fig. 4).

4.4. Organic carbon (C_org)

The down-core variation in C_org is similar to CaCO₃ (Fig. 4). A small decrease (0.3 %) in C_org is observed from the bottommost section until 4.39 ka (128 cm). The C_org gradually increased by 1.7 %, from 4.37 ka (127 cm) until 0.60 ka (15 cm). C_org remained uniform in the top 15 cm section (0.68 kyr) of the core. A few minor variations are also observed within the gradual increase in C_org.

4.5. Organic carbon/nitrogen (C_org/N)

The organic matter in marine sediments accumulates from both the land and marine sources. The terrestrial and marine organic matter has a distinct carbon to nitrogen ratio (C_org/N) (Calvert et al. 1995) and thus is used to understand the change in the relative contribution of these two sources. C_org/N increased from the bottommost section until 4.51 ka (135 cm). A continuous decrease in C_org/N, from 10.56 to 7.76, is observed from 4.23 ka (122 cm) onwards to the core top (Fig. 4). Within this gradual decreasing pattern, a few prominent lows (4.02 ka, 102 cm and 1.36 ka, 32 cm) are also observed.

4.6. Planktic foraminiferal abundance (specimen/g sediment)

After a minor decrease from the bottommost section until 4.34 ka, the planktic foraminiferal abundance increased from 4.22 ka to 1.98 ka (∼41 cm). Subsequently, the planktic foraminiferal abundance decreased until 0.25 ka (∼7 cm). The abundance was again high in the top 7 cm section (0.29 kyr) of the core (Fig. 5). A few minor fluctuations in planktic foraminiferal abundance are also observed within the general trend stated above. The abundance varied from a minimum of 1945 specimen/g sediment at 4.22 ka (124.5 cm) to a maximum of 10 110 specimen/g sediment at 1.98 ka (41 cm).

Fig. 5. The absolute abundance of planktic foraminifera normalized to 1 g dry sediment and the relative abundance of upwelling indicator species Globigerina bulloides in core SSD004 GC02. The yellow shaded bar is the Northgrippian Age.

4.7. Relative abundance of Globigerina bulloides (%)

Globigerina bulloides is a widely accepted upwelling indicator planktic foraminifer. The relative abundance of G. bulloides was low in the bottommost section of the core (average 18.0 % between 4.6 ka and 4.9 ka). The relative abundance increased and remained high (average 22.6 %) until 1.5 ka (32 cm). A prominent decrease in G. bulloides relative abundance was observed in the top 25 cm section (1.17 kyr) of the core (Fig. 5).

5. Discussion

The total carbon, CaCO₃ as well as C_org, was very low towards the end of the Northgrippian Age covered in the studied core section (4.2–4.9 ka) and the beginning of the Meghalayan Age. Following these low sedimentary carbon values at the Northgrippian–Meghalayan transition, both the organic and inorganic carbon increased throughout the Meghalayan Age. The beginning of the increase in C_org (4.4 ka) preceded that in CaCO₃ (4.2 ka) and resulted in higher total carbon burial in the Gulf of Mannar during the Meghalayan Age (Figs 6, 7). The low CaCO₃ as well as C_org during the Northgrippian–Meghalayan transition is attributed to weaker monsoon-induced productivity. The weaker summer monsoon at the Northgrippian–Meghalayan boundary has also been inferred from the terrestrial records (Enzel et al. 1999; Dixit et al. 2014; Kotlia et al. 2015). The persistent increase in total carbon, CaCO₃ as well as C_org, stabilized between 2.7 ka and 3.7 ka as well as in the top ∼1.2 kyr section of the core.

Fig. 6. A comparison of C_org variation in the Gulf of Mannar (SSD004 GC02) during the last 5 kyr with that in different parts of the eastern Arabian Sea (SO90-39 KG/SO130-275 KL, Böll et al. 2014; SO90-63 KA, Burdanowitz et al. 2019; SK291 GC15, Saravanan et al. 2019; SN-6, Nagoji & Tiwari, 2017; SK237 GC04, Naik et al. 2017). The yellow shaded bar is the Northgrippian Age.

Fig. 7. A comparison of CaCO₃ wt % variation in the Gulf of Mannar (SSD004 GC02) during the last 5 kyr with that in different parts of the eastern Arabian Sea (SO90-39 KG/SO130-275 KL, Böll et al. 2014; SO90-63 KA, Burdanowitz et al. 2019; SK291 GC15, Saravanan et al. 2019; SN-6, Nagoji & Tiwari, 2017; SK237 GC04, Naik et al. 2017). The yellow shaded bar is the Northgrippian Age.

In the SE Arabian Sea, CaCO₃ content in sediments is strongly coupled with monsoon (Guptha et al. 2005; Narayana et al. 2009). Thus, the increased C_org, CaCO₃ and nitrogen during the Holocene has often been used to infer higher productivity due to stronger summer monsoon in the eastern Arabian Sea (Kessarkar et al. 2010). A substantial fraction of the CaCO₃ is biogenic, comprising calcareous shells. Amongst a huge variety of calcareous marine organisms, foraminifera and coccolithophores contribute the largest fraction of the biogenic carbonate flux in the ocean (Ramaswamy & Gaye, 2006; Langer, 2008). The increase in CaCO₃ is thus mainly due to an increased abundance of foraminiferal shells. The diversity and abundance of foraminifera depends on the ambient conditions, especially food availability (Schiebel et al. 2001). In the northern Indian Ocean, monsoon influences the C_org flux, the food for foraminifera. Both summer and winter monsoons affect the primary productivity in the northern Indian Ocean by bringing nutrients from the land by terrigenous influx as well as through upwelling and convective mixing (Madhupratap et al. 1996; Sreeush et al. 2018). The enhanced terrigenous supply to the Arabian Sea and a subsequent increase in biological productivity during the summer monsoon has been confirmed by sediment trap studies (Nair et al. 1989). Many zooplanktons feeding on primary producers have a calcareous skeleton. The calcareous skeletons also sequester a substantial fraction of carbon and bury it in the ocean sediments. Thus the CaCO₃ in the sediments is influenced by productivity, dissolution of CaCO₃ and dilution by terrigenous material (Naidu, 1991; Pattan et al. 2019). The shift in monsoon-induced evaporation–precipitation during the Holocene was synchronous with a change in surface productivity, planktic foraminiferal abundance and coarse sediment fraction (Saraswat et al. 2016). In modern times, the higher primary productivity during the later phase of the summer monsoon is attributed to the coastal upwelling and river runoff bringing nutrients to the surface waters (Jyothibabu et al. 2008).

The increasing CaCO₃ thus suggests that the monsoon began to intensify at 4.0 ka, which would have enhanced the upwelling-induced productivity in the Gulf of Mannar (Naidu, 1991). The findings are in line with the records from the northern India and eastern Arabian Sea. Dixit et al. (2014) reported that the weak monsoon phase at the Northgrippian–Meghalayan transition lasted for only 200 years and further that the monsoon recovered to the present level by 4 ka. A clear monsoon intensification trend beginning at 4 ka is also observed in core 63KA recovered from the northern Arabian Sea (Staubwasser et al. 2003). A similar consistent increase in CaCO₃ in the SE Arabian Sea, during the late Holocene, was also attributed to the strengthened monsoon (Sarkar et al. 2000). The increasing trend in CaCO₃ in the Gulf of Mannar matches with another core (AAS-VI/GC-05) collected off Cochin from the SE Arabian Sea (Pattan et al. 2019), suggesting a strong regional feature (Fig. 7). However, the high primary productivity during the winter season could also drive the increase in CaCO₃. The increased Indus runoff during the late Holocene was attributed to strengthened winter monsoon precipitation (Staubwasser et al. 2003).

The higher primary productivity may not always result in increased biogenic carbonate flux. The bottom water conditions, including the dissolved oxygen concentration and grain size of the sediments, strongly influence the burial of both the C_org and biogenic carbonate. Incidentally, the increase in both the primary productivity and denitrification during the Northgrippian and Meghalayan (since ∼7 ka) in the eastern Arabian Sea was reported to be coeval with an increase in CaCO₃ dissolution, as evident from the low CaCO₃ concomitant with low shell weight and prominent dissolution features in the shells, suggesting a significant regional bias in preservation (Naik et al. 2014). The decreased CaCO₃ was contemporaneous with the lowest dissolved oxygen levels in the bottom waters (Naik et al. 2014). Thus, the preservation of foraminiferal carbonate in the ocean sediments resulting in long-term carbon burial, varies regionally (Naik et al. 2017) and depends on sedimentation rate (Agnihotri et al. 2003 b) and bottom water conditions (pH, CaCO₃ compensation depth). Earlier, similar higher productivity and extreme suboxic condition during the late Holocene (5.5 ka to present) were reported from the SE Arabian Sea (Pattan et al. 2019). Therefore, the consistent increase in CaCO₃ in the Gulf of Mannar suggests persistence of conditions favouring carbon burial in the sediments. The stabilization of carbon burial in the Gulf of Mannar in the last 1.5 kyr suggests weakening of the upwelling during the summer monsoon, as also inferred from the terrestrial records (Sanwal et al. 2013). The weakened upwelling signature in the Gulf of Mannar during the last 1.5 kyr is, however, opposite to that of the western Arabian Sea (Gupta et al. 2003). The response of this region to summer monsoon winds is different than that of the western Arabian Sea (Bassinot et al. 2011). It should, however, be noted here that the carbon burial in the upper section of the core might also be influenced by anthropogenic activities.

5.1. Organic carbon (C_org) contribution

The C_org increased throughout the Meghalayan Age, with the exception of the interval between 2.7 ka and 3.7 ka, as well as the top 1.2 kyr section (Fig. 6). A few other records from the SE Arabian Sea also have an increasing C_org burial during the Meghalayan Age, suggesting increased productivity mainly controlled by the summer monsoon (Diniz et al. 2018). A similar progressive increase in productivity is observed in the SE Arabian Sea, throughout the Holocene, with a sharp jump at 5.4 ka (Naik et al. 2017). The increase in C_org in the SE Arabian Sea since the mid-Holocene has been attributed to better preservation facilitated by the higher sedimentation rate and reducing conditions (Nagoji & Tiwari, 2017). The principal reason for the increased C_org accumulation in the sediments is the organic matter flux from the highly productive surface waters. Both, the strong upwelling due to intense summer monsoon (Sarma et al. 2007) and convective mixing during the winter season (Madhupratap et al. 1996) increase the primary productivity and the subsequent organic matter flux. Thus, a strong monsoon leads to an increase in productivity and subsequent higher C_org flux to the seafloor (Sreeush et al. 2018). The strong influence of monsoon on the biological productivity in the Arabian Sea has been confirmed from the sediment trap studies. The biological productivity is generally high during the summer monsoon (Nair et al. 1989). Additionally, increased productivity, lower than that during the summer season but higher than that in the non-monsoon months, is observed during the winter season (Nair et al. 1989; Guptha et al. 1997). The past records also demonstrate a strong influence of productivity on organic matter flux. A substantial increase in C_org content and CaCO₃ in the eastern Arabian Sea has been reported during the Holocene, and attributed to the increased productivity (Thamban et al. 1997; Agnihotri et al. 2003 a). Thus the increased C_org accumulation during most of the Meghalayan Age in the Gulf of Mannar can be attributed to increased productivity. However, it is to be noted that the quick burial and dissolved oxygen concentration at the sediment–water interface strongly modulate C_org preservation.

The core is located towards the lower boundary of the oxygen minimum zone. Therefore, the temporal variation in carbon burial is likely to be influenced by the changes in bottom water oxygen concentration. In the eastern Arabian Sea, the higher C_org concentration coincides with the oxygen minimum zone (OMZ; 150–1500 m), suggesting the strong influence of anoxic bottom waters in preserving the organic matter (Paropkari et al. 1992). However, from the subsequent studies covering a broader zone, it was found that the C_org and nitrogen are maximum between 200 and 1600 m depth and the lowest dissolved oxygen is at 200 and 800 m depth (Calvert et al. 1995). Thus, although the deficient oxygen is one of the factors favouring better organic matter preservation, variation in supply, dilution by other sedimentary components and texture of the sediment also strongly influence the amount of organic matter buried in the sediments (Calvert et al. 1995). The C_org and CaCO₃ show a similar trend throughout the core, except for a few minor short-term deviations (between 2.6 ka and 3.6 ka). The CaCO₃ fraction in the sediment decreased whereas C_org increased from 3.0 ka to 3.4 ka. The subsequent abrupt increase in the CaCO₃ fraction at 2.9 ka was contemporaneous with a decrease in C_org. The short-term opposite trend of C_org and CaCO₃ concentration is due to diagenetic effects, mainly sulphate reduction and other associated processes. The sulphate reduction dissolves the CaCO₃, but a higher sedimentation rate and increased clay content lead to a better preservation of C_org (Bhushan et al. 2001).

5.2. Upwelling indicator planktic foraminifera

The relative abundance of G. bulloides was higher than average (20.0 %), between 4.50 ka and 1.50 ka and lower in the last 1.47 kyr. A few planktic foraminifera thrive in upwelled colder, nutrient-rich waters with plenty of food. Globigerina bulloides is one such planktic foraminifera species abundant in cold, organic-matter-rich waters and thus is used as an indicator of upwelling (Auras-Schudnagies et al. 1989; Saraswat & Khare, 2010; Naik et al. 2017). From the foraminiferal distribution during the summer monsoon, it was suggested that the change in the relative abundance of G. bulloides could be used to trace the palaeo-upwelling intensity (Prell & Curry, 1981) and thus the associated summer monsoon strength and productivity in both the Arabian Sea and Bay of Bengal (Naidu et al. 1999). The relative abundance of G. bulloides in the Gulf of Mannar is comparable with that in typical summer-monsoon-wind-induced upwelling-affected regions like the western Arabian Sea (Gupta et al. 2003). Therefore, the higher relative abundance of G. bulloides between 1.50 ka and 4.50 ka (average 22.6 %) suggests intense upwelling and thus strengthened summer monsoon during the larger part of the Meghalayan Age, as compared to the last 1.5 kyr (Fig. 8). The change in the relative abundance of G. bulloides in the Gulf of Mannar is, however, different than that in the western Arabian Sea (Gupta et al. 2003; Saravanan et al. 2019). The abundance consistently decreases in the western Arabian Sea from ∼5 ka until ∼1.5 ka (Gupta et al. 2003), whereas we see a higher relative abundance of G. bulloides between 1.50 ka and 4.50 ka, but no consistent increasing or decreasing trend. The difference is attributed to the varying response of the different parts of the Indian Ocean to the summer-wind-intensity and direction-induced upwelling (Bassinot et al. 2011). The change in the relative abundance of G. bulloides between 1.5 ka and 4.5 ka slightly matches with that in a core (SK291/GC15) collected from the central eastern Arabian Sea (Saravanan et al. 2019). However, we need to mention that core SK291/GC15 was collected from a very weak upwelling area (Saravanan et al. 2019). A similar higher G. bullodies during most of the Meghalayan Age was also reported in another core (MD77-191) collected from the nearby SE Arabian Sea (Bassinot et al. 2011). The last ∼1.5 kyr record is missing from MD77-191, hampering the comparison of our late Meghalayan Age record with this core.

Fig. 8. The relative abundance of monsoon wind forced upwelling-induced cold nutrient-rich water indicator Globigerina bulloides in the Gulf of Mannar (SSD004 GC02), Oman Margin (ODP 723A, Gupta et al. 2003) and central eastern Arabian Sea (SK291 GC15, Saravanan et al. 2019). The yellow shaded bar is the Northgrippian Age.

It is interesting to note that although the relative abundance of G. bulloides was consistently high during the early Meghalayan Age, we did not see any persistent increasing trend. The lack of any increasing trend in G. bulloides relative abundance suggests additional factors facilitating increased productivity and thus C_org flux. It is likely that the increased carbon burial was due to the increased productivity during the winter season coupled with better preservation. The Gulf of Mannar receives precipitation during the winter season. The rivers debouching in the Gulf of Mannar bring nutrient-rich turbid waters during the winter season (Chandramohan et al. 2001). Additionally, the nutrient-rich water advects from the Palk Bay, in the SW Bay of Bengal, to the Gulf of Mannar (Jyothibabu et al. 2021). Therefore, the zooplankton biomass and chlorophyll-a concentration is high during both the summer and winter monsoon seasons in the Gulf of Mannar, suggesting increased productivity (Jagadeesan et al. 2013). The nearly fourfold increase in planktic foraminifera abundance from 1945 specimen/g sediment at 4.22 ka to 10110 specimen/g sediment at 1.98 ka suggests an overall increase in productivity, during this interval. As the summer-monsoon-induced upwelling, although it was high, did not increase consistently, the increase in productivity between 4.22 ka and 1.98 ka, as inferred from planktic foraminifera abundance, is attributed to intense winter monsoon.

On the finer timescale, from the relative abundance of G. bulloides, we report a weaker monsoon between 2.65 ka and 2.28 ka and a subsequent strengthening of the monsoon between 2.05 ka and 1.50 ka. The weak monsoon followed by the strengthened monsoon phase between 2.05 ka and 1.50 ka in the Gulf of Mannar core is similar to the weakening of the monsoon at 2 ka and the subsequent wet phase, inferred from the eastern Arabian Sea (Khare et al. 2008). However, the duration of the wet phase in the two records is different, likely because the region off Goa is mainly influenced by the summer monsoon, while the Gulf of Mannar receives substantial precipitation during both the summer and winter seasons. A part of the difference may also be because of the different radiocarbon age calibration methods followed in these records. The decreased G. bulloides abundance during the last 1 kyr, however, suggests a weakened summer monsoon. The impact of weakened summer monsoon during the last 1 kyr is evident in deceased C_org and CaCO₃ accumulation in the Gulf of Mannar. However, the Gupta et al. (2003) G. bulloides percentage data would suggest an increase in upwelling and thus strengthened summer monsoon winds during the last 1 kyr. The different trend in G. bulloides in the Gulf of Mannar and the western Arabian Sea is attributed to the differential response of these two regions to the summer winds (Bassinot et al. 2011).

5.3. Terrestrial versus marine organic carbon contribution

The C_org/N increased during the Northgrippian but consistently decreased from 4.0 ka onwards until recent times. Although, the overall decrease in C_org/N during the Meghalayan Age was ∼2, the trend was very prominent. In addition to the primary productivity, a substantial fraction of the organic matter in the ocean, especially the marginal marine regions, is also of terrestrial origin, brought by the river runoff as well as winds. The marine and terrestrial contribution of the organic matter is delineated with the help of C_org/N (Calvert et al. 1995). The marine organic matter has a relatively lower C_org to nitrogen ratio as compared to terrestrial plants. The C_org/N ratio is thus an index to determine the relative contribution of marine or terrestrial organic matter. The terrigenous organic matter generally has a high C_org/N (>20), whereas marine origin organic matter has low C_org/N (5–8) (Jasper & Gagosian, 1989). The C_org in core SSD004 GC02 is increasingly of marine origin, based on C_org/N ratio (Fig. 9). The gradually decreasing C_org/N ratio throughout the core confirms the increase in marine organic matter contribution to the C_org throughout the Meghalayan Age. The decreasing terrestrial organic matter contribution is attributed to the increasing distance of the core site from the Thamirabarani River due to a >10 m rise in sea-level since the beginning of the Meghalayan Age (Grant et al. 2014), as well as the increase in marine productivity. The increasing marine contribution to the organic matter, despite there being no such trend in upwelling indicator G. bulloides, further supports our inference of higher primary productivity due to winter monsoon in the Gulf of Mannar.

Fig. 9. The change in total carbon, CaCO₃, C_org, C_org/N and relative abundance of Globigerina bulloides during the Meghalayan Age, compared with the sea-level changes (Grant et al. 2014) and atmospheric CO₂ concentration (Bereiter et al. 2015). The yellow shaded bar is the Northgrippian Age. The intervals of significant change are marked by grey shaded regions.

5.4. Comparison with regional high-resolution records

We compared the SSD004 GC02 CaCO₃ and C_org with other similar high-resolution records from the eastern Arabian Sea (SO90-39 KG/SO130-275 KL, Böll et al. 2014; SO90-63 KA, Burdanowitz et al. 2019; SN-6, Nagoji & Tiwari, 2017; SK237 GC04, Naik et al. 2017). All these records are from the continental slope region. A couple of these records (SO90-39 KG/SO130-275 KL and SO90-63 KA) are from the NE Arabian Sea. The primary productivity is very high in the NE Arabian Sea due to convective mixing during the winter (Madhupratap et al. 1996) and advection of high-nutrient water from the western Arabian Sea during the summer (Saraswat et al. 2020). A similar high productivity during both the summer and winter seasons is also observed in the Gulf of Mannar, although the physical mechanisms are different than those in the NE Arabian Sea. A large difference is observed in both the trend and absolute abundance of CaCO₃ during the last 5 kyr in these two regions (Fig. 7). The Gulf of Mannar CaCO₃ record is different than other SE Arabian Sea records (Nagoji & Tiwari, 2017, Naik et al. 2017), but strikingly similar to another core, AAS-VI/GC-05, collected from a depth of 280 m in the SE Arabian Sea (Pattan et al. 2019). The SE Arabian Sea is influenced by upwelling-induced primary productivity during the summer season. Amongst these regions, the highest CaCO₃ was in the Gulf of Mannar during the Meghalayan Age. We report a significant increase in CaCO₃ between ∼1 ka and 3 ka, as compared to a consistent decrease in several other records. The records, however, match in a constant CaCO₃ during the last millennium. Interestingly, C_org consistently increased throughout the Meghalayan Age in all the records (Fig. 6). A similar C_org but different CaCO₃ in various parts of the eastern Arabian Sea is intriguing. A relatively high CaCO₃ in the Gulf of Mannar is attributed to its better preservation, due to a less intense oxygen-deficient zone as compared to the NE Arabian Sea (Sarma et al. 2020). The increased microbial respiration in a C_org-rich environment decreases the pH, leading to dissolution of biogenic carbonates (Cai et al. 2011). A similar C_org in all records suggests comparable productivity as well as its burial in sediments throughout the eastern Arabian Sea.

5.5. Factors affecting carbon burial and preservation

The preservation of both the biogenic CaCO₃ and organic matter in the sediments depends on ambient conditions. One of the major factors influencing organic matter preservation is the grain size. The finer grains preserve more organic matter (Bergamaschi et al. 1997). The sediments in the oceans are mainly brought by rivers or wind. The distance of a marine region from the river mouth thus affects the size and volume of the sediments being brought into an area. The sea-level also controls the grain size. The transgressing sea-level inundates and the regressing sea-level exposes the continental shelf. In the case of the marine regions close to the river mouth, the change in sea-level significantly shifts the point of sediment debouchment by rivers (Phillips & Slattery, 2006). Therefore, sea-level changes significantly affect the size as well as the volume of sediments being brought into the marginal marine regions and deeper waters through channels, and thus affect the carbon burial in marine sediments. The terrigenous dilution and/or dissolution increases the amplitude of the carbonate cycle but reduces CaCO₃ concentration in the Indian Ocean (Olausson, 1971; Naidu, 1991). The sea-level has increased by >10 m since the beginning of the Meghalayan Age (Grant et al. 2014; Fig. 9). The transgressing sea-level completely inundated the extensive, gently sloping relict carbonate platform along the southern margin of India (Hashimi et al. 1982). A significantly consistent CaCO₃ between 2.7 ka and 3.7 ka, matches with a decreased rate of sea-level rise during the same interval, as against a comparatively rapid rate of sea-level increase in the early Meghalayan Age. The inundation of the carbonate platform thus increased the distance of the core site from the nearby Thamirabarani River and thus reduced the coarse fraction supply to the area. However, the quantity and grain size of the sediments supplied by the river to the core site can also change due to the delta lobe avulsion as it is closely linked with the sea-level change (Chadwick & Lamb, 2021). The substantial change in terrigenous fraction and CaCO₃ was also highlighted from the western margin of India (Khare, 2018). In the SE Arabian Sea, the change in C_org during the glacial–interglacial interval was attributed to the variation in sediment texture as a result of higher terrigenous supply leading to increased dilution (Narayana et al. 2009). The resultant increase in finer fraction facilitated better preservation of both the organic matter and the biogenic carbonate.

The seawater pH at the sediment–water interface as well as of the pore water, also strongly modulates carbon burial (Keil, 2017; LaRowe et al. 2020; Freitas et al. 2022). As the core site lies in the oxygen-deficient zone, the pH at the sediment–water interface is likely to strongly modulate carbon burial in the sediments, before the remineralization of both the organic matter and CaCO₃. All such factors also have to be evaluated to fully understand the temporal changes in carbon burial in this region. With the available data, it is clear that both the intense monsoon-induced primary productivity and increasing sea-level facilitated the persistent increase in both the inorganic and carbon burial in the Gulf of Mannar, throughout the Meghalayan Age. However, the uniform carbon concentration in the topmost section of the core is most likely influenced by anthropogenic activities.

5.6. Carbon burial during the last millennium

A distinct shift in almost all the parameters in the top ∼25 cm section representing the last ∼1.17 kyr is intriguing. We want to state that this section has a chronological uncertainty as the age of the top section of the core was interpolated based on the sedimentation rate between the subsequent radiocarbon-dated intervals (39.5 cm and 49.5 cm). The total carbon remains uniform in this section, mainly due to the similar trend in CaCO₃. The increasing trend in C_org also flattens in the top ∼15 cm section of the core. The atmospheric CO₂ concentration increased rapidly during the later part of this interval, driven by increased fossil fuel usage since the beginning of the industrial revolution (Fig. 9; Bereiter et al. 2015). The low relative abundance of G. bulloides in this section suggests a reduced upwelling and thus weaker summer monsoon. A weaker summer monsoon during the Late Meghalayan Age has also been reported from the terrestrial records (Srivastava et al. 2017). Thus the flattening of the total carbon burial during the last 1.17 kyr was driven by the weakening of the summer monsoon. It should, however, be noted here that the findings are in contrast with the high relative abundance of G. bulloides during the similar interval in the western Arabian Sea, suggesting a stronger monsoon-induced upwelling (Gupta et al. 2003). The differential response of these two regions is attributed to the difference in the orientation of the coastline to the wind direction and thus upwelling (Bassinot et al. 2003). Thus, it is clear that the summer-winds-driven upwelling-induced productivity decreased during the last 1.17 kyr in the Gulf of Mannar.

5.7. Implications for carbon cycling

The increase in carbon burial during the Meghalayan Age in the eastern Arabian Sea coincides with ∼10 ppmv increase in the global atmospheric CO₂ (Fig. 9; Bereiter et al. 2015). It is intriguing that the atmospheric CO₂ increased during times of enhanced carbon burial. The opposite trend between the atmospheric CO₂ and carbon burial in the eastern Arabian Sea can be explained by several factors. First and the foremost is the regional nature of the carbon burial, with the ambient conditions in the Gulf of Mannar being favourable for carbon burial during the Meghalayan Age. Another factor contributing CO₂ is the carbonate counter pump, whereby precipitation of CaCO₃ by the marine organisms increases the CO₂ concentration in the surface waters and its subsequent efflux to the atmosphere (Salter et al. 2014). The increase in planktic foraminifera abundance during the early Meghalayan Age must have contributed CO₂ to the surface waters and subsequently to the increasing atmospheric CO₂, through the carbonate counter pump.

6. Conclusions

We report a persistent increase in the total carbon, CaCO₃ and C_org in the Gulf of Mannar, throughout the Meghalayan Age, except the bottommost (4.9 ka to 4.2 ka, 120–150 cm) and the topmost section (1.17 kyr, 0–25 cm) of the core. The increase in CaCO₃ and C_org is concomitant with a phase of high relative abundance of the upwelling indicator G. bulloides in the early Meghalayan Age, suggesting intense upwelling in response to the strong monsoon. The increasing planktic foraminifera abundance and decreasing C_org/N ratio suggest higher marine productivity due to winter monsoon. Thus the high CaCO₃ and C_org content in the Gulf of Mannar during most of the Meghalayan Age is attributed to high primary productivity influenced by the consistent summer and strong winter monsoon. The corresponding increase in sea-level during the early phase of the Meghalayan Age facilitated better preservation of both the C_org and CaCO₃, thus leading to increased carbon burial in the SE Arabian Sea. The uniform carbon content in the top section of the core is attributed to the weakening of the summer monsoon.