Among various climate drivers, direct evidence for the Intertropical Convergence Zone (ITCZ) control of sediment supply on the millennial scale is lacking, and the changes in ITCZ migration demonstrated in paleoclimate records need to be better investigated. Here, we use clay minerals and Sr–Nd isotopes obtained from a gravity core on the Ninetyeast Ridge to track the corresponding source variations and analyze the relationship between terrestrial material supply and climatic changes. On the glacial–interglacial scale, chemical weathering weakened during the North Atlantic cold-climate periods and falling sea level hindered the transport of smectite into the study area due to the exposure of Andaman and Nicobar Islands. However, the influence of the South Asian monsoon on the sediment supply was not obvious on the millennial scale. We suggest that the north–south migration of the ITCZ controlled the rainfall in Myanmar and further directly determined the supply of clay minerals on the millennium scale because the transport of smectite was highly connected with the ITCZ location; thus, the regional shift of the ITCZ induced an abnormal increase in the smectite percentage during the late Last Glacial Maximum (LGM) in our records. The smectite percentage in the studied core is similar to distinct ITCZ records but different in some periods, revealing that regional changes in the ITCZ were significantly obvious, the ITCZ is not a simple north–south displacement, and closer connections occurred between the Northern–Southern Hemisphere in the eastern Indian Ocean during the late LGM.
Deposited sediments are essential recorders of the paleoclimate and paleoceanographic conditions since the climate is tied to the whole sedimentation process from weathering and transport to the deposition of sediments on land. The terrestrial materials of “source–sink” systems are supplied to marine environments under the combined effects of multiple climate-related driving forces and ocean processes (Li et al., 2018; Yu et al., 2019), and understanding these effects is crucial for reconstructing the coevolutionary relationship of the paleoenvironment with the paleoceanographic conditions and paleoclimate. Various factors may control the formation and transport of terrestrial materials at low latitudes, such as the northeastern Indian Ocean. Recently, the South Asian monsoon has been revealed to be the main driving force of terrestrial material supply in Bangladesh and of hydrological changes in the Bay of Bengal (BoB, Dutt et al., 2015; Gebregiorgis et al., 2016; Joussain et al., 2017; Li et al., 2018; Liu et al., 2021). Moreover, the Intertropical Convergence Zone (ITCZ) is an important climate-driving force in low-latitude regions (Deplazes et al., 2013; Ayliffe et al., 2013), which has a pivotal role in heat transportation on Earth (Schneider et al., 2014), and the north–south shift of the ITCZ is thought to connect the climates in the Northern and Southern Hemisphere (Huang et al., 2019; Zhuravleva et al., 2021). Because monsoon dynamics are shaped by large-scale meridional temperature gradients and an ITCZ shift in tropical monsoon areas (Mohtadi et al., 2016), there are hopeful opportunities to analyze sediment responses to ITCZ or monsoon variations. The paleoclimate breakthroughs mentioned above enable us to analyze the response of sedimentary records to the ITCZ shift in the BoB more accurately. However, evidence for direct control of terrestrial sediment supply by the ITCZ remains lacking, which is an obstacle to understanding the response of the depositional environment to the ITCZ shift.
As the main deposition area for vast amounts of weathered Himalayan
materials, the BoB accumulates numerous Himalayan terrestrial materials that
are loaded by the Ganges–Brahmaputra (G-B) River (Goodbred and Kuehl, 2000)
and forms the largest subaqueous fan, the Bengal Fan (3000 km long from
north to south, 1400 km wide from east to west, with an area of
Recent studies have revealed that clay minerals can be used to effectively track changes in source areas in the source–sink system of the BoB due to the great differences in clay mineral components among the source areas around the BoB (Joussain et al., 2016; Li et al., 2017; Liu et al., 2019; Ye et al., 2020). Moreover, Sr–Nd isotopes have been widely reported to track the variations in sediment provenance in the BoB (Ahmad et al., 2005; Colin et al., 1999, 2006).
In this study, we measured clay minerals and Sr–Nd isotopes in a deep-sea
gravity core obtained from the southeastern BoB (Fig. 1) to reconstruct
variations in the sources of sediments in the Ninetyeast Ridge and to
further explore the climate forces that affected the supply of terrestrial
materials during the past 45 kyr. Core 17I106, located above the abyssal plain
at
Geographical setting of the BoB. The locations of cores 17I106
(red asterisks) and U1452 (orange diamond) are shown.
Gravity core 17I106 (90.0040
Age–depth model of core 17I106 in the northeastern Indian Ocean.
Carbon-14 and calibrated calendar ages of mixed planktonic foraminifera measured in core 17I106 in the northeastern Indian Ocean.
Clay minerals (
A total of 22 samples (
The age model is built based on 10 radiocarbon dates from core 17I106. The
top age is 3.8 ka, and the bottom age is 44.9 ka; thus, this core
covers a continuous sedimentary succession of the last
Comparison of clay mineral and Sr–Nd isotope data in the northeastern Indian Ocean with paleoclimate records.
The lower sedimentation rates (3–5 cm ka
The relatively high illite percentages measured in core 17I106 indicate that
the weathered Himalayan materials carried by the G-B River system are the
primary source of sediments in the study area (Fig. 4a). Compared with the
large amounts of materials loaded by the G-B River system, the weathered
areas and runoff volumes of the Indo-Burman ranges are relatively small, and
consequently, their sediment contributions are limited in the study area,
although their sediments are also characterized by relatively high illite
percentages (Joussain et al., 2016). Evidence of surface sediments in the
BoB further reveals that the smectite percentages of sediments in the
central region are significantly lower than those in the eastern and western
regions (Li et al., 2017; Liu et al., 2019), indicating that sediments of
Indian Peninsula origin are difficult to transport into the eastern BoB
through the central BoB. Because the limited weathering area of Andaman and
Nicobar Islands cannot provide a large amount of smectite according to
provenance studies (Ali et al., 2015), the Myanmar materials characterized
by high smectite percentages have the advantage of shorter transport
distances compared to those sourced from the Indian Peninsula as the main
source area of smectite around the BoB. Therefore, the most important source of smectite in the study area is the Myanmar region. In marine
environments, kaolinite is preferentially deposited in estuary areas due to
mineral segregation (Gibbs, 1977) and thus may not be transported over long
distances, so the kaolinite in the study area was most likely sourced from
neighboring Sumatra (Fig. 4a, Liu et al., 2012). The Sr–Nd isotopes
measured in the studied core are close to those measured in the
Irrawaddy, Indo-Burman ranges, and Sumatra source regions (Fig. 4b), indicating
that terrestrial materials with diameters
Sediment provenance of core 17I106 in the northeastern Indian Ocean.
In the northeastern BoB, the southwest monsoon turns southward into the Andaman Sea, resulting in the transport of sediments from the Indo-Burman Range and Irrawaddy River to the central Andaman Sea (Colin et al., 2006). The location of core 17I106, drilled on the Ninetyeast Ridge, was above the abyssal plain, and the terrestrial materials deposited to the west of this location are difficult to resuspend and deposit on the ridge under the force of bottom currents or turbidity currents. In fact, the G-B-loaded materials are mainly carried eastward by surface ocean currents in summer to the Andaman Sea, where the seasonal surface currents load materials from the Himalayan and Indo-Burman ranges into the Andaman Sea through the northern strait (NS) (P. Liu et al., 2020; Rayaroth et al., 2016). These G-B River sediments can also be transported southward along the west side of the Andaman and Nicobar Islands (Fig. 5), and a westward ocean surface current in the middle strait (MS) loads sediments of the Irrawaddy River southwest into the study area (Chatterjee et al., 2017).
Map showing dispersal patterns of the BoB clay minerals for core 17I106. The locations of core 17I106 (red asterisks) and of the reference core and sites are shown: SK 170/2 in the northern BoB, SK-168, SK-234-60, NGHP Site 17 in the western Andaman Sea, Mawmluh Cave in northeastern India, and Tengchong Qinghai Lake in China are represented by orange diamonds. The orange, purple, and red arrows represent the main dispersal directions of illite, smectite, and kaolinite when the fluvial sediments were discharged into core 17I106. The white and red arrows denote the SW and NE monsoon currents, respectively. In the western BoB, the East Indian Coastal Current (EICC) reverses annually with the monsoon wind (Schott and McCreary, 2001). In the lower-latitude regions of the BoB, monsoon-driven currents flow eastward in summer to form the summer monsoon current (SMC) and westward in winter to form the winter monsoon current (WMC) (Shankar et al., 2002). The elevation legend is shown to the right of this figure.
In general, illite is the major mineral produced during the strong physical erosion of metamorphic rocks and granite rocks as well as during the reprocessing of sedimentary rocks (Chamley, 1989; Winkler et al., 2002), while smectite is the secondary mineral produced during the chemical weathering of parent aluminosilicate and iron–magnesium silicate under warm and humid climate conditions (Chamley, 1989; Hillier, 1995). The climatic forces from the North Atlantic are thought to extensively impact the tropical eastern Indian Ocean (EIO) and surrounding areas of the BoB (Sun et al., 2011; DiNezio and Tierney, 2013; Dutt et al., 2015; Gautam et al., 2020; Mohtadi et al., 2014; Liu et al., 2021), whose climate signals can be transmitted via the tropical Atlantic bipolar surface sea temperature (SST) anomaly and associated southward shift of the ITCZ (Marzin et al., 2013), westerlies teleconnection and sea ice (Sun et al., 2011), or the reorganization of the Hadley circulation (Mohtadi et al., 2014). During the North Atlantic cold-climate periods (Heinrich events and YD period, Fig. 3h), when rainfall and temperatures decreased in the South Asian monsoon region (An et al., 2011; DiNezio and Tierney, 2013; Gautam et al., 2020), physical weathering was enhanced in the Himalayas (Joussain et al., 2016), which made illite percentages at core 17I106 relatively high during these cold-climate periods, but chemical weathering weakened in Myanmar, and the smectite percentage thus decreased in the source area before these cold periods and continued to increase after these periods. The increasing (decreasing) trend of illite (smectite) percentages before cold-climate periods and the decreasing (increasing) trend of illite (smectite) percentages after cold-climate periods in our records suggest that the weathering degree in the source area influenced the supply of clay minerals during these cold-climate periods.
Sea level fluctuation is also critical in controlling the supplementation of
terrestrial materials, especially clay minerals (Li et al., 2018; Liu et
al., 2019), by changing the transport paths and/or distances as well as the
further input of sediments into the study area. The changing trends of the
sea level in seas adjacent to the BoB (Fig. 3i, Waelbroecka et al., 2002;
Grant et al., 2014; Hanebuth et al., 2000; Thompson and Goldstein, 2006) are
well correlated with the smectite percentages measured in core 17I106,
especially during 35–21 ka, when the smectite percentages declined
continuously. Since the Andaman and Nicobar Islands connecting the Andaman
Sea and the BoB have continuously expanded as the sea level has continuously
declined, the strait width has been consistently reduced, thereby preventing
the entrance of terrestrial materials into the Andaman Sea and the further
continuous decline in smectite percentages in the study area. Here, we
suggest that the variations in the measured illite percentages were mainly
caused by changes in smectite deposition because the sedimentary records
obtained from the northern BoB do not support the controlling effect of sea
level on illite percentages over the past 50 kyr (Joussain et al., 2016; Li
et al., 2018; Liu et al., 2019). The relative exposure of 200 km from
the current Irrawaddy River delta may affect the deposition process on the
continental shelf or further deposition of the sediments delivered to the
deep ocean, but core 17I106 is formed by the long-distance transport of
large amounts of fine-grained terrestrial material, indicating that these
sediments can be transported over long distances, and the
The South Asian summer monsoon is normally thought to be an important factor affecting weathering conditions around the BoB (Dutt et al., 2015; Gebregiorgis et al., 2016; Joussain et al., 2017; Li et al., 2018; Rashid et al., 2011; Zhang et al., 2020; Zorzi et al., 2015). Stalagmites in Mawmluh Cave record variations in river runoff in the surrounding area; these variations are determined by the impacts of surface sea temperature (SST) and water vapor transport paths (Dutt et al., 2015). In fact, the Mawmluh Cave records of the South Asian monsoon strength are driven by temperature gradients that drive changes in winds and moisture transport into the BoB (Dutt et al., 2015), not just responding to the rainfall amount. The smectite percentage changes measured in core 17I106 were slightly correlated after Heinrich event 1 (H1) but were irrelevant before H1 (Fig. 6b). This indicated that the combination of temperature and moisture failed to play a crucial role in smectite transport to core 17I106, although weathering features in the source area may be shaped by the South Asian monsoon. Moreover, the view could be confirmed by the smectite record obtained from the studied core not being well correlated with records previously obtained in the Andaman Sea (Fig. 6c, d, Gebregiorgis et al., 2016) or with a sporopollen record obtained in southwestern China (Fig. 6e, f, Zhang et al., 2020), especially before the LGM. The consistency of salinity and SST in core SK 168 (Fig. 6c, d) as well as moisture and temperature index (Fig. 6e, f) in southwestern China reveal that the hydroclimate in the South Asian monsoon region might have been influenced by SST in the Indian Ocean. All these inconsistencies between the smectite percentage in core 17I106 and monsoon records indicate that smectite supplementation may be mainly controlled by rainfall rather than by chemical weathering due to thermodynamic differences between sea and land environments (S. Liu et al., 2020).
Comparison of smectite percentages in core 17I106 with paleoclimate records.
During the late LGM, the smectite percentage increased abnormally in core 17I106, and this increase cannot be explained by dry and cold weathering conditions, a lower sea level, or a weakened summer monsoon at that time. In contrast, this abnormal change may have been attributed to an increase in the smectite input in sediments from the Burman source area or to a decrease in the amounts of sediment input from the Himalayas. Under the influence of the winter monsoon during the LGM, the denuded sediments on the Irrawaddy estuary shelf may have been transported southward through the west side of Andaman Island (Prajith et al., 2018), as was confirmed in previous work showing that the winter monsoon led to an increase in terrestrial materials from the Irrawaddy River to the Ninetyeast Ridge during the Heinrich event (Ahmad et al., 2005). However, the winter monsoon was strong in the western part of the study area from 21 to 15 ka (Fig. 6g), and the sea level remained relatively low during that period (Gautam et al., 2020). The smectite percentages in the studied core increased significantly from 21 to 19 ka and dropped rapidly after 19 ka. This inconsistency contradicts the conclusion that the increased smectite percentage in the source area was caused by a strong winter monsoon. Moreover, the changes in the sediment compositions measured in the Himalayan source area were probably related to variations in regional glaciers. During the LGM period, the increased glacial cover may have reduced surface runoff and enhanced the transport of physical weathering products, while the increased amount of ice meltwater may have transported more illites following glacial melt. However, the reduced glacial area in the Himalayas during 18–15 ka did not occur simultaneously with the increased illite percentage (Yan et al., 2020; Weldeab et al., 2019, Fig. 6h). Therefore, the abnormal changes measured in the smectite percentage during the late LGM period were caused by other climate-driven mechanisms, and the millennium-scale smectite percentage fluctuations that occurred before the LGM require a more reasonable explanation.
Changes in rainfall and the corresponding runoff are generally utilized to explain short-term variations in clay minerals. In the EIO, rainfall is controlled by monsoon activities (An et al., 2011; Beck et al., 2018; Gebregiorgis et al., 2016) and/or ITCZ migrations (Deplazes et al., 2013; Stoll et al., 2007; Tan et al., 2019). Glacial–interglacial monsoon precipitation changes at the regional scale are shaped by dynamics (changes in the wind fields) and temperature (McGee, 2020). The wind fields may be driven by the relative dominance of the northern low-pressure and southern high-pressure systems (An et al., 2011) as well as cross-equatorial moisture transport (Clemens et al., 2021), while the SST in the eastern Indian Ocean (Zhang et al., 2020) and western Indian Ocean (Wang et al., 2022), surface and subsurface temperature changes (Tierney et al., 2015), and temperature gradients (Weldeab et al., 2022) also play an important role in South Asian rainfall. At the same time, as a climate-driving force in low-latitude regions, ITCZ migrations may be the main factor responsible for regional hydrological changes (Deplazes et al., 2013; Weber et al., 2018) since the shift in the ITCZ was considered to control rainfall distribution and intensity in central India over geological timescales (Zorzi et al., 2015) and to cause summer temperature and moisture fluctuations in southwestern China during the last deglaciation (Zhang et al., 2019).
During the glacial–interglacial period, the ITCZ migrated north–south and balanced thermal differences by transferring atmospheric heat; this process represents an indispensable climate-regulating power on Earth (Broccoli et al., 2006; McGee et al., 2018; Schneider et al., 2014). In the Cariaco Basin and Arabian Sea (Fig. 7a–b), tropical rainfall is highly correlated with the North Atlantic climate, and sea ice variations in the North Atlantic affect the north–south shift of the ITCZ in low-latitude regions through atmospheric circulation and ocean processes (Deplazes et al., 2013). The smectite particles measured in core 17I106 mainly came from the Myanmar source area; in this area, rainfall is greatly affected by the seasonal shift of the ITCZ. Before the LGM, the smectite percentages in the study core were well matched with the ITCZ record in the Arabian Sea, where the supplementation of smectite percentages reached the peak when the ITCZ shifted significantly northward (Fig. 2b; Deplazes et al., 2013). During cold-climate events, when the ITCZ moved significantly southward, rainfall decreased, and the smectite percentages decreased correspondingly in the source area. Therefore, we suggest that these changes in the smectite percentages in the studied core are correlated with ITCZ migration and that rainfall is an important factor determining the smectite percentage from the source area of Myanmar on the millennial scale. If precipitation induced by wind and temperature of the South Asian monsoon has an intense impact on the source area, the source area monsoon indicators, for example, foraminifera, sporopollen, stalagmite (Fig. 6) and other indicators, would correspondingly change, but our record failed to catch these variations in monsoon indicators in the BoB. We suggest that every factor affecting precipitation induced by wind and temperature of the South Asian monsoon, as mentioned above, may have made it difficult to cause millennial-scale fluctuations similar to the ITCZ shift during the MIS3 period. The South Asian monsoon is indeed the result of combined factors that may contribute to the heterogeneity of monsoon rainfall in the BoB, which were also influenced by the north–south shift of the ITCZ. In core 17I106, the corresponding variations in the relatively high smectite percentages and the northward shift of the ITCZ indicate that the northward movement of the ITCZ is the most important factor influencing the incremental changes in river sediment load corresponding to the increased smectite percentages in the Myanmar region. Here we emphasize that the northward and southward ITCZ shifts bring about rainfall increases and decreases relative to other rainfall forces. The changes in clay minerals reflect changes in clay mineral supply in the source area, and it is these relative increases and decreases in rainfall that lead to changes, which is a response to environmental changes. The sporopollen evidence suggested a cold and wet period during MIS3 in Yunnan, China (Zhang et al., 2020), which may have been caused by the frequent northward movement of the ITCZ during this period.
Comparison of smectite percentages with ITCZ north–south shift
records.
Although the changes in smectite percentages in the study area are associated with ITCZ shifts before the LGM, the ITCZ shift in the Indo-Pacific warm pool (IPWP) was more “regional” than those in the Arabian Sea and the Cariaco Basin (Deplazes et al., 2013). During the late LGM, when the ITCZ did not move extensively in the Arabian Sea, the ITCZ gradually shifted northward in the IPWP from 21 to 18 ka (Fig. 7d, Ayliffe et al., 2013). However, the smectite percentage increased significantly in the study area, and we have excluded the possibility that the winter monsoon or meltwater influenced these changes. Further comparisons with IPWP records reveal that the ITCZ changes agree well with the smectite percentage variations during the late LGM, indicating that the northern migration of the ITCZ induced high smectite percentages in core 17I106 (Fig. 7c, d). These results suggest that the clay minerals of core 17I106 are inextricably linked to ITCZ shifts on the millennial scale. In summary, our smectite record shows that before the LGM, the ITCZ was in a relatively southerly position in the Myanmar area, while during the late LGM, the northward movement of the ITCZ in the BoB led to increased rainfall in the Myanmar source area and an increased supply of smectite. At the same time, the ITCZ was not significantly shifted in the Arabian Sea region either pre-LGM or post-LGM, which is what the Arabian Sea record shows (Deplazes et al., 2013).
The smectite percentage in the studied core is different from the ITCZ records in some periods, such as the late LGM, revealing that regional changes in the ITCZ were significantly obvious and that the ITCZ is not a simple N–S displacement. This consistency may indicate that the regional extension of the north–south thermodynamic gradient in the EIO exceeded that in the Arabian Sea and that the north–south shift of the ITCZ caused the climate systems of the Northern and Southern Hemisphere to be more closely connected in the EIO during the late LGM (Huang et al., 2019; Zhuravleva et al., 2021). A recent study considered less northward migration of the summer ITCZ position in the western BoB than in the eastern BoB during Heinrich Stadials HS1 and HS5 (Ota et al., 2022), which indicated that regional ITCZ variations in the BoB may be very common. These factors may be correlated with observed variations in regional air–sea interactions, such as the exposure of the Sunda Shelf (DiNezio and Tierney, 2013), the effect of the thermocline in the EIO (Mohtadi et al., 2017), and even a potential El Niño-like mode (Thirumalai et al., 2019) and Indian Ocean dipole (IOD) (Abram et al., 2020) changes, which may make the ITCZ shift more dramatic or keep the ITCZ position in the Northern Hemisphere longer. Thus, the regional variations in the ITCZ should be fully considered when studying climate change, especially in low-latitude regions that are sensitive to climatic and environmental changes, such as the EIO (Niedermeyer et al., 2014).
We reconstructed the variations in sediment sources on the Ninetyeast Ridge over the past 45 kyr. The main source areas comprise the Himalayas transported by the G-B River and Irrawaddy River; sediments were stably supplied from these regions throughout the studied core. When North Atlantic cold events (Heinrich and YD) occurred, chemical weathering weakened and physical weathering increased; correspondingly, the smectite percentage decreased and the illite percentage increased. From 35 to 21 ka, the falling sea level led to an increase in the exposed area of the Andaman and Nicobar Islands and further hindered the entrance of smectite from the Andaman Sea into the study area. At the same time, the influence of the South Asian monsoon on the sediment supply was not obvious. The time-phase mismatches observed among records excluded the influence of Burman shelf sediment erosion forced by the winter monsoon or of variations in G-B River sediments induced by ice meltwater on the abnormal increases observed in the smectite percentages during the late LGM. The smectite record of core 17I106 is consistent with the ITCZ changes recorded on the millennial scale, indicating that the ITCZ controls the rainfall in the Burman source area and, further, the clay mineral variations in the study area. The inferred ITCZ shift recorded in the studied core coincided with the global ITCZ change that occurred before the LGM, but during the late LGM, the core record was consistent with the change in the regional ITCZ recorded by the IPWP. This revealed that regional changes in the ITCZ were very significant, and the ITCZ is not a simple N–S displacement at the same time. Thus, the regional variations in the ITCZ should be fully considered when studying climate change, especially in low-latitude regions that are sensitive to climate and environmental changes.
The entire dataset is available on the Science Data Bank
(
JL and YH conceived and designed the experiment. XX wrote the paper with contributions from all authors. LZ and LY provided the ages of planktonic foraminifera, and SL, YY, LC, and LT helped to analyze the measured data and discuss the related relevant topics in this paper.
The contact author has declared that neither they nor their co-authors have any competing interests.
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
We thank Hui Zhang for the help with Sr–Nd isotope measurements. We also thank
the editor Pierre Francus, Michael E. Weber, and another anonymous reviewer for their constructive comments. Core sediment samples were
collected on board the R/V
This research has been supported by the National Natural Science Foundation of China (grant nos. 42176075, 41576044, and 42130412), the Key Special Project for Introduced Talents Team of the Southern Marine Science and Engineering Guangdong Laboratory (Guangzhou) (grant no. GML2019ZD0206), the Bureau of Frontier Sciences and Education, Chinese Academy of Sciences (grant no. XDB42000000), and the Open Fund of the Key Laboratory of Submarine Geosciences, Ministry of Natural Resources (grant no. KLSG2102).
This paper was edited by Pierre Francus and reviewed by Michael E. Weber and one anonymous referee.