Keyword

1 Introduction

Sediment deposition in the Bay of Bengal (BOB) is governed by the dominantly detrital input even though there are other two sources such as biogenic (e.g. calcareous ooze, carbonate ooze, petropods, Coccolithophores, microfossils) and diagenetic (deformation of sedimentary layers, solution structures, occurrence of pyrite crystals etc.). Its input reveals the nature of the physical and chemical erosion of the terrain and it is strongly dependent on the monsoon intensity, especially the southwest or summer monsoon (Bhushan et al. 2007; Sarin et al. 1979). Nath et al. (2005) used sedimentological and geochemical proxies in order to show the evidence of Himalayan erosional events in a sediment core from the equatorial Indian Ocean. Rare earth elements (REE) and other geochemical data (Kurian et al. 2008) suggest a strong influence of the Irrawaddy River as a major continental source of the Andaman Backarc Basin (ABB) sediments. Kessarkar et al. (2005) studied sediment characteristics and radiogenic isotope of the detritus in sediment cores from the distal Bengal Fan and reported changing sedimentary environment at about 12,000 years BP and Pleistocene hemiturbidites, older than 13,000 years BP, with major source contributions from the Himalayan drainage systems. The hemiturbidites were overlain by calcareous pelagic sediments in the past 12,000 years BP. The change in litho facies suggests that the sediment deposition by turbidity current activity ceased in the distal Bengal Fan at ~12,000 years BP, and this was perhaps due to rapid sea level rise during the melt water pulse of the Early Holocene period.

A detailed review on monsoonal shifts (Nigam and Hashimi 1995; Gupta and Thamban 2008; Staubwasser and Weiss 2006; Kuppusamy and Ghosh 2012; Ponton et al. 2012; Patnaik et al. 2012; Achyuthan et al. 2017) indicate that limited studies have been carried out in the BOB as compared to the Arabian Sea. Rashid et al. (2011) reconstructed the sea-surface temperature (SST) and salinity from paired δ18O and Mg/Ca measurements in planktonic foraminifera Globigerinoides ruber from the western BOB core VM29-19. This indicated that the monsoon was stronger in the Bølling/Allerød period and weaker during the Younger Dryas. They observed; post ~5000 years BP the Indian summer monsoon significantly weakened, indicating less dilution of the sea surface waters by the Ganga-Brahmaputra-Meghna outflow and/or less direct rainfall (i.e. more saline sea surface waters). This was due to the prevailing late Holocene dry climate such as observed in the subsequent abandonment of the settlements of the great Indus Valley Civilizations. These studies highlight the utility of sediment cores from the BOB to study palaeomonsoon and its variability over different time scales and resolution (Ponton et al. 2012; Giosan et al. 2018). It is however appropriate to mention here that marine sediment cores can provide monsoon precipitation related information if the oceanic salinity changes can be calculated after delineating independently determined temperature from the oxygen isotope values of the planktic foraminifera. Likewise, surface run-off proxies such as the terrestrial sedimentary fractions can also provide precipitation related information. These are proxies to monsoon winds and as well as actual rainfall on the land and it depends on the proxies used to explain the different processes. This aspect has often been missed in interpreting higher upwelling to higher rainfall. To add on; an interesting study of the ancient texts such as Arthashastra, Nirukta, Ramayana etc., along with modern scientific data, it was observed that the periods of water stress as inferred from the texts seem to correlate with the periods of low monsoon winds as inferred from the oceanic records (Pandey et al. 2003).

The present work aims to study the influence of various global climate phenomena, such as Roman Warm Period (RWP, 2200–1800 years BP) (Desprat et al. 2003; Vollweiler et al. 2006; Wang et al. 2013), the Medieval Warm Period (MWP, 1100–700 years BP) (Hughes and Diaz 1994; Mann 2002a; Sinha et al. 2007), and the Little Ice Age (LIA, 600–200 years BP) (Mann 2002b) on the SWM during the mid to the late Holocene using marine sediment core collected off the Landfall Island, Andaman Islands.

During the monsoon season, depressions and cyclones, mainly originate in the BOB and are the chief rainfall producing synoptic weather systems over India (Dash et al. 2004). The Andaman Islands are located in the prime monsoon zone, sensitive to depression and cyclone formation and hence marine sediment cores retrieved from here could be very good palaeoclimate archives. The BOB receives large quantities of fresh water from the adjacent coastal mountain rivers (1.6 × 1012 m3yr−1, Subramanian 1993) and over 1350 million tonnes of suspended sediment every year (Milliman and Meade 1983; Subramanian 1985). The lithogenic sediments in the BOB derived from erosion of the Himalayas and Indo-Burma ranges, and brought by the Ganga-Brahmaputra and Irrawaddy rivers into the basin holds the signature in understanding monsoon variations, weathering patterns, and sediment provenance (Colin et al. 1999). Hence it has attracted a fair attention with regard to its sedimentation geology notably the origin and history of the Bengal sediments (Stewart et al. 1965), however the geochemical investigation of this vast expanse is limited and confined to the deep (Ramesh and Ramasamy 1997). Moreover, till date, no, high-resolution evidence on sediment weathering, sediment source, and their palaeoclimatic implications from the Andaman Forearc Basin, off the Landfall Island, BOB has been presented. In the BOB, Landfall Island is an ideal and suitable site to retrieve marine sediment core to study and reconstruct the relationships between continental weathering and palaeomonsoonal shifts. The main aim and objectives of the present study are: (a) To reconstruct a high resolution palaeoclimate record from Middle to late Holocene period. (b) To delineate the rate of sedimentation varying with southwest monsoonal variation.

2 The Study Area

The Landfall Island is located north of the Andaman Island forearc basin (Fig. 1). The geology of the area consists of conglomerate, sandstone and shale of the Mithakari Group (Pal et al. 2003) are underlain by volcanic rocks (the ophiolite suite). The ophiolitic suite is represented by andesite and dacite where plagioclase, orthopyroxene, clinopyroxene, olivine, titano-magnetite, hornblende minerals are the most common minerals with sometime some appearance of quartz grains (Ray et al. 2011). The area near the Landfall Island receives dominantly SWM and less intense rains during NE monsoon period. Meteorological data from Port Blair shows that the present-day climate is humid with a mean annual precipitation of 3168 mm (average of 30 years: 1951–1980 CE). The rain mainly occurs between May and October due to SWM (Laskar et al. 2011). The long-term mean annual surface temperature and humidity are 26.6 °C and 82%, respectively. The BOB is an ideal site to decipher the palaeoclimate change in the Indian subcontinent (Chauhan and Suneethi 2001), since the BOB receives the largest freshwater influx and sediments from the Himalayas and the Indian peninsular rivers during the NE and the SWMs (Chauhan and Vogelsang 2006). This is because the clays and the sediment detritus are the weathering products of parent rocks and their formation and the discharge is dependent upon the geology, drainage and the prevalent climate (Weaver 1989) of a given region.

Fig. 1
figure 1

Map showing the sediment core location (SN SC 03 2008) near the Landfall Island with the regional hydrography (prepared after Shetye et al. 1991, 1993; Varkey et al. 1996; Chauhan and Vogelsang 2006 and the site of the marine sediment core (NGHP-16A) collected in the Godhavari delta studied by Ponton et al. (2012)

Full size image

The mineral characteristics of the sediment influx from the Himalayan Rivers are markedly different in their composition and abundance than those dumped by the peninsular rivers and hence the mineralogy of the sedimentary inputs varies with climatic conditions and monsoonal intensity across the BOB. The circulation of SW and NE monsoon over the BOB is presented in Fig. 1.

3 Materials and Methods

A marine sediment gravity core nearly 124 cm long was retrieved from the BOB using a gravity corer, from the North of Andaman (13° 47′ 15.05″ N, 92° 53′ 47.96″ E), near the Landfall Island from a depth of 250 m (Fig. 1). The sediment core was collected during the Sagar Nidhi cruise organized by the National Institute of Ocean Technology (NIOT), Chennai, and the Institute of Ocean Management (IOM), Anna University, Chennai in June 2008. The sediment core was opened, described for its colour using the Munsell colour chart and also the grain size distribution and sediment texture (Fig. 2). The sediment core was sub- sampled at an interval of 2 cm, then dried at 50 °C, scooped in a zip lock plastic cover and preserved for further analyses.

Fig. 2
figure 2

Lithounits of the marine sediment core with variations in sand, silt and clay content with calibrated age. Note that the rate of sedimentation is non-linear

Full size image

The sediment samples (n = 21) were analysed for textural composition and grain size determination after treating the sediment samples with mild acetic acid (0.5 M) to remove the carbonate coatings and further treated with H2O2 to remove the organic matter. The washed, cleaned and dried samples were then analysed for grain size composition following pipette method of Krumbein and Pettijohn (1938). The grain size variations were then calculated for sand, silt and clay content to its weight percentage (Fig. 3). The same twenty-one samples that were selected for textural studies representing the entire sediment core was further analysed for clay mineral composition. For this purpose, clay particles were separated using Stoke’s settling velocity principle. The separated clay fractions were made free of carbonate and organic matter by treating the samples with 5 ml acetic acid and 10 ml hydrogen peroxide, respectively. The slides were prepared by pipetting 1 ml of the concentrated clay suspensions onto glass slides. Extreme care was taken for the uniform distribution and to avoid size sorting (Rao and Rao 1995). These slides were then air dried and then kept in a desiccator for one hour under Ethylene glycol vapours at 100 °C. X-ray diffraction spectra were then carried out on these slides from 3° to 30° at 20/min on a Rigaku X-ray diffractometer using nickel-filtered Cu Ka radiation. Clay minerals were then identified and quantified following the semi-quantitative method of Biscaye (1965) (Fig. 4). Detailed clay mineral analyses were carried out in the sedimentology laboratory of the National Institute of Oceanography (NIO), Goa. Organic matter (OM) content was analysed following the protocol developed by Loring and Rantala (1992) and Calcium carbonate (CaCO3) content was determined by calculating the mass difference after treating the sediment samples 10% HCl to the known weight of the sediment sample following Carver (1971). Globigerinoides ruber (G. ruber) was handpicked from all the sediment samples of the sediment core. For this purpose, the sediment samples were soaked in water overnight,

Fig. 3
figure 3

Down core variation of sand, silt, clay, CaCO3, organic matter (OM) (in %), δ18O and δ13C (‰) with calibrated ages in yrs BP

Full size image
Fig. 4
figure 4

Down sediment core variations in clay minerals, K/C and C/I

Full size image

then washed and passed through an ASTM 150 µm sieve mesh. The retained fractions on this sieve were oven dried at 60 °C and set aside for microscopic examination. About 10–12 tests of planktonic foraminifera (G. ruber) 250–315 µm size was selected for stable oxygen (O) and carbon (C) isotopic measurements. Microscopic examination revealed that the tests were well preserved with no apparent signs of dissolution and taphonomy. All the selected foraminiferal tests were cleaned ultrasonically in methanol to do away with clay coatings. Oxygen (δ18O) and carbon (δ13C) isotopic measurements were carried out at the National Geophysical Res Institute, Hyderabad, India, using a Delta plus advantage Isotope Ratio Mass Spectrometer (IRMS) coupled with a Kiel-IV automatic carbonate device following Ahmad et al (2008). The results of the Isotopic compositions are reported in δ notations as per mil deviation from VPDB standard. Analytical precision was better than 0.10‰ for δ18O and 0.05‰ for δ13C (Fig. 3). Calibration to the VPDB standard was achieved by repeated measurements of international reference standards NBS-19 and NBS-18. Twelve sediment samples were selected at regular intervals covering the entire sediment core for phytolith study. The extraction of phytoliths was undertaken in the laboratory by using standard methods with removal of carbonates and nitrates followed by heavy density separation (Piperno 1988). Slight modifications were required at times depending on the nature of the sediment type. Simultaneous extraction of Phytoliths, Diatoms and Sponge Spicules was undertaken. Up to 200 Phytoliths, inclusive of other microfossils from each sample were observed and counted so as to generate a quantified data (Figs. 5, 6). Observations were also made on the physical characters of phytoliths so as to get an idea about its general preservation condition. The classification used for phytolith analysis is a combination of—phytolith shape, anatomical origin and classification based on grass families.

Fig. 5
figure 5

Photomicrographs of various types of phytoliths, spongy spicules and diatoms of the present study (light Microscopic Photographs, scale = 10 µm)

Full size image
Fig. 6
figure 6

Down sediment core variation of phytolith and nanoplankton assemblage

Full size image

The phytoliths were observed under ‘Olympus’ Res microscope and photomicrographs were taken under 45× magnifications. The basic phytolith classification adapted here is derived from grass families as suggested by Twiss (2001) with further modifications by including Phytoliths of anatomical origin and other non-diagnostic silicified cells, such as woody elements and perforated cells ensued from shrubs and trees (Eksambekar 2002). For the nanoplankton assemblages, a pinch of the powdered sample material was taken and kept in a covered crucible in which a few drops of distilled water were poured to make an even suspension by stirring it. With the help of a clean dropper, a smear of sediment suspension was spread on pre-heated slide. Two slides, one containing the fines, and the other containing comparatively coarse fraction of the suspension were prepared and allowed to dry on a hot plate. Drops of mounting medium (Canada Balsam) were poured with the help of a glass rod and a cover slip of 22 × 40 mm dimension was used to cover the slide containing dried suspension film. The cooked slide with the cover slip was picked with the help of a pincer and allowed to cool on a flat surface. The air bubbles were allowed to leave the surface. The cover slip was even pressed after taking out the slide from the hot plate so that extra Canada balsam may leave the slide while the slide is cooling and made permanent. These slides were examined for the presence of nanofossils under Leitz polarizing microscope (LM) with ×10 or ×12.5 oculars and an ×100 oil immersion objective. Photomicrographs were taken by using gypsum plate which helped in a precise identification of crystal elements of critical species. Nannofossils were identified and a checklist was prepared. The recovered nannofossils were compared with global marker charts and microphotographs from available publications for verification and confirmation (Figs. 7 and 8a–f). In the present study, organic carbon rich five samples were radiocarbon dated following the acid-base-acid pre-treatment by liquid scintillation counting at the Department of Geosciences, University of Arizona, Tucson, USA. Calibrated and reservoir corrections (400 years) and calibrated using Calib 6.0 (Stuiver et al. 2010) following Stuiver and Reimer (1993) was applied and presented in Table 1 and Figs. 2 and 3.

Fig. 7
figure 7

Downcore variations of Elemental ratios with CaCO3%, δ13C, δ18O reveals variations wetter and drier conditions owing to the intensity of SWM

Full size image
Fig. 8
figure 8figure 8figure 8figure 8figure 8figure 8

a Photomicrographs of nanoplanktons under plane and crossed nicols. b Photomicrographs of nanoplanktons under plane and crossed nicols. c Photomicrographs of nanoplanktons under plane and crossed nicols. d Photomicrographs of nanoplanktons under plane and crossed nicols. e Photomicrographs of nanoplanktons under plane and crossed nicols. f Photomicrographs of nanoplanktons under plane and crossed nicols

Full size image
Table 1 Conventional and Calibrated 14C ages of at selected depth intervals
Full size table

4 Results

The sediment core samples were radiocarbon dated to have a continuous record of the Holocene past and these studies were conducted to reconstruct the mid—late Holocene palaeoenvironment variations. and the ages range in age from middle Holocene to late Holocene (5680 ± 170; −22.8‰ δ13C to 2085 ± 120 years BP; −23.3‰ δ−13C) (Tables 1, 2, 3 and 4; Fig. 2). An age of 1658 years BP (−23.3‰ δ13C) was obtained from the core at the depth 15 cm. Based on the reservoir corrected ages the average sedimentation rates were calculated and the sedimentation rate varies 0.35 mm y−1 from 111 to 87 cm (6078 years BP to 5402 years BP), 0.17 mm y−1 from 87 to 69 cm (5402 years BP to 4376 years BP), 0.28 mm y−1 from 69 to 35 cm (4376 years BP to 3179 years BP), 0.13 mm y−1 from 35 to 15 (3179 years BP to 1658 years BP) and 0.09 mm y−1 from 15 cm to the sediment core surface (1658 years BP to present) indicating a non-linear phases of deposition (Figs. 2 and 3). Interpolations of ages as a function of depth were calculated by considering the sedimentation rates between the obtained radiocarbon ages (Table 1). Overall the sediment core exhibited sediment layers in shades of olive black to olive grey. Based on sediment texture and colour, the sediment core was divided into three distinct litho-units. Unit I (124–100 cm) is olive black (10Y 3/2) coloured, sandy to fine silt, Unit II (100–45 cm) is olive grey (10Y 4/2) coloured dominantly fine to medium fine grained fine sandy silt. Fine silty layers were observed within this unit. Unit III (45–0 cm) is olive grey coloured (10Y 5/2), with the clayey fine silt of fine laminae and minor ripples were observed in the uppermost layers of this unit (Figs. 2 and 3). Moreover, there is no significant variation in δ13C values as it varies marginally from −23.3 to −24.3‰, showing a difference of nearly −1.0‰ only, since 4376–1658 years BP (Figs. 2, 3).

Table 2 Weight percentages of sand, silt, clay and organic matter (OM) content in the sediment core
Full size table
Table 3 Clay mineral composition and their ratio of the sediment core
Full size table
Table 4 δ18O and δ13C values of Globigerina ruber
Full size table

In this sediment core grain size analysis reveals a predominance of silt and clay over the sand content (Fig. 3; Table 2), indicating relatively deeper and calm environment of deposition. The sediment core is primarily composed of silt (41.88–68.05%) followed by clay (15.90–40.32%) and sand (8.28–22.04%) (Table 2). Based on Shepard (1954) classification using the ternary diagram that is divided into ten classes of sediment texture, the sediment samples of the present study are clayey silt in nature (Fig. 3). On the contrary, to the fine silt content, the sand content is higher at the depths 120–106 cm (18–17%; 6360–5937 years BP), 90–88 cm (22%, 5486 years BP) and 42–40 cm (20%, 3425 years BP) and least between 60 and 52 cm (8%, 4059 to 3848 years BP). The clay percentage is higher at the depth 36 cm (39%, 3285 years BP), and 40% at depth 6–4 cm (663 years BP) and exhibits a low at 42–40 cm depth (15.9%, 3425 years BP) (Table 2). The variation in clay composition, content, K/C and C/I ratios for all the samples studied is presented in Table 3. Detailed clay mineral analyses indicate the dominant occurrence of smectite (range ~25–49%) and illite (range ~21–43%) followed by kaolinite (range ~12–19%) and chlorite (range ~9–17%). The down sediment core clay composition variation reveals smectite (S) (49%) and illite (I) (42%) peaks at the depth 108 cm and 18 cm respectively, while kaolinite (K) (19%) and chlorite (C) (16%) occur at the depths 118 cm and 48 cm respectively. Smectite shows an inverse trend with illite and chlorite. K/C ratio (0.93) shows a minimum value at the depth 66-68 cm while the C/I ratio (0.56) is maximum at the same depth (Table 3, Fig. 3). At the depth 112–114 cm K/C ratio is the maximum (2.03) followed by a minimum of C/I ratio (0.28) at the depth 8–10 cm (Fig. 4). The δ13C and δ18O isotopic records of planktonic foraminifera (G. ruber) shows that the δ13C in G. ruber vary significantly from 0.70 to 1.33‰ with an average of 1.06‰ (Table 4). Planktonic δ18O values vary between −4.18 and −2.33‰. An average δ18O value for the mid to late Holocene is −3.13‰ (Table 4; Fig. 3).

Different phytolith morphotypes were noted that included subtypes within each group. The phytolith morphotypes were grouped as follows: (1) Panicoid (e.g. dumbbells) (2) Chloridoid (e.g. saddles) (3) Festucoid (e.g. spherical, square) (4) Elongate (e.g. Rods and tracheids) (5) Trichome (e.g. epidermal hairs) (6) Bulliform (e.g. bulky phytoliths like fan shape) (7) Silicified cells (e.g. woody elements) (Fig. 5). Phytoliths observed in each sample were compared with the ‘in house’ phytolith database of phytolitharium collection housed at the “Phytolith Research Institute”, Pune and other published references.

The phytolith morphotypes from the marine sediment core based on their anatomical origin and structural character suggest the presence of following vegetation in the observed samples. High percentage of Panicoid (14%) with low Chloridoid (2.5%) was noted at 96–94 cm depth around 5599 years BP, a low Panicoid (2%) and Festucoid (9%) was found at 42–40 cm depth approximately around 3355 years BP with high content of Trichome (27.5%) (Fig. 6; Table 5). While high Festucoid was found at 70–68 cm depth around 4433 years BP. It was also observed that at 28–30 cm depth in the sediment core around 2646 years BP the Chloridoid (2.5%), Elongate (5%) and Festucoid (9%) content is low, with a higher percentage of Bulliform (37.5%) and Woody elements (19.5%). The low percentages of Bulliform (2.5%) and Trichome (3%) with high Spongy spicules (45.5%) occur at the depth 10–8 cm around 884 years BP (Table 5; Fig. 6).

Table 5 Types of phytoliths in percentage with climate index (Ic), phytolith index (Iph) and cold/warm ratio
Full size table

Diatoms were observed to be high at 124–122 cm around 6416 years BP with low percentage of woody elements (5%) and Spongy spicules (5%). The low occurrence of diatoms was observed at the depth 54–52 cm ~3778 years BP. In the present study most, sponges and spicules were composed of calcium carbonate or silica (Figs. 5, 6).

In this study siliceous sponge spicules have been separated into two major categories based on their size as megascleres and microscleres. The siliceous spicules studied here are mainly of two types: megascleres (larger skeletal elements, >100 μ in length, existing either free or as part of a framework) and microscleres (smaller, 10–110 μ, always occurring as free elements). These siliceous spicules are formed of concentric layers of opaline silica.

In the present study the diatoms have been differentiated using full forms that are centric, i.e. circular, and pennate (Fig. 5), i.e. having bilateral form. The term ‘pennate’ denotes feather-like structures. Diatoms have been divided into solitary and colonial forms. Diatoms have been further sub-divided according to whether they lack a raphe (a median line or slot in the cell wall), a pseudoraphe, or completely lack a raphe. Diatoms commonly found in the marine plankton have been divided into the centric diatoms including three sub-orders based primarily on the shape of the cells, the polarity and the arrangement of the processes.

The nanoplankton data of all the samples (n = 62) from the marine sediment core of the present study show a significant variation in the productivity, presented as a down core variation (Table 6; Figs. 6, 7). It contains a consistent record of complete or disaggregated spicules (star-shaped spicules), belonging to the Didemnidae family, at intermediate depths. These prosobranch ascidians are essentially composed of shallow benthic coastal colonies, though same species have been recorded from abyssal depths. The nanoplankton identified and their abundance from the present study is shown in Fig. 6 and Table 6.

Table 6 List of the nanoplankton species identified in the marine sediment
Full size table

The dominant species in the calcareous nannofossil assemblage were Gephyrocapsa oceanica, small Gephyrocapsa, Emiliania huxleyi, Ascidian spicules and the species of Calcidiscus leptoporus, Helicospahaera carteri, Ceratolithus telesmus, C. cristatus, C. acutus are also abundant. Subordinate number of H. hyalina, H. inversa, H. pavimentum are also seen. Rare Syracosphaera pulchera, Neosphaera ccolithomorpha are in the assemblage and very rare Scyphospheres, Rhabdoliths and Discoasters Braarudosphaera bigelowii were also present in some samples (Fig. 8a–f). No specimen of Florisphaera profunda, which is a lower photic zone species, was observed. When nutricline is shallow, coccolithophore production in the upper euphotic zone is enhanced and the abundance of F. profunda is minimal. The species richness of C. leptoporus, H. carteri, Gephyrocapsa spp., which are typical to the upper photic zone with high nutrient availability are also present in the assemblage. The ecological preference of H. carteri indicates its preference for nutrient enriched waters.

5 Discussion

5.1 Provenance of the Mid-Holocene Marine Sediments

SWM controls precipitation in BOB (Ding 1994) and the strengthened winter monsoon causes deep chemical weathering of the bedrocks (Wang 1999) resulting in the intense weathering of the nearby continental margins of the Andaman Islands. Thus, the terrestrial detritus in the sediments in northern Andaman Islands exhibit stronger chemical weathering signals (Wei et al. 2003). The chemical, textural and isotopic signatures (C and O) observed at the Andaman Forearc Basin (AFB) sediment core near the Landfall Island reflects a mixture of sediments distinctly from a nearby source, i.e. from the Andaman terrain of ophiolite basalt suites and sediments from Irrawaddy, Salween and Sittang rivers that transport weathering products from the land masses to the AFB. This inference also supports the argument suggested by Moore et al. (1982) that the sediments are transported from the Irrawaddy delta to the south of Sumatra along the Andaman Fore arc basin. However, the sediment fluxes vary during the dry, warm/wet phases. High content of silt occurs at the depth 60–55 cm (4000–3900 years BP), while laminar clays and fine silt are deposited at the depth 35 cm (3200 years BP). Sediment texture and clay mineral composition reveal low clay content (16%), with higher coarser components at the 40 cm depth and this indicates a high energy event, unaffected chemical weathering of the Landfall Island bed rocks (conglomerates, sandstone, shale and the ophiolite suite of rocks). This clearly shows that the rate of sedimentations has been non-linear. It is also noted that the decrease in the sedimentation rate (Fig. 2) post 4300 years BP that points to probably a thicker vegetation canopy in the hinterland and with low carbonate content indicating strong chemical weathering of the Island bed rocks. The low values of calcium carbonate and organic matter also points to low productivity due to reduced SWM. Probably a very weak monsoon could have also resulted in reduced sediment input (Fig. 4, 5 and 9).

Fig. 9
figure 9

SWM shifts inferred based on sediment rate, sand and CaCO3 content, ⅀Nd, δ18O and chemical weathering Index. It is noted that each parameter has responded in different way to the SWM variability

Full size image

Studies carried out on the clay minerals (Rao 1983) near the study area reveal that the clays and the very fine sediments are rich in kaolinite, chlorite and illite whereas montmorillonite (smectite group) is less. Rao (1983) found two distinct clay mineral assemblages: (i) Fe-rich chlorite in the eastern side, (ii) Fe-poor chlorite in the western side of the Andaman Islands, however this study was not quantified.

In the present study smectite and illite are the two principal clay minerals present in the sediment core other than the kaolinite and chlorite (Fig. 4; Table 3). The high smectite content in the sediment core suggests that the sediments are derived mainly from nearby basic rock source (Griffin et al. 1998; Siby et al. 2008). Illite and chlorite are often derived mainly by physical erosion or degradation of metamorphic and granitic rocks or from the erosion of sedimentary rocks (Liu et al. 2005). The catchment area of the rivers Irrawaddy and Salween (which discharge from Myanmar and have catchment in the NE Himalayas) comprise metamorphic and intrusive rocks (Chauhan et al. 1993). Thus, it can be inferred that a considerable amount of fine grained illite and chlorite sediment deposition near the Landfall Island is controlled by the Irrawaddy and Salween Rivers.

Clay minerals such as chlorite and illite are often produced under the arid/cold climate while the smectite and kaolinite are formed under humid/warm conditions causing strong weathering coupled with intense SWM (Griffin 1968; Liu et al. 2004; Weaver 1989; Chauhan and Suneethi 2001; Chauhan et al. 2004). Occurrence of varying amounts of kaolinite, smectite, chlorite, illite, kaolinite/chlorite (K/C) and chlorite/illite (C/I) ratio indicates periods of intense wetter conditions (intense SWM) with incursions of shorter weaker monsoonal conditions from mid to late Holocene. High kaolinite, K/C ratio with reduced chlorite and low C/I ratio around 6500–6000 years BP (124–111 cm depth) points to intense SWM causing strong weathering of the bedrocks in the hinterland (Fig. 4). The reduced CaCO3 content (72.32–70.63%) and occurrence of kaolinite, around mid-Holocene 3300 years BP and 6000 years BP (depth 38–40 cm and 108–110 cm) respectively indicates strong weathering of bed rocks, intensified SWM that led to an upwelling resulting in higher surface biogenic productivity. This observation is also supported by the large coarser component influx at these depths. In the studied sediment core an overall fluctuating occurrence in the abundance of smectite, kaolinite, illite and chlorite supported by K/C and C/I ratio is observed from 6000 to 2000 years BP (111–20 cm depth) indicating an overall weakening of SWM. However, within this 4000-year period a short incursion of extreme weakening of SWM occurred from 4400 years BP to 4200 years BP i.e. from 70 to 65 cm depth as revealed by high smectite content, reduced kaolinite and illite clay minerals with high C/I ratio (Fig. 4). The characteristic clay mineral association of smectite and illite also implies dry to semi-drier conditions around 4400–4200 years BP. Subsequently an amelioration of SWM is observed with high percentage of kaolinite and chlorite with a higher C/I ratio from 3900 years BP to 3600 years BP (depth 55–48 cm). An occurrence of arid phase is noticed around ~2000–1700 years BP (20–15 cm) due to the presence of higher illite content and reduced C/I ratio (Fig. 4). Similar observations were made by Chauhan and Vogelsang (2006) using the clay mineral composition of the sediments from the western BOB. They reported four arid phases during the Holocene period (10,000–9600, 7300, 5800–4300 and 2200–1800 years BP) and our inferences gains support from the observations made by Chauhan and Vogelsang (2006) and Staubwasser and Weiss (2006).

5.2 Sediment Texture, CaCO3, OM, Stable Isotopes, Phytolith, Diatom and Nanoplankton Data

In the present study, high sand content in the sediment core occurring at depths 114–112 cm (~6191 years), 90-88 cm (~5486 years) and 42–40 cm (~3425 years BP) (Table 2) with less CaCO3 content indicates a terrigenous influx of sediments and wet conditions (intense monsoon) (Figs. 3, 7). The increasing OM content also point towards wetter conditions. Moreover, a virtual increase in sand component (from 8.28 to 20.02%) and the decrease in clay content (from 23.67 to 16.00%) between ~4059 and 3425 years BP suggests high terrigenous input from the surrounding islands and an unaffected chemical weathering of the Landfall Island bed rocks during the mid and late Holocene period. The sand layers with less silt component during this period have also pointed to the river influx such as Irrawaddy, Sittang and the Salween rivers. On the contrary, from the depths such as 96–94 cm (~5655 years), 60–58 cm (~4059 years) and 24–16 cm (~2038 years) the sediments are dominantly fine silt with a decreasing trend of sand content and elevated CaCO3 percentage indicating warm and dry conditions (Figs. 3, 7).

The δ13C of G. ruber varies from 0.70‰ (16–18 cm; ~1700 years BP) to 1.33‰ (112–114 cm; ~6100) years BP (Fig. 3; Table 4). These variations in the δ13C are not coherent with the Holocene variations, however, we attribute the shifts in δ13C to the variations in carbon isotope composition of surface waters due to the strengthening or weakening of SWM and wind induced variability. During enhanced high wind velocity period, subsurface water with low δ13C might have been brought from to the surface, resulting in depletion of δ13C values and vice versa (Ahmad et al. 2012). Our δ13C values conform well with the values of planktonic foraminifera 0.74‰ to 1.25‰ from the sediment core SK157-14 (Ahmad et al. 2012) data.

Duplessy (1982) analysed the variation in the stable isotope G. ruber (δ18O) from the marine sediment cores collected from the BOB, Andaman Sea and the Arabian Sea. The data ranges between −3.70‰ and −1.58‰ for the Holocene times and for the glacial times it varied from +0.55 to −1.17‰. Chauhan (2003) also observed the heavy δ18O values of G. ruber during 20,000–15, 000 years BP (−0.9‰) and lighter δ18O value of G. ruber in the BOB around −3.38‰ in the Holocene period, which is much lighter compared to the Last Glacial Maxima (LGM). A comparison of stable isotope data reveals that the beginning of the Holocene is marked with much lighter δ18O values which persisted for the entire Holocene, with two pauses at ~5000–4300 years BP and ~2000 years BP when compared to the last glacial maximum (LGM) (magnitude 2.44‰ for G. ruber).

The sediment influx data also corroborate well with the stable isotope data on the G. ruber. δ18O isotope data on the G. ruber reveals fluctuations in the monsoonal regime from 6000 to 4300 years BP (−4.03 to −3.39‰). More positive δ18O value occurs at ~5500 years BP and around 4200 years BP (Fig. 3) indicating weakened SWM during this period. However around 4400–4200 years BP; weakened monsoon with cooler conditions is discerned by the values of δ18O moving from −3.39 to −2.33‰ (values becoming more positive). The lighter values of δ18O in the species G. ruber could be due to local factors such as changes in sea surface salinity (SSS) and by higher fluvial influx from the Ganga-Brahmaputra system, and/or higher sea surface temperature (SST) in the BOB. Further, there is a slight amelioration in the monsoons as revealed by the δ18O values ranging from −2.76 to −3.43‰ (~3425 years BP). Higher δ18O values of G. ruber and an increase in its abundance (Fig. 3) can be strongly linked to an increase in SSS caused by a decrease in riverine runoffs strongly associated with reduced SWM.

An integration of the down core sediment texture, clay mineralogy and stable isotope data on G. ruber points that the decrease in sediment influx into the Bay of Bengal near the Landfall island reflects weakening SWM and higher salinity conditions of the sea water during (~6000–4400 years BP) (Figs. 7, 9). An increasing occurrence of kaolinite and illite indicate an intense chemical alteration of island bed rocks from ~4400 to 1700 years BP. The arid/dry episodes of 5000–4300 and ~2000 years BP largely coincide with similar events reported from the northern Indian Ocean (Von Rad et al. 1999; Chauhan et al. 2004; Chauhan and Suneethi 2001).

A high-resolution study of mineralogy and major element geochemistry combined with Sr, Nd and Oxygen isotopes (δ18O) was conducted by Colin et al. (1999, 2006) in two marine sediment cores collected off the Irrawaddy river mouth (MD77-180) in the Bay of Bengal and in the Andaman Sea (MD77-169). Pedogenic clays (smectite and kaolinite) to primary mineral (feldspar, quartz, illite and chlorite) ratios showed strong precessional cycles, suggesting a control due to the past changes in the summer monsoon intensity (Colin et al. 1999, 2006). They also demonstrated wet periods of summer monsoon reinforcement corresponding to an increase in the weathering of the Irrawaddy plain soils with a decrease in 87Sr/86Sr ratio.

Planktonic diatoms generally decrease with climate warming because of reduced nutrient redistribution and increasing sinking velocities. Climate warming exhibits a selection pressure on the diatom cell size, this favour small-sized diatoms that are able to out compete larger sized cells and expand under intensified stratification. This observation was made in this sediment core recovered from the Bay of Bengal High frequency of diatoms is found in lower zones as compared to the upper levels. This points also to a hydrodynamic condition that created a calmer environment, with strong sedimentation of clay particles with abundant plant remains such as branches, roots and leaves as indicated by phytolith morph types and silicified woody elements, forming an organic clay sedimentary deposit, colonized by small shrubs and vegetation. These environmental conditions probably contributed to the retention of organic muddy sediments that favoured the dominance of diatom species. The presence of freshwater sponge spicules in sediments obviously indicates an aquatic environment, but their gemmoscleres/ megasclare besides allowing for species identification, point to seasonal changes in temperature and sea water column levels.

Siliceous sponge spicules provide a basis for palaeoecological interpretation. Sponges require relatively clear, non-turbid water for living conditions because muddy waters clog their pores (Eksambekar 2002). Thus, it appears that only some spicules were from in situ sponges while the rest were probably re-deposited. The low spicule abundance in the sediment core resulted from local current conditions and the dilution effect through input of terrestrial sediment (Figs. 6, 7). Other possible explanations for the varying spicule abundance in the sediment core are due to the difference in the local fauna, such as coral reefs which usually have high diversity and abundance of sponges.

Picoplanktons (<2 µm) and nanoplankton (2–20 µm) can also contribute downward carbon flux (Gorsky et al. 1999), other than diatoms and the dinoflagellates (two chief component of phytoplankton) (Severdrup et al. 1942), Hence the calcareous nanoplankton is also used as one of the most important climate indicators among other proxy such as foraminifera, alkenones, dinoflagellate cysts, especially in the Mediterranean Sea (Triantaphyllou et al. 2009).

The nanoplankton assemblage in the studied marine sediment core (0–124 cm) reveals the abundance of Emiliania huxleyi and Gephyrocapsa Oceanica rare of Pleistocene to recent age (Bukry 1971). It is also noted that the presence of some important species of nanoplanktons of the Pliocene (Discoaster brouweri) and Miocene (Coccolithus pelagicus) suggests that the sediments are derived from the reworked sediment of Mio-Pliocene age (Fig. 8a–f; Table 6).

The nanoplankton data from the BOB sediment core (Table 6) revealed fluctuations in its abundance for the past 5680 years BP (Figs. 5, 6 and 7). This variation is associated with the fluctuations in the nutrient enriched/depleted water and changes in the thermocline/nutricline depth. A significant increase in the surface water productivity during the deglaciation period is evident from the phytolith proxy data.

The overall correlation between ascidian spicule abundance and the relative abundances of upper photic zone calcareous nannoplankton reflect a close connection between the enrichment in nutrients in the upper photic zone and an increment in bottom productivity. This supports the idea of benthic communities in this area strongly benefiting from surface water productivity (bentho-pelagic coupling). According to Xiao et al. (2006) CaCO3 is a good indicator of temperature variation. Higher content CaCO3 (75.56%) percentage indicates warmer and dry conditions in the hinterland (around 66-64 cm, 4270 years BP) (Fig. 3).

High content of sand with reduced silt, OM, and nanoplankton productivity ~6300–6000 years BP, 5600–5200 years BP suggests intense SWM. It is also observed that ~3600–3300 years BP the sand and silt content was higher with reduced nanoplankton and clay reveals high river influx due to the SWM. In contrast, two major phases of weakening SWM were found around 4200–3800 years BP and 2000–1800 years BP inferred from the reduced sand with increasing percentage of silt, OM and nanoplankton productivity (Fig. 7). It was in agreement with Jochem (1989), that the primary production of nanoplankton was lower in winter monsoon whereas it is higher during the summer season as a dominating size fraction which can, contribute 70–100% of primary production compared to the other plankton (Pico and microplankton).

The basis of palaeomonsoon reconstruction in the tropics is the amount effect (Yadava and Ramesh 2005). They observed significant amount effect in the plains of central India. However, exceptions were also observed in the tropical to subtropical SWM rainfalls in the hilly terrains of NE India. (Breitenbach et al. 2010; Laskar et al. 2013a) and some parts of Southern India (Yadava et al. 2007). Variations in the amount of past rainfall have been owing to factors, such as changes in the δ18O of the ocean (‘‘ice volume’’ effect), shifts in the source of moisture Bar-Matthews et al. (1999), Fleitmann et al. (2003, 2007), Fleitmann and Matter (2009), and changes in the seasonality of precipitation (e.g., change in the proportion of winter and summer precipitation).

5.3 Reconstruction of Palaeomonsoon and a Comparison with Other Records from the Region

It has been noted that in the last 4000 years BP, there is no significant change in the ocean δ18O, that can cause changes in rainwater and hence sedimentation. In the present study, a strong reduction in the monsoon during 4400–4000 and 2200–1800 years BP is observed (Figs. 7, 8). The period 2200–1800 years BP coincides with the RWP, previously thought to be restricted to higher latitude regions (Vollweiler et al. 2006; Desperat et al. 2003; Martinez-Cortizas et al. 1999; Laskar et al. 2013a, b). The BOB sediment core data indicate that the effect of RWP is evident in the tropics also. Weakening of the monsoon has also been observed around 1500 and 400–800-years BP, the latter period is the transition from the MWP to LIA. A significant reduction in the SWM during this period was also reported by several others (Fleitmann et al. 2004; Sinha et al. 2007).

Middle Holocene is an important period in Indian Prehistory as during this period river valley civilization, and establishment of rain fed agriculture got stabilized. Middle Holocene terrestrial and marine palaeoclimate proxy records overlap to a large extent indicating erosion from the hinterland and being deposited in the oceans (Patnaik et al. 2012). Till date marine record from Arabian Sea has the most consistent and continuous record of the monsoon (Sirocko et al. 1993; Kudrass et al. 2001; Thamban et al. 2002) whereas limited studies have been carried out in the BOB (Duplessy 1982; Chauhan et al. 1993, 2004; Rashid et al. 2011).

Based on the marine and terrestrial records, Patnaik et al. (2012) advocated that the Holocene period is characterized by a strong monsoonal phase known as “Holocene Climatic Optimum (HCO)” i.e. the middle Holocene period, that subsequently was followed by weak monsoonal phases around ~2500–1500, 1000, 650–450 years BP and the Little Ice Age (LIA) (1450–1850 AD). Chauhan et al. 2004 and Chauhan and Vogelsang (2006) found an intense weakening of SWM ~4300 and 2200–1800 years BP. Sarkar et al. (1990, 2000) and Yadava and Ramesh (2005) found an increase in the intensity of SWM around ~3200 years BP. Weakened monsoon during the Roman warm period is observed and this may be owing to cause and lag effect of the monsoons. However, it is generally observed that both in climate modelling and proxy data settings, monsoon increases during warm periods.

SWM significantly affected and fluctuated during the RWP. The reduction in the SWM during the RWP is evidently observed in the abundances of Globigerina bulloides (Gupta et al. 2003) and δ18O of G. ruber and G. Sacculifer (Tiwari et al. 2006) in sediment cores from Arabian Sea. A significant decrease in monsoon strength is also observed during the period between 800 and 400 years BP, i.e. the transition period from the MWP to the LIA (Figs. 9, 10) in all the studies. A relatively dry episode at the beginning and a wet phase at the end of the LIA are also reported from the Himalayan region (Kotlia et al. 2012). The monsoon was significantly stronger as compared to the present during 1200–800 cal years BP, the MWP. Evidence for a stronger monsoon during this interval is also found in the records of varve thickness (Von Rad et al. 1999) and stalagmite δ18O from the Indian subcontinent (Sinha et al. 2007). The SWM was similar to the present without any major change in the source, extent, and seasonality before 6000 years BP except two short-term arid phases around 4400–4000 years BP and 2000–1700 years BP (Sinha et al. 2007).

Fig. 10
figure 10

A comparison of palaeomonoonal reconstruction based on several proxies prepared after Sarkar et al. (2015)

Full size image

Xu et al. (2002) studied plant cellulose δ8O variations in sediment cores retrieved from peat deposits at the northeastern edge of the Qinghai- Tibetan Plateau in China. Following the decline of the RWP, their data revealed the existence of three particularly cold intervals centered at approximately 500, 700 and 900 AD during the DA (cold period).

Yang et al. (2002) used nine separate proxy climate records, derived from peat, lake sediment, ice core, tree ring and other proxy sources, to compile a single weighted temperature history for China spanning the past two millennia. Their composite record revealed five distinct climate epochs: a warm stage from 0 to 240 AD (the tail-end of the RWP), a cold interval between 240 and 800 AD (the DA), a return to warm conditions from 800 to 1400 AD (which included the MWP between 800 and 1100 AD), a cool interval between 1400 and 1820 (LIA), and the current warm regime (MWP) that followed the increase in temperature that began in the early 1800s. Another important finding of this study was that warmer temperatures of the past two millennia were observed during the second and third centuries AD near the end of the RWP.

Ponton et al. (2012) reconstructed the Holocene palaeoclimate in the core monsoon zone (CMZ) of the Indian peninsula using a sediment core raised from the offshore; from the mouth of the Godavari River (Fig. 1). Carbon isotopes of sedimentary leaf waxes indicated a gradual increase in aridity-adapted vegetation from ~4000 until 1700 years ago and was subsequently trailed by the arid-adapted plants. A comparison with the oxygen isotopic composition of planktonic foraminifer Globigerinoides ruber revealed high salinity peaks in the Bay of Bengal over the last 3000 years, and especially after 1700 years ago, signifying that in the CMZ aridification intensified in the late Holocene through a series of sub-millennial dry episodes (Ponton et al. 2012).

Probable variances in seasonal monsoonal moisture sourcing during the 5000–4200 yrs BP transition are evident in other terrestrial (speleothem) records from the region. A gradual increase in δ18O values over the Holocene from the Qunf record in Oman, which receives moisture exclusively from the Arabian sea reflect the decrease in monsoon strength due to changes in solar insolation (Fig. 10). A comparison of the Mawmluh cave record in NE India, which is closer to the BOB also receives the Arabian Sea sourced ISM moisture during the summer months with a mean δ18O of −7.2‰ and also that the BO B moisture would affect NE India (Breitenbach et al. 2010). The transition from early Holocene wet to late Holocene dry conditions in the BOB was not gradual, but characterised by relatively rapid changes in hydroclimate conditions, with at times bringing about a complete failure of the summer monsoon especially between 5000 and 4200 yrs BP.

A mid-Holocene transition has been recognized throughout the Asian monsoon region, nonetheless its nature (i.e., abrupt vs. gradual) has been highly debated. Morrill et al. (2003) discussed an abrupt (century scale) change in ISM while Fleitmann (2007) argued for gradual weakening of the ISM during the Holocene. Recent leaf wax based stable isotopic records from the Indian Ocean basin (Bay of Bengal) revealed a similar overall pattern but with gradual vegetation and hydrological changes during mid-Holocene (Ponton et al. 2012; Contreras-Rosales et al. 2014), reflecting the decrease in monsoon strength due to changes in solar insolation. This strongly affected the river valley civilisations causing cultural changes and decline that occurred across the Indian subcontinent as the climate became more arid after ~4000 years. Sedentary agriculture took hold in the drying central and south India, while the urban Harappan civilization collapsed in the already arid Indus basin. The establishment of a more variable hydroclimate over the last ca. 1700 years may have led to the rapid proliferation of water-conservation technology in south India also (Fig. 10).

To check whether our marine sediment core based chronology reflect local or broader regional rainfall variability, we compared our data with other marine and continental climate proxy records from some of the areas receiving SWM (Chauhan et al. 2004; Chauhan and Vogelsang 2006; Chauhan and Suneethi 2001; Mathien and Bassinot 2008; Rashid et al. 2007; Chauhan et al. 2006; Sharma et al. 2006; Saxena et al. 2013; Masud Alam et al. 2009; Enzel et al. 1999; Lückge et al. 2001; Sarkar et al. 2000; Thamban et al. 2002; Patnaik et al. 2012; Kuppusamy and Ghosh 2012; Thamban et al. 2007; Ponton et al. 2012; Laskar et al. 2013a, b) (Figs. 10, 11). An integration of all our proxy results shows that the variations in the sand/clay ratio influenced the ratio of the benthic/planktonic diatom species, sponge spicules and phytolith and this probably was caused by local hydrodynamic changes in SWM and also due to the fluctuations in the marine water column, temperature and salinity. Our data points to an overall weakening of SWM from 5000 years BP to 1200 AD (Figs. 9, 10). The Early middle Holocene period (~6500 years BP to 5000 years BP) experienced strong SWM that eventually tapered. The period from ~5000 to 3600 years BP witnessed a weakening of SWM rains, followed by fluctuations in its intensity from 3600 to 2400 years BP. Further the SWM conditions did not improve, but continued to weaken from 2400 years BP to 800 years BP. However, considerable variability exists in the continental records about the widespread climate of the Himalayas during the mid Holocene period. For example, in the Gangotri area, the Shivling Stage of glaciations has been reported around 5100 years BP (Sharma and Owen 1996) and the pollen records from the Dokini Glacier-periglacial peat indicate prevalence of arid climate during 4000–3500 years BP (Phadtare 2000).

Fig. 11
figure 11

A comparison of monsoonal shifts of the present study with global climate events, Bond events (shaded) and other palaeoclimate records, reviews from Arabian Sea, Bay of Bengal, Andaman Sea and Indian sub-continent. [References (1) Chauhan et al. (2004), (2) Chauhan and Vogelsang (2006), (3) Chauhan and Suneethi (2001), (4) Mathien and Bassinot (2008), (5) Chauhan et al. (2000), (6) Rashid et al. (2007), (7) Laskar et al. (2013a, b), (8) Lückge et al. (2001), (9) Sarkar et al. (2000), (10) Thamban et al. (2002), (11) Sharma et al. (2006), (12) Saxena et al. (2013), (13) Masud Alam et al. (2009), (14) Singh et al. (1990), (15), Enzel et al. (1999), (16) Patnaik et al. (2012), (17) Kuppusamy and Ghosh (2012), (18) Thamban et al. (2007) and 19, 20, 21 is our present data of phytolith, clay mineralogy and Geochemistry

Full size image

In our records this phase terminates at around 4300 years BP. Phytolith records from Deoria Tal (2770 m asl; Uttarkashi), Sharma et al. (1995) reported an occurrence of a humid period around 4000 years BP and a persistence of drier climate between 3200- and 1700-years BP. By and large, the variations observed in the marine sediment record are similar to the other proxy records, except with some mismatches and sharp changes that may be attributed to uncertainty in the chronologies and relative differences in sampling resolution.

The most common feature in all these records is the reduction in the SWM intensity during the RWP and MWP and this may be owing to a reduction in solar radiation reaching the Earth, which may have triggered a change in atmospheric circulation patterns (Wang et al. 2003) Further a reconstruction of ocean current patterns, based on the granularity of deep ocean sediment, Bianchi and McCave (1999) concluded that there was a Roman Warm Period, which peaked around AD 150. An analysis of oxygen isotopes in mollusk shells analysed from an Icelandic inlet indicated that Iceland experienced an exceptionally warm period from 230 BC to AD 40 (Patterson et al. 2010). Recent Res (Werner et al. 2019), based on a much larger dataset using palaeoclimate proxies has revealed that the RWP, along with other periods such as the MWP and the LIA were regional phenomena and not globally clear episodes (Fig. 10).

Our present study results based on the sediment core collected from the BOB near the Landfall Island clearly demonstrates that there exists an agreement of climate change in the region as observed in North Atlantic Ocean, distant Himalayas and East Asia and provides evidence for RWP, DA (Dark Ages), MWP (Fig. 11).

6 Conclusions

A marine sediment core about 124 cm retrieved at a depth of 250 m depth, near Landfall Island, North Andaman, Bay of Bengal, was studied using various proxies such as clay mineralogy, oxygen (δ18O) and carbon (δ13C) stable isotope, neodymium (Nd), geochemistry, phytolith and nanoplankton assemblage. Based on the data collected and integration of results following conclusions are drawn:

  1. 1.

    The sediment texture is predominantly represented by clayey silt. The sediment core reveals layers of coarser sand flux since ~6500–6000 and ~3300 years BP, that reflects a strengthened SWM in an overall weak SWM of the middle to late Holocene.

  2. 2.

    The study of grain size and isotopic data of these sediments show alternate dry/warm and wet (humid) conditions prevailing in the southern tropical areas since ~6500 years BP with weaker southwest monsoon during 4300–4000 years BP, strengthened SWM conditions ~3300 years BP and an amelioration in climate since ~2000 years BP. Due to the intense SWM, the short, wet conditions resulted in intense chemical weathering of the Island rocks. The arid event with an upwelling record during 3500 years BP of the western Arabian Sea as reported by Naidu et al. (2012) is not prominent in the Bay of Bengal

  3. 3.

    The clay minerals and nannoplankton assemblages suggest that the sediments deposited in this region have been contributed by two major provenances. A nearby source, i.e. from Andaman sediments predominately derives from the Andaman terrain of ophiolite suites and sedimentary rocks, the nanoplanktons are reworked, followed by sediments from the Irrawaddy, Salween and Sittang rivers from Myanmar.

  4. 4.

    Smectite and illite are the two dominant clay minerals in this sediment core. High content of smectite reflects the weathering product of mafic rock contributed from the nearby island whereas illite is predominantly coming from the rivers such as Irrawaddy and Salween.

  5. 5.

    An overall weakening of SWM is observed from ~6000 to 2000 years BP, and this is inferred from the high smectite, high C/I ratio and low K/C ratio. Within this period; an incursion of a more intense weakening of SWM is noted from the elevated C/I ratio and lesser K/C ratio from 4400 to 4200 years BP, which is also supported by the δ18O data of G. ruber (−3.39 to −2.33‰). The occurrence of high kaolinite percentage and K/C ratio points to strong weathering during 3900 to 3600 years BP and subsequently an arid phase is observed between 20 and 15 cm (~2000–1700 years BP). A major wet phase also found reaching its maximum at 3400–3200 years BP and amelioration in climate since ~2000 years BP to present. The intensity of SWM tends to repeat itself.

  6. 6.

    The phytolith analysis shows a significant variation in the SWM from ~6500 to present, with weakening of the SWM from 5000 years BP. Variations in the sand/clay ratio influenced the ratio of the benthic/planktonic diatom species, sponge spicules and phytolith and this probably was caused by local hydrodynamic changes in SWM and also due to the fluctuations in the marine water column, temperature and salinity.

  7. 7.

    Application of various proxies in present study reflects the significant signatures of intense and weakened SWM such as the wet phase during 6500 to 5600 years BP (HCO), 3400–3200 years BP (SBO) and warm, dry phase ~2200–1800 years BP (RWP) and 1000–800 years BP (MWP) with an extreme weakened monsoon ~4400–4000 years BP (Sub—Boreal Optimum).