Holocene evolution of the Banni Plain at the north‐east margin of the Arabian Sea: Constraints from a ca 50 m long sediment core

Holocene evolutionary history of the Banni Plain in the Great Rann of the Kachchh Basin is reconstructed from a subsurface sediment core of ca 50 m. Detailed data on textural and lithofacies variations, grain‐size analysis, environmental magnetism and accelerator mass spectrometry 14C dates on seven samples were generated on the sediment core retrieved from the Banni Plain near Berada. A high‐resolution record extending back to 10 ka has been reconstructed from the top ca 40 m of the core section comprising shallow marine sediments. The core is divisible into five depositional units. The basal part is a fluvial depositional unit followed upward by estuarine, sub‐tidal, intertidal and supra‐tidal environments. The sediment accumulation rate is highest in the sub‐tidal to intertidal facies (1.9 cm year−1) and decreases towards the supra‐tidal facies to 0.09 cm year−1. Environmental magnetic analysis, χlf coupled with the S‐ratio, indicates high magnetic mineral concentrations during the Early Holocene, suggesting a wet period accompanied by high monsoon precipitation. This is followed by the onset of semi‐arid conditions in the Great Rann of the Kachchh Basin as indicated by the low values of the χlf and S‐ratios. A westward and northward shift in the shoreline towards the deeper part of the basin is suggested during the Late Holocene, which is coupled with aridity and reduced monsoonal conditions. The change in depositional pattern from the retrogradational deposit of fluvial (Unit 1) to estuarine sediment (Unit 2), progressing to sub‐tidal (Unit 3), is attributed to sea‐level transgression followed by regressive intertidal (Unit 4) to supra‐tidal deposition (Unit 5), culminating in complete withdrawal of the sea, aided by tectonic uplift, during the Late Holocene. The results reveal that the sediment accumulation rates and depositional environments changed over time in response to changes in sea level from minima to maxima and then eventually to the present level.


| INTRODUCTION
Since the last glacial period, global sea level has increased with several short-term events when the rise in sea level accelerated or decelerated (Carlson & Clark, 2012;Saito et al., 1998;Tanabe, 2020;Yokoyama et al., 2018).The sea-level changes through the past 10 kyr have significantly controlled the overall coastal evolution of marginal marine basins around the world (Boyd et al., 1992;Hori et al., 2002b;Stanley & Warne, 1994).Holocene sea-level rise significantly affected landward movement and initiation of deltas/estuaries (Giosan et al., 2006;Stanley & Warne, 1994).It has been established that worldwide delta formation was initiated during the Early Holocene, a period of decelerated sea-level rise (Stanley & Warne, 1994).Detailed analysis of sediment cores revealed that the formation of Asian deltas was linked to a reduced rate of sea-level rise around 8 ka.Prior to this deceleration, there was a rapid rise in sea level between 9.0 ka and 8.2 ka (Bird et al., 2007;Collins et al., 2021;Goodbred & Kuehl, 2000;Hori & Saito, 2007;Li et al., 2012;Saito et al., 2001;Song et al., 2013;Xu et al., 2020).The exact timeframes are noted from other continents, namely Africa and Europe (Amorosi et al., 2012(Amorosi et al., , 2020;;Bruno et al., 2017;Filip & Giosan, 2014;Milli et al., 2013).
A similar approach was adopted here on the sediments of a ca 50 m deep, continuous sediment core obtained near Berada village in the Banni Plain (Figure 1A).The Banni Plain forms the southern part of the tectonically controlled Great Rann of Kachchh (GRK) Basin located at the north-western margin of the Arabian Sea.The GRK Basin has been described as a partially landlocked marginal marine basin that is connected to the Arabian Sea at its western end (Maurya et al., 2013).Presently, the central and northern half of the GRK Basin comprise a vast supratidal flat, while the southern half forming the Banni Plain is free of marine influence owing to its slightly higher elevation (Maurya et al., 2013;Padmalal et al., 2019).Kumar et al. (2021) demonstrated that in the central, deeper part of the GRK Basin, to the north of the Banni Plain, shallow marine conditions prevailed throughout the Holocene.However, the Banni Plain, which is closer to the rocky uplands in the south, is expected to show transitional conditions.This study tries to assess the sedimentary evolution of the Banni sub-basin based on a reconstruction of depositional conditions, palaeoenvironments and the role of eustatic and relative sea-level changes throughout the Holocene.To characterise sedimentary facies, structures, and the nature of contacts, radiographs of the core and core cut sections have been studied and supplemented by particle size analysis.Environmental magnetism was performed to understand the sedimentary packages and their implications for palaeoclimate on a temporal scale.A model of the sedimentary evolution of the Banni Plain during the Holocene is reconstructed vis-à-vis sea-level changes, based on the stratigraphic framework and accelerator mass spectroscopy (AMS) radiocarbon chronology.

| REGIONAL SETTING
The GRK Basin is bounded by the seismically active Kachchh Mainland Fault (KMF) to the south and the Nagar Parkar Fault (NPF) to the north (Figure 1A,B).The southern part of the GRK Basin forms a distinct geomorphic entity, the Banni Plain, which is at a higher level (5-12 m above sea level, m a.s.l.) than the salt encrusted, marine tidal inundation-prone, supra-tidal flat of the central and northern part of the GRK Basin (2-4 m a.s.l.; Kar, 1993;Merh, 2005;Maurya et al., 2008Maurya et al., , 2009Maurya et al., , 2016;;Padmalal et al., 2019;Figure 1C).At the present time, there are parts of the Banni Plain inundated only by monsoonal rains.The southern edge of the Banni Plain is covered with fan shaped sandy alluvial sediments deposited by ephemeral streams flowing from the Kachchh mainland (Chowksey et al., 2011).During the monsoon season (July-September), the low-lying areas of the Banni Plain get ponded by freshwater from precipitation and northerly draining rivers from the Northern Hill range to the south.Previous investigations show that the Banni Plain and Great Rann are contiguous terrains formed over Holocene shallow marine sediments deposited in partially landlocked palaeo-gulf conditions (Khonde et al., 2017;Maurya et al., 2013;Merh, 2005).
The flat surface with extremely gentle gradients is the most distinguishing characteristic of the Banni Plain.Annual submergence during monsoons is controlled by subtle topographic variations related to subsurface structural elements (Maurya et al., 2016).Topographic variations show four surfaces: the higher Banni surface, the middle Banni surface, the lower Banni surface and the saline/salt encrusted surface, including depressions (Maurya et al., 2016).The linear depression running eastwest through the Banni Basin comprising several water bodies and stretches of salt encrusted saline surfaces formed due to drying up of submerging waters.The middle Banni surface consists of radial drainage patterns and a network of shallow fluvial channels.
Structurally, the Banni Basin is described as a half graben (Biswas, 1974(Biswas, , 1987) ) bounded by the KMF to the south and the Banni Fault (BF) at its northern margin (Figure 1B).The Banni Basin comprises a Precambrian basement overlain by thick Mesozoic and Cenozoic sediments (Biswas, 1993(Biswas, , 2005)).A thick pile of ca 250-300 m of Quaternary sediments overlies the Tertiary sediments (Biswas, 1993).The Quaternary sediment thickness is ca 300 m near the southern margin of the KMF, forming an asymmetrical syncline in the subsurface called the Banni Syncline (Biswas, 1993).An uplifted subsurface horst block is present below the region around Bhirandiyala (Biswas, 1993).The present elevations of the Banni Plain and the Great Rann surface is attributed to differential tectonic uplift along subsurface structures, accompanied by seismic activity during the last ca 2 kyr (Biswas, 1974;Maurya et al., 2013Maurya et al., , 2016;;Merh, 2005;Padmalal et al., 2019;Roy & Merh, 1981).Geomorphological, archaeological and historical accounts show that the entire Banni-Great Rann Basin was submerged under a shallow sea until ca 2 ka (Biswas, 1974(Biswas, , 1993;;Gaur et al., 2013;Glennie & Evans, 1976;Maurya et al., 2013Maurya et al., , 2016;;Roy & Merh, 1981).However, the sedimentary evolution, palaeoenvironments and impact of sea-level changes are not documented from the Banni Plain.To generate data on these aspects, a continuous sediment core of ca 50 m depth was obtained from Berada, located in the southern part of the Banni Plain.

Berada core
A ca 50 m deep sediment core was raised at Berada located in the Banni Plain of the GRK Basin.Core recovery was generally more than 90%.In the laboratory, the cores were x-rayed immediately after recovery followed by core cutting-splitting, photography, physical observations on colour and textural changes followed by high resolution sub-sampling at regular intervals of ca 2 cm (Maurya et al., 2013).To identify sedimentary structures, radiograph images were acquired for all cores (Material S1: Figure S1).One half of the split cores were preserved at a controlled temperature as an archive.The radiographs helped identify shells and fossil materials, rootlets, sedimentary structures and fossil plants, as well as distinguish deposits without disturbing them.Furthermore, the photographs taken after splitting the cores were combined with the corresponding radiograph to identify sedimentary structures (Material S1: Figure S1), colour variation and estimate changes in sedimentary characteristics (grain size, nature of laminas, thickness, nature of contact, abrupt/or gradual changes in sediment nature, etc.).

| Chronology
The Berada core is chronologically constrained using seven radiocarbon ages.Amongst these, four AMS dates (BRD/AMS/PRL/1,2,3,4) based on inorganic carbon were carried out at the AMS facility of Physical Research Laboratory (PRL), Ahmedabad during the present study.The other three AMS dates (two on the inorganic and one on the organic fraction) were obtained from the NSF AMS Facility, University of Arizona, USA and published previously by Khonde et al. (2017).The strategies and limitations for the datable materials used in the present study are provided in Material S1.Micropalaeontological analysis of the core samples carried out previously, at low-resolution, showed the occurrence of shallow marine planktonic and benthonic foraminifera (Khonde, 2014).However, the foraminifera in general showed extremely poor diversity and abundance due to the highly stressed environment (Khonde, 2014;Khonde et al., 2011).In view of the very low foraminiferal abundance and their significantly reduced size, small single and broken gastropod and bivalve shells were also used for dating (Material S1: Figure S2).The shells were collected from the 2 cm slice of split core.To prepare sufficient material for dating, a few shell fragments from the same slice were added as required to avoid or minimise dating errors (Material S1: Figure S2).All shell fragments showed negligible abrasion as a result of transport as evidenced by sharp angular edges and margins with intact ornamentation (Material S1: Figure S2).Such features on the small fragile shells suggests minimal transport before rapid burial under terrigenous sediment deposited at a high sedimentation rate.Details of the samples, including microphotographs of the shell samples used for radiocarbon chronology in the present study are given in Material S1.All samples yielded radiocarbon ages in sequence and within acceptable error limits.Several studies have demonstrated that the dating of fragmented shells chosen carefully can be used for interpreting palaeoenvironmental and sea-level changes (Craw et al., 2021;Finstad et al., 2013;Vacchi et al., 2017).
The radiocarbon dates were calibrated using mixed marine calibration IntCal20 and Marine20: Calib 8.2 software (Reimer et al., 2009).The marine contribution for the same was-50%; ΔR = −30 ± 106 (Bhushan et al., 1994;Dutta et al., 2001) and are reported in radiocarbon kilo-annum (1000 years ago; Table 1).Table 1 lists the materials used and other details of core dates.Based on the 14 C dates, the age model was produced using the Bayesian method (Haslett & Parnell, 2008) with the 'Bchron package' in the R program corresponding to 1σ uncertainty (95.4% probability; Figure 2).
The top of the core could not be dated due to the absence of datable material.However, the age of the Banni Plain surface is approximately 1.4 + 0.02 ka.This age is based on organic matter obtained from a shallow trench at Chachi (Figure 1), located about 10 km north of the present core site (Pillai et al., 2017).This compares well when extrapolating ages obtained from the central and bottom parts of the cores acquired during the present study.Comparable ages are obtained from the surface sediments at various other far-off places in the GRK Basin (Makwana et al., 2018;Sharma et al., 2021;Tyagi et al., 2012).This age for the surface of the Banni Plain is used later in the section on sea-level change to explain the emergence during the Late Holocene.

| Grain-size analysis
Particle size analysis of the shallow marine sediment section in the core was only possible to depths of 40 m as the sediments below 40 m depth are fluvial and consist of very coarse-grained sediments, gravels, cobbles and pebbles not suitable for laser particle size analysis.Laser diffraction particle size analysis (Beckman Coulter) was carried out on 89 samples at equal intervals of ca 45 cm at the Birbal Sahni Institute of Palaeosciences (BSIP), Lucknow.For each sample, around 5 g of sample was mildly crushed and air-dried before acid treatment.Carbonate in the sample was removed by 10% hydrochloric acid (HCl) while organic matter was subsequently removed using a 30% hydrogen peroxide (H 2 O 2 ) solution.After each acid treatment, sample pH was neutralised by thorough rinsing with de-ionised water.Before particle size analysis, samples were treated with a 2-3 mL solution of sodium hexametaphosphate (Na₆[PO₃]₆) to disperse the particles and avoid coagulation.Each sample was shaken with a sonicator before analysis to avoid settling of large particles.
To determine mean, median, mode, skewness and sorting, the particle size analysis results were analysed using Gradistat software (Material S2: Table S1; Blott & Pye, 2001).The textural classification was established following the scheme proposed by Blott and Pye (2012), thus, sediment texture analysis was used to determine the downcore lithological variation of the Berada core (Figure 3A,B,C).To further demonstrate the change in energy conditions during sediment deposition, linear discriminate function (LDF) analysis was considered.To distinguish the process and environment of deposition LDF of Y1, Y2, Y3 and Y4 was employed following the technique of Sahu (1964), where Y1 distinguishes between shallow agitated water/beach, Y2 shallow marine/beach, Y3 shallow marine, fluvial and Y4 turbidity current and deltaic deposition (Material S2: Table S2).

| Environmental magnetism
A total of 196 samples collected at ca 20 cm intervals were analysed using standard environmental magnetic parameter techniques (Walden et al., 1999).For analysis, 10 g of air-dried sample was tightly packed in 10 cc non-magnetic sample holders.Using the Bartington Susceptibility Meter (Model MS2B; noise level ca 3 × 10 −9 m 3 kg −1 for a 10 g sample), low frequency (0.47 kHz) magnetic susceptibility (χlf) was measured.Anhysteretic remnant magnetisation (ARM) was induced in the samples using an ASC Scientific D-2000T alternating field demagnetiser in a constant dc biasing field of 0.05 mT superimposed on a decaying alternating field (a.f.), with a peak intensity of 100 mT at a decay rate of 0.001 mT per half-cycle.The susceptibility of ARM (χARM) was calculated by dividing the mass specific ARM by the size of the DC biasing field (0.05 mT = 39.74A/m; Walden et al., 1999).Isothermal remnant magnetisation (IRM) was induced in the samples at different field strengths of 20, 50, 70, 100, 200, 300, up to 1000 mT and back fields up to-300 mT using an ASC Scientific IM-10-30 impulse magnetiser.The remnance was measured in a JR-6 spinner magnetometer of AGICO.The interparametric ratios used for this study are S-ratio, SIRM/χlf and χARM/SIRM, ARM/χlf, soft IRM and hard IRM.The isothermal remnance induced at 1000 mT was considered as the saturation isothermal remnant magnetisation (SIRM = IRM 1000mT ).The S-ratio was calculated using the expression (IRM −300mT /SIRM).
The environmental magnetism analysis of the sediment core indicates wide-ranging magnetic imprints linked to the palaeomonsoon and palaeoclimate (Shankar et al., 2006).The χARM is sensitive to single domain ferrimagnetic minerals (SD, 0.03-0.07μm; Maher, 1988), while χfd% indicates fine viscous grains near the boundary between SP and SD (Maher, 1988).The χARM/SIRM and χARM/χlf ratios (Material S3: Table S1) indicate variations in ferrimagnetic grain size and serve as a magnetic mineral grain-size indicator (Banerjee et al., 1981;Maher, 1988).Particle size correlation is also shown with χARM (Material S3: Table S1; Zhang et al., 2007).The χlf is influenced by organic matter dilution and diamagnetic minerals, while the S-ratio is not, so combining the two can infer environmental processes, erosion and sediment flux (Geiss & Banerjee, 1997).

| Chronology
The chronological sequence for the Berada core was established based on the mean sedimentation rate between the two adjacent calibrated ages, using linear interpolation for the samples which is general practice (Joshi et al., 2023;Khan et al., 2019;Lacourse & Gajewski, 2020;Reimer et al., 2009).There is a 95% confidence interval (1σ) for the calibrated ages, while the weighted means are given in Table 1.Based on the available seven AMS radiocarbon dates from the Berada core, the bottom samples at 39.1 m and 38.8 m depth yielded dates of ca 9.8 ka and 9.5 ka, respectively.A low sedimentation rate of 0.06 cm year −1 is recorded from a depth of ca 39-38 m, thereafter, followed by an increase in sedimentation rate to 0.9 cm year −1 and 0.7 cm year −1 between ca 38-33 m and ca 33-19 m, respectively.
Based on the Bayesian sedimentation model derived from the 14 C dates, the core shows that the Early Holocene period extends to a depth of ca 23 m.The sedimentation rate is enhanced significantly during the Middle Holocene period when compared to the Early Holocene.This change is noted between depths of ca 19 m and 11 m, where the sedimentation rates range between 1.9 cm year −1 and 1.7 cm year −1 , which is the highest in the entire core.The sedimentation rate in the latter half of the Middle Holocene between 11 m and 5.5 m is reduced to 0.4 cm year −1 at ca 6 ka.
The sedimentation rate for the top section of the core between ca 5 m and 0 m depth is not interpreted as these sediments did not yield datable material.However, a significant drop in sedimentation rate is obvious when compared to the rest of the core.This inference is confirmed if we extrapolate the age of 1.4 + 0.02 ka from surface sediments in a nearby short core retrieved at Chachi (Figure 1A; Pillai et al., 2017) to the core site for comparison purposes.Doing so, yields a sedimentation rate of 0.09 cm year −1 , which is the lowest in the entire core.

| Sedimentary characteristics of the core sediments
The study of radiographs and a visual examination of the split Berada cores allowed for documentation of sedimentary features with depth.Material S1: Figure S1   The sediments from 50 to 47 m depth are characterised by the presence of unconsolidated, brown coloured, highly angular and unsorted coarse sand of fluvial origin.The brown coloured, compact, ferruginous sandy silt changes to ash coloured hard clay at a depth of 46-42 m.At a depth of 44 m, a cross bedded sedimentary structure is observed.From 42 to 39.6 m depth, the core consists of unconsolidated brownish-white coloured coarse sands with larger pebble-sized fragments.Stiff muds frequently occur in this section of the core (Figure 4A). 4.2.2 | Depth 39.6-30 m A sharp contact separates the coarse grained fluvial sediments below 39.6 m from the overlying dominantly fine-grained sediments that extend to a depth of 30 m (Figure 4B).At 38.8 m, which is just above the contact, a thin peat layer is noted along with bioturbated mud containing complete shells as well as broken shells of gastropods and cerithiids.Additionally, complete and broken gastropod shells are found between 38 m and 36 m depth.The section is characterised by the high organic matter content.The shell fragments show sharp edges and margins indicating minimal transport before they were covered in rapidly accumulating sediment.The fine grained and finely laminated nature of the sediments also suggest a low energy environment.In addition, organic matter also occurs as scattered particles ca 1-2 mm in size between depths of 36-35 m.The section also contains bioturbated mud deposits caused by borrowing activity at a depth of 34-32.5 m (Figure 4C). 4.2.3 | Depth 30-19 m   The contact of this unit with the underlying unit is gradational.A few complete gastropod shells are seen at a depth of ca 30 m (Figure 4D).A few lensed shaped accumulations of organic matter are present at around 28 m (Figure 4E).The sediments at a depth of 28 m show the presence of fining upward sand mud couplets along with lenticular to wavy sedimentary structures (Figure 4E).Sediments throughout the section show a uniform dark green to green colour with some greyish variation towards the top of the core.A 2-3 cm thick mudball is present within a deformed structure at a depth of around 24 m (Figure 4F).The sediments from this unit show distinct sedimentary structures and features, like clusters of organic matter, when compared to the overlying unit. 4.2.4 | Depth 19-7 m   This unit comprises highly disturbed and non-homogenous sediments that are characterised as bio-turbidites with extensive burrowing activity at depths of ca 19 m and 16 m (Figure 4G; Madsen et al., 2007;Sorrel et al., 2010).The presence of faint lamination and cross lamination is evident at a depth of 18.5 m, where broken shell fragments are also found.The sediments at ca 18 m depth show the presence of rootlets.This section marks a sharp erosional contact with the underlying sediment at a depth of ca 17 m.Although the sharp contact is at ca 17 m depth (Figure 4G), sedimentologically the sediments at ca 19 m depth are different to the underlying sediments.Hence, this section is demarcated from a depth of 19 m onwards.Nonhomogenised depositional conditions are evident from the sediments at the 17 m contact point.The moisture content of this section is high compared to the sediments below.Additionally, the sediment changes colour from dark greenish to light brownish green at the point of contact.

| Depth 7-0 m
The section clearly marks a sharp contact at ca 7 m with the underlying sediments (Figure 4H).The variation from light greenish to brown coloured sediments at ca 7 m indicates a change in depositional conditions.The section contains laminations and tidal bundles at ca 3 m and 2 m (Figure 4I,J).Dispersed grains of gypsum and anhydrite are common at 2.5 m and above.As a whole, this section shows sediment deposition under emergent/regressive conditions.The unit also shows the presence of terrestrial sediments (oxidised or exposed) along with salt crystals and anhydrite minerals.

| Grain-size analysis
Grain-size variations, textural recognition and their interparametric relationships were examined based on grainsize analysis carried out on the sediments of the Berada core.The bottommost Unit 1 comprises coarse grained fluvial sand.This is overlain by a marine section (0-40 m depth) in the Berada core, which is predominantly fine grained and is divisible into four sedimentary units (Units 2-5) based on grain-size characteristics (Material S2: Table S1).Grain-size data of sediments comprising Units 2-5 show five textural classes according to the scheme given by Blott and Pye (2012).Very slightly sandy slightly clayey silt (62%) texture dominates the core followed by slightly sandy slightly clayey silt (19%), slightly clayey silt (5%), slightly clayey sandy silt (3%) and very slightly clayey sandy silt (2%) texture.More specifically, silt size particles dominate (55%-86%), while the upper half of the core contains a relatively higher sand content ranging between 2% and 40%.Unit 5 shows the highest sand content followed by Unit 3, with Units 2 and 4 showing the lowest (Figure 5).
Amongst the samples analysed, 58.4% were unimodal, 35.9% were bimodal, while trimodal grain sizes contributed to 3.3% of samples and polymodal 2.2%.According to the Gradistate software the trimodal and polymodal samples are categorised as very poorly sorted, while unimodal samples show poor sorting (Blott & Pye, 2001).The computed values of Y1, Y2, Y3 and Y4 range between −0.4 to −15, 407.7 to 269.8, −19 to −39.0 and 12.1-7.9,respectively (Material S2: Table S2).The scatter plot of Y1 versus Y2 shows that 4% of the total samples were deposited within beach/shallow agitated environments (Figure 6A), whereas the rest appear to have been deposited under aeolian/shallow agitated conditions (Figure 6B).The scatter plot of Y3 versus Y4 suggests the Early Holocene period samples show transition from fluvial to turbidite depositional conditions.
The results of grain-size analysis are summarised in Table 2 while the detailed description of the sedimentary units is given in Material S2.Unit 1 (50-39.6m) is dominated by coarse to very coarse sand, deposited under fluvial conditions.The sediments are yellow to brown in colour, angular, unsorted and poorly sorted, with pebbles  ) is dominated by silt, followed by clay and negligible fine sand, with rhythmic fluctuation patterns between 8% and 17% clay distributed across this unit (Figure 5).Unit 3 (30-19 m) shows an increase in sand content compared to the underlying unit to ca 6%, while the silt contribution is ca 82% and clay is ca 11% (Table 2).Unit 4 (19-7 m) predominantly shows mud, with the sand contribution varying between 3% and 12% at different depths (Table 2).An intertidal environment is suggested for this unit based on the grain-size assemblage.Unit 5 (7-0 m) is dominated by silt, with a significant increase in sand content (1%-40%) at a depth of 1 m (Table 2).In general, the fine-grained sediments of Units 2-5 suggest deposition within temporally variable shallow marine environments with the top unit marking the transition from intertidal to supratidal to present continental conditions in the Banni Plain (Table 2).

| Magnetic mineral parameters
The magnetic mineral results reveal significant variations in various parameters, such as χlf, χARM, SIRM and interparametric ratios, including SIRM/χlf, χARM/SIRM, χARM/χlf, S-ratio, soft IRM and hard IRM.The core is divisible into three units as indicated by the plots of each magnetic parameter, which includes the line of average or median value (Figure 5).The most noteworthy trend observed is the significant and contrasting change at the boundary between each Holocene Period, with sudden spikes in magnetic parameter values.From 39.6 to 23 m depth (Unit 1), the concentration of magnetic minerals (χlf) is highest at ca 36 m with an average value of ca 13 (10 −8 m 3 kg −1 ), which then dip towards the lower side of the average value between 33 m and 35 m.The S-ratio displays some interesting changes from the bottom of the unit, ranging from 39.6 to 23 m (Unit 1).Two low peaks are observed at ca 39 m and 35 m, indicating the lowest value of less than 0.4.Afterward, the S-ratio gradually increases reaching higher values of ca 1 at a depth of ca 37 m and 33 m.Between depths of 23-11 m (Unit 2), magnetic concentration-dependent parameters χlf, χARM show suppressed magnetic mineral concentration with average values of 11.5 (10 −8 m 3 kg −1 ) and 9.9 (10 −8 m 3 kg −1 ), respectively, while the S-ratio shows a slightly increasing positive trend between 15 m and 20 m depth with a value of 0.8.The shift in the χARM/SIRM parameter, observed between depths of ca 19-12 m, from low to high values (from 0.06 to 0.4 10 −3 mA −1 ), indicates a transition from multidomain (MD) to single-domain (SD) magnetic grain size.From 11 to 0 m depth (Unit 3), significantly higher values in all magnetic parameters, including magnetic grain size and magnetic minerals, are encountered.Interestingly, anti-correlation between trend values of χlf and χARM is noted at ca 3 m depth, although they have a generally positive trend throughout this unit.At a depth of ca 6 m, the S-ratio shows a high value of ca 1.This value then follows the average trend line and remains consistent until the end of this unit.Positive values are displayed by all magnetic parameters (χlf, χARM, χARM/SIRM) from ca 1 m to present, except for the S-ratio.A detailed description of the results of the environmental magnetism analysis, including down core variation of χlf, χARM, SIRM by unit and interparametric ratios is included in Material S3.
Figure 5 shows that the plots of environmental magnetism parameters correlate well with the grain-size variation, especially for the silt sized fraction throughout the core.

| INTERPRETATIONS
The present study on a ca 50 m deep sediment core has generated a high-resolution dataset of sedimentological and environmental magnetic parameters that contributes to understanding the Holocene sedimentary evolution of the Banni sub-basin.The five facies and their depositional units observed in the Berada core include the fluvial to transgressive facies, estuarine to marshy facies, the sub-tidal zone facies, the intertidal facies type and the supra-tidal facies (Figures 7 and 8).The Banni-Great Rann Basin is a challenging area in which to interpret sea-level changes due to its tectonically and seismically active nature (Maurya et al., 2006(Maurya et al., , 2013(Maurya et al., , 2016)).The sedimentary evolution of the Banni Plain is discussed with respect to the sea-level changes and palaeoenvironmental conditions.

| Sediment accumulation curve and chronology
Accumulation curves based on age-depth plots of two cores, Berada core (this study) and Dhordo core (Kumar et al., 2021) are shown in Figure 9.The accumulation curve provides an inter and intra age-depth relationship indicating the rate and period at which the sedimentary environments (facies) such as fluvial, sub-tidal and intertidal occurred/co-occurred and passed through one core to another in an area.A close relationship is observed between the sedimentary facies and accumulation rates.The accumulation rate of the fluvial (Berada core: depth ca 50-40 m) and estuarine facies (Berada core: depth ca 40-30 m) is limited to the southern periphery of the basin, which is not encountered in the Dhordo core (central basin) until ca 60 m depth (Figures 9 and 10).The Dhordo core shows high sedimentation rates of 8.71 cm year −1 followed by a drop to 0.46 cm year −1 (ca 60-37 m) during sub-tidal depositional conditions (Early Holocene), while the Berada core shows a comparatively low sedimentation rate of 0.06-0.9cm year −1 (ca 30-19 m) for the same period.
A high sedimentation rate during the Early Holocene is also noticeable from the sea-level curve of Hashimi et al. (1995) from the western continental margin of India.Moreover, owing to the present sub-aqueous stratigraphy and topographic difference between the central (Dhordo) and southern portions (Berada) of the GRK Basin, the accommodation space created at Berada would have triggered huge sediment infills towards the centre of the basin compared to that at the southern margin.The sheltered regime of the Berada area must have also contributed to the comparatively suppressed sedimentation rate.
The sub-tidal conditions observed in both cores marks the highest shoreline extension into the GRK Basin, which correlates with a period of rapidly rising sea levels during the Middle Holocene.This was followed by a regression phase during the latter part of the Middle Holocene, leading to the establishment of intertidal conditions over a wide area (Berada core: depth ca 19-7 m; Dhordo core: depth ca 27-9 m).The sedimentation rate in both cores during this phase is similar and ranges between 1.7 cm year −1 and 0.64 cm year −1 .The establishment of supra-tidal conditions, attributed to regressive conditions throughout the basin during the  F I G U R E 9 Accumulation curves of Berada core (present study) and Dhordo core (Kumar et al., 2021) with inferred sedimentary units, and comparison with reconstructed sea-level curves and data from other regions during the Holocene.Sea-level data from other regions.East coast of India (Banerjee, 2000), Western continental margin of India (Hashimi et al., 1995).

| facies and depositional environments in the Banni Plain
Five different depositional environments (facies) for sedimentary Units 1-5 are observed in the Berada core based on physical observations of the cut sections (sediment colour, sedimentary structures, nature of contact, organic matter distribution), particle size data (texture, mean, mode, median, skewness, kurtosis), fossil shell content and bioturbation activity. 5.2.1 | Unit 1: Fluvial to transgressive facies (50-39.6 m) Unit 1 forms the bottom part of the Berada core and is characterised by the presence of coarse, pebbly, poorly sorted massive sands that occur between ca 50-39.8m depth (Figure 4A).The colour of the coarse sandy part of this unit varies from yellow to brownish red indicating deposition in oxidising, shallow water conditions (de Boer, 1998; Figures 4 and 7A).Quartz, feldspars and dark minerals are the primary constituents of these sediments, which marks the fluvial regime at the core site.These sediments appear to originate from local/proximal sources, that is from the Mesozoic sandy formations that are exposed in the Kachchh Mainland Hill range to the south (Maurya et al., 2013). 5.2.2 | Unit 2: Estuarine to marshy facies (39.6-30 m) These shallow marine sediments show a sharp erosional contact with the underlying fluvial facies.The peat/organic matter layer encountered at 38 m, along with the presence of cross lamination or inclined lamination between 38 m and 39 m, indicates a sharp change in depositional conditions from fluvial to shallow marine (Figures 4 and  7B).The occurrence of micro-fossils, complete and broken shell fragments at ca 39 m indicates deposition during marine flooding (Khonde, 2014).Abundant organic F I G U R E 1 0 Schematic correlation along line A-B (shown in Figure 1B) of depositional environments of Berada core (present study) and Dhordo core (Kumar et al., 2021).Note that the depositional environments inferred in the southern margin (Berada core) and in the central part (Dhordo core) of the GRK Basin complement each other.material is common in the estuarine/marshy setting and is often delivered by freshwater flux as organic debris/bulk organic matter that gets preserved in a stagnant water column.(Song et al., 2013).This is also evident from the scatter plot Y4 versus Y3, indicating sediment deposition in a shallow marine environment (Figure 6B).Particle size data for this unit shows >95% mud with a low sand content, which is typical in estuarine/marshy settings comprising tidally influenced moderate to low energy environments (Dalrymple et al., 1990;Li, Chen, et al., 2000;Li, Wang, et al., 2000).The moderately well sorted nature of the sediments, with an average skewness value of ca 0.3 φ, suggests dominance of finer grained sediment (Figure 3; Material S3: Table S1).The scatterplot between skewness versus sorting shows deposition in a sheltered/ closed environment, which is consistent with the geomorphic set-up of the basin (Figure 6).
The occurrence of organic matter within the upper part of the unit, together with the formation and preservation of a peat layer, suggests that calm and low energy deposition was initiated during the Early Holocene at ca 9.5 ka (Figures 4B and 7B).An upward decrease in sand content and a colour change from yellowish brown in the preceding section to greyish to blackish bioturbated sediments is indicative of an increased water column and reduced energy.This supports the interpretation of advancing marine conditions at the core site forming estuarine to marshy conditions.
F I G U R E 1 1 Schematic models showing the Holocene evolutionary history of the Banni Plain.(A) Pre-Holocene-alluvial deposits in the Banni Plain in response to low sea level.(B) Early Holocene-sea-level rise and shoreline shift into the Banni Basin leading to estuarine-tidal to subtidal deposition.(C) Middle Holocene-subtidal followed by intertidal depositional environment due to initiation of regression in later part.(D) Late Holocene: regression leading to development of supratidal conditions followed by progressive shift to continental conditions. 5.2.3 | 3: Sub-tidal zone (30-19 m)   This unit is marked by an overall in grain size (coarsening upward cycle) compared with the underlying unit.The sand proportion increases to 15 from 2% in the preceding unit but drastically reduces in the upper part (Figures 5 and 8; Material S3: Table S1).Based on lithological characteristics, structure/texture and a grain size coarsening upward trend with lenticular to wavy structures, the presence of laminated sand mud couplets (Figure 7E), deposition within a sub-tidal environment is suggested.The couplet of sand and mud at 28 m, along with the lenticular to wavy structure, is marked by an increase in the sand-mud ratio.The sand percentage reaches up to 15% at a depth of 25 m, along with the appearance of cross-bedding, indicating varying energy conditions in sub-tidal settings (Tanabe et al., 2003).Flaser, lenticular bedding and mud drapes suggest that these facies were deposited under a tide dominated environment (Reineck & Singh, 1980).Higher energy conditions are found in the sub-tidal zone, resulting in coarser sediments (sand) during ebb tides (Allen & Posamentier, 1993;Dalrymple et al., 2012). 5.2.4 | Unit 4: Intertidal (mixed tidal zone to mud tidal zone, 19-7 m) Unit 4 is characterised by a fining up sequence between 24 m and 20 m that distinguishes it from the underlying unit.The average mean value of this unit is 6.7 φ where ca 85% is unimodal and ca 15% bimodal.A lower sandmud ratio at around ca 24 m is accompanied by strongly laminated layers of clay and sand, visible in the x-ray image (Figures 4 and 7F).This unit also marks a change in colour from light green to dark green in the core cut sections (Figure 4G).Broken to complete shells of gastropods are encountered at 19 m along with bioturbated massive clay, which makes a gradual contact with the overlying sediment (Figure 7F).The homogenous, structureless sedimentation with negligible cross-lamination indicates negligible change in depositional energy (Semeniuk, 2002;Wang et al., 2012).The bioturbated sediment encountered at 16 m grades into an undisturbed, homogenous muddy deposit (Figures 4G and 7F).The presence of laminated mud at 13 m is interpreted as a change in environment from mixed flat to mud flats.The mud flats are typically strongly bioturbated mud along with mud lamination (Dalrymple & Makino, 1989).The sand-mud ratio from this depth also remains more than 9.The characteristics of organic particles and the sediments suggest deposition of terrigenous clay in a mixed to muddy tidal flat environment (Duane, 1964; Figure 4). 5.2.5 | Unit 5: Supra-tidal (7-0 m; 5-6 m m a.s.l.) A gradual change in the colour contrast from greyish green to earthy-light brown is encountered at 7 m which is found to correspond to a change in sand percentage to 40% (Figures 4 and 7G).Scattered rootlets and organic debris are identified from the core radiographs between 2.25 m and 3.6 m (Figure 7H).Such fine layers of organic-rich sediment are common in a vegetated floodplain or salt-marsh environment (Hori et al., 2002a).The transformation towards the present day plain with gentle gradients is attributed to enhancement of land-like conditions during this period (Pillai et al., 2017(Pillai et al., , 2018)).The occurrence of anhydrite minerals like gypsum is evidence of semi-arid and hypersaline conditions, which point to the existence of emergent/ regressive conditions.

| Implication for palaeoenvironment
The magnetic mineral analysis reveals notable fluctuations in several parameters, including χlf, χARM, SIRM and interparametric ratios such as SIRM/χlf, χARM/ SIRM, χARM/χlf, S-ratio, soft IRM and hard IRM.Based on the chronological sequence of the Holocene epoch, the core is partitioned into three units: Units 1, 2 and 3 corresponding to the Early, Middle and Late Holocene, respectively.

Early Holocene (10-8.2 ka)
During the Early Holocene period (ca 9.7-8.8ka), sediments at a depth of 40-35 m, display moderate to lower values for the S-ratio, which are less than 0.7 (Material S3: Table S1).However, the χlf values in the same sediments contrast with the S-ratio values.This discrepancy can be explained by the trend in hard IRM, which shows a peak reading, confirming the presence of haematite/goethite magnetic minerals (Thompson & Oldfield, 1986).The presence of high coercivity hard minerals suggests a higher terrigenous flux compared to the biogenic elements, indicating that the core location was closer to the landward part of the estuarine/marshy land during this period and was influenced by high tides extending beyond the study area.The Early Holocene is a period of stronger monsoons, with fluctuating but gradually building monsoonal conditions and the input of coarse to finer sediments.The core shows high to low χARM/SIRM values, along with a S-ratio (0.6-0.5) which ranges 0.8 and 0.9 towards the end of this suggesting moderate influx of finer sand sediments, possibly due to the gradual build-up of monsoons during 8.5-7.5 ka (Shankar et al., 2006;Warrier et al., 2011).Palaeoclimate records from Lunkaransar dry lake (Enzel et al., 1999) and the eastern Arabian Sea (Overpeck et al., 1996;Thamban et al., 2002) suggest fluctuating monsoons during the Early Holocene period, with monsoon winds strengthening during 7.5-8.8ka (Sirocko et al., 1993).The mineralogical change noted from the S-ratio after 9 or 8 ka indicates the influence of strengthening monsoonal conditions.

Middle Holocene (8.2-4.2 ka)
The sediment from this period is mainly unimodal where the sand percentage remains less than 5% (Figure 5; Material S2: Table S1).The dominance of mud (silt and clay) advocates deposition in low energy conditions typical of intertidal to mud flat zones.The magnetic mineral concentration (χlf) shows a consistent but transitioning trend towards average values coupled with the S-ratio, SIRM, χARM and SIRM/χlf data.Above average values from 27 to 20 m depth indicates enhanced weathering and high sediment flux (Evans & Heller, 2003).At ca 7.5 ka, a peak in the χlf values and S-ratio reflects the input of ferrimagnetic minerals, with the highest values noted from soft IRM coupled with a high peak in values of χARM further advocates the input of finer ferrimagnetic minerals.These findings are consistent with the clay mineralogy results that indicate high sediment flux in the study area as a result of chemical weathering (Khonde, 2014;Khonde et al., 2017).The granulometric parameter χARM/SIRM shows a transition towards higher values, demonstrating the input of finer magnetic minerals under enhanced monsoonal conditions.The Middle Holocene is known for increased precipitation.Overall, magnetic signatures from sediments of this period indicate strong monsoonal conditions.Other palaeoclimate studies from the Banni Plain also report an increase in monsoonal conditions (Juyal et al., 2006;Makwana et al., 2018).Strong monsoon precipitation during the Early to Middle Holocene is noted from the Thar Desert (Roy & Singhvi, 2016).The data from this study also collaborates pollen evidence from the Himalayan region, which predicts strong monsoons during the Middle Holocene (Phadtare, 2000).

Late Holocene (4.2 ka to present)
This unit contains the highest percentage of sand (40%) in the entire core and the sediments range from clayey silt to sandy silt at ca 2 m (Figure 5; Material S2: Table S1), indicating deposition in shallow but fluctuating energy conditions.The magnetic mineral data in the Late Holocene shows high values of χlf and an increase in sediment grain size under varying climates, while the S-ratio shows a decreasing trend along with χARM/SIRM and an increment in hard IRM, indicating the input of coarser haematite/goethite minerals (Figure 5).The presence of gypsum and anhydrite at ca 3 m and 2 m suggests deposition under drying conditions, with decreased chemical weathering implying oxidising conditions (Khonde et al., 2017).The sediment particle size coarsens towards the top due to regressive conditions.The fluctuations in magnetic parameters at a depth of ca 3.8 m (ca 4 ka) reflect the globally known arid event that occurred in India, including the Thar Desert and mainland Gujarat, at 4.2 ka (Dixit et al., 2014;Gupta et al., 2003;Prasad et al., 2014;Railsback et al., 2018;Staubwasser et al., 2003).The Medieval Warm Period, which occurred from approximately 2-1 ka, is marked by a decrease in magnetic susceptibility and an increase in organic matter at a core depth of ca 1.5 m (Figure 5).The Little Ice Age, which occurred from approximately 1-5 ka, is evidenced by a decrease in sediment grain size and an increase in the S-ratio at a depth of ca 0.5 m (Figure 5).These events demonstrate the sensitivity of the sedimentation in the Berada core to global climate changes, and their potential impacts on local environmental conditions.

| Influence of sea-level changes on marine sedimentation in the Banni Plain
The relationship between sediment accumulation curves and sea-level changes during the Holocene is particularly evident in marginal marine basins that are subject to terrestrial sediment inputs (Tanabe et al., 2003).In these regions, linkage between changes in sea level and sedimentation rates can be observed in the sediment cores extracted from marginal basins (Ta, Nguyen, Tateishi, Kobayashi, Saito, et al., 2002;Ta, Nguyen, Tateishi, Kobayashi, Tanabe, et al., 2002).This approach is extensively used by researchers to provide first estimates of relative and eustatic sea-level changes and their relationship with the sediment accumulation history of the marginal marine basins (Hori et al., 2002b;Ta, Nguyen, Tateishi, Kobayashi, Tanabe, et al., 2002).Since the Berada sediment core did not yield any reliable sea-level markers, the observed sedimentation pattern was interpreted in terms of the available sea-level data from around the globe.
During the Holocene epoch, global sea level has experienced changes due to a combination of factors including thermal expansion of sea water and ice sheet melting (Church et 2013;Masson-Delmotte et 2021).Based on available data, the sea-level curve shows a general pattern on a global scale (Figure 9).A rapid rise in sea level during the Early Holocene is observed, followed by a slower rate of rise during the Middle Holocene, and a relatively stable sea level during the Late Holocene until the last few centuries when sea level has risen rapidly (Düsterhus et al., 2021;Kopp et al., 2009;Peltier, 2004;Rohling et al., 2020;Spratt & Lisiecki, 2016).For example, studies by Dutton and Lambeck (2012) indicate a rapid rise of about 60-70 m in global sea level during the Early Holocene, followed by a slower rise of about 10-20 m during the Middle Holocene, and relatively stable sea level during the Late Holocene until the last few centuries when sea level has risen by about 1-2 m.A recent study by Kopp et al. (2016) found that the current rate of global sea-level rise is faster than at any time in at least the past 2.8 kyr.

| Early Holocene
The analysis conducted by Camoin et al. (2004) indicates that sea level in the western Indian Ocean experienced its most rapid rise during the Early Holocene period until ca 7.5 to 7 ka (Figure 9).A similar steady rise in sea level during this period is recorded from the Maldives and the east coast of India (Banerjee, 2000;Kench et al., 2009).However, data from Singapore and the wider south-east Asian region, demonstrates that the Early Holocene rapid rise of sea level was not continuous, but shows an inflection (Figure 10) which is linked to the global climate event known as the 8.2 ka cooling event, which caused a temporary slowdown in the rate of sea-level rise (Bird et al., 2007(Bird et al., , 2010)).Hashimi et al. (1995) suggested that sea-level rose rapidly during the Early Holocene on the western continental margin of India (Figure 9).This palaeoenvironmental data shows that marine sedimentation (Unit 2) commenced at ca 10.2 ka (Figure 8) as a consequence of the Early Holocene sea-level rise in the GRK Basin.The marine submergence from the north-west and north led to the establishment of estuarine conditions in the Banni Plain.The estuarine sediments show a high sedimentation rate (Figure 2), which conforms to the rapidly rising sea level (Figure 10).The central part of the GRK Basin further north also witnessed high sedimentation rates, but within a subtidal environment during this time (Kumar et al., 2021), which correlates well with data from the Berada core.This phase is recorded in the study area as estuarine conditions, which may have been established due to rising sea level choking off the fluvial sediments that had earlier flowed into the basin, before sea transgression stopped the flow of the river and established estuarine conditions (Chowksey et al., 2011;Maurya et al., 2013).

| Middle Holocene
A general relative slowdown in the rate of sea-level rise during the Middle Holocene (Figure 10) was identified by several researchers using marine limiting, index point and terrestrial limiting data points (Banerjee, 2000;Bird et al., 2007Bird et al., , 2010;;Camoin et al., 2004;Hashimi et al., 1995;Kench et al., 2009;Mann et al., 2016).According to Bird et al. (2010), the rate of sea-level rise slowed down between 7.8 and 7.4 ka, followed by a renewed rise of 4 or 5 m that was complete by 6.5 ka (Figure 9).Camoin et al. (2004) also observed a slow down in the rate of rising sea level during this period.However, Mann et al. (2016) showed a maximum rate of sea-level rise of ca 1.5 mm year −1 between 5 or 4 ka.A Middle Holocene sea-level highstand of 0.50 ± 0.1 m occurred at ca 2 or 4 ka as evidenced by the elevated microatolls on South Maalhosmadulu (Kench et al., 2009).A gradual rise in sea level during the Middle to Late Holocene is also observed on the east coast of India (Banerjee, 2000).The Berada core shows a relatively higher sedimentation rate of ca 1.9 cm year −1 (Unit 3), up to ca 7 ka (Figure 2), followed by a low sedimentation rate of 0.4 cm year −1 (Unit 4).The change in sedimentation rate is coeval with the change in depositional environment from subtidal to intertidal (Figures 2, 8 and 10) which is also observed in the fining of grain size.The relatively high sedimentation rate during the early part of the Middle Holocene and the relatively lower sedimentation rate in the later part of the Middle Holocene relates to the deceleration of sea-level rise at it approached the present level.It is therefore possible to infer that variations in the rates of sea-level rise exerted a major control on sedimentation accumulation in the GRK Basin with other factors playing a minor role.A similar pattern, where sea-level variation influenced sedimentation, is shown in other regions (Banerjee, 2000;Bird et al., 2007Bird et al., , 2010;;Kench et al., 2009).

| Late Holocene
The sea-level data from different regions suggest a largely stable sea level with minor fluctuations (Figure 9).In the Maldives, Kench et al. (2009) found that sea level was relatively stable during the Late Holocene, with only minor fluctuations.In Singapore and the wider south-east Asian region, Bird et al. (2010) observed a gradual rise in sea level from 6 ka until the present day.In the central Indian Ocean, al. (2016) showed that level was relatively during the Late Holocene until the last few centuries, when there was a rapid rise.In the western Indian Ocean, Nandasena et al. (2011) observed a similar pattern of relative sea-level stability during the Late Holocene with a significant rise in the last few centuries.On the west coast and eastern coast of India, the sea level was relatively stable during the Late Holocene, with minor fluctuations (Hashimi et al., 1995;Nigam et al., 2009).Studies from various regions, including the Maldives, south-east Asia, the central and western Indian Ocean, and the eastern coast of India, suggest that sea level was relatively stable during this period, with only minor fluctuations or a gradual rise until the last few centuries (Horton et al., 2005;Woodroffe & Horton, 2005).
In general, the stable sea-level conditions during the Late Holocene broadly corresponds to a lower rate of sediment accumulation (Camoin et al., 2004;Mann et al., 2016).The Berada core also shows the lowest rates of sedimentation (Unit 5) during this period.The same pattern is observed from the central part of the GRK Basin (Kumar et al., 2021).The topmost sediment column (Unit 5) of the Berada core shows that supratidal conditions prevailed in the Banni Basin during most of the Late Holocene, indicated by a change in sand percentage (40%; Figure 4) and coarsening upward deposition along with the cross lamination present at 3 m (Figure 8H).The presence of a weak and less dynamic tidal system with a smaller tidal range in the central and northern part of the basin (Padia et al., 2022;Sharma et al., 2021;Tyagi et al., 2012) suggests a supratidal environment in the Banni Plain, which progressively led to termination of marine influenced sedimentation.
A precise age for the cessation of sedimentation in the Banni Plain could not be determined as the upper part of the core (above ca 5 m depth) did not yield sufficient datable material.Previously, ca 2 ka was generally considered as the approximate time frame for the final withdrawal of the sea from the GRK Basin (Glennie & Evans, 1976;Merh, 2005;Roy & Merh, 1981).However, recent studies from shallow pits and trenches spread all over the GRK Basin provide tentative constraints on the age of the surficial sediments (Makwana et al., 2018;Pillai et al., 2018;Sharma et al., 2021;Tyagi et al., 2012).From Vigukot, in the north-western part of the GRK Basin, an optically stimulated luminescence (OSL) age of 1.0 + 0.2 ka is obtained from sediments at a depth of 40 cm (Tyagi et al., 2012), while the sediments at 1.5 m depth in the central part are dated to 0.8 + 0.4 ka (Sharma et al., 2021).A radiocarbon age of 0.6 + 0.30 ka is reported from the sediments at a depth of 25 cm at Luna in the westernmost part of the Banni Plain, while at the same depth, an age of 1.4 + 0.02 ka is obtained from Chachi (Figure 1) in the central part of the Banni Plain (Pillai et al., 2017).Sediments at a depth of 1 m near Bhitara (Figure 1) at the north-western margin of the Banni Plain gave an age of 0.9 + 0.1 ka (Makwana et al., 2018).
Of the above mentioned locations, Chachi is located ca 10 km north of the present core site at Berada (Figure 1).All other locations are >70 km away from the core site.The distance in the GRK Basin is emphasised as the ages of near surface sediments show a pattern.It is observed that the surface sediments show progressively younger ages towards the west, north and north-west which conform to the directions of progression of the regressing shorelines during the emergence of the GRK Basin.Maurya et al. (2013) pointed out that various geomorphologic components of the GRK Basin emerged at different times due to differential uplift along subsurface structural elements.The present-day elevation and submergence characteristics of the GRK Basin suggests that the Banni Plain was the first to emerge followed by the Bet zone and the supra-tidal salt flat, which still gets regularly submerged by marine waters from the west (Maurya et al., 2013).The present morphology and geomorphological setting of the Banni Plain suggest that the shoreline regressed westward, northward and northwestward due to neotectonic activity along subsurface structural elements and the KMF, the southern basin bounding fault of the GRK Basin (Maurya et al., 2016(Maurya et al., , 2017)).In view of this pattern of emergence and proximity to the core site, the age of 1.4 + 0.02 ka obtained from Chachi at ca 25 cm depth (Pillai et al., 2017), appears to be the most plausible age for the surface of the Berada core.

| DEPOSITIONAL MODEL OF THE BANNI PLAIN
The cross-section A-B (Figure 10) shows a diagrammatic representation of the Holocene palaeoenvironmental conditions encompassing the southern margin (Berada core) and the central part (Dhordo core) of the GRK Basin.The currently active tidal zone and the coastline are located ca 80 km west of both core locations.The stratigraphy is established based on the chronological demarcation of the sediment core and the isochron lines relating to the sedimentary facies succession.The cross-sectional comparison of the Dhordo and Berada cores demonstrate the lithological variation and the radiocarbon dates corresponding to their respective depths (Figure 10).The facies succession and the marked units from the core are used to reconstruct the stratigraphy, relating sea-level rise from low stand to subsequent high stand, until its withdrawal.The evolutionary model of the palaeoenvironmental changes in the Banni Plain through the Holocene as revealed by the present study of the Berada core is shown in Figure The lithology at the base the Berada core marks the unconformity the fluvial deposit (Unit 1; Depth 50-39.6 m) during sea-level low stand (Figures 10 and 11A).Deposition of fluvial sediments during intervals of low sea level are limited to the southern part of the basin, whereas the base in the central basin (Kumar et al., 2021) shows the presence of marine influenced sub-tidal deposits.
With rising sea level, the fluvial sediments in the Berada core were overlain by marshy/estuarine (Unit 2; Depth 39.6-30 m) deposits during the Early Holocene.The peak in magnetic susceptibility (χlf) at ca 35 m depth indicates a shift to terrestrial sediment caused by the rising sea level and developing monsoon during that period.These estuarine or marshy sediment deposits are attributed to transgressive sea level (Figure 11B; Maurya et al., 2013;Oomkens, 1974;Stanley & Warne, 1994).The estuarine sediments were limited to the Berada core area at the basin margin, while the central part received sediments under sub-tidal conditions (Kumar et al., 2021).This argument also supports observations from the Berada core and further suggests a high tide limit reaching the Banni Plain at around ca 10 ka.During the high sea level of the Middle Holocene at ca 7 ka, the estuarine sediments of the Berada core were overlain by sub-tidal deposits (Unit 3; Depth 30-19 m), denoting further shoreline migration inland up to the highlands in the south (Figure 11C).It is therefore possible to infer high sea-level stillstands in the GRK Basin from ca 8 to 6 ka which led to sub-tidal deposition all over the basin.During this period, the coupled effects of sea level and enhanced monsoonal conditions resulted in highly enhanced sediment input (1.9 cm year −1 ) as demonstrated by the high peaks of environmental magnetic parameters such as S-ratio, χlf, and χARM.Upwards, both cores show a regressive facies succession of intertidal deposits (Unit 4; Depth 19-7 m) in response to the relative lowering of sea level.With further retreat of the shoreline during the Late Holocene, supra-tidal flat deposits (Unit 5; Depth 7-0 m) are found to overlie the intertidal deposits, indicating deposition under regressive conditions (Figure 11D).The progressive withdrawal of the sea since ca 4 ka is related to regional tectonic uplift reported from the Banni Plain and other parts of the Kachchh Basin (Biwas, 1974;Maurya et al., 2016Maurya et al., , 2017)).The minor peaks in magnetic data are indicative of suppressed monsoonal conditions leading to progressive termination of sediment input and establishment of present continental conditions.

| CONCLUSIONS
(i) The sediment sequence in the Berada core shows at least five major depositional environments through the Holocene that is fluvial, estuarine, sub-tidal, intertidal and supra-tidal.(ii) The Berada core, located in the southern margin of the GRK Basin suggests that basinwide marine conditions were firmly established at ca 10 ka.(iii) Marine sedimentation commenced under estuarinetidal conditions in the Early Holocene, followed by an intertidal to subtidal environment in the Middle Holocene and a regressive supratidal environment during the Late Holocene.(iv) The temporal variation in rates of sediment input correlates with variable rates of Holocene sea-level rise.In general, higher rates of sediment input are found to correspond with relatively rapid rates of sea-level rise and relatively intensive monsoons.(v) The Banni Plain witnessed developing monsoons during the Early Holocene until ca 7 ka under rapidly rising sea levels, whereas the Banni Plain during the Middle Holocene displays intensified monsoons.Aridification of the Banni Plain was initiated at ca 4 ka, a time that also marks the progressive migration of the shoreline to its present location in the west.

F
I G U R E 1 (A) Outline map of India, with red colour solid box showing the location of Kachchh.(B) Geomorphological map of the GRK Basin.Major fault lines are also shown.The locations of the Berada core (present study) and the Dhordo core (Kumar et al., 2021) are marked by solid black arrows.The location of Chachi (Pillai et al., 2017) is shown in solid red colour square box, and the location of Bithara (Makwana et al., 2018) is shown in solid yellow colour square box.Note that the Chachi location is just ca 10 km from the Berada core location.(A) and (B) The alignment of the section shown in Figure 10.(C) N-S topographic profile across the Great Rann Basin showing the elevation differences in the geomorphic divisions.Location of the Dhordo core is marked on the profile.Vertical scale is highly exaggerated.The elevation data is based on the SOI topographical maps (survey years 1960-1966).BE, Bela Island; BF, Banni Fault; IBF, Island Belt Fault; KH, Khadir Island; KMF, Kachchh Mainland Fault; NPF, Nagar Parkar Fault; PA, Pachham Island.

F
I G U R E 3 (A) Litholog of the Berada core from Banni plain with AMS radiocarbon ages.The bottom sediments from ca 50-40 m are fluvial while the section upwards of ca 40 m comprises shallow marine sediments.The side panel line graph with colour code and numeric code denotes various textural variations of marine sediments.(B) Pie chart showing distribution of the textural attributes of marine sediments in Berada core excluding the basal fluvial unit.(C) Pie chart showing distribution of the textural characteristics of the Berada core after including the basal fluvial unit.Both pie charts show dominance of very slightly sandy slightly clayey silt in the Berada core.
represents a full length view of 15 out of 17 core pipe sections acquired at Berada in the Banni Plain.Selected photographs of the core are shown in Figure4.

F
I G U R E 5 Curves showing downcore variations in various environmental magnetic parameters along with grain-size parameters.Time divisions are to the extreme right.The period of Holocene climate optimum between ca 8-6 ka and ca 5-4 ka are represented with grey shading and pink shading, respectively.Medieval Warm Period (ca 2-1 ka) and the Little Ice Age (ca 1-0.5 ka) are shown in blue and yellow, respectively.F I G U R E 6 (A) Linear Discrimination Function (LDF) binary plot of Y3 versus Y4 of sediments from the Berada core showing different energy and depositional conditions (after Sahu, 1964).(B) LDF binary plot of Y1 versus Y2 (after Sahu, 1964).(C) Graphical representation of mean, sorting and skewness along with depth.Short black arrows along the curve of mean graph shows the general trend of the mean grain-size variation.T A B L E 2 Summary of down core variations of grain-size characteristics and interpreted depositional environments of the Berada core.Textural classes of the dominantly fine-grained sediments from Units 2-5 are based on Blott and Pye (2012).

F
Selected photographs and x-ray radiographs of the core pipe obtained from Berada cores at various depth.(A) Fluvial; Unit 1 (44.3-43.2m): hard compact brownish oxidised clay with laminations.(B) Estuarine; Unit 2 (39.6-39.1 m): peat layer within clayey sediments having sharp contact with the underling homogenous undisturbed sediment deposit.(C) Estuarine, Unit 2 (39.1-38.1 m): increasing upwards mud and sand intercalation.(D) Estuarine, Unit 2 (35.8-35.6 m): extensive burrowing activity with scattered organic matter.(E) Sub-tidal flat; Unit 3 (28-25.60m): x-represents lens shaped organic matter, note the upwards pointing arrow represent the increasing sand mud couplets; lenticular to wavy structure with increase in sand content.(F) Intertidal flat; Unit 4 (17.6-15.8m): high mud content, dotted line separates the highly bioturbated sedimentary unit overlying homogenous deposition, high mud content.(G) Supra tidal; Unit 5 (4.75-6.50m).Colour alternation of greenish grey to brown clayey sand, dotted line is drawn to represent the lithological colour variation.(H) Supra-tidal; Unit 5 (1.75-2.30m); Brown sandy clay with poorly laminated sediments.Late Holocene, is reflected in the uppermost part of both cores.Extrapolation of the 1.4 + 0.02 ka age determined for surface sediments in a nearby trench at Chachi (Figure1;Pillai et al., 2017) to the Berada core site yields sedimentation rates of 0.09 cm year −1 .This compares well with the rate of 0.14 cm year −1 in the Dhordo core (depth ca 9-0 m) in the central part of the GRK Basin, reported byKumar et al. (2021).

F
Litholog of Berada core with sand content variation and inferred depositional environments.

S. no Sample name Material dated Depth range (cm) Laboratory name Lab code pMC value Radio-carbon age (years) Calibrated age range (1 Sigma) years bp
Table showing 14 C AMS radiocarbon ages obtained at various depths from Berada core.