Variable monsoons and human adaptations: Archaeological and palaeoenvironmental records during the last 1400 years in north-western India

Abstract

We present the first systematic evaluation of the relationship between the archaeological and palaeoclimatic record from north-western India during the past millennium, from the urban site of Chandravati. The rarity of Medieval sites, systematic excavations and multi-disciplinary work in the subcontinent obscure the impact of two distinct climate anomalies − the ‘Medieval Warm Period’ (‘MWP’, 740 − 1150 CE), followed by the ‘Little Ice Age’ (‘LIA’, 1350 − 1850 CE). The finds from the archaeological site indicate the presence of winter and summer crops, suggesting the region was likely warm and mild humid during pre-Medieval period (ca. 600 − 800 CE). During Medieval times (between ca. 800 − 1300 CE), a diversification of the crop assemblage suggests that the region was under a warm and humid climate, corresponding to the ‘MWP’, driving increased monsoon precipitation. During the post-Medieval period (ca. 1350 − 1800 CE), drought-resistant millets and other summer pulse crops indicate the region probably experienced weak SW monsoon precipitation coinciding with globally recognised ‘LIA’. These interpretations are supported through phytolith data from the archaeological deposit broadly indicating two phases, the first being a period of diversified agricultural/anthropogenic activity (ca. 600 − 1350 CE), followed by a period dominated by drought-resistant crops (ca. 1350 − 1800 CE). Pollen data from a proximal lake corroborate the warm and humid phase ca. 800 − 1400 CE, with strong representation of warm−humid favouring tropical forest taxa, followed by non-arboreal indicators of a drier more open landscape ca. 1500 − 1800 CE. These environmental changes may have combined with other historic and institutional factors that led to the ultimate abandonment of the city. These changing cropping patterns, vegetation and cultural developments provide insight into past human response to climate change as well as important lessons for modern societies in exploring sustainable agricultural strategies to future climate change.

Keywords

agriculture Chandravati Little Ice Age Medieval Warm Period north-west India palaeoclimate past millennium

Introduction

Examining economic and environmental changes in the archaeological and palaeoclimatic records provide some of the most compelling insights into the growth, collapse or resilience of large urban agglomerations (Bar-Oz et al., 2019; Cookson et al., 2019; Penny et al., 2019). In South Asia, studies of this nature have tended to focus on fluctuations of the Indian Summer Monsoon (ISM) as a driver of the urbanisation and de-urbanisation of the prehistoric Harappan Civilisation in the northern subcontinent, ca. 2500–1700 BCE (Dixit et al., 2018; Giosan et al., 2012, 2018; Prasad et al., 2014) Building on the works of Weber (1999) and Madella and Fuller (2006), archaeobotanical studies characterised variations in economic plant assemblages as indicative of strategies aimed to mitigate the impacts of climate change, including crop diversification or shifting patterns of summer/winter cultivation contextualised against changes in material culture or settlement pattern (Petrie et al., 2017; Pokharia et al., 2017). The entanglement of ecological, economic and cultural factors indicate that for ongoing social stability to be maintained, strategies for resilience to environmental stresses are crucial.

The impacts of historic period climate events in South Asia, such as the ‘Medieval Warm Period’ (‘MWP’) and ‘Little Ice Age’ (‘LIA’) during are relatively understudied, as research on social and political change has tended to focus on the fragmentation of centralised Indian Dynasties, invasions from Central and West Asia and the eventual integration of the Mughal Empire (Chandra, 2003).

Here we examine the growth, expansion and eventual abandonment of the urban site of Chandravati in Rajasthan, NW India. We present new on-site archaeological data as well as proximal environmental data to link changing local environmental conditions to global climate events, namely, the ‘MWP’ and ‘LIA’. We consider the impacts of historic climate events such as the ‘MWP’ and ‘LIA’ on the ISM. Climate variations are compared with social and economic changes in the region, and provide context for understanding the ways in which resilience of the population was maintained through economic strategies such as selective seasonal cultivation or crop diversification.

Timing and impacts of the ‘MWP’ and ‘LIA’ in Asia

Early syntheses of historical accounts and climate proxies have tended to focus on Europe, allowing the identification of a period of warming between the 10th and 13th centuries, followed by a cooling anomaly of up to 1°C until the 19th century (Lamb, 1965). Though the timing of the onset of cooling conditions is still debated, this general chronology has allowed for archaeologists and historians to consider economic and social changes in Europe as a response to climate cycles (Aberth, 2012; Fagan, 2000). More recently, studies have focussed on the timing and impact of the ‘MWP’ and ‘LIA’ in Asia, in particular, the spatial distribution and forcing factors involved of fluctuations in the East Asian Summer Monsoon (Chen et al., 2015; Kamae et al., 2017). The onset of the ‘LIA’ in China from the mid-16th century has been linked to famines, north–south migrations and social unrest that had severe implications for the stability of the Ming and Qing Dynasties (Wu et al., 2018; Xiao et al., 2015).

In northwest India, Agnihotri et al. (2002) correlated fluctuating biogenic carbon and nitrogen discharge into the Arabian Sea against insolation variation, finding a relationship between increased surface productivity and higher summer insolation. These trends indicated a general increase between 900 and 1400 CE, followed by a decline interpreted as the onset of the ‘LIA’. Drawing on this study, Uberoi (2012) has speculated on the impact of solar forcing on the decline of the Mughals, in particular famines during the reign of emperors Akbar (1556–1605), Shah Jahan (1628–1659) and Aurangzeb (1659–1707) correlating closely with the Maunder Minimum (but see Berkelhammer et al., 2010 regarding linkages between insolation and ISM fluctuation).

In a recent review of changing hydroclimatic conditions in South Asia during the last 1000 years, Dixit and Tandon (2016) have synthesised multiple tree ring, lacustrine, coastal and speleothem records for a detailed reconstruction of precipitation regime changes relating to both the ISM and the Westerly Disturbances. While these records tend to show a major shift in effective moisture availability between the ‘MWP’ and ‘LIA’, the most pertinent studies for examining ISM shifts in the arid zones of north-western India are Arabian Sea, coastal and marine cores (Agnihotri et al., 2002; Von Rad et al., 1999), lake records from the Thar Desert (Prasad et al., 1997), as well as pollen studies from Madhya Pradesh (Chauhan and Quamar, 2012; Quamar and Chauhan, 2014). Highly resolved speleothem records (e.g. Kathayat et al., 2017) prove data for fluctuating regional rainfall patterns, which when compared with other proxies may allow insights into localised ecological, cultural or economic responses. In general, local-scale environmental studies have tended to not focus on the ISM during the recent Holocene, and the timing and impacts of the ‘MWP’ and ‘LIA’ are only briefly discussed (Rawat et al., 2015a, 2015b; Trivedi and Chauhan, 2009).

Site location and context

Rajasthan is the largest and westernmost state in modern India and can be divided into four physiographic regions: the Thar Desert in the west; the Aravalli hills running from the SW to NE; the eastern alluvial plains and the southeast plateau (Swain et al., 2012). The state is more arid than other regions of India, with the Thar Desert receiving the extreme lowest rainfall in modern India (below 100 mm), and the rest of the state receiving less than 600 mm precipitation per year (Singhvi and Krishnan, 2014). Modern surface observation data from the 100 years between 1901 and 2016 suggest an average summer peak of 40°C in May, prior to the onset of the ISM, and a low of around 8 to 9°C during the post-monsoonal winter in December to January (Indian Meteorological Department (IMD), 2018). Agriculture is primarily rain-fed (Gupta, 2016) and major cereal crops are pearl millet, wheat, maize and sorghum. Chickpea, moth and mung beans are the main pulses cultivated in the region. These are typically grown in a bi-seasonal system, with wheat and chickpea as winter/rabi crops, and millets and other pulses as summer/kharif types (Swain et al., 2012).

Chandravati (24°26′17.8″N, 72°44′32.2″E) is located in the foothill zone of Mt Abu in the southern Sirohi district of Rajasthan, among the Aravalli hills (Figure 1). The ancient settlement is situated close to the confluence of the Sevarni and Banas rivers. The larger Banas is oriented roughly north–south with fertile terrain on both banks, entering Gujarat at Mawal about 4 km downstream of Chandravati. Both rivers are comparatively shallow and cut into underlying bedrock. The Sevarni has a flood plain about 300 m broad at river-bends, above which are undulating plains. Between the two rivers, one of these plains forms a sub-triangular area with its base towards the mountains and apex towards the Banas, suitable for human settlement.

Figure 1.

Location of Chandravati and Mandovari Lake in Sirohi district, Rajasthan.

Chandravati was the capital of the Abu branch of the Parmar dynasty and the major city of the Sevarni–Banas area, possibly spread up to 50 ha in size. Believed to be founded around the 6th century, the site reached its apogee during the 10th century under the Parmar rule of Sindhuraja. Being the major metropolis of the region, Chandravati was subject to multiple attacks by the Ghaznavids in the 11th century, and was previously believed to be abandoned following raids by the Mughals in the 15th century. The site was excavated between 2014 and 2017 by Professor JS Kharakwal, in collaboration with Rajasthan State Archaeology Department (Kharakwal et al., 2016). In addition to these excavations, an environmental trench was dug to extract pollen and sediment size data from the now dry Mandovari Lake bed, 4 km east of the site.

Materials and methods

Archaeological deposit: Botanical macroremains and phytoliths

Macroremains

Sediment from cultural layers, pits, floor levels and so on was collected systematically and subjected to wet flotation during the course of the excavation. All buoyant carbonised material was passed through a 30-grade mesh (0.5 mm) geological sieve. In total 2235 L of sediment was floated. Plant remains were identified using reference specimens at the BSIP Herbarium. Absolute counts of plant taxa, ubiquity (Popper, 1988) and diversity index (DI) or Shannon–Weaver (Pearsall, 2000) were used to analyse the data (Supplementary Table 1, available online). Relative abundances of economic crops were standardised as densities (following Weber, 1999) across all phases by dividing the total number of samples by the volume of sediment floated (Supplementary Table 2, available online). Calorific contributions of cereal crops were undertaken following methods described by Jennifer Bates (2016; Bates et al., 2017; details in Supplementary Table 2, available online).

Phytoliths

Samples from archaeological trench P35 were collected for phytolith analysis at regular depths measured from the top of the trench. Samples were dried and phytoliths were extracted following standard techniques in Piperno (2006) and Prasad et al. (2007). Ten grams of the dried sediment were treated with 10% hydrochloric acid (HCl) and 20% hydrogen peroxide (H₂O₂) to remove carbonate and organic content. For heavy liquid flotation, a solution of cadmium iodide and potassium iodide (at specific gravity 2.3) was used. At least, 350 to 400 phytoliths were counted from each sample and phytoliths were grouped using standard classifications (e.g. Fredlund and Tieszen, 1994; Mulholland and Rapp, 1992; Twiss, 1992; Twiss et al., 1969). The phytolith types were classified according to International Code for Phytolith Nomenclature (Madella et al., 2005). Cluster analysis was undertaken using CONISS in Tilia software (Grimm, 1987). A total of 18 phytolith morphotypes were identified, including grass and non-grass phytoliths.

Lacustrine deposit: Pollen and sediment size

Pollen analysis

A 1.90 m deep trench was dug into the profile of Mandovari Lake, a natural water body modified into a reservoir during the Medieval period and now desiccated. Thirty-eight samples (150 g) were collected at 5 cm intervals for pollen analysis. Extraction of pollen generally followed protocols outlined in Faegri and Iverson (1975). Ten grams of the sample were boiled with 10% aqueous KOH (potassium hydroxide) solution to deflocculate the sample and remove humic colloids. This was followed by treatment with 40% HF solution to digest the silica content. Acetolysis was undertaken a 9:1 mixture of acetic anhydride and concentrated sulphuric acid. Samples were mounted in 50% glycerin solution for microscopic examination. Identification of the recovered pollen and spores was made through consultation of reference pollen slides available at the BSIP Herbarium as well as photographs and descriptions in the published literature (Chauhan and Bera, 1990; Nayar, 1990) and a target minimum of 100 palynomorphs were counted. Relative abundances were plotted in Tilia 2.0.41 and pollen zones identified on total sum of squares of absolute abundances (Grimm, 1987).

Sediment particle size

Sampling of lake sediments took place at 3 to 5 cm intervals. Carbonates were dissolved with 10% HCl and organic debris was oxidised with 20% H₂O₂ Samples were processed using a CILAS 1190 Laser Particle Size Analyzer. Samples were transferred to 50 ml centrifuge tubes and sonicated using an ultrasonic vibrator at 600 rpm, then 0.5 mg of sample was transferred to the Laser Particle Size Analyzer for sediment distribution reading. Output data were analysed using the Gradistat v. 8.0 macro for Microsoft Excel (Blott, 2010; Supplementary Figure 1, available online).

Radiocarbon dating

Radiocarbon measurements on charcoal from all archaeological phases were carried out using liquid scintillation counting (LSC) at Birbal Sahni Institute of Palaeosciences (lab code: BS), Lucknow, India (Rajagopalan et al., 1978; Table 1). In addition, four accelerator mass spectrometry (AMS) dates were returned from carbonised seeds/grains, dated at Direct AMS (lab code D-AMS), Seattle, USA (Table 1). Though returned dates were acquired from the different trenches across the site, each of these trenches represented discrete archaeological phases, therefore, their relative relationship allowed for the construction of a contiguous Bayesian sequence using OxCal 4.3.2 (Ramsey, 2017; Reimer et al., 2013; Supplementary Table 3, available online).

Table 1.

AMS and LSC ¹⁴C radiocarbon dates of wood charcoal and carbonised grains/seeds from the archaeological site Chandravati, dated at Birbal Sahni Institute of Palaeosciences (lab code: BS), Lucknow, India and Direct AMS, Bothell, (lab code: D-AMS), USA (68.2% probability).

Archaeological provenience (context/depth from datum point in cm)	Sample type	Lab ref.	¹⁴C date (yrs. BP)	Calibrated age (CE) (1σ probability)	Cultural phase
I9/3/595	Charcoal	BS-3817	350 ± 110	1447–1645	Phase III
ZRR68/1/664	Carbonised seed	D-AMS 031201	545 ± 26	1396–1422
I9/4/635	Charcoal	BS-3816	590 ± 110	1291–1422
K13/1/291	Carbonised seed	D-AMS 032378	637 ± 23	1281–1404
LL18/2/244	Carbonised grain	D-AMS 031202	700 ± 27	1273–1295	Phase II
P35/5/750	Carbonised seed	D-AMS 031203	837 ± 27	1169–1224
P35/8/830	Carbonised grain	D-AMS 031204	913 ± 30	1045–1097
I9/5/663	Charcoal	BS-3815	1110 ± 190	783–1018	Phase I
DD19/2/615	Charcoal	BS-3900	1510 ± 70	530–619
ZA14/3/570	Charcoal	BS-3820	1620 ± 170	357–539

AMS: accelerator mass spectrometry; BP: before present; CE: common era.

Three bulk sediment samples from the Mandovari Lake profile were also AMS dated at the Gliwice Radiocarbon Laboratory (lab code: GdA), Poland (Table 2). A Bayesian age–depth model based on the three dates returned from the lake profile was produced using the Bacon package in R (Blaauw and Christen, 2011; Supplementary Figure 2, available online).

Table 2.

AMS radiocarbon dating of sediment samples from Mandovari Lake (68.2% probability).

Sample ID	Depth below surface (cm)	Lab ref.Gliwice radiocarbon laboratory (GdA), Poland	Sample type	Radiocarbon age BP	Calendar age CE range (1 σ probability)
ML/103-106	102	GdA-4811	Sediment	319 ± 20	1631–1670
ML/151-154	152	GdA-4812	Sediment	1576 ± 53	321–428
ML/187-190	190	GdA-4813	Sediment	1231 ± 61	658–729

AMS: accelerator mass spectrometry; ID: identification; BP: before present; CE: common era.

Atmospheric circulation and precipitation

Data were drawn from the CMIP5 project, REMO–ESM model output from the Earth System Grid Federation (ESGF) data portal. Calculations for circulation change through time were modelled using Climate Data Operator (CDO) and visualised though the Grid Analysis and Display System (GrADS). Circulation was computed for periods 850 to 1350 CE and 1400 to 1850 CE as annual, summer (JJAS) and winter (DJF) data (Supplementary Figures 3 and 4, available online).

Results

Contrary to past assumptions that Chandravati was abandoned following attacks by the Mughals, archaeological excavation and radiocarbon dates indicate that the site was inhabited up to the late 17th century. Collapsed walls from the final archaeological phase indicate that the site was likely deserted following a major earthquake (Kharakwal et al., 2016). Reports from the Gazetteer of the Bombay Presidency (Government of Bombay, 1874) report that by 1824, the site was virtually abandoned, providing a terminal date of occupation by ca. 1800. Descriptions of archaeological phases, material culture and architecture have been previously reported (Kharakwal et al., 2016) and based on these as well as new AMS dates (Table 1), as well as known historical and epigraphic evidence (Kharakwal et al., 2016), a sequenced chronology may be presented as a Pre-Medieval Phase I (ca. 600–800 CE), Medieval Phase II (ca. 800–1300 CE) and Post-Medieval Phase III (ca. 1300–1800 CE). Charred macrobotanical recovered from all phases and contexts across the site, and phytoliths from all phases in archaeological trench P35 indicate major changes in plant economy and ecology at Chandravati.

Macrobotanical remains: Large-grained cereals versus small-grained millets

In total, 24,357 plant remains were recorded and identified, including cultivated crops, agricultural weeds, as well as wild species. In general, the cultivated plants include both winter growing crops (Hordeum vulgare, Triticum aestivum/durum, Pisum sativum, Cicer arietinum, Lens culinaris, Lathyrus sativus and Linum usitatissimum) and summer crops (Figure 2) (Oryza sativa, Sorghum bicolor, Pennisetum glaucum, Panicum miliaceum, Setaria italica, Setaria sp., Brachiaria sp., Echinochloa sp., Vigna unguiculata, Vigna radiata/mungo, Macrotyloma uniflorum, Sesamum indicum and Gossypium arboreum/herbaceum; Supplementary Table 1; Supplementary Figures 5–7, available online).

Figure 2.

Botanical macroremains from Chandravati: (a) Hordeum vulgare and (b) Macrotyloma uniflorum (Phase I), (c) Triticum cf. aestivum and (d) Oryza sativa (Phase II), (e) Panicum sp., (f) Sorghum bicolour and (g) Setaria sp. (Phase III). Scale bars = 1 mm.

In Pre-Medieval Phase I (600–800 CE), winter crops compose 15% of the total crop assemblage, dominated by barley (Hordeum vulgare 10%). Summer season crops (ca. 85% of assemblage) are primarily pulses (Vigna radiata 59%), with cereals in proportions under 10% (Oryza sativa 2%), Setaria cf. italica 7%, Panicum miliaceum 1%). Large-grained cereals accounts for 59% in comparison with small-grained cereals (41%).

During Medieval Phase II (800 − 1300 CE), the assemblage diversifies significantly (DI = 1.00; Supplementary Table 1, available online). Winter crops comprise for 34% of the assemblage, and summer crops accounts for 67%. Barley (n = 88) and rice (n = 13) are the numerically dominant cereals. In all, large-grained cereals accounts for 86%, whereas small-grained cereals accounts for 14%.

Post-Medieval Phase III (1300 − 1800 CE) sees general crop continuity with earlier phases, with the notable disappearance of Oryza sativa (rice). The abundance of winter crops is meagre (2%) in comparison with summer season crops which account for 98%. Among the summer season crops, the drought-resistant millets dominate (72%).

Across these phases, calculation of the calorific contribution of cereals in a botanical assemblage indicates a significant dietary shift towards small-grained summer millets during cultural Phase III (Figure 3). While these cereals dominate the assemblage numerically during this phase, large-grained winter barley and summer rice still contribute a significant portion of calories in the assemblage, suggesting that despite changing environmental conditions, economic diversity could be maintained during this period. The density of summer pulses in the archaeological assemblage during this phase along with significant densities of winter chickpea supports the notion of a balanced summer/winter economy during archaeological Phase III (Supplementary Table 2, available online).

Figure 3.

Cereal proportions from Chandravati. (a) Relative abundances of large/small grain types and (b) relative calorific contributions of all types.

Phytolith data

Grass silica short cells (GSSCs) included bilobate, cross, saddle, trapezoid and rondel types. Other grass morphotypes include elongate dendritic and elongate psilate type and bulliform phytoliths. Multi-celled grass panels were present in most of the samples, along with crop and cut phytoliths. Family/species-specific phytoliths types include achene phytoliths from Cyperaceae, cuneiform bulliforms and double-peaked glume cells from Oryza sp., Hordeum sp. and Triticum sp. Other phytolith types recovered are trichomes and cylindric-sulcate (tracheids) type. Cluster analysis based on frequency distribution identified two zones, Zone I broadly corresponds with cultural Phases I and II; Zone II corresponds with cultural Phase III (Figure 4).

Figure 4.

Phytolith spectra from archaeological trench P35.

Zone I (cultural Phases I and II: 600–1300 CE)

This zone shows a dominance of GSSC phytoliths, ranging from 50% to 65%. Among GSSCs, the percentage of bilobate phytoliths is highest (29–41%) followed by saddle (7–11%), trapezoid (4–7%), rondel (4–6.5%) and cross (3–7%). While long cell phytoliths account for 2% to 8% and 7% to 11% for elongate dendritic and elongate psilates, respectively. Bulliform phytoliths account for 3% to 8%, trichomes range from 2.5% to 5%. Crop phytoliths recovered include millets (2–4%), wheat (2.5–5%), barley (2–5.5%) and rice husk (3–4.5%). Cyperaceae phytoliths range from 1% to 4%.

Zone II (cultural Phase III: 1300–1800 CE)

Zone II continues to be dominated by GSSCs (54–67%), comprising bilobates (19–35%) saddle (11–19%), trapezoid (7–11%), rondel (4.5–8%) and cross (2–4%) types. The percentage of bulliform phytoliths range from 3% to 8% as in Zone I, while trichomes percentage was higher than earlier ranging from 1% to 9%. Multi-celled grass panels, cut phytoliths and crop phytoliths were also retrieved. Crop phytoliths include millets (2–6%), wheat (1.5–4.5%), barley (2–7%) and rice husk (1.5–2.5%). Cyperaceae phytoliths are almost completely absent in this zone (Figure 5).

Figure 5.

Pollen spectrum of top 95% of occurring taxa in Mandovari Lake.

Pollen and sediment results

Total pollen counts ranged between 81 and 302 (mean 154 ± 56). Cluster analysis on absolute pollen abundances indicated two pollen zones between 187 and 115 cm and 115 and 0 cm. Relative abundances of the top 95% of pollens are presented in Figure 5. Of the three dated bulk sediment samples, GdA-4812 appears to be an inversion (Table 2); however, this was left in the age–depth model, as Bacon is able to deal with inversions by presenting a wider range of uncertainty. At the base of the pollen record, uncertainty range is ± ∼200 years, however, this range narrows to ± ∼100 years close at the boundary of the two pollen zones, allowing the timing of this transition to be interpreted with a good degree of certainty.

Pollen Zone I (187–115 cm, ca. 800–1500 CE)

Arboreal taxa are well represented by average abundances of Bombax (5%), Lagerstroemia (4%) and Sapotaceae (5%). Total arboreal pollens comprised an average 20 ± 10% during this period, while shrubs are dominated by Croton (mean 15 ± 7%). Non-arboreals comprised an average 62 ± 17%, dominated by Tubuliflorae (48%) and Liliaceae (4%). In general, sediment size data from this zone are dominated by medium and fine silts 80–90%, with fine sand ranges between 2% and 7% and clay between 8% and 15%. Sediments in this zone are generally classified as poorly sorted coarse silts (Supplementary Figure 6, available online).

Pollen Zone II (115–0 cm, ca. 1500–1850 CE)

Tropical forest type pollens decline sharply in this period, with Sapotaceae (4 ± 4%) being the only type with average representation above 2%. Croton also falls to an average under 5%. The pollen spectrum is dominated by herbaceous taxa, namely, Tubuliflorae types, Liliaceae and Artemisia. Slight increases in Cerealia and anthropogenic weeds, such as Apiaceae, Brassicaceae, Amaranthaceae and Caryophyllaceae may indicate more intensive land-use. Sediments in this zone are poorly sorted medium silts and outside of the top 30 cm of the trench, the sand fraction is generally negligible. Clay proportions are also low with the exception of some fluctuations at 75 cm and 25 cm.

Atmospheric data

Figure 6 presents atmospheric circulation and modelled summer monsoon (JJAS) and winter (DJF) precipitation delivery for the post-Medieval period, derived from long-term modelling for the periods 800 to 1350 CE and 1400 to 1800 CE (Supplementary Figures 3 and 4, available online).

Figure 6.

(i & ii) Annual, (iii & iv) summer and (v & vi) winter circulation and precipitation 1400 − 1800 CE.

The study region appears to be controlled by monsoonal circulation with higher summer precipitation. During the post-Medieval period (Figure 6), modelling over the study area indicates the site would have received winter and summer precipitation below the long term average.

Discussion

Climate change: ‘MWP’ and ‘LIA’

The variability of South Asian climate records during the last two millennia suggests that past climate shifts and their environmental and socio-economic responses may be best studied at a local scale. The archaeological, macro- and micro-botanical data from archaeological site Chandravati and Mandovari Lake present a good opportunity for reconstructing these processes. A synthesis of these data seem to indicate two broad environmental phases around the Aravalli hills. The first of these periods spans ca. 600 to 1350 CE comprising both archaeological Phases I and II as well as Pollen Zone I. Though Pollen Zone I spans only the period ca. 800–1500 CE, inferences from phytolith and macrobotanical data from Chandravati allow for climate inferences covering the first 200 years of this period. Tropical forest taxa such as Bombax, Sapotaceae and Lagerstroemia, Croton shrubs, and ferns are well represented between 800 and 1500 CE. The higher representation of these taxa is dependent on higher available moisture in the form of precipitation, and is therefore interpreted as a period of warm−humid conditions and strengthened ISM delivery (Chauhan and Bera, 1990; Nayar, 1990). Higher precipitation may have also been a driver of coarser sediment influx in the particle size data. The archaeobotanical assemblage of the first period comprises primarily winter cereals, which are generally dependent on summer precipitation, and the presence of Cyperaceae phytoliths along with a higher proportion of Panicoideae grass phytoliths would also suggest a warm and mesic climate in this phase.

The second pollen zone is characterised by a reduction in tropical forest, shrubs and ferns between ca. 14500 and 1850 CE and may be indicative of a weakened summer monsoon precipitation around Chandravati. The dominance of non-arboreal taxa along with scattered trees suggests warm but relatively dry climatic conditions and the expansion of taxa such as Liliaceae may be indicative of a more open, dry environment. We interpret low rainfall as driving the transformation of forest vegetation to open grassland vegetation. The almost total disappearance of sand in the sediment data may also be an indicator of weakened precipitation. The increase in saddle type phytolith, representative of Chloridoideae grasses also indicates a drier climate after ca. 1300 CE. In addition, the decrease in Cyperaceae phytoliths is evidence for lower available moisture locally.

The above proxies suggest two broad environmental phases, a warm−humid phase with higher precipitation between ca. 600 and 1300 CE and a warm arid phase from ca. 1400 to 1800 CE with a transmission period of around one century (Figure 7). This interpretation is consistent with the CMIP5 REMO–ESM circulation modelling, suggesting a linkage between local conditions, ISM delivery and climate teleconnections. In their review of climate and precipitation change over the last millennia, Dixit and Tandon (2016) note that there is good coherence across all speleothem records in South Asia, while tree ring records provide the only reliable rainfall reconstruction data for winter precipitation over the Himalaya. As a result, we compare our interpretations with the most recent speleothem record from the central ISM zone in the Himalayan foothills (Kathayat et al., 2017) and a recent reconstruction of boreal precipitation from Western Himalayan tree rings (Yadava et al., 2016; Figure 8). Oxygen isotope data from the Sahiya Cave speleothem (Kathayat et al., 2017) suggests a period of increasing summer precipitation to ca. 1200 CE, followed by a long-term drying trend and arid event ca. 1600 CE. Yadava et al. (2016) also detect an arid event over the Western Himalaya ca. 1500 CE, suggesting that both climate systems experienced weakened precipitation related to the ‘LIA’. The data in this study suggest a transition beginning ca. 1300 CE, with the onset of fully arid conditions by ca. 1500 CE. Drier conditions from 1300 CE have been linked to numerous famines across the Indian subcontinent (Dixit and Tandon, 2016) and were likely a driver of the shift towards aridity tolerant millet agriculture at Chandravati. Arid events after ca. 1500 CE have been linked to severe famines, as well as the abandonment of the Fatehpur Sikri as the Mughal capital due to the failure of water supply (Uberoi, 2012). Similar conditions may have precipitated the decline of the urban settlement at Chandravati, despite technological attempts to maintain resilience.

Figure 7.

Environmental data. (a) Summary phytolith data from archaeological contexts, (b) summary pollen data and (c) sediment particle size data from Mandovari Lake.

Figure 8.

Comparison of (a) combined study data with (b) tree ring proxy for Westerly precipitation in Himalayas (Yadava et al., 2016) and (c) speleothem δ18O proxy for ISM precipitation, Sahiya Cave (Kathayat et al., 2017).

Agriculture and archaeology: Settlement expansion and decline

During the Pre-Medieval phase (600–800 CE), areas of Rajasthan may have been integrated into a number of shifting statelets, often of Central Asian nomadic origin, including Indo-Scythians, Kushan–Sassanians or Hunas, however, much of the region was stabilised under the expansion of the Gupta Empire from the Ganges Basin to most of the northern Indian subcontinent, considered to be the first administrative empire ‘proper’ on the subcontinent since the Mauryans. Though historians such as Chattopadhyaya (1994) caution against historiography conflating stability of state and economy during this period, the rise of large administrative institutions seems to suggest the ability to maintain an agricultural surplus. While the relationship between Chandravati and these historical–political entities are unclear, the thickness of the archaeological deposit during this phase suggests a period of stability at the site and the presence of quality finished metal goods and lapis lazuli beads may indicate the development of a secondary economy and long-distance trade (Kharakwal et al., 2016).

Cereal agriculture at and around Chandravati during this period mainly consists of rabi barley, followed by wheat. As well as dominating the botanical assemblage numerically, these two crops also contribute 91% of the calorific value of the botanical remains from the site, with Echinochloa and foxtail millets contributing 4% each and broomcorn millet the remaining 1%. During this period, Chakravarti (2008) describes the development and proliferation of the iron plough, as well as improvements in manuring techniques as key to the expansion of agriculture across northern India. As rabi crops are generally sown after the monsoon, the development of these technologies may have been important for efficient preparation of wheat and barley fields around Chandravati during this period.

The expansion of the settlement at Chandravati during cultural Phase II is contemporary with the consolidation of Rajput control over Rajasthan and Gujarat between the 10th and 14th centuries (Chandra, 2003). The construction of three forts at Chandravati during this period (Kharakwal et al., 2016) seems to be consistent with Rajput building and land redistribution programmes (Chattopadhyaya, 1994). Social and technological changes during this time may have also impacted the production of food around Chandravati. Chakravarti (2008) notes the proliferation of epigraphic mentions of vapi (step wells) and the newer technology of araghatta (bucket carrying water wheels) across the arid areas of Rajasthan and Gujarat from ca. 1000 CE. These technologies as well as the expansion of stable state institutions for the storage and distribution of water allowed for the development of intensive bi-seasonal agriculture during this period, also attested to in local records as well as Persian, Arabic and Chinese accounts (Chakravarti, 2008).

The conversion of Mandovari Lake into an artificial reservoir during this period may be evidence for the development of these water management systems at Chandravati, a construction consistent with periods of agricultural intensification and diversification, as well as expansions of bureaucratic power and financial endowment (Bauer and Morrison, 2016; Morrison, 1993). Higher precipitation during the ‘MWP’ increased agricultural productivity, providing a surplus which in turn may have allowed state institutions to implement further systems of water management and distribution. The rise of the thakkuras, a class of feudal land managers under the Rajputs, during this period may have also been related to these changes (Upadhyaya, 2016).

Whether as a result of climate or socio-technological circumstances, the botanical remains from cultural Phase II at Chandravati have the highest DI values for winter and summer cereals, as well as summer pulses. In addition, winter barley provides its largest calorific contribution (84%) to diet of all three phases at the site while large-grained cereals in total make up 87% of the assemblage. Notably, summer rice makes up a significant part of these crops and contributes 10% of the calorific content. Unlike summer millets, rice has significantly higher water and labour requirements, indicating a period of both favourable moisture availability as well as available resources for paddy preparation and cultivation. This appears to be consistent with descriptions of the expansion of rice agriculture in the region described in historical sources (Chakravarti, 2008).

Diminished rainfall during Phase III at the site seems to have led to major changes in agricultural production. For the first time, the assemblage is dominated by small-grained summer millets, which are generally very hardy crops adapted to wide range of temperatures and moisture regimes (Pokharia et al., 2014). These cereals are an important component of dryland agriculture and can be grown on wide range of soils and harsh climatic conditions with low rainfall. The tolerances made this crop an important food in the past and also in the present era of erratic climate, global warming and reduced water availability. The reduction in rice, the other major summer cereal suggests diminished water availability for paddy cultivation. Of the two large-grained winter cereals, absolute counts of wheat are almost negligible, while the more aridity tolerant barley is better represented. Analysis of calorific values of these crops also shows that barley still makes a significant dietary contribution (15%) despite relatively low numerical representation. Foxtail millet, previously making up no statistically significant proportion of food energy now dominates at 75%, while rice (4%) and other millets (1–3%) make up the remaining fraction.

During this period, the archaeobotanical and archaeological evidence from the site indicates that economic and social resilience was maintained in part through the selective shifting of cropping strategies despite climate deterioration. Drawing on late historic and 20th century historical and ethnographic accounts from the Aravalli hills, Rosin (1994) provides descriptions of the way in which shifting patterns of short season, rain-fed millet agriculture have supported various patterns of transhumant pastoralism, particularly around the Mt Abu region. These agro-pastoral strategies may have allowed for resilience in food production, however, they would have shifted the socio-economic focus around Chandravati away from the urban centre and onto the hinterland.

By cultural Phase III, Chandravati was no longer a major capital city, though continued occupation at the site and the maintenance of several monumental structures (Kharakwal et al., 2016) suggest that the site was still an important centre for the region. From ca. 1300 CE onwards, incursions by nomadic confederations from Central Asia into South Asia took place with increasing frequency, possibly because of failing Westerly precipitation disrupting traditional agro-pastoralism in Central Asia (Yadava et al., 2016). The continuity of occupation and productive economy in the archaeological deposit indicates that the site was not abandoned as a result of these conflicts ca. 1600 CE, as previously believed (Kharakawal et al., 2016). Ongoing occupation was supported by technological adaptation in the form of shifting cultivation; however, it is apparent that a combination of climate stress, conflict and the diversion of resources away from Chandravati by elites would have contributed to a protracted decline of the settlement. The entanglement of institutional, environmental and economic factors meant that following a calamity such as the 17th-century earthquake, rehabilitation of the city was not feasible, leading to wholesome abandonment of Chandravati as an urban centre.

Conclusion

We have provided new evidence for warm−humid and warm arid environmental phases in northwest India, ca. 800 to 1300 CE and 1400 to 1800 CE, respectively, and linked these to the ‘MWP’ and ‘LIA’ global climate stages. From the archaeological evidence, these environmental changes had a major impact on agricultural productivity and the development of social institutions at the urban site of Chandravati. The ultimate abandonment of the settlement appears to be because of a single calamity, however, this followed a period of decline in which climate stresses, conflict and institutional neglect undermined the long term resilience of the urban centre.

Supplemental Material

sj-pdf-1-hol-10.1177_0959683620919976 - Supplemental material for Variable monsoons and human adaptations: Archaeological and palaeoenvironmental records during the last 1400 years in north-western India

Supplemental material, sj-pdf-1-hol-10.1177_0959683620919976 for Variable monsoons and human adaptations: Archaeological and palaeoenvironmental records during the last 1400 years in north-western India by Anil K Pokharia, Jeewan Singh Kharakwal, Shalini Sharma, Michael Spate, Deepika Tripathi, Ashok Priyadarshan Dimri, Xinyi Liu, Biswajeet Thakur, Sadhan Kumar Basumatary, Alka Srivastava, Kamalesh S Mahar and Krishna Pal Singh in The Holocene

Footnotes

Acknowledgements

The authors thank the Director, Birbal Sahni Institute of Palaeosciences, Lucknow, for granting permission for this collaborative work. AKP (Principal Investigator) thanks the Department of Science & Technology, Govt. of India, New Delhi, for granting financial support under SERB-DST Project No. EMR/2015/000881. Thanks are also due to technical assistants, photographers and young archaeologists for their help at the site during the course of excavations. The authors also wish to thank the two anonymous reviewers and Professor Francis Mayle for their helpful comments.

Author contributions

AKP and JSK designed the research; JSK and KPS excavated the site; AKP, JSK, SS and KPS undertook field work and collected samples; AKP, SS, DT, BT, SKB, AS and MS undertook analysis, identification and interpretation of data; AKP, MS and XL wrote the paper; APD modelled past precipitation and circulation data; KSM provided technical supports. All authors discussed the results and provided input on the manuscript.

Funding

The author(s) received no financial support for the research, authorship and/or publication of this article.

ORCID iDs

Anil K Pokharia

Shalini Sharma

Deepika Tripathi

Ashok Priyadarshan Dimri

Kamalesh S Mahar

Supplemental material

Supplemental material for this article is available online.

References

Aberth

(2012) An Environmental History of the Middle Ages: The Crucible of Nature. London; New York: Routledge.

Agnihotri

Datta

Bhusan

, et al. (2002) Evidence for solar forcing on the Indian monsoonal during the last millennium. Earth and Planetary Science Letters 198: 521–527.

Bar-Oz

Weissbrod

Erickson-Gini

, et al. (2019) Ancient trash mounds unravel urban collapse a century before the end of Byzantine hegemony in the southern Levant. Proceedings of the National Academy of Sciences of the United States of America 116: 8239–8248.

Bates

(2016) Social organisation and change in Bronze Age South Asia: A multi-proxy approach to urbanisation, deurbanisation and village life through phytolith and macrobotanical analysis. PhD Thesis, University of Cambridge.

Bates

Petrie

Singh

(2017) Cereals, calories and change: exploring approaches to quantification in Indus archaeobotany. Archaeological and Anthropological Sciences 10(7): 1703–1716.

Bauer

Morrison

(2016) Water management and reservoirs in India and Sri Lanka. In: Selin

(ed.) Encyclopaedia of the History of Science, Technology, and Medicine in Non-Western Cultures. Dordrecht: Springer, pp. 1–12.

Berkelhammer

Sinha

Mudelsee

, et al. (2010) Persistent multidecadal power of the Indian Summer Monsoon. Earth and Planetary Science Letters 290: 166–172.

Blaauw

Christen

(2011) Flexible paleoclimate age-depth models using an autoregressive gamma process. Bayesian Analysis 6: 457–474.

Blott

(2010) GRADISTAT Version 8.0: A Grain Size Distribution and Statistics Package for the Analysis of Unconsolidated Sediments by Sieving or Laser Granulometer. Shinfield: Kenneth Pye Associates Ltd.

10.

Chakravarti

(2008) Agricultural technology in early Medieval India (c. A.D. 500–1300). The Medieval History Journal 11: 229–258.

11.

Chandra

(2003) Essays on Medieval Indian History. New Delhi, India: Oxford University Press.

12.

Chattopadhyaya

(1994) The emergence of the Rajputs as historical process in early Medieval Rajasthan. In: Schomer

Erdtman

Lodrick

, et al. (eds) The Idea of Rajasthan: Explorations in Regional Identity. New Delhi: Manohar Publishers & Distributors; American Institute of Indian Studies, pp. 161–191.

13.

Chauhan

Bera

(1990) Pollen morphology of some important plants of tropical deciduous Sal forest, district Sidhi, Madhya Pradesh. Geophytology 20(1): 30–36.

14.

Chauhan

Quamar

(2012) Pollen records of vegetation and inferred climate change in southwestern Madhya Pradesh during the last ca. 3800 years. Journal of the Geological Society of India 80: 470–480.

15.

Chen

Feng

, et al. (2015) Hydroclimatic changes in China and surroundings during the medieval climate anomaly and little ice age: spatial patterns and possible mechanisms. Quaternary Science Reviews 107: 98–111.

16.

Cookson

Hill

Lawrence

(2019) Impacts of long term climate change during the collapse of the Akkadian Empire. Journal of Archaeological Science 106: 1–9.

17.

Dixit

Tandon

(2016) Hydroclimatic variability on the Indian subcontinent in the past millennium: review and assessment. Earth-science Reviews 161: 1–15.

18.

Dixit

Hodell

Giesche

, et al. (2018) Intensified summer monsoon and the urbanization of Indus Civilization in northwest India. Scientific Reports 8(1): 4225.

19.

Faegri

Iverson

(1975) Textbook of Pollen Analysis. Copenhagen: Munksgaard.

20.

Fagan

(2000) The Little Ice Age: How Climate Made History 1300-1850. New York: Basic Books.

21.

Fredlund

Tieszen

(1994) Modern phytolith assemblages from the North American Great Plains. Journal of Biogeography 21: 321––335.

22.

Giosan

Clift

Macklin

, et al. (2012) Fluvial landscapes of the Harappan civilization. Proceedings of the National Academy of Sciences of the United States of America 109: E1688–E1694.

23.

Giosan

Orsi

Coolen

(2018) Neoglacial climate anomalies and the Harappan metamorphosis. Climate of the Past 14(11): 1669–1686.

24.

Government of Bombay (1874) Gazetteer of the Bombay Presidency: Cutch , Palanpur, and Mahi Kantha, 1880. Mumbai, India: Government Central Press.

25.

Grimm

(1987) CONISS: a FORTRAN 77 program for stratigraphically constrained cluster analysis by the method of incremental sum of squares. Computers & Geosciences 13: 13–35.

26.

Gupta

(2016) Politics of Water Conservation: Delivering Development in Rural Rajasthan, India. Cham: Springer.

27.

Indian Meteorological Department (IMD) (2018) Rainfall Statistics of India – 2017. Report No. ESSO/IMD/HS/RainfallReport/01(2018)/24. New Delhi, India: Hydromet Division, IMD.

28.

Kamae

Kawana

Oshiro

, et al. (2017) Seasonal modulation of the Asian summer monsoon between the Medieval Warm Period and Little Ice Age: a multi model study. Progress in Earth and Planetary Science 4(1): 22.

29.

Kathayat

Cheng

Sinha

, et al. (2017) The Indian monsoon variability and civilization changes in the Indian subcontinent. Science Advances 3(12): e1701296.

30.

Kharakwal

Pokharia

Sharma

, et al. (2016) Preliminary observations of excavation at Chandravati, Sirohi, Rajasthan. Sodh Patrika 266–269: 20–54.

31.

Lamb

(1965) The early Medieval warm epoch and its sequel. Palaeogeography, Palaeoclimatology, Palaeoecology 1: 13–37.

32.

Madella

Alexdendre

Ball

(2005) International Code for Phytolith Nomenclature 1.0. Annals of Botany 96: 253–260.

33.

Madella

Fuller

(2006) Palaeoecology and the Harappan civilisation of South Asia: a reconsideration. Quaternary Science Reviews 25(11–12): 1283–1301.

34.

Morrison

(1993) Supplying the city: the role of reservoirs in an Indian urban landscape. Asian Perspectives 32: 133–151.

35.

Mulholland

Rapp

(1992) A morphological classification of grass silica-bodies. In: Rapp

Mulholland

(eds) Phytolith Systematics Emerging Issues. Boston, MA: Springer, pp. 65–89.

36.

Nayar

(1990) Pollen Flora of Maharashtra State, India. New Delhi, India: Today & Tomorrow’s Printers and Publishers.

37.

Pearsall

(2000) Paleoethnobotany: A Handbook of Procedures. 2nd Edition. San Diego, CA: Academic Press.

38.

Penny

Hall

Evans

, et al. (2019) Geoarchaeological evidence from Angkor, Cambodia, reveals a gradual decline rather than a catastrophic 15th-century collapse. Proceedings of the National Academy of Sciences of the United States of America 116: 4871–4876.

39.

Petrie

Singh

Bates

, et al. (2017) Adaptation to variable environments, resilence to climate change: investigating Land, water and settlement in Indus northwest India. Current Anthropology 58(1): 1–30.

40.

Piperno

(2006) Phytoliths: A Comprehensive Guide for Archaeologists and Paleoecologists. Oxford: AltaMira Press.

41.

Pokharia

Agnihotri

Sharma

, et al. (2017) Altered cropping pattern and cultural continuation with declined prosperity following abrupt and extreme arid event at ∼4,200 yrs BP: evidence from an Indus archaeological site Khirsara, Gujarat, western India. PLoS ONE 12: e0185684.

42.

Pokharia

Kharakwal

Srivastava

(2014) Archaeobotanical evidence of millets in the Indian subcontinent with some observations on their role in the Indus civilization. Journal of Archaeological Science 42: 442–455.

43.

Popper

(1988) Selecting quantitative measurements in paleoethnobotany. In: Hastorf

Popper

(eds) Current Paleoethnobotany. Chicago, IL: The University of Chicago Press, pp. 53–71.

44.

Prasad

Anoop

Riedel

, et al. (2014) Prolonged monsoon droughts and links to Indo-Pacific warm pool: a Holocene record from Lonar lake, Central India. Earth and Planetary Science Letters 391: 171–182.

45.

Prasad

Kusumgar

Gupta

(1997) A mid to late Holocene record of palaeoclimatic changes from Nal Sarovar: a palaeodesert margin lake in western India. Journal of Quaternary Science 12(2): 153–159.

46.

Prasad

Phartiyal

Sharma

(2007) Evidence of enhanced winter precipitation and the prevalence of a cool and dry climate during the mid to late Holocene in mainland Gujarat, India. The Holocene 17: 889–896.

47.

Quamar

Chauhan

(2014) Signals of Medieval Warm Period and Little Ice Age from southwestern Madhya Pradesh (India): a pollen-inferred Late-Holocene vegetation and climate change. Quaternary International 325: 74–82.

48.

Rajagopalan

Vishnu-Mittre Sekar

(1978) Birbal Sahni Institute radiocarbon measurements II. Radiocarbon 20(3): 398–404.

49.

Ramsey

(2017) Methods for summarizing radiocarbon datasets. Radiocarbon 59(6): 1809–1833.

50.

Rawat

Gupta

Sangode

, et al. (2015a) Late Pleistocene–Holocene vegetation and Indian summer monsoon record from the Lahaul, Northwest Himalaya, India. Quaternary Science Reviews 114: 167–181.

51.

Rawat

Gupta

Srivastava

, et al. (2015b) A 13,000 year record of environmental magnetic variations in the lake and peat deposits from the Chandra valley, Lahaul: implications to Holocene monsoonal variability in the NW Himalaya. Palaeogeography, Palaeoclimatology, Palaeoecology 440: 116–127.

52.

Reimer

Bard

Bayliss

, et al. (2013) IntCal13 and Marine13 radiocarbon age calibration curves 0–50,000 Years cal BP. Radiocarbon 55: 1869–1887.

53.

Rosin

(1994) Locality and frontier: securing livelihood in the Aravalli Zone of Central Rajasthan. In: Erdtman

Lodrick

Lloyd

(eds) The Idea of Rajasthan: Explorations in Regional Identity. New Delhi, India: Manohar Publishers & Distributors, pp. 30–63.

54.

Singhvi

Krishnan

(2014) Past and the present climate of India. In: Kale

(ed.) Landscapes and Landforms of India. Dordrecht: Springer, pp. 15–23.

55.

Swain

Kalamkar

Ojha

(2012) State of Rajasthan Agriculture 2011-12. AERC Reports No. 145. New Delhi, India: Ministry of Agriculture.

56.

Trivedi

Chauhan

(2009) Holocene vegetation and climate fluctuations in Northwest Himalaya, based on pollen evidence from Surinsar Lake, Jammu region, India. Journal of the Geological Society of India 74: 402–412.

57.

Twiss

(1992) Predicted world distribution of C3 and C4 grass phytoliths. In: Rapp

Mulholland

(eds) Phytolith Systematics Emerging Issues. Cham: Springer, pp. 113–128.

58.

Twiss

Suess

Smith

(1969) Morphological classification of grass phytoliths. Soil Science Society of America Proceedings 33(1): 109–115.

59.

Uberoi

(2012) Little Ice Age in Mughal India: solar minima linked to droughts? Eos, Transactions American Geophysical Union 93(44): 437–438.

60.

Upadhyaya

(2016) Ṭhakkura: its varying significance in early Medieval northern India. Indian Historical Review 43: 11–24.

61.

Von Rad

Schaaf

Michels

, et al. (1999) A 5000-yr record of climate change in varved sediments from the oxygen minimum zone off Pakistan, Northeastern Arabian Sea. Quaternary Research 51(1): 39–53.

62.

Weber

(1999) Seeds of urbanism: palaeoethnobotany and the Indus civilisation. Antiquity 73: 813–826.

63.

Zhou

Zhang

, et al. (2018) Spatial-temporal variations of natural disasters during the Ming & Qing Dynasties (1368–1912AD) in Chaohu Lake Basin, East China. Quaternary International 467: 242–250.

64.

Xiao

Fang

Zheng

, et al. (2015) Famine, migration and war: comparison of climate change impacts and social responses in North China between the late Ming and late Qing dynasties. The Holocene 25(6): 900–910.

65.

Yadava

Brauning

Singh

, et al. (2016) Boreal spring precipitation variability in the cold arid western Himalaya during the last millennium, regional linkages, and socio-economic implications. Quaternary Science Reviews 144: 28–43.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

3.17 MB