Journal of Archaeological Science: Reports 21 (2018) 183–199

Contents lists available at ScienceDirect

Journal of Archaeological Science: Reports

journal homepage: www.elsevier.com/locate/jasrep

Enamel isotopic data from the domesticated animals at Kotada Bhadli, Gujarat, reveals specialized animal husbandry during the Indus Civilization T ⁎ Kalyan Sekhar Chakrabortya, , Supriyo Chakrabortyb, Petrus Le Rouxc, Heather M.-L. Millera, Prabodh Shirvalkard, Yadubirsingh Rawate a Department of Anthropology, University of Toronto, Canada b Center for Climate Change Research, Indian Institute of Tropical Meteorology, Pune, India c Department of Geological Science, University of Cape Town, Cape Town, South Africa d Department of A.I.H.C and Archaeology, Deccan College Postgraduate and Research Institute, Pune, India e State Department of Archaeology and Museum, Gandhinagar, Gujarat, India

ARTICLE INFO ABSTRACT

Keywords: During the Mature period (2600–1900 BCE) of the Indus Civilization, Gujarat was occupied by two types of South Asia settlements, the settlements with predominantly Classical Harappan traits and the settlements with pre- Indus Civilization dominantly Sorath traits. Most reconstructions of the data conclude that the settlements with Classical Harappan Sorath Tradition traits in Gujarat were engaged in the production and trade of non-perishable Harappan-style goods, while the Agro-pastoralism settlements with Sorath traits focused on agro-pastoralism. To date, no major attempt has been made to un- Carbon isotopes derstand the nature of the agro-pastoralism practiced in these Sorath settlements. This pilot study based on the Oxygen isotopes ff Strontium isotopes oxygen and carbon isotope data from cattle/bu alo, sheep, and sheep/goat teeth from the Sorath settlement of Kotada Bhadli (2400–2000 BCE) suggests that this settlement was indeed engaged in specialized seasonal herding of the domesticated animals. Comparing these data with published data on cattle/buffalo, sheep and goat from the nearby Harappan settlement of Bagasra indicates that a distinct form of sedentary animal hus- bandry was practiced at the site of Kotada Bhadli, the nature of which is considerably different from that practiced at the permanent settlement of Bagasra. At Kotada Bhadli both wild flora and agricultural waste were utilized as fodder, depending on their seasonal availability. Such seasonal availability, and the diversification of fodder selection may suggest a regional adaptation to the local climate by the agro-pastoralists during the Late phase of the Mature period (2300–1900 BCE) of the Indus Civilization. The strontium isotopes from the tooth enamel suggest the domesticates consumed at Kotada Bhadli were most likely raised locally, with a grazing catchment not extending far from the settlement. The results also suggest that at least in Gujarat during the Mature period of the Indus Civilization, carbon isotope values along with tooth morphology might be used to distinguish between sheep and goats, due to human practices involving the deliberate use of millets as fodder for sheep but much less so for goats.

1. Introduction pastoralism played in the regional economy, through the analysis of carbon, oxygen and strontium isotopes from the tooth enamel of cattle/ Gujarat, the southern frontier of the Indus Valley Civilization buffalo and sheep/goat from the Sorath settlement of Kotada Bhadli, (Fig. 1) (henceforth Indus Civilization) was a hub for Harappan-style located on mainland Kachchh, Gujarat. The results will be compared craft activities and trade due to its naturally available resources and its with the published results of biogenic isotopes from the site of Bagasra, location near to the sea that helped in facilitating trade networks. The the only Harappan site from Gujarat that has published isotope data economy of Gujarat during the Indus Civilization was not only limited from domesticated animals. We also compare the strontium isotope to the production and trade of Harappan style objects; agro-pastoralism values of the domesticated animals consumed at the site of Kotada also played an important role in the regional economy of Gujarat. In Bhadli with the model of modern biologically available strontium va- this paper, we try to evaluate the nature of pastoralism and its relation lues of Gujarat, as developed by (Chase et al., 2018), to determine the to the regional agricultural production and climate, as well as the role mobility of the domesticated animals and availability of grazing land.

⁎ Corresponding author. E-mail address: [email protected] (K.S. Chakraborty). https://doi.org/10.1016/j.jasrep.2018.06.031 Received 5 February 2018; Received in revised form 10 April 2018; Accepted 24 June 2018 Available online 19 July 2018 2352-409X/ © 2018 Elsevier Ltd. All rights reserved. K.S. Chakraborty et al. Journal of Archaeological Science: Reports 21 (2018) 183–199

Fig. 1. Shaded area shows the approximate extension of the Indus Civilization during the Mature period (2600–1900 BCE), and the location of major sites and settlements mentioned in the text.

All the conclusions and discussions in this paper are based on a very 1992; Sonawane, 2004). While a few settlements with Sorath traits may limited number of animals (n = 23) from a single site, which was oc- have participated in the acquisition and distribution of raw materials cupied for no more than a few centuries. Therefore, this paper does not for the production of Harappan-style goods (Law, 2013), the majority of generalize the data from this settlement to all similar settlements with Sorath settlements contain remains demonstrating their primary in- Sorath traits; on the contrary, this site is used to argue for the existence volvement with agriculture and pastoralism (Bhan, 2011; Momin, 1984; of multiple forms of economic specialization and a very local level of Reddy, 1994, 2003; Rissman, 1985). Agriculture and pastoralism were environmental adaptation, which may have varied even from settle- major economies of the Indus Civilization and surpluses were produced ment to settlement. for the trade and exchange of perishable goods, as well as to support the During the Mature period of the Indus Civilization craft specialists who were not food producers (Wright, 2010). During (2600–1900 BCE), two types of settlements were found in Gujarat, one the Mature period of the Indus Civilization, various regional and inter- with predominantly Classical Harappan type traits, similar to those regional network systems developed to facilitate trade and exchange. found in the region of the Indus Valley and its tributaries, the other with These regional networks were utilized for the exchange of raw materials predominantly Sorath type traits that are argued to be developed re- and finished Harappan-style goods, and likely human services and food gionally (Possehl, 1980, 1992; Possehl and Herman, 1987; Possehl and grains (Kenoyer, 1995). Regionally specific studies also indicate that Raval, 1989; Sonawane, 2002, 2004; Verma and Menon, 1999). the networks were used for the movement of plants, animals, and Broadly, the economic activities of the settlements in Gujarat with marine products (Belcher, 2005; Chase et al., 2014b; Reddy, 1994). predominantly Sorath traits during the Mature period have been asso- Both the types of settlements that occupied Gujarat during the ciated with agriculture and pastoralism (Possehl, 1992; Sonawane, Mature period differ from contemporaneous settlements found on the 2004). These settlements lack evidence for the production and trade of Indus alluvial plain. Unlike those in the Indus, the settlements in non-perishable Harappan-style goods, separating them from the inter- Gujarat are small, stone built, fortified, and located along trade routes regional and international trade that typically characterizes settlements and near the sources of raw materials that were utilized during the in Gujarat with predominantly Classical Harappan type traits (Possehl, Indus Civilization (Ajithprasad and Sonawane, 2011; Possehl, 1980,

184 K.S. Chakraborty et al. Journal of Archaeological Science: Reports 21 (2018) 183–199

1992; Sonawane, 2002, 2004). Comparing these two types of settle- Harappan traits (Possehl, 1992; Sonawane, 2002). Recent excavations ments, the sites with Classical Harappan traits in Gujarat are larger with indicate that Sorath traits have expanded into Kachchh particularly extensive fortifications. The material remains from these settlements during the Late phase of the Mature period between 2300 and show considerable similarities with the settlements on the Indus allu- 1900 BCE, when many settlements with predominantly Sorath traits vial plain (Possehl, 1992). In contrast, the settlements with pre- started to appear in Kachchh, for example, Desalpur (Uesugi et al., dominantly Sorath traits are smaller; while fortified, they lack the forms 2015) and Kotada Bhadli (Shirvalkar and Rawat, 2012). Settlements of architecture observed in the settlements with Classical Harappan which earlier contained Classical Harappan type materials, now started traits, and the material remains seldom include Indus style objects to exhibit the predominance of Sorath type remains, such as period IIB (Ajithprasad, 2008; Possehl and Raval, 1989). Earlier archaeologists and III at Kanmer (Uesugi and Meena, 2012), period II at Shikarpur categorized the Sorath Tradition as a regional manifestation of the (Ajithprasad, 2009), Stage VI at Dholavira (Bisht, 1999) and period III Indus Valley Tradition, built by the migrants from the Indus alluvial at Bagasra (Sonawane et al., 2003). Classical Harappan materials con- region (Possehl and Herman, 1987; Possehl and Raval, 1989). More tinued to be used at these settlements during the Late phase of the recently, Shirvalkar (2013) suggests it to be a regional development of Mature period, but in a lesser proportion than the preceding phase Indus Valley Tradition that emerged locally. Still others totally deny the (Kharakwal et al., 2012; Sonawane et al., 2003). Similarly, the craft Harappan identity of this Sorath Tradition and see it as a purely re- production activities at these settlements continued, but in a dispersed gional tradition (Verma and Menon, 1999). manner, as observed by Lindstrom (2013) at Bagasra. Particularly in Kachchh, the appearance of Sorath traits during the 1.1. The settlement under study Late Phase of the Mature period is not only marked by the presence of Sorath type ceramics, but also by a change in domestic structures and a Kotada Bhadli (23°20′ N and 69°25′ E) is a small village settlement, re-orientation of settlement patterns (Bisht, 2015; Kharakwal et al., measuring about 3.11 ha, and located at the confluence of two rivers in 2012). This sudden appearance of Sorath traits in Kachchh during the the foothills of the northern hill ranges in Nakhatrana Taluka of District Late phase of the Mature period makes it a very important period to Kachchh, Gujarat (Fig. 1). This settlement was excavated from 2010 to study. The settlement of Kotada Bhadli (Shirvalkar and Rawat, in press; 2013 by Deccan College and the Gujarat State Archaeological Depart- Shirvalkar and Rawat, 2012), with predominantly Sorath type pottery ment (Shirvalkar and Rawat, in press; Shirvalkar and Rawat, 2012). The and without evidence of craft production and trade beyond household most prominent feature of this site is a huge stone-built wall sur- levels, seems to have had a rural economy based on agriculture and rounding the settlement (Shirvalkar and Rawat, 2012). A single re- pastoralism. In general, during the Late phase of the Mature period of sidential complex with multiple interconnected rooms was excavated in the Indus Civilization, the overall production of non-perishable crafts the central area of the site. The space between the inner face of the declined, and by the Late period of the Indus Civilization in Gujarat southern fortification wall and the central residential complex was (post 1900 BCE), pastoralism may have become one of the primary found to be devoid of any architectural features. The architectural economies, as was hypothesised by Bhan (2011) based on the dis- patterns, construction materials, and ceramic assemblage from this site tribution of numerous camp-sites throughout Gujarat and on the lack of indicate its association with the settlements with predominantly Sorath evidence indicating production and trade of Harappan style goods. traits (Shirvalkar and Rawat, 2012). Therefore, it is important to understand the nature of this pastoralism The material remains recovered from Kotada Bhadli are devoid of and the inter-relationship between agriculture and pastoralism during exotic or Classical Harappan-style objects. There is no indication of any the Late phase of the Mature period (post 2300 BCE), in order to pro- form of craft manufacturing or trade activities apart from a few crucible vide a baseline for future studies of the role of pastoralism during the fragments and some copper objects, such as thin rods, bangles, broken subsequent Late period of the Indus Civilization, especially to evaluate chisels and an unknown copper object with golden tip, indicating small whether it became the predominant economy in Gujarat. scale copper-bronze object production was probably taking place at this settlement (Shirvalkar and Rawat, 2012). 1.2. Pastoralism and agriculture in Gujarat during the Mature period of the The two seasonal rivers adjacent to the settlement of Kotada Bhadli Indus Civilization remain dry during most of the dry season, but contain small pockets of fresh water suitable for animal and human consumption. During the Cattle, along with sheep goats and buffalo were the primary do- months of monsoon, however, these rivers may have provided sufficient mesticated animals of the Indus Civilization; however, cattle out- water to facilitate agriculture and grazing. Therefore, the location of number any other animals in most of the Harappan settlements this settlement could potentially benefit both agriculture and pastor- (Meadow and Patel, 2003; Possehl, 1977; Thomas, 2002). Different alism, which seem to have been the primary economies of this settle- forms of pastoralism may have been practiced during the Indus period, ment, and makes this site an interesting location to study the inter- such as household level sedentary animal husbandry, semi-sedentary relation between agriculture, pastoralism, and the local ecology. If pastoralism, and also mobile pastoralism (Wright, 2010). Mughal agriculture was favored over pastoralism at this settlement, we may (1997) identified a few such camps of mobile pastoral groups on the expect agricultural fodder to play a significant role in the selection of Ghaggar-Hakra plain. In Gujarat, a semi-sedentary form of pastoralism fodder, whereas if the residents of Kotada Bhadli favored pastoralism was practiced at the site of Oriyo Timbo, which was seasonally occu- over agriculture, we can expect wild plants to play a considerable role pied during the Late Harappan period (Reddy, 1994; Rissman, 1985). In as fodder. There is also the possibility that both agriculture and pas- contrast, a household level sedentary form of animal husbandry was toralism were favored equally (see Chase et al., 2014b; García-Granero probably practiced at Bagasra during the Mature Harappan period. The et al., 2016 for archaeological case studies on agro-pastoralism situa- evidence from Bagasra shows that all the animals were pastured in the tions in Gujarat during the Indus Civilization). region with no seasonal migration; however, unlike sheep/goat, which Considering the architectural pattern, the settlement of Kotada was raised from birth locally at Bagasra, many of the cattle/buffalo Bhadli appears to have been permanent, but it was occupied for only a were acquired at a young age from different locations, but probably short period around 2300–1950 BCE (Chakraborty et al., in prep). The from within the same region (Chase et al., 2014b). A specialized station radiometric dates from the excavations at Rojdi and Rangpur indicate of cattle breeders at Valabhi was also identified from Gujarat (Mehta, that the Sorath traits were present in Gujarat, particularly in the region 1984). of Saurashtra (another district of Gujarat), as early as 2500 BCE Towards the Late Phase of the Mature period (2300–1900 BCE),

(Possehl, 1992; Possehl and Herman, 1987). During this time, the re- millets, which follow a C4 photosynthesis pathway, have become the gion of Kachchh was occupied by settlements exhibiting Classical predominant crop in Gujarat (Pokharia et al., 2011, 2017; Weber and

185 K.S. Chakraborty et al. Journal of Archaeological Science: Reports 21 (2018) 183–199

Kashyap, 2016). Foxtail millet (Setaria viridis, Setaria glauca), finger composition, which would indicate a C4 grass enriched diet as grazers, millet (Eleusine coracana), little millet (Panicum sumatrense), Italian when compared against goats that would primarily have a C3 enriched millet (Setaria italica) and sorghum millet (Sorghum bicolor) were cul- diet because of their browsing nature. Although Gujarat was primarily a tivated in large quantities at sites such as Rojdi and Kanmer (Pokharia C3 rich environment, C4 grasses played a major role in animal diet due et al., 2014). Millets were not only cultivated for human consumption, to deliberate human feeding of millets to the domesticates as fodder, but the biogenic isotopic evidence from the animal teeth at the site of based on the evidence from Bagasra (Chase et al., 2018). In this study, Bagasra indicate their usage as primary animal fodder as well (Chase we tried to evaluate whether a clear demarcation of the δ13C value et al., 2014b). Chase et al. (2014b) conclude that the cattle/buffalo that exists between sheep and goats and whether δ13C can provide addi- were consumed at Bagasra were predominantly fed with agricultural tional information for the discrimination between sheep and goats. This millets throughout the year. On the other hand, the sheep/goats con- would assist in separating the roles played by the different categories of sumed a mixed diet, although millets still played a considerable role as animals in the diet and economy. Due to our very limited sample size, fodders for them as well (Chase et al., 2014b). This study also suggests our conclusions are preliminary indications of possible directions of there was a provision for year-round storage of millets for fodder at the work, and further analysis is required for any establishment of princi- site of Bagasra. ples. Unfortunately, there are limited data available on agro-pastoralism and the relation between agricultural and pastoralism in Gujarat, par- 2.2. Sample preparation and cleaning ticularly during the Late phase of the Mature period of the Indus Civilization (2300–1900 BCE), when millets became the predominant Individual teeth were first sonicated for 15 min (×2) to remove any crops (Pokharia et al., 2017). The increased dependency on millets by adherent sediments. A thin layer of enamel surface was removed using a the end of the Mature period in Gujarat (Pokharia et al., 2014, 2017; rotary drill bit to get rid of any degraded external enamel surface. Weber, 1998; Weber and Kashyap, 2016) may indicate an adaptation to Sequential samples weighing approximately 5 mg were obtained in changes in regional climate. Such regional adaptation by alteration of bands perpendicular to the growth line from cervix to the cusp, using a ‘proportional exploitation’ of diversified agricultural production, a term diamond drill bit (Fig. 3). Each enamel sample was then numbered and used by Petrie and Bates (2017), may also have affected the modes of stored in centrifuge tubes. pastoralism. A detailed study of the agro-pastoralism from this period is therefore crucial for our understanding of land use, specialized pro- 2.3. Sample preparation and analysis for carbon and oxygen isotopes duction, exchange networks, and the nature of regional adaptation to the ecological settings during the Mature and Late periods of the Indus The enamel powders were first treated with 3% NaOCl for 24 h Civilization. (0.1 ml/mg), followed by 0.1 M acetic acid for 4 h (0.1 ml/mg) as pro- posed by Balasse et al. (2002). After each step, the samples were rinsed 2. Materials and methods 5 times with double distilled water, and after the final step, the samples were dried at room temperature. Ten cattle/buffalo teeth, 5 mandibular second molars (M2) and 5 The analysis was carried out at the Indian Institute of Tropical mandibular third molars (M3), along with 13 sheep/goat mandibular Meteorology, Pune, India. Isotopic measurements were done using a second and third molars, were collected from throughout the site Delta V Plus Isotope Ratio Mass-Spectrometer, where CO2 was gener- (Table 1). All the teeth collected were attached to the mandibles and ated from enamel by reacting with 100% phosphoric acid at 72 °C. NBS only one tooth was collected from each mandible to avoid sampling 19 and PRL Lab Standard were used to standardize our isotopic mea- multiple teeth from the same individual. In this paper, we primarily surement (Chakraborty and Ramesh, 1992). The result is reported in δ discuss the dietary regime and mobility of cattle/buffalo and sheep/ terminology with respect to PDB for carbon and SMOW for oxygen, goat in their first two years of life, as the formation of M2 in both the with a precision of typically ± 0.1‰ for δ13C and ± 0.2‰ for δ18O. animals completes during the first year of life, whereas the formation of The VPDB scale of oxygen was converted to SMOW using the equation M3 completes at the age of two years, and the isotopic composition of proposed by Coplen (1988). the tooth enamel does not change thereafter. (δ18 O=×+ 1.03091 δ18 O 30.91) Enamel formation is sequential, progressing from cusp to cervix SMOW PDB over a period of 12–18 months. Sequential samples along the length of the crown from the cervix to cusp were collected whenever possible to 2.4. Sample preparation and analysis for strontium isotopes assess dietary changes and changes in the hydrological environment (Fig. 3). On average, 3–8 samples were collected from a single sheep/ Enamel samples were digested in a concentrated HNO3 for 1 h at goat tooth, and 6–12 samples were collected from a single cattle/buf- 140 °C. The strontium separation for enamel samples was carried out by falo tooth, bringing our sample size to 151. Due to the limited size of following the method proposed by Pin et al. (1994). The strontium samples, multiple samples along the length of each tooth were grouped fraction was analyzed for 87Sr/86Sr on a NuPlansma HR in the MC-ICP- together to obtain the average strontium value of each tooth. MS facility, housed in the Department of Geological Science, University of Cape Town, Rondebosch, South Africa. NIST SRM987 was used as 2.1. Morphological separation of sheep from goats reference standard with a normalizing value of 0.710255 for 87Sr/86Sr. All Sr isotope data were corrected for Rb interference using the mea- Using the morphological criteria proposed by Halstead and Collins sured signal for 85Rb and the natural 85Rb/87Rb, and the correction of (2002), we tried to distinguish the sheep from the goats in our assem- instrumental mass fractionation was done using the exponential law blage to understand the foddering pattern of sheep, goats and cattle/ and 87Sr/86Sr value of 0.1194. buffalo. This would help us determine if there were any special fod- dering patterns for any group of animals. The criterion M3.1, M3.2, 3. Results and interpretation M3.3, and M3.6 (see Halstead and Collins, 2002 for the detailed de- scription of each criterion) indicated that one tooth (KTB-32) out of 13 Due to the unavailability of the completed faunal analysis from the sheep/goat teeth belonged to a sheep. The rest of the teeth appeared site of Kotada Bhadli, a preliminary estimation of age at death of the 23 ambiguous, and therefore were categorized as sheep/goat. According to individual animals under study was done based on the wear pattern

Balasse and Ambrose (2005),inaC4 grass rich environment, identifi- (Grant, 1975, 1982), in order to have a general estimation of the cation of sheep is possible through looking into their carbon isotope mortality pattern. It appears that 60% of the cattle lived into their

186 K.S. Chakraborty et al. Journal of Archaeological Science: Reports 21 (2018) 183–199

Table 1 Details of the samples from Kotada Bhadli included in this paper. R/L indicates Right or Left and M1–3 indicates tooth position of the Molar tooth.

Specimen no Taxon Side Age Length (mm) Sample no Distance from the cervix δ13C‰ (PDB) δ18O‰ (SMOW) 87Sr/86Sr

KTB-22 Cattle/buffalo LM2 2–6 months 48.9 A11* 7.68 −1.14 29.70 0.7095 A10* 11.52 0.19 30.93 A08 19.19 0.00 29.80 A07* 23.12 −0.06 30.12 A06 27.56 0.03 29.45 A05 30.77 0.07 29.12 A04* 34.57 −0.51 28.55 A03* 37.96 0.04 28.21 A02* 40.98 0.02 26.97 A01 48.16 0.59 27.08 KTB-23 Cattle/buffalo LM2 6–8 years 41.67 A29 5.25 −0.46 27.61 0.7096 A28 8.25 0.48 28.02 A27* 10.19 0.77 27.72 A26 14.17 −0.34 26.02 A25* 17.66 −0.74 26.35 A24 20.71 0.76 26.48 A23* 24.06 0.94 27.20 A22 28.95 0.15 25.22 A21* 32.76 0.16 25.63 A20 37.22 0.03 27.94 KTB-24 Cattle/buffalo LM2 6–8 years 40.26 A42 2.12 −1.30 28.33 0.7104 A41 5.6 −1.92 27.34 A40 8.61 −1.51 25.95 A39 13.28 −0.74 25.11 A38 16.39 1.03 26.92 A37* 19.76 −1.06 26.36 A36* 22.15 0.48 26.56 A35* 24.65 −0.75 27.54 A34 27.08 1.01 30.10 A32 33.64 0.92 31.23 A31 36.73 0.25 29.30 KTB-25 Cattle/buffalo LM2 6–8 years 39.78 A61 4.01 −1.21 26.72 0.7096 A60* 7.87 −2.97 25.52 A59 11.42 −1.08 25.96 A58 14.87 −2.17 24.84 A57 18.04 −0.66 24.86 A56 20.58 −1.15 25.06 A55 24.41 −0.59 23.84 A54 27.38 0.63 25.38 A53* 31.45 −0.47 24.73 A52 34.88 0.72 24.27 A51* 37.75 0.07 24.43 KTB-26 Cattle/buffalo LM2 2–3 years 40.77 A82* 2.55 1.46 29.87 0.7095 A81* 5.05 1.00 29.10 A80 8.27 0.72 28.95 A79 11.42 0.66 28.67 A77 17.65 −0.31 29.29 A76* 20.9 0.01 29.02 A74* 26.77 −0.48 27.92 A73* 30.98 −1.14 26.46 A72 34.57 −1.66 27.00 A71 37.77 −1.83 27.23 KTB-27 Cattle/buffalo LM3 6–8 years 40.77 A99* 5.23 −0.10 27.68 0.7095 A98* 9.3 −0.01 27.40 A97* 13.68 −0.92 27.27 A96 17.36 −1.40 26.06 A95 21.07 −1.20 25.64 A94 24.48 −1.37 25.69 A93* 28.13 −2.50 25.79 A92* 32.53 −1.79 26.10 KTB-28 Cattle/buffalo LM3 2–3 years 49.9 B08 23.23 −0.59 27.21 B07 26.45 −0.46 28.02 B06 28.9 −0.21 27.66 B05 32.4 0.14 27.76 B04 36 0.42 28.05 B03 40.12 −0.18 28.40 B02 42.22 −0.73 26.85 B01 46.7 −1.50 26.14 (continued on next page)

187 K.S. Chakraborty et al. Journal of Archaeological Science: Reports 21 (2018) 183–199

Table 1 (continued)

Specimen no Taxon Side Age Length (mm) Sample no Distance from the cervix δ13C‰ (PDB) δ18O‰ (SMOW) 87Sr/86Sr

KTB-29 Cattle/buffalo RM3 8–12 years 35.47 B17* 3.03 0.47 27.11 0.7094 B16 5.72 0.15 26.55 B15* 8.87 0.61 27.11 B14 12.4 −0.51 25.70 B13* 15.86 −1.22 27.03 B12 19.36 −0.28 26.63 B11* 23.58 −0.12 26.14 KTB-30 Cattle/buffalo LM3 6–8 years 48.99 B32 5.49 0.71 27.66 0.7095 B31* 9.94 0.82 29.25 B30 13.26 1.17 30.45 B29* 17.08 0.90 30.75 B28 20.75 1.56 31.97 B27* 24.85 1.10 32.90 B26 28.77 0.52 32.13 B25* 32.13 1.50 33.13 B24 36.19 0.83 32.38 B23* 38.69 0.60 32.16 B22 42.49 1.39 32.73 B21 45.95 1.29 32.36 KTB-31 Cattle/buffalo LM3 2–3 years 49.27 B46 26.31 0.85 30.57 0.7094 B45* 29.57 0.08 30.74 B44* 33.5 0.62 30.66 B43* 37.25 0.22 29.32 B42* 41.13 0.39 30.61 B41* 45.19 0.36 29.55 KTB-32 Sheep RM3 8 years 19.86 B56* 7.36 −2.38 29.50 0.7096 B55* 10.06 −1.82 28.05 B54* 12.36 1.83 32.25 B53* 13.71 −0.30 28.38 B52* 16.36 −0.83 27.56 B51* 18.09 −1.03 28.04 KTB-33 Sheep/goat LM2 1–2 years 28.25 B64* 2.15 −4.83 24.65 0.7096 B63* 4.33 −5.17 26.46 B62 9.42 −6.15 27.64 B61 12.25 −9.72 27.62 KTB-34 Sheep/goat LM2 4–6 years 29.11 B75 3.51 −7.26 30.01 B74 7.25 −7.31 31.11 B73 12.51 −8.29 30.30 B72 18.41 −7.99 29.32 B71 23.82 −7.31 28.14 KTB-35 Sheep/goat LM2 1–2 years 28.63 B87 5.03 −6.49 27.15 B86 9.03 −3.13 26.74 B85 12.33 −3.71 26.29 B84 15.13 −2.89 25.31 B83 19.53 −3.37 25.92 B82 23.13 −3.60 26.45 B81 25.13 −5.07 28.31 KTB-36 Sheep/goat LM2 1–2 years 24.65 B98 3.05 −9.02 32.43 B97 5.25 −8.95 31.06 B96 7.75 −6.66 29.77 B95 11.05 −5.55 28.57 B94 14.25 −4.00 31.87 B93 16.65 −5.26 31.84 B92 20.05 −6.86 31.77 B91 22.25 −7.81 32.39 KTB-37 Sheep/goat LM3 1–2 years 32.25 C08 5.55 −3.85 27.91 C07 9.55 −3.38 27.90 C06 13.15 −3.63 29.37 C05 16.95 −2.42 33.43 C04 19.75 −4.10 32.65 C03 23.25 −5.72 31.91 C02 26.45 −5.17 29.51 C01 30.05 −3.71 27.71 KTB-38 Sheep/goat LM2 6–12 months 34.85 C14 7.75 −5.34 25.77 0.7097 C13* 16.15 −5.88 26.68 C12* 19.55 −6.81 26.33 C11 23.55 −6.46 26.44 KTB-39 Sheep/goat LM2 4–6 years 26.5 C30 4.4 −7.50 29.80 0.7095 C27* 11 −8.43 28.40 C26* 12.9 −10.67 27.08 C25* 15.1 −9.29 28.53 C24* 17.2 −8.60 27.55 C23* 19.5 −7.22 28.14 C22* 21.5 −7.81 28.66 C21* 24.3 −8.16 28.54 (continued on next page)

188 K.S. Chakraborty et al. Journal of Archaeological Science: Reports 21 (2018) 183–199

Table 1 (continued)

Specimen no Taxon Side Age Length (mm) Sample no Distance from the cervix δ13C‰ (PDB) δ18O‰ (SMOW) 87Sr/86Sr

KTB-40 Sheep/goat RM2 6–12 months 27 C34 17.4 −3.95 29.85 0.7096 C33* 20 −4.23 29.78 C32* 22.2 −4.86 29.55 C31* 24.9 −5.51 29.38 KTB-41 Sheep/goat LM2 6–12 months 25.5 C43 2.4 −2.30 28.72 0.7095 C42* 9.9 −4.82 28.90 C41* 12.7 −4.90 27.68 KTB-42 Sheep/goat LM2 6–12 months 26 C58* 0.1 na na 0.7095 C57* 2.7 na na C56* 5.8 na na C55* 9 na na C54* 10.9 na na C53* 14.4 na na C52* 17.7 na na C51* 20.9 na na KTB-43 Sheep/goat RM2 1–2 years 18 C68* 0.4 na na 0.7095 C67* 2.6 na na C66* 4.9 na na C65* 7.1 na na C64* 9.6 na na C63* 11.3 na na C62* 13.7 na na C61* 16 na na KTB-44 Sheep/goat LM3 1–2 years 28.6 C78* 7.1 na na 0.7093 C77* 10.4 na na C76* 12.2 na na C75* 15.3 na na C74* 18.3 na na C73* 20.6 na na C72* 22.8 na na C71* 26.2 na na

Asterisk (*) after individual samples indicates the samples that were grouped together to obtain the average strontium isotope values from each teeth. adulthood, whereas 30% of the cattle/buffalo were slaughtered when growth occurred in water-stressed conditions, it will be around −6.5 to 13 they were sub-adults and only 10% were slaughtered when they were −7‰; in contrast, for animals consuming a C4 dominated diet, the δ C yearlings (age at death below 1 year). In the case of sheep/goat, 50% of will be around 2.5‰ (Cerling et al., 1997; Sternberg et al., 1984a, the animals were slaughtered when they were sub-adults, and 33% 1984b; Zhang et al., 2012). Additional factors also contribute to re- when they were yearlings. Only 17% of the sheep/goats survived into gional values of δ13C, including variation in precipitation, relative hu- their adulthood (Fig. 4). On the other hand, the single sheep (KTB-32) midity, light, height and the density of canopy cover (Heaton, 1999). 13 in the sample survived into adulthood, suggesting that this particular As the δ C values of both C3 and C4 plants from the region of animal was probably used for purposes other than meat. Similarly, Gujarat fall well within the overall global range (Lancelotti et al., 2013; Meadow (1989) also observed that at Harappa, half of the sheep lived Reddy, 1994), a δ13C value of around 2.5‰ of the enamel bioapatite passed their adulthood, and suggested that they may have been used for should indicate a C4 dominant diet. The carbon isotope data from the wool rather for meat. At Bagasra, most of the cattle lived into adulthood tooth enamel of the cattle/buffalo from Bagasra, who predominantly and beyond, suggesting cattle were raised primarily for the secondary consumed a C4 diet, indicate a similar value of 2.5‰ or higher. In products, such as milk, traction, and leather, before they were con- contrast, the sheep/goat from Bagasra, who predominantly consumed a sumed for their meat; most of the Bagasra sheep/goats were slaugh- mixed diet that includes both C3 and C4, with a slightly higher pro- tered during their early adulthood, indicating their primary use as meat portion of C3, indicate much lower values as compared to the cattle/ (Chase et al., 2014b). buffalo, even as low as −7‰ (Chase et al., 2014b). According to Reddy

(1994), the soil organics from Oriyo Timbo and Babar Kot indicate a C3 dominated environment in Gujarat during the Late period of the Indus 3.1. Using carbon isotopes to determine diet and fodder selection from tooth Civilization. This is also observed by Pokharia et al. (2017) at the site of enamel Khirsara during the Mature period, with the slightly increased indica- tion of C towards the Late phase of the Mature period, due to the in- The carbon isotopic value (denoted as δ13C) of tooth enamel is re- 4 creased cultivation of millets. These observations of soil organics may lated to the δ13C value of the ingested food, and can be used to dis- indicate the possible availability of wild C vegetation in the vicinity of tinguish between diets based on C and C plants in herbivores 3 3 4 the settlement for animal consumption, whereas the primary source of (Sternberg et al., 1984a, 1984b; Zhang et al., 2012). In Gujarat, during C would be from agricultural millets through specialized foddering. At the Indus Civilization as well as in modern times, both C and C plants 4 3 4 the same time, it is very unlikely that the residents of Kotada Bhadli did were used as fodders for domesticated animals. For the present study, not exploit the diverse wild grasses available in the famous Banni δ13C from the tooth enamel of cattle/buffalo and sheep/goat from the grassland which is located very near to the settlement. This ‘grass and site of Kotada Bhadli was used to determine the role played by these shrub savanna’ ecosystem of Banni grassland may have provided the two categories of plants in their respective diets. residents with ample C fodder, as during the time Kotada Bhadli was Generally, the δ13C value of modern C plants is between −20‰ to 3 3 occupied the Banni grassland was predominantly C (Pillai et al., 2017). −35‰ (vs VPDB), with an average of −27‰. For C plants, it is be- 3 4 A certain portion of the C signature might possibly come from barley tween −9‰ to −17‰ with an average of −13‰. For large mammals 3 and rice plant fodder since these were cultivated in a very limited prior to the Industrial Revolution, the δ13C value of tooth enamel for an quantity during the Late Mature period in Gujarat (Pokharia et al., animal consuming a C3 dominated diet is approximately −11.5‰;if

189 K.S. Chakraborty et al. Journal of Archaeological Science: Reports 21 (2018) 183–199

2014; Weber et al., 2010). characterized by spring and autumn growing seasons (Ehleringer et al., The δ13C carbon values of the cattle/buffalo teeth from Kotada 1997; Niu et al., 2005). In the absence of irrigation, millets in present- Bhadli (Table 1) indicate that unlike Bagasra where all the cattle day Gujarat are generally sown to take advantage of the Indian summer throughout their lives consumed mostly C4 grasses, individual cattle/ monsoon and harvested during October and November (Reddy, 1994; buffalo from Kotada Bhadli exhibit considerable variation within and Weber and Kashyap, 2016). Alternatively, there is an additional mode between individuals. Unlike Bagasra, the cattle/buffalo of Kotada of cultivation using the floodplains along the river; millets are oppor-

Bhadli consumed a mixed diet consisting of both C4 and C3 vegetation. tunistically sown once the monsoon is over and the floodplains are The δ13C values of cattle/buffalo from Kotada Bhadli match quite well available for cultivation, a present-day practice in parts of Southern with the sheep/goats from Bagasra, who also consumed a mixed diet, India (Reddy, 1994). Using either of these cultivation methods in Gu- and both had a C4-rich mixed diet, rather than a C3-rich mixed diet jarat during the Indus Civilization, millets would only have been (Fig. 5). For the animals consumed at Kotada Bhadli, the estimated available as fodder once they were harvested in the post-monsoon carbon isotope values among the cattle/buffalo vary from −2.97‰ to season, when the δ18O values of meteoric water are higher than the 1.56‰, suggesting that these animals consumed a varied proportion of months of monsoon. If the residents of Kotada Bhadli depended on the

C4 during the formation of their enamel (Fig. 6, and Table 1). In con- seasonal availability of fodder, we may expect higher dependence on trast, the sheep/goats, with their browsing habits, have carbon isotope millets during the post-monsoon season, and dependence on C3 plants values varying from −10.67‰ to −2.30‰, indicating a predominant during the months of monsoon, prior to the harvesting of millets. As diet of C3 leaves with occasional intake of C4 (Fig. 6, and Table 1). observed by Reddy (1994), however, in modern Gujarat certain types of Interestingly, and as expected the only designated sheep in our as- millets are predominantly grown as animal fodder, which may have semblage consumed more C4 compared to the sheep/goats and followed been true during the Harappan period as well, and therefore, we cannot a very similar diet to the cattle/buffalo, who are also grazers in nature. rule out the possibility of the availability of certain proportion of millets The δ13C values of the enamel from the sheep are from −2.37‰ to even during the months of monsoon, which were deliberately shown for

1.82‰, indicating only occasional intake of C3 in their diet (Fig. 6, and animal consumption. Using both carbon and oxygen isotopes from an- Table 1). Although the majority of the cattle/buffalo from Bagasra imal tooth enamel, we can investigate how the availability of fodder consumed predominantly C4 vegetation, the cattle/buffalo from Kotada affects the dietary pattern of the animals at Kotada Bhadli. Bhadli consumed a fairly mixed diet throughout the year, with a higher The mean oxygen value of cattle/buffalo from Bagasra as observed portion of C4. Even in the case of sheep/goats, at least two of them by Chase et al. (2014a, 2014b) is 33.1‰, whereas the same for Kotada 18 indicate a predominant C3 diet considering a water-stressed environ- Bhadli is 27.61‰. For sheep/goat from Bagasra, the mean δ Ois ment, but for the rest of animals a certain portion of C4 also played a 36.60‰ and for Kotada Bhadli, the mean value for sheep/goats is role, which must have come from millet fodder, or from wild C4 grasses. 28.54‰ and for the proposed sheep is 28.04‰ (Fig. 7, Table 1). Such a difference in values between these two sites could be expected because 3.2. Using oxygen isotopes to determine the hydrological weather and the of their locations; the site of Bagasra is located near the Arabian Sea, relation between the climate and fodder selection whereas the site of Kotada Bhadli is located inland in the foothills of the northern hill ranges (Figs. 1 and 2). Similar overall δ18O values of all Oxygen isotopes from tooth enamel have long been used to re- the animals from Kotada Bhadli indicate that they were raised in close construct paleoclimate, inter-annual climatic variation, seasonality of proximity to each other. The range of oxygen isotope values of in- birth, and even movement through the landscape (Balasse et al., 2012; dividual teeth from Kotada Bhadli (Fig. 8) indicate that almost all the Chase et al., 2014b; Lightfoot and O'Connell, 2016; Towers et al., 2014; teeth have experienced considerable climatic variation during their Zin Maung Maung et al., 2011). Oxygen isotope values of mammalian formation. Such seasonal variation was also observed by Chase et al. tooth enamel are directly related to ingested water, which in turn is (2014a, 2014b). However, the range of variation observed by Chase and related to precipitation, controlled by local precipitation and moisture his team among their cattle/buffalo is 5.67‰, and among sheep/goats sources, and modified by evaporative effects (Balasse et al., 2002; Kohn is 6.12‰, which is considerably smaller than the range we observed, et al., 1996; Zhang et al., 2012; Zin Maung Maung et al., 2011). Isotopic composition of meteoric water is controlled by several atmospheric variables, such as temperature, rainfall, humidity, circulation, and outgoing longwave radiation (Chakraborty et al., 2016). The δ18O va- lues of precipitation from Hissar in north-western India indicate that during the months of the Indian summer monsoon (June–September), the δ18O for precipitation is lower than in the months of early summer (i.e. pre-monsoon) and winter (post-monsoon) (Sarkar et al., 2016). There is considerable distance between Hissar in Haryana and Kachchh in Gujarat (Fig. 1), which are also two geologically different regions. However, in terms of rainfall, both these regions receive maximum rainfall during the monsoon and receive little or no rainfall during the pre- and post-monsoon seasons. Therefore, even in Gujarat, the δ18O values of precipitation will be lower during the months of monsoon and will be higher during pre-and post-monsoon. Both the rivers adjacent to the settlement of Kotada Bhadli are seasonal and primarily rainfed, and therefore, the water that was available to the animals of Kotada Bhadli were directly related to the precipitation. Other minor factors, such as latitudes, aridity and evaporation processes may alter the isotopic composition of precipitation; in addition, as suggested by Zin Maung Maung et al. (2011), water conservation factors and metabolic factors modify the δ18O value of tooth enamel of the large mammals.

In general, the geographical distribution of C4 plants is associated with higher average annual temperatures and the summer growing Fig. 2. The geology of Kachchh and Saurashtra indicating the location of Kotada Bhadli and Bagasra (modified from Sen et al., 2009, Fig. 1). season, as opposed to the C3 plants, which are predominant in areas

190 K.S. Chakraborty et al. Journal of Archaeological Science: Reports 21 (2018) 183–199

6

4

2

0

-2

-4

C‰(PDB) -6 13 δ -8

-10

-12

-14 Sheep Bagasra Bagasra Sheep/Goat Sheep/Goat Cattle/Buffalo Cattle/Buffalo Kotada Bhadli Kotada Bhadli Kotada Bhadli

Fig. 5. Comparison of δ13C values of tooth enamel from domesticated animals consumed at both Kotada Bhadli and Bagasra, the data from Bagasra was ac- quired from Chase et al. (2014b).

the site likely ran dry, and the animals only had access to the small pockets of water at the river beds which are enriched in δ18O. The inter-annual variation of both δ13C and δ18O indicate a sea- sonally dependent diet for the majority of the animals tested from Kotada Bhadli, except 3 cattle/buffalo (KB22, KB-30, and KB-31) that show a stable diet without any indication of seasonal dependency. The majority of the cattle/buffalo (KB23, KB24, KB27, KB28, and KB29) exhibit depleted carbon isotope values when the oxygen values are also

depleted, indicating consumption of proportionately more C3 during Fig. 3. KTB-29 showing the sequential sampling along the enamel. the summer monsoon, and consumption of more C4 during the hottest and driest months. Alternatively, KB25 and KB26 indicate consumption of a slightly higher proportion of C grasses and/or millets during the which is 9.06‰ in case of cattle/buffalo and 8.78‰ in case of sheep/ 4 months of monsoon, and C vegetation during the dry and warm goats (Fig. 7, Table 1). Such a range of variation can be expected at 3 months (Fig. 9). Like the majority of the cattle/buffalo, the possible Kotada Bhadli because the primary source of water for the animals was sheep from Kotada Bhadli (KB-32) indicates a C4 rich diet with a higher precipitation. During the months of monsoon, they had access to the portion of C4 during the hot and dry months and slightly more C3 rain-water, whereas during pre-and post-monsoon both the rivers along during the wet and cold months (Fig. 10). In contrast to the majority of Cattle/buffalo Sheep/goat Yearling Sub-adult Adult Yearling Sub-adult Adult

10% 17%

33%

30% 60%

50%

Fig. 4. Age profile of cattle/buffalo and sheep/goat from the site of Kotada Bhadli.

191 K.S. Chakraborty et al. Journal of Archaeological Science: Reports 21 (2018) 183–199

Fig. 6. Carbon isotope values from Kotada Bhadli, indicating the nature of fodder selection. Lines connect the sample from individual tooth as shown in Fig. 3.

42 the cattle/buffalo and the possible sheep, but similar to the observa- 40 tions of Chase et al. (2014a, 2014b), all the sheep/goats from Kotada Bhadli consumed more C4 during the wet and cold months, whereas 38 during the hot and dry months, they relied more on C3 vegetation (Fig. 10). 36

)WOMS(‰O 34 3.3. Using strontium isotope to determine mobility 32 Strontium isotope ratios (87Sr/86Sr) from human and animal skeletal 30 remains provide a useful tool to determine migration and mobility ff

18 28 across di erent geological landscapes (Bentley, 2006). Strontium is a δ trace element with a similar chemical structure to calcium, which is a 26 major constituent of bioapatite; through food and water, strontium 24 enters the body and gets stored by replacing a portion of calcium in bioapatite (Frei, 2012). Although enamel strontium values primarily 22 reflect the local geological values, various factors such as whole-rock 20 average, animal diet, ocean spray, atmospheric dust, and post-deposi- tional alteration may affect the enamel strontium isotopes from ar-

Sheep chaeological remains (Frei, 2012; Frei and Frei, 2011; Price et al., Bagasra Bagasra 2002). Therefore, in order to determine the extent of migration and Sheep/Goat Sheep/Goat Cattle/Buffalo Cattle/Buffalo Kotada Bhadli Kotada Bhadli Kotada Bhadli movement of animals on the landscape, it is important to create a Fig. 7. Comparison of the δ18O isotope values of the domesticated animals baseline value of bioavailable strontium (Bentley, 2006; Chase et al., consumed at both Kotada Bhadli and Bagasra, the data from Bagasra was ac- 2018; Frei, 2012; Frei and Frei, 2011; Price et al., 2002). Fortunately, quired from Chase et al. (2014b). baseline values of bioavailable local strontium from a considerable portion of Gujarat are available to us from the work of Chase et al.

192 K.S. Chakraborty et al. Journal of Archaeological Science: Reports 21 (2018) 183–199

Fig. 8. Oxygen isotope values from Kotada Bhadli. Lines connect the sample from individual tooth as shown in Fig. 3.

(2018). 2018). Geologically Kachchh is not very different from Saurashtra, where The strontium isotope values of cattle/buffalo from Bagasra are the site of Bagasra is located. A recent study by Chase et al. (2018) varied; although local, the values indicate widespread grazing land or indicates that both these regions produce very similar values for bio- movement of cattle/buffalo. In contrast, the sheep/goats have limited logically available strontium isotopes. The observed values from Saur- variation in strontium isotope values, indicating restricted movement ashtra range from 0.7081–0.7103, whereas for Kachchh they range during their lifetime (Chase et al., 2014a, 2014b, 2018). Due to the from 0.7085–0.7103; the slightly enriched values in Kachchh may be varied geology of Kachchh, if the cattle/buffalo from Kotada Bhadli due to greater coverage of Mesozoic and Tertiary sedimentary rock covered a vast grazing land or moved between landscapes, we may along with greater inputs of radiogenic Himalaya-derived sediments expect varied average strontium isotope values within the population, (Chase et al., 2018). Mainland Kachchh is dominated by East-West as was observed at Bagasra. Also, with the help of the baseline map trending highlands and islands, surrounded by a residual depression created by Chase et al. (2018), we should be able to trace this move- (Merh, 1995). Unlike Saurashtra, which is mostly dominated by the ment, if any movement can be observed. Unfortunately, due to the Deccan Trap, Kachchh is formed by Mesozoic rock, Deccan Trap, Ter- limited size of the samples from Kotada Bhadli, we are unable to de- tiary rock and Quaternary sedimentation, all packed close to each other termine if there was any inter-seasonal mobility of the animals con- (Chase et al., 2018; Merh, 1995; Sen, 2001; Sen et al., 2009). The site of sumed at Kotada Bhadli, as we had to combine multiple samples along Kotada Bhadli is located on the Mesozoic rock formation; the north of the length of each tooth to acquire the average strontium value from the site is dominated by Quaternary sedimentation, the south is domi- each animal tooth, which would only indicate the large-scale overall nated by Deccan Trap and Tertiary formations, all within a close range movements of the animals between regions. (Fig. 2). Therefore, even minimal mobility around the site of Kotada The average strontium isotope values of all animals examined from Bhadli would produce varied strontium values. The biologically avail- the site of Kotada Bhadli ranges between 0.7093 and 0.7096, excluding able strontium values for the Mesozoic/Tertiary formation ranges be- one cattle/buffalo (KB-24) with a slightly higher strontium value of tween 0.7091 and 0.7099, whereas the basalt formation of the Deccan 0.7104 (Table 1). Excluding the outlier, the average strontium isotope Trap indicates a slightly lower range between 0.7087 and 0.7091. The value for the remaining animals is 0.7095 (Fig. 11), which is most likely Quaternary formation to the north of mainland Kachchh had con- the local value for the site, considering that all the animals were raised siderably higher values, greater than 0.7099, indicating considerable at the site or locally before they were consumed at Kotada Bhadli. input from the radiogenic sediments from the Himalayas (Chase et al., Unlike Bagasra, there are no considerable differences in the strontium

193 K.S. Chakraborty et al. Journal of Archaeological Science: Reports 21 (2018) 183–199

Fig. 9. Intra-tooth variation of carbon (δ13C, triangles) and oxygen (δ18O, squares) isotope ratios of the tooth enamel bioapatite of Cattle/Buffalo from the site of Kotada Bhadli. values between, cattle/buffalo and sheep/goat; rather, the similar however, that the strontium data obtained for this study indicates the average values of sheep/goats and cattle/buffalo from Kotada Bhadli average value for each tooth based on accumulative samples collected indicate that they were most like raised locally or possibly both these from multiple locations along the length of every individual tooth, so it categories of animals had a similar mobility pattern. The outlier cattle/ is not possible to determine inter-seasonal movement. It is possible that buffalo (KTB-24) from our assemblage might have been acquired from this average strontium value is the accumulative average strontium somewhere else prior to consumption at Kotada Bhadli, or might have value of the animals' movement during the formation of their enamel, had a slightly different grazing pattern than the rest of the assemblage. with high and low values from seasonal movement canceling each other As already mentioned, it is necessary to take into consideration, out. Alternatively, it is also possible that the animals moved between

194 K.S. Chakraborty et al. Journal of Archaeological Science: Reports 21 (2018) 183–199

Fig. 10. Intra-tooth variation of carbon (δ13C, triangles) and oxygen (δ18O, squares) isotope ratios of the tooth enamel bioapatite of Sheep and Sheep/Goats from the site of Kotada Bhadli. regions that have similar strontium values. Therefore, although it is archaeological values of the animal teeth from the site of Kotada Bhadli quite likely that the animals from Kotada Bhadli were not acquired from fall well within the range of biologically available strontium values a region that is geologically quite different from Kachchh, and so has a from modern animal dung from the region, assessed by Chase et al. very different strontium value, it is possible that the animals may have (2018), which is between 0.7091 and 0.7099. This overlap further experienced inter-seasonal movement in search of wild vegetation. It is supports our argument that if not all, at least the majority of the do- also quite unlikely that the residents of Kotada Bahadli did not exploit mesticates were raised locally, probably with seasonal movement. Un- the Banni grassland, which is located very near to the settlement, but fortunately, bioavailable strontium values from modern animal samples had a very different biologically available strontium isotope value, with are not available at present from Kotada Bhadli, and therefore, it is not an average around 0.7101 (Chase et al., 2018). The range of the possible to assign a range of values that can be expected for the

195 K.S. Chakraborty et al. Journal of Archaeological Science: Reports 21 (2018) 183–199

Fig. 11. Average 87Sr/Sr86 isotope data of each individual tooth from cattle/buffalo, sheep, and sheep/goats from Kotada Bhadli, compared against the average 87Sr/ Sr86 values of each individual tooth of the faunal remains from Bagasra obtained from Chase et al. (2014b). The upper, blue line indicates the average value of 87Sr/ Sr86 from the archaeological samples at the site of Kotada Bhadli, and the lower, red line indicate the average value of 87Sr/Sr86 from the archaeological samples at the site of Bagasra mentioned in (Chase et al., 2014b). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) settlement itself. Fortunately Chase et al. (2018) model provide a range help us identify the existence of such inter-regional contact. of values expected for the Mesozoic-Tertiary deposits from Kachchh, based on modern animal dung that help us to estimate a broad range of 4. Discussion bioavailable strontium values from the region where the site is located. The one sample KTB-24 with a higher average strontium value of It is now clear that the cattle/buffalo consumed at the site Kotada 0.7104, may indicate that this animal was acquired from a region with a Bhadli were raised slightly different from those at Bagasra. The cattle/ higher average strontium value than the region around Kotada Bhadli. buffalo from Bagasra were predominantly fed with C grasses, primarily This does not necessarily mean a long distance move, however, as 4 agricultural products, indicating year-round supply of agricultural Chase et al. (2018) have noted that the Quaternary deposit a little north millets as fodder. At Kotada Bhadli, although C grasses played a major of the settlement in the Banni grassland region is slightly more radio- 4 role, a portion of seasonally available C vegetation was also included genic than the Kotada Bhadli region, and this region is close enough for 3 as fodder, primarily during the wet and cold months when C vegeta- seasonal grazing. In future, analysis with samples large enough to ac- 3 tion was widely available in the vicinity of settlement. Some of the quire sequential strontium values is needed to resolve the issue of the cattle/buffalo also consumed a slightly higher proportion of C vege- seasonal movement of the animals consumed at the Kotada Bhadli. 3 tation even during the hot and dry months and C during the wet and The average strontium values of the archaeological samples from 4 cold months, probably to supplement agricultural fodder. At least 3 out Kotada Bhadli is 0.7095, with a range between 0.7093 and 0.7096, of 10 cattle/buffalo indicate no clear seasonal dependency of diet; ra- which is slightly higher than the average values of archaeological ther, they indicate a mixed diet including both C and C vegetation samples from Bagasra, which is 0.7093. When we compared the 3 4 throughout the year. Such a varied diet for cattle/buffalo at the site of average strontium values of each individual tooth from Bagasra against Kotada Bhadli indicates that although their diet was controlled, it was the strontium values from Kotada Bhadli, it was interesting to observe not entirely dependent on agricultural production; instead, the seasonal that many cattle/buffalo consumed at Bagasra have a strontium value availability of wild vegetation at the vicinity of the settlement also similar to the range of values observed from Kotada Bhadli (Fig. 11). It played a role. The consumption of a large portion of C grasses is possible that, as was argued by Chase et al. (2018), some cattle/ 4 throughout the year, suggests the use of agricultural millets as fodder buffalo consumed at Bagasra may have been grazing a little inland from was very likely, and also suggests that there was a provision for the that settlement, or were acquired from somewhere in Central Saur- year-round storage and/or supply of such millet fodders for animal ashtra. Nevertheless, we cannot rule out the possibility that at least consumption. Therefore, the residents of Kotada Bhadli not only herded some of these cattle/buffalo may have arrived from some regions of animals but were also engaged in the production, storage and/or ex- Kachchh with a similar strontium value as that observed at Kotada change of millets, likely for both human and animal consumption. The Bhadli. Chase and his team also acknowledge the possibility of such availability of millets for the consumption of domesticated animals inter-regional contact between Kachchh and Saurashtra during the during the months of the monsoon when they were just sown and were Harappan period (Chase et al., 2014a), which may not have been re- still green indicate that at least some portion of millets were probably stricted to non-perishable goods, but also included perishable items for cultivated especially as animal fodder. Reddy (1994) has observed such human consumption. Further studies from the region of Kachchh would practices even in present-day Gujarat. We cannot also rule out the

196 K.S. Chakraborty et al. Journal of Archaeological Science: Reports 21 (2018) 183–199 possibility that some of the domesticated animals ran into the fields 2017). when the millets were still green. It is interesting to observe that unlike Bagasra, at Kotada Bhadli, 5. Conclusion seasonal availability of fodders played a major role in the rearing of both cattle/buffalo and sheep/goats. Although the year-round supply of The results from this pilot study indicate that different modes of agricultural fodder seems to have been available at the site of Kotada pastoralism were practiced in Gujarat during the late third millennium Bhadli either through storage and/or through exchange, wild fodder BCE. The mode of pastoralism practiced at Kotada Bhadli indicates a also played a considerable role in the animal's diets, at least when seasonally variable diet, which is considerably different from the pas- seasonally available. Year-round storage and higher dependency on toralism practiced at Bagasra. The results from this study also challenge agricultural millets as fodder was probably essential due to the pro- the idea of nomadism being associated with pastoralism from the Indus posed weakening of the Indian Summer Monsoon during 3rd millen- Era. Rather, with access to agricultural fodder and seasonally available nium BCE (Sarkar et al., 2016). Both Possehl (1986) and Pokharia et al. wild vegetation, a sedentary form of agro-pastoralism is possible. The (2017 and 2014) argued that the weakening of the monsoon was most strontium data may suggest economic interaction between Saurashtra likely one of the factors responsible for the increasing cultivation of and Kachchh, the form of exchange of domesticated animals for food or millets in Gujarat during the latter part of the 3rd millennium BCE. In other uses. Smaller settlements like Kotada Bhadli may have played a contrast, Petrie et al. (2017: pp. 18) have observed that diversified vital role in those regional networks through supplying domesticated subsistence practices encompassing both summer and winter crops animals or animal-based goods subsistence goods to manufacturing were already in practice at least in Northwestern India, even before the centers like Bagasra. Our data also suggest that it may be possible to proposed weakening of the summer monsoon, and the “proportional apply Balasse and Ambrose's (2005) model to separate sheep from exploitation of winter and summer crops” made it possible for the re- goats, based on the δ13C values of tooth enamel. sidents to adjust to the weakening of the monsoonal rainfall around The question is, why there are such differences between Bagasra and 2200–2100 BCE. At the same time diversification of fodder through the Kotada Bhadli? Was the only difference the use of an opportunistic incorporation of wild grasses may have been a strategy to maintain seasonal foddering practice or were there both social and economic large herds during the period of the weakening of monsoonal rainfall, reasons behind such differences? It is possible that, as these settlements when the settlement of Kotada Bhadli was occupied. It is also possible were engaged in different economic activities, they may also have that incorporation of wild grasses, particularly when they were green, different levels of engagement with agriculture and animal husbandry. might have had a positive impact on the health of the herd. Also, the The year-round supply of and higher dependence on agricultural fodder depletion of grasses near the major sites may have been another reason by the animal owners from both Bagasra and Kotada Bhadli during the for the incorporation of agricultural products into the animal diet. Mature Harappan period may indicate a form of adaptation to the de- Sheep/goats from Kotada Bhadli on the other hand, consumed cline in rainfall during the 3rd and 2nd millennia BCE. At the same mostly C3 leaves, but probably also consumed a small portion of millets time, the diversification of fodder through the incorporation of sea- and/or C4 grasses as evidenced by the indication of C4 in their diet, sonally available wild plants as seen at Kotada Bhadli may indicate a mostly during the months of monsoon when the grasses and millets specialized animal husbandry that was practiced at the settlement, were green and tender. The availability of C4 grasses including both probably along with at least part-time farming and other economic millets and wild grasses to both cattle/buffalo and sheep/goats parti- activities, in the absence of craft specialization, as an alternative cularly during the months of monsoon when the millets were not har- adaptation to these environmental changes. The existence of various vested is likely. A phytolith study is underway to determine the ex- forms of regional adaptation to changing climate has been proposed by ploitation of grasses at the settlement of Kotada Bhadli and will provide Petrie et al. (2017). Along with the diversification of agricultural pro- further information on the nature of fodder exploitation at this settle- duction, diversification of fodder selection by the pastoralists may have ment. been a form of regional adaptation to the regional climate and ecology. In the classification by Khazanov (1994), pastoralists practicing a Differences between Bagasra and Kotada Bhadli are visible in this semi-sedentary form of animal husbandry predominantly feed their preliminary study in terms of their agro-pastoral reconomy, confirming animals with wild grasses, and agricultural fodders usually only play a that these settlements do not only show differences in their material role when the climate is either favorable for agricultural production or culture (Classical Harappan vs. Sorath respectively) but also, their as- when wild grasses are not readily available. The major question for us is sociation with different types of economic specialization. This pilot the extent of migration associated with this form of pastoralism. The study indicates that towards the end of the Mature period of the Indus apparent permanent nature of the settlement at Kotada Bhadli, the Civilization in Gujarat, various types of economic specialization were consumption of both agricultural and wild fodder throughout the year, evident; these specializations were not only confined to the production and the average strontium values of the domesticated animals falling and distribution of non-perishable Harappan goods but were also within the range of the biologically available strontium value of the visible in interdependent agricultural and pastoral practices Kotada region near Kotada Bhadli, all suggest that the animals were probably Bhadli was one such settlement, where a seasonally dependent oppor- raised locally. Based on the year-round storage and/or supply of agri- tunistic mode of specialized pastoralism was practiced, which may have cultural fodder throughout the year, it appears that the pastoralists at been important in Gujarat to support specialized craft producers with Kotada Bhadli were probably sedentary, and their settlement was oc- food, and animal and human transport or labour. Future studies with cupied by at least some members of the group throughout the year. data from small settlements of this type will help to examine the role of Therefore it is most likely that the residents of Kotada Bhadli practiced these smaller settlements in the movement of domesticated animals a sedentary animal husbandry that was supplemented with agricultural between settlements, regions, and traditions through a regional net- production and other specialized economic activities during the Late work. phase of the Mature Harappan period. Due to our incomplete strontium data, however, we cannot completely rule out the possibility of inter- Acknowledgment seasonal movement of animals in search of pasture land, and the pos- sibility of the existence of wild C4 grasses in the vicinity of the settle- We would like to thank Archaeological Survey of India and Gujarat ment is yet to be determined. The likely exploitation of Banni grassland State Department of Archaeology for letting us analyze the archae- by the residents of Kotada Bhadli may have provided the animals with ological material. We would also like to thank Prof. Vasant Shinde, both C3 and C4 grasses, with a higher proportion of C3 particularly Vice-Chancellor of Deccan College, Pune for letting us access the ma- during the period when the settlement was occupied (Pillai et al., terials stored at the repository. The funding to carry out fieldwork and a

197 K.S. Chakraborty et al. Journal of Archaeological Science: Reports 21 (2018) 183–199 portion of analysis were provided by the Department of Anthropology Denmark. Danish J. Archaeol. 1, 113–122. https://doi.org/10.1080/21662282.2012. and the School of Graduate Studies University of Toronto, and from 760889. ffi Frei, K.M., Frei, R., 2011. The geographic distribution of strontium isotopes in Danish Prof. Heather M.-L Miller through research funds provided by the o ce surface waters - a base for provenance studies in archaeology, hydrology and agri- of the Dean, University of Toronto Mississauga, and the funding for rest culture. Appl. Geochem. 26, 326–340. https://doi.org/10.1016/j.apgeochem.2010. of the analysis was covered by Indian Institute of Tropical Meteorology, 12.006. García-Granero, J.J., Lancelotti, C., Madella, M., Ajithprasad, P., 2016. Millets and Pune. We would also like to extend our gratitude to our colleagues, and Herders: the origins of plant cultivation in Semiarid North Gujarat (India). Curr. friends, Nitesh Sinha, Amey Datye and Naveen Gandhi for their tech- Anthropol. 57, 149–173. https://doi.org/10.1086/685775. nical help with IRMS, Sourav Mukhapadhyay for helping us with GIS, Grant, A., 1975. The use of tooth wear as a guide to the age of domestic animals. In: ff Dr. Pankaj Goyal for zooarchaeological analysis, Dr. Genevieve Dewar, Cunli e, B. (Ed.), Excavation at Portchester Castle. Society of Antiquaries, London, pp. 179–245. Dr. Adam Green, Dr. Greg Braun, Dr. Anil K Pokharia, Prof. Vijay Sathe, Grant, A., 1982. The use of tooth wear as a guide to the age of domestic ungulates. In: Prof. Veena Mushfi, and Dr. Rajesh S.V. for their continuous support Wilson, R., Grigson, C., Payne, S. (Eds.), Ageing and Sexing Animal Bones From with sampling, proofreading and discussion. Finally, I would like to Archaeological Sites. British Archaeological Reports, British Series BAR, Oxford, pp. 91–108. thank my parents, my wife Moumita and my friends S. Udaykumar, Halstead, P., Collins, P., 2002. Sorting the sheep from the goats: morphological distinc- Sutapa Lahiri, Esha Prasad, Tejal Ruikar, Bharat Dighe and Devdutta tions between the mandibles and mandibular teeth of adult Ovis and Capra. J. – Phule for their help and support during the fieldwork and analysis. Archaeol. Sci. 29, 545 553. https://doi.org/10.1006/jasc.2001.0777. Heaton, T.H.E., 1999. Spatial, species, and temporal variations in the 13C/12C ratios of C3 plants: implications for palaeodiet studies. J. Archaeol. Sci. 26, 637–649. https://doi. References org/10.1006/jasc.1998.0381. Kenoyer, J.M., 1995. Interaction system, specialised crafts and culture change: the Indus valley tradition and the Indo-Gangetic tradition in South Asia. In: Witzel, A.W. (Ed.), Ajithprasad, P., 2008. Jaidak (Pithad): a Sorath Harappan site in Jamnagar district, Indian Philology and South Asian Studies. Walter de Gruyter, Berlin and New York, Gujarat and its architectural features. In: Osada, T., Uesugi, A. (Eds.), Occasional pp. 213–257. Paper 4: Linguistics, Archaeology and the Human Past. Indus Project. Research Kharakwal, J.S., Rawat, Y.S., Osada, T., 2012. Excavation at Kanmer: 2005-06–2008-09. – Institute for Humanity and Nature, Kyoto, Japan, pp. 83 100. Indus Project. Research Institute for Humanity and Nature, Kyoto, Japan. – Ajithprasad, P., 2009. Excavation at Shikarpur: 2008 2009, Excavation Report for Khazanov, A.M., 1994. Nomads and the Outside World, 2nd ed. University of Wisconsin Archaeological Survey of India. Maharaja Sayajiroa University of Baroda, Vadodara. Press, Madison. Ajithprasad, P., Sonawane, V.H., 2011. The Harappan culture in North Gujarat: a regional Kohn, M.J., Schoeninger, M.J., Valley, J.W., 1996. Herbivore tooth oxygen isotope paradigm. In: Osada, T., Uesugi, A. (Eds.), Occasional Paper 12: Linguistics, compositions: effects of diet and physiology. Geochim. Cosmochim. Acta 60, Archaeology and the Human Past. Indus Project. Research Institute for Humanity and 3889–3896. https://doi.org/10.1016/0016-7037(96)00248-7. – Nature, Kyoto, Japan, pp. 223 269. Lancelotti, C., Caracuta, V., Fiorentino, G., Madella, M., Ajithprasad, P., 2013. Holocene Balasse, M., Ambrose, S.H., 2005. Distinguishing sheep and goats using dental mor- monsoon dynamics and human occupation in Gujarat: stable isotope analyses on phology and stable carbon isotopes in C4 grassland environments. J. Archaeol. Sci. plant remains. Herit. J. Multidiscip. Stud. Archaeol. 1, 288–300. – 32, 691 702. Law, R., 2013. The important stone and metal resources of Gujarat during the Harappan Balasse, M., Ambrose, S.H., Smith, A.B., Price, T.D., 2002. The seasonal mobility model period. Herit. J. Multidiscip. Stud. Archaeol. 1, 319–343. for prehistoric herders in the south-western cape of South Africa assessed by isotopic Lightfoot, E., O'Connell, T.C., 2016. On the use of biomineral oxygen isotope data to – analysis of sheep tooth enamel. J. Archaeol. Sci. 29, 917 932. https://doi.org/10. identify human migrants in the archaeological record: intra-sample variation, sta- 1006/jasc.2001.0787. tistical methods and geographical considerations. PLoS One 11. https://doi.org/10. Balasse, M., Obein, G., Ughetto-Monfrin, J., Mainland, I., 2012. Investigating seasonality 1371/journal.pone.0153850. and season of birth in past herds: a reference set of sheep enamel stable oxygen Lindstrom, K.E., 2013. Pottery Preferences and Community Dynamics in the Indus – isotope ratios. Archaeometry 54, 349 368. https://doi.org/10.1111/j.1475-4754. Civilization Borderland. University of Wisconsin-Madison. 2011.00624.x. Meadow, R.H., 1989. Prehistoric wild sheep and sheep domestication on the eastern Belcher, W.R., 2005. Marine exploitation in the third millennium BC - the eastern coast of margin of the Middle East. In: Crabtree, P.J., Campana, D.V., Ryan, K. (Eds.), Animal – Pakistan. Paléo 31, 79 85. Domestication and Its Cultural Context. University Museum, Univerity of Bentley, R.A., 2006. Strontium isotopes from the earth to the archaeological skeleton: a Pennsylvania, Philadelphia, pp. 24–36. – review. J. Archaeol. Method Theory 13, 135 187. https://doi.org/10.1007/s10816- Meadow, R.H., Patel, A.K., 2003. Prehistoric pastoralism in northwestern South AIsa from 006-9009-x. the Neolithic through the Harappan period. In: Weber, S.A., Belcher, W.R. (Eds.), Bhan, K.K., 2011. Pastoralism in Late Harappan Gujarat, western India: an ethnoarch- Indus Ethnobiology: New Perspective From the Field. Lexingtion Books, New York, aeological approach. In: Psada, T., Uesugi, A. (Eds.), Occassional Paper 10: pp. 65–94. Linguistics, Archaeology and the Human Past. Indus Project. Research Institute for Mehta, R.N., 1984. Valabhi - a station of Harappan cattle-breeders. In: Lal, B.B., Gupta, – Humanity and Nature, Kyoto, Japan, pp. 1 26. S.P. (Eds.), Frontiers of the Indus Civilization. Books and Books, New Delhi, pp. ff Bisht, R.S., 1999. Dholavira and Banawali: two di erent paradigms of the Harappan Urbis 227–230. – Forma. Puratattva 29, 14 37. Merh, S.S., 1995. Geology of Gujarat. Geological Society of India, Bangalore. – – Bisht, R.S., 2015. Excavations at Dholavira (1989 90 to 2004 2005). Archaeological Momin, K.N., 1984. Village Harappans in Kheda district of Gujarat. In: Lal, B.B., Gupta, Survey of India, New Delhi. S.P. (Eds.), Frontiers of the Indus Civilization. Books and Books, New Delhi, pp. Cerling, T.E., Harris, J.M., MacFadden, B.J., Leakey, M.G., Quadek, J., Eisenmann, V., 230–234. Ehleringer, J.R., 1997. Global vegetation change through the Miocene/Pliocene Mughal, M.R., 1997. Ancient Cholistan: Archaeology and Architecture. Ferozsons, – boundary. Nature 389, 153 158. Rawalpindi. fi 18 13 Chakraborty, S., Ramesh, R., 1992. Climatic signi cance of d O and d C variations in a Niu, S., Yuan, Z., Zhang, Y., Liu, W., Zhang, L., Huang, J., Wan, S., 2005. Photosynthetic banded coral (Porites) from Kavaratti, Lakshadweep Islands. In: Desai, B.N. (Ed.), responses of C3 and C4 species to seasonal water variability and competition. J. Exp. Oceanography of the Indian Ocean. Oxford and IBH publication (P) Ltd., New Delhi, Bot. 56, 2867–2876. https://doi.org/10.1093/jxb/eri281. – pp. 473 478. Petrie, C.A., Bates, J., 2017. “Multi-cropping”, intercropping and adaptation to variable Chakraborty, S., Sinha, N., Chattopadhyay, R., Sengupta, S., Mohan, P.M., Datye, A., environments in Indus South Asia. J. World Prehist. 30, 81–130. https://doi.org/10. 2016. Atmospheric controls on the precipitation isotopes over the Andaman Islands, 1007/s10963-017-9101-z. Bay of Bengal. Sci. Rep. 6, 19555. https://doi.org/10.1038/srep19555. http://www. Petrie, C.A., Singh, R.N., Bates, J., Dixit, Y., French, C.A.I., Hodell, D.A., Jones, P.J., nature.com/articles/srep19555#supplementary-information. Lancelotti, C., Lynam, F., Neogi, S., Pandey, A.K., Parikh, D., Pawar, V., Redhouse, Chakraborty, K.S., Bestel, S., Miller, H.M.-L., Shirvalkar, P., Rawat, Y.S., n.d. AMS Dates D.I., Singh, D.P., 2017. Adaptation to variable environments, resilience to climate and Phytolith Data From Kotada Bhadli, Kachchh, Gujarat. change: investigating land, water and settlement in Indus Northwest India. Curr. Chase, B., Ajithprasad, P., Rajesh, S.V., Patel, A., Sharma, B., 2014a. Materializing Anthropol. 58, 1–30. https://doi.org/10.1086/690112. Harappan identities: unity and diversity in the borderlands of the Indus Civilization. Pillai, A.A.S., Anoop, A., Sankaran, M., Sanyal, P., Jha, D.K., Ratnam, J., 2017. Mid-late – J. Anthropol. Archaeol. 35, 63 78. https://doi.org/10.1016/j.jaa.2014.04.005. Holocene vegetation response to climatic drivers and biotic disturbances in the Banni Chase, B., Meiggs, D., Ajithprasad, P., Slater, P.A., 2014b. Pastoral land-use of the Indus grasslands of western India. Palaeogeogr. Palaeoclimatol. Palaeoecol. 485, 869–878. Civilization in Gujarat: faunal analyses and biogenic isotopes at Bagasra. J. Archaeol. https://doi.org/10.1016/j.palaeo.2017.07.036. – Sci. 50, 1 15. https://doi.org/10.1016/j.jas.2014.06.013. Pin, C., Briot, D., Bassin, C., Poitrasson, F., 1994. Concomitant separation of strontium Chase, B., Meiggs, D., Ajithprasad, P., Slater, P.A., 2018. What is left behind: advancing and samarium-neodymium for isotopic analysis in silicate samples, based on specific interpretation of pastoral land-use in Harappan Gujarat using herbivore dung to extraction chromatography. Anal. Chim. Acta 298, 209–217. https://doi.org/10. examine bioshphere strontium isotope (87Sr/862 Sr) variation. J. Archaeol. Sci. 92, 1016/0003-2670(94)00274-6. – 1 12. Pokharia, A.K., Kharakwal, J.S., Rawat, Y.S., Osada, T., Nautiyal, C.M., Srivastava, A., Coplen, T.B., 1988. Normalization of oxygen and hydrogen isotope data. Chem. Geol. Isot. 2011. Archaeobotany and archaeology at Kanmer, a Harappan site in Kachchh, – Geosci. 72, 293 297. https://doi.org/10.1016/0168-9622(88)90042-5. Gujarat: evidence for adaptation in response to climatic variability. Curr. Sci. 100, Ehleringer, R.J., Cerling, E.T., Helliker, R.B., 1997. C4 photosynthesis, atmospheric CO2, 1833–1846. – and climate. Oecologia 112, 285 299. https://doi.org/10.1007/s004420050311. Pokharia, A.K., Kharakwal, J.S., Srivastava, A., 2014. Archaeobotanical evidence of Frei, K.M., 2012. Exploring the potential of the strontium isotope tracing system in millets in the Indian subcontinent with some observations on their role in the Indus

198 K.S. Chakraborty et al. Journal of Archaeological Science: Reports 21 (2018) 183–199

Civilization. J. Archaeol. Sci. 42, 442–455. https://doi.org/10.1016/j.jas.2013.11. Shirvalkar, P., Rawat, Y.S., n.d. Excavation at Kotada Bhadli, Taluka: Nakhatrana, 029. District: Kachchh, Gujarat (2010–2014). Archaeological Survey of India, New Delhi. Pokharia, A.K., Agnihotri, R., Sharma, S., Bajpai, S., Nath, J., Kumaran, R.N., Negi, B.C., Sonawane, V.H., 2002. Post-urban Harappan cultures of Gujarat. In: Setter, S., Korisettar, 2017. Altered cropping pattern and cultural continuation with declined prosperity R. (Eds.), Protohistory: Archaeology of the Harappan Civilization. Manohar, New following abrupt and extreme arid event at ~4,200 yrs BP: evidence from an Indus Delhi, pp. 159–172. archaeological site Khirsara, Gujarat, western India. PLoS One 12, 1–17. https://doi. Sonawane, V.H., 2004. Harappan civilization in Western India with special reference to org/10.1371/journal.pone.0185684. Gujarat. J. Interdiscip. Stud. Hist. Archaeol. 1, 63–78. Possehl, G.L., 1977. Pastoral nomadism in the Indus Civilization: an hypothesis. In: Sonawane, V.H., Ajitprasad, P., Bhan, K.K., Krishnan, K., Prathapachandran, S., Taddei, M. (Ed.), Fourth International Conference of the Association of South Asian Majumdar, A., Patel, A.K., Menon, J., 2003. Excavation at Bagasra-1996–2003: A Archaeologist in Western Europe. University of Pennysylvania Museum, Istituto preliminary report. In: Man Environ. XXVIII, pp. 21–50. Universitario Orientale, Naples, pp. 537–551. Sternberg, L.O., Deniro, M.J., Johnson, H.B., 1984a. Isotope ratios of cellulose from plants Possehl, G.L., 1980. Urban and post urban Harappan settlement patterns in Gujarat. In: having different photosynthetic pathways. Plant Physiol. 74, 557–561. https://doi. Indus Civilization in Saurashtra. B. R. Publishing Corporation, New Delhi, pp. 49–79. org/10.1104/pp.74.3.557. Possehl, G.L., 1986. African millets in South Asian prehistory. In: Jacobson, J. (Ed.), Sternberg, L.O., Deniro, M.J., Ting, I.P., 1984b. Carbon, hydrogen, and oxygen isotope Studies in the Archaeology of India and Pakistan. Oxford and IBH publishing co., ratios of cellulose from plants having intermediary photosynthetic modes. Plant American Institute of Indian Studies, New Delhi. Physiol. 74, 104–107. https://doi.org/10.1104/pp.74.1.104. Possehl, G.L., 1992. The Harappan civilization in Gujarat: the Sorath and Sindhi Thomas, P.K., 2002. Investigation into the archaeofauna of Harappan sites in Western Harappans. East. Anthropol. 45, 117–154. India. In: Settar, S., Korisettar, R. (Eds.), Protohistory: Archaeology of the Harappan Possehl, G.L., Herman, C.F., 1987. The Sorath Harappan: a new regional manifestation of Civilization. Indian Council of Historical Research, New Delhi, pp. 409–420. the Indus urban phase. In: Taddei, M., Callieri, P. (Eds.), Ninth International Towers, J., Gkedhill, A., Bond, J., 2014. An investigation of cattle birth seasonality using Conferecen of the Association of South Asian Archaeologists in Western Europe. δ13C and δ18O profiles within first molar enamel. Archaeometry 56, 208–236. Istituto Italiano Per IL Medio Ed Estremo Oriente, Fondazione Giorgio Cini, Island of https://doi.org/10.1111/arcm.12055. San Giorgio Maggiore, Venice. Uesugi, A., Meena, S., 2012. Pottery. In: Kharakwal, J.S., Rawat, Y.S., Osada, T. (Eds.), Possehl, G.L., Raval, M.H., 1989. Harappan Civilization and Rojdi. Oxford and IBH, New Excavation at Kanmer: 2005-06–2008-09, Kanmer Archaeological Research Project Delhi. an Indo-Japanese Collaboration. Indus Project. Research Institute for Humanity and Price, T.D., Burton, J.H., Bentley, R.A., 2002. The characterization of biologically avail- Nature, Kyoto, Japan, pp. 219–480. able strontium isotope ratios for the study of prehistoric migration. Archaeometry 44, Uesugi, A., Rajesh, S.V., Abhayan, G.S., Chase, B., Rawat, Y.S., Patel, A., Kumar, A., 117–135. https://doi.org/10.1111/1475-4754.00047. Gadekar, C., 2015. Indus ceramics from Desalpur, Kachchh Gujarat. Herit. J. Reddy, S.N., 1994. Plant Usage and Subsistance Modeling: An Ethnoarchaeological Multidiscip. Stud. Archaeol. 3, 180–218. Approach to the Late Harappan of Northwest India. Dep. Antrhopology, University of Verma, S., Menon, J., 1999. The development of “Harappan Culture” as an archaeological Wisconsin-Madison (unpublished). lebel: a case study of Kathiawar. Indian. Hist. Rev. 36, 1–22. Reddy, S.N., 2003. Discerning palates of the past: an ethnoarchaeological study of crop Weber, S.A., 1998. Out of Africa: the initial impact of millets in South Asia. Curr. cultivation and plant usage in India. In: Ethnoarchaeological Series International Anthropol. 39, 267–274. https://doi.org/10.1086/204725. Monographs in Prehistory, USA. Weber, S., Kashyap, A., 2016. The vanishing millets of the Indus Civilization. Archaeol. Rissman, P.C., 1985. Migratory Pastoralism in Western India in the Second Millennium Anthropol. Sci. 8, 9–15. https://doi.org/10.1007/s12520-013-0143-6. B.C.: The Evidence From Oriyo Timbo (Chiroda). Anthropology, University of Weber, S., Kashyap, A., Harriman, D., 2010. Does size matter: the role and significance of Microfilms International (Unpublished). cereal grains in the Indus Civilization. Archaeol. Anthropol. Sci. 2, 35–43. https:// Sarkar, A., Mukherjee, A.D., Bera, M.K., Das, B., Juyal, N., Morthekai, P., Deshpande, doi.org/10.1007/s12520-010-0025-0. R.D., Shinde, V.S., Rao, L.S., 2016. Oxygen isotope in archaeological bioapatites from Wright, R.P., 2010. The Ancient Indus: Urbanism, Economy, and Society. Cambridge India: implications to climate change and decline of Bronze Age Harappan civiliza- University Press, New York. tion. Sci. Rep. 6, 26555. https://doi.org/10.1038/srep26555. http://www.nature. Zhang, C., Wang, Y., Li, Q., Wang, X., Deng, T., Tseng, Z.J., Takeuchi, G.T., Xie, G., Xu, Y., com/articles/srep26555#supplementary-information. 2012. Diets and environments of late Cenozoic mammals in the Qaidam Basin, Sen, G., 2001. Generation of Deccan trap magmas. Proc. Indian Acad. Sci. Earth Planet. Tibetan Plateau: evidence from stable isotopes. Earth Planet. Sci. Lett. 333–334, Sci. 110, 409–431. https://doi.org/10.1007/BF02702904. 70–82. https://doi.org/10.1016/j.epsl.2012.04.013. Sen, G., Bizimis, M., Das, R., Paul, D.K., Ray, A., Biswas, S., 2009. Deccan plume, litho- Zin Maung Maung, T., Takai, M., Uno, H., Wynn, J.G., Egi, N., Tsubamoto, T., Thaung, H., sphere rifting, and volcanism in Kutch, India. Earth Planet. Sci. Lett. 277, 101–111. Aung Naing, S., Maung, M., Nishimura, T., Yoneda, M., 2011. Stable isotope analysis https://doi.org/10.1016/j.epsl.2008.10.002. of the tooth enamel of Chaingzauk mammalian fauna (late Neogene, Myanmar) and Shirvalkar, P., 2013. Harappan migration: a perspective about the Gujarat Harappans. its implication to paleoenvironment and paleogeography. Palaeogeogr. Herit. J. Multidiscip. Stud. Archaeol. 1, 301–318. Palaeoclimatol. Palaeoecol. 300, 11–22. https://doi.org/10.1016/j.palaeo.2010.11. Shirvalkar, P., Rawat, Y.S., 2012. Excavation at Kotada Bhadli, District Kachchh, Gujarat: 016. a perliminary report. Puratattva 42, 182–201.

199