Crops
Rabi Crops. Wheat and barley were the staple cereals of rabi cultivation. The Harappans cultivated various types of wheat: a little emmer and einkorn, along with three kinds of bread wheat, of which shot wheat (Triticum aestivum sphaerococcum) was the most common in the Mature Harappan period. Barley was more important than wheat at some sites, including the Indus outpost at Shortugai on the Amu Darya and the Baluchi site of Miri Qalat. The Harappans grew three or four varieties of barley, including both naked and hulled types. This range of crop varieties allowed them to exploit the different properties of the various types of land suitable for cultivation. At Rojdi in Gujarat, barley was very poorly represented in the extensive collection of botanical remains and was not cultivated after period A (2500-2200 BCE), and in the Kachi plain bread wheat was more important than barley. Oats (Avena sp.) were present at Mehrgarh in the fourth millennium and have also been recovered from Pirak and Late Harappan Hulas. Oats seem generally to have been present in early archaeological contexts as a weed of cultivation that invaded stands of wheat and barley, rather than being deliberately cultivated: This fits with their sporadic appearance in South Asian botanical samples.
A number of other crops were cultivated during the rabi season, the majority—including lentil, pea (or field pea, Pisum sativum var. arvense), chickpea, and Linum—probably originally domesticated in West Asia. Peas are known from sites such as Kalibangan, Chanhu-daro, and Harappa, chickpea at Kalibangan, and lentils at Nausharo and Rojdi, though peas were not cultivated at the latter site until the Late Harappan period when all three were also cultivated at Hulas; chickpeas are also known at Pirak. Lentils and peas were among the plants cultivated at Shortugai, the Indus outpost in northern Afghanistan. Another pulse, also possibly derived from the west, was grasspea (chickling vetch, Lathyrus sativus). Although remains of this plant are known
A pit from the Ravi period at Harappa containing seeds of wheat and barley. These were the main staples of agriculture in the northwest until the last centuries of the Mature Harappan period. (Harappa Archaeological Research Project, Courtesy Department of Archaeology and Museums, Government of Pakistan)
From a number of Harappan sites and at Late Harappan Hulas, it was probably not grown for human consumption, since it is poisonous to humans if eaten in large quantities. Its recent consumption has been as a famine or desperation food, and it may have been grown for animal feed.
Millets. During the third millennium, a number of indigenous cereals were brought under cultivation by the Indus civilization or by contemporary South Asian cultures. Little millet (Panicum sumatrense) was common at Mature Harappan Rojdi, Oriyo Timbo, and Babar Kot in Gujarat, and present at Harappa around 3000 BCE, and browntop millet (Brachiaria ramosa) was also grown at Rojdi. A small amount of Setaria sp. was cultivated at Surkotada and Rojdi: This may have been S. verticillata, bristley foxtail millet, also domesticated in South India during the third millennium, or S. pumila, yellow foxtail millet, both native species. Foxtail millet (Setaria italica), known in the Late Harappan period, is thought possibly to be a local domesticate but was more probably introduced. It was a major crop in China, having been brought under cultivation in the seventh millennium BCE, and was being grown as far west
As Tepe Gaz Tavila in southeast Iran by the sixth millennium. Seeds of another indigenous millet, Job's tears (Coix lacrima-jobi), have been found at Harappa and at the contemporary Ahar-Banas settlement of Balathal, in both cases as beads, a common use for these seeds.
Broomcorn (or common) millet (Panicum miliaceum) was probably brought under cultivation in southern Central Asia (as well as in China) and might have reached the Indus civilization via their trading outpost at Shortugai, which was situated in the region adjacent to southern Turkmenia, where broomcorn millet was an important crop. A wild ancestor of broomcorn millet exists in South Asia, so it may alternatively have been a local domesticate. Several species of Panicum were present at Rojdi, and it is possible that broom-corn millet was among them. The first certain occurrence of this millet in South Asia is at Pirak, in the early second millennium.
During the early second millennium, a number of plants of African origin appeared in Gujurat and were incorporated into the range of crops grown by the local Harappans. These included three kinds of millet: jowar (sorghum or Guinea corn, Sorghum bicolor), bajra (pearl millet, Pennisetum typhoides), and ragi (finger millet, Eleusine coracana). Abundant ragi was reported at Rojdi during the earlier part of the Mature Harappan period, from about 2500 BCE onward, as well as possible ragi phytoliths in bricks and sherds at Harappa, but its presence this early is unlikely. Dorian Fuller (2001, personal communication), an ar-chaeobotanist with a detailed knowledge of South Asian plants, cautions that it is likely that some claimed occurrences of ragi are based on a misidentification of Setaria spp., Echinochloa colona (Sawa millet), or Brachiaria ramosa (browntop millet), all native South Asian millets; a native weedy grass (Eleucine indica) was also abundant at Rojdi. Later there was ragi in Cemetery H levels at Harappa and in Late Harappan Hulas to the east, and Fuller himself has identified a grain of ragi at Hallur in South India, dated after 1800 BCE.
Bajra may have been present in late third-millennium Babar Kot in Saurashtra; it is known later at Rangpur and also reached South India after 1800 BCE. Jowar was a major crop at Rojdi during the early second millennium, and, in the Posturban period, jowar is also reported at Pirak and at Late Harappan Hulas. The timing of the appearance of these higher-yielding African millets in Gujarat coincides with the period of considerable increase in settlements in the region and the expansion of cultivation into areas of moisture-retentive soils. Other local millets, kodon (Paspalum scrobiculatum) and sawa millet, were also added to the range of crops in the period after 2000 BCE.
Although the evidence is very uncertain, it is possible that some African crops were under cultivation in Oman (ancient Magan) during the third millennium, and sorghum and another African millet may also have been cultivated in the Yemen at this time, though their identification is not certain. These plants are assumed to have spread due to contacts across the Red Sea between southwest Arabia and East Africa, regions between which there were close communications in later times. Unfortunately the third-millennium archaeology of both areas is poorly known, and the earliest record of these crops in
Africa greatly postdates their appearance elsewhere, although there is much earlier evidence of their exploitation as wild plants. It is also possible that some of these plants were part of the original flora of the Yemen and were taken into cultivation there. From southwest Arabia the domestic plants probably spread via local exchange networks through the southern coastal region of Arabia as far as Oman. Alternatively, the crops may have been carried as provisions by seafarers plying the sea lanes of this coast, who on their return home handed over the residue as novelties that were used in experiments. If these commodities were carried by organized sea traders, the latter may have come from either Oman or the Indus.
Rice. Rice is indigenous to parts of South and East Asia, including the Indus region and the Ganges Valley. The history of its cultivation is complex and probably involved a number of different centers of domestication. Genetic evidence has recently established that rice was brought into cultivation in at least two separate areas: domestication of a perennial wild rice in East Asia produced the short-grained japonica variety whereas domestication, probably in several regions of South Asia, of an annual wild rice gave rise to the longgrained indica variety, which also spread through Southeast Asia and China.
Rice cultivation began in the middle Ganges region during the third millennium and somewhat later in eastern India. The cultures growing rice in Southeast Asia had close cultural connections with the inhabitants of eastern India, Bangladesh, and intervening regions, indicated by shared artifact types such as cord-marked pottery and distinctive shouldered axes.
Rice grew wild in Gujarat. Charred rice husks and impressions of rice husks and leaves in Harappan pottery have been found in this region, at Lothal and Rangpur. These have been studied by Naomi Miller, who has established that they are unlikely to reflect rice cultivation. Instead it is probable that rice was among the wild plants consumed by grazing cattle resulting in rice husks being present in their dung, which was used for fuel and a tempering agent in pottery. Rice husks and phytoliths have also been found in pottery and bricks at Harappa. Rice, probably wild, is known from Early Harappan Balu in Haryana and Kunal. In Swat rice appears at Ghaligai before 2000 BCE as grain impressions in sherds of Late Kot Diji pottery: These may have been from either domestic or wild rice. By the early second millennium, however, rice was certainly being grown in the eastern Indus region. It was among the cultivated plants at the Late Harappan site of Hulas, where both wild and cultivated indica rice were identified.
Japonica rice was grown in China's Yangtze Valley by the sixth millennium BCE, and its cultivation spread from there to other parts of China and into Southeast Asia. From northern China, rice cultivation spread to Manchuria and Korea. It was possibly from here that rice cultivation reached Kashmir, a region that had a number of links with China: Rice began to be grown at Gufkral in Kashmir during the first half of the second millennium. Rice, apparently japonica, was the principal crop of the settlement at Pirak in the Kachi plain, an arid region where irrigation would have been required.
Other Food Plants. South Asia had a number of native pulses that were locally domesticated. These included green gram (Vigna radiata) and black gram (Vigna mungo), which were grown at a number of Mature Harappan sites and at contemporary Balathal in Rajasthan. Horsegram (Macrotyloma uniflorum) was domesticated in South India during the same period and is known from Late Harappan Hulas. During the early second millennium, two further pulses, of African origin, were added: hyacinth bean (Lablab purpureus) and cowpea (Vigna unguiculata), the latter being grown at Hulas and both appearing in South India after 1800 BCE. All varieties of pulse were more important in peripheral regions such as Gujarat than in the Indus Valley heartland.
Very few other Harappan cultivated plants have been recovered. There is evidence, however, of the widespread cultivation of a species of Brassica, brown mustard (Indian rape), and of gourds in the Mature Harappan period, and later of ivy gourd, while okras were grown at Balathal in the neighboring Ahar-Banas region. Jujube (ber, Zizyphus jujuba), an edible red berry, was known at Mehrgarh from the earliest period, though it was probably gathered rather than cultivated; this may also have been true of its later use. Melons were cultivated at Shahr-i Sokhta in adjacent Seistan and probably by the Harappans. Other fruits that may have been grown or collected locally include caper, mango, and sugarcane, and adjacent regions may also have supplied fruits, vegetables, and nuts, including cucumbers, pistachios, almonds, and walnuts, all known from sites farther west; walnuts have also been recovered from Hulas, along with the fruit of the pipal tree (Ficus religiosa).
The stones of dates, a high-calorie fruit that is three-quarters sugar, were found in early Mehrgarh; date palms occurred wild in Baluchistan and they may have been cultivated from early times. Many date stones were recovered from Nausharo and at Mohenjo-daro, and it is likely that dates were transported to parts of the Indus region where they were not grown.
Grapes were being grown in the Kachi plain by the early third millennium, as well as in adjacent Baluchistan and Seistan. Grape pips were found at Mehrgarh and Nausharo and later at Pirak I; they were also common at Shortugai, the Indus outpost in northern Afghanistan.
Herbs and spices, such as garlic, turmeric, ginger, cumin, and cinnamon, are likely to have been grown or collected too, but the only trace yet identified is of coriander at Miri Qalat in Baluchistan.
Sesame, native to South Asia, was probably the principal plant grown for its oil: It is known from a number of Harappan sites, including Chanhu-daro and Harappa, and contemporary sites in the Indo-Iranian borderlands such as Miri Qalat. By 2250-2200 BCE, sesame was under cultivation in Mesopotamia, presumably first brought there by Harappan traders. Castor, another Indian oilseed, was cultivated at Late Harappan Hulas.
Fibers. Oil could also be obtained from linseed (Linum usitatissimum), which was found at Miri Qalat and a number of Harappan sites, including Nausharo and Rojdi. Alternatively it may have been grown for its fiber, flax. The latter was being used to manufacture linen cloth during this period on the Iranian
Plateau; however, no linen has been identified from Harappan sites. There is evidence of cotton cloth at Mohenjo-daro and probably Harappa: The production of cotton textiles may have meant that linen was of no interest to the Harappans. Cotton may have been cultivated at Mehrgarh by the fifth millennium, though, like Linum, it may also have been grown for its oil-rich seeds. In the Mature Harappan period it was grown in both the Indus Valley and Baluchistan. Locally available plants, such as indigo and turmeric, were probably used as dyes; indigo is among the plants recovered from Rojdi, and the use of madder root is attested to by the presence at Mohenjo-daro of cloth dyed red with madder.
Water and Irrigation
Irrigation Works. In Baluchistan the sparse winter rainfall, though important, could not be relied on to water the crops raised in the generally limited areas of suitable soil. Water could be obtained from wells and springs in some cases, but by the early third millennium, if not before, the inhabitants of the region also developed small-scale dams (bunds and gabarbands) to retain some of the water that flowed in seasonal streams and small rivers (nais) after the rains. In some cases, for example at Early Indus Diwana on the upper Hab River, a dam was designed to impound water, which could be released or channeled onto fields as required. In other cases dams and channels led the floodwater into embanked fields, where they deposited silt and provided enough soil moisture for the growing crops. One type of dam consisted of small walls built to jut out into the bed of a stream or river so that some of its water was diverted onto the ground behind the wall, depositing fertile silt that formed a small field. Settlements in the Kulli area (southern Baluchistan) seem invariably to be associated with dams; this area also received some unreliable summer rainfall.
On the edges of the Kachi plain and along the western piedmont of Sindh, as in Baluchistan, small dams and occasional small channels were created to retain and distribute the seasonal runoff from mountain streams and rivers. These nais carried their greatest flow in July and August but also brought minor floods in January and February from the limited winter rainfall, which also provided supplementary water during the growing season. Some nais, fed by springs, had a small perennial flow.
The Kachi plain is a hot arid region where agriculture relied on the limited winter rainfall, heavier but not always reliable monsoon rains, and the water provided by the Bolan, Mula, and Nari Rivers. Here also dams and channels were necessary to make best use of the water supplies, and only in this area of the lowlands have irrigation channels been found.
Canal irrigation is attested to at Shortugai, the Indus outpost in northern Afghanistan at the confluence of the Amu Darya and Kokcha Rivers: A canal has been traced that drew off water from the Kokcha. This might be taken to indicate that the Indus people brought canal irrigation technology with them when they settled here; however, the Namazga culture in adjacent southern Turkmenia, from whom it is likely that the inhabitants of Shortugai acquired
The broomcorn millet that they cultivated, had long experience of canal irrigation that may have inspired the inhabitants of Shortugai.
Water Supply. Unlike the situation in the mountains and foothills of the Indo-Iranian borderlands, there is little evidence that major irrigation works were used or required over most of the Indus region. Groundwater, rivers, lakes, streams, and especially floodwaters sufficed.
The Indus floods in Sindh came largely in July and August, providing water throughout the summer for kharif crops while winter crops were sustained by the water retained in streams, channels, lakes, and dhands (seasonal lakes), supplemented by water brought down in January or February by the nais flowing off the mountains of Baluchistan. The Indus plains had a variety of zones suitable for agriculture. The margins of dhands and oxbow lakes, the latter formed by abandoned meanders of the Indus, allowed cultivation from year to year. The active flood plain of the rivers provided excellent arable land, its fertility renewed annually by the silts deposited by the floodwaters, the coarse sediments closest to the river being richest in nutrients. Patches of deeper sediment reflected the unpredictable distribution of channels cut by the river's floodwaters: These had to be searched for, but they provided the best agricultural land, cultivable without plowing. In western Sindh, Lake Manchar flooded an enormous area during the inundation, and the retreating floodwaters left fertile ground highly suitable for cultivation: Today this is around 8,000 hectares in extent. While the productivity of the Indus in Sindh is very high, it is not reliable. About one year in four brings abnormally high or low quantities of water; the river floods unevenly, depending on where it breaks its banks; and it changes its course frequently. This combination of high but unpredictable productivity must have made it advantageous to develop storage practices and facilities.
Farther north in the Punjab, farming was confined to the alluvial soils in the valleys, annually flooded by the five rivers, with large areas of uncultivated higher ground between them. This region received monsoon rain in the summer, and some winter rainfall, particularly in the western portion. The area southeast of the Punjab, the eastern part of the ancient Saraswati system, is prime agricultural land well watered by numerous seasonal watercourses and by heavy summer monsoon rains. Although today the numerous rivers and streams carry only seasonal flow in their upper reaches and are dry farther west, in the Indus period this river system carried far more water and flowed at least to the Fort Derawar area in Cholistan, if not beyond, with perennial flow in the larger rivers. Dense settlement along branches of the Saraswati River indicates that it was one of the most productive regions of the Harappan realms and underlines the significance of the progressive reduction in the volume of water carried by the Saraswati River system in the Late Harappan period. Agriculture was supported by the floodwaters of the rivers, with their burden of alluvium; since the rivers were fed by water from the Siwaliks rather than the Himalayas, the volume of water they carried was considerably less than that in the Indus and its tributaries, with proportionally less violent floods. In Cholistan, water for agriculture could also be obtained from shallow
Wells tapping the high water table of the river valleys, and there was also some summer and winter rainfall.
Complex irrigation systems have been sought, but it seems likely that none were required, agricultural settlement being confined to the riverine environments where simple means of water provision were adequate, in contrast with the highlands where water conservation was essential. In some parts of the Indus realms, particularly Sindh, small channels were probably dug to bring water from dhands or streams into fields and to carry away excess water from swampy areas. Neither irrigation nor drainage channels have been located, but this does not mean that they did not exist. The annual deposition of alluvium filled in many irregularities of the plain's surface, which would have included artificial channels, and the unpredictable distribution of the inundation waters meant that the location of fields would often have changed. These factors mean that new channels would have had to be dug each year, rather than cleaning out old ones, and would have made it inappropriate to invest effort in constructing major irrigation canals. Any surviving traces of such channels must by now be deeply buried beneath four millennia's alluvium.
Most Harappan farming settlements in Gujarat were located in Saurashtra. In the Mature Harappan period, these were confined to locations along the rivers and streams, and particularly along the Nal Depression, which retained floodwater through the winter months. Only in the Late Harappan period did farming settlements spread onto the moisture retentive, black cotton soils in other parts of Saurashtra, where kharif crops could be raised, watered by rainfall brought by the summer monsoon. The number of settlements in the region expanded at least fourfold in this period. Kutch, to the north of Saurashtra, was an island in the Indus period. Today the brackish subsoil water and poor rainfall provide little support for arable agriculture, but in Indus times, when a considerable flow of river water entered the Ranns, the underground water was probably sweet and could have been accessed for irrigation by digging wells.
Wells and reservoirs also supported the inhabitants of Dholavira on Khadir Island in the Great Rann. Wells here and in other regions could provide ample water for growing crops. Drawing water from them would have been a laborintensive activity, requiring considerable animal power, though in areas subject to summer flooding only shallow wells were needed to reach the high water table. A masonry well at Allahdino may have been used for irrigation: It was situated on higher ground, from which the water could run down to the fields. The fine examples of wells in Indus towns show the high level of Harappan competence in constructing them. In the central region, Sindh, the Indus-Ganges doab, and perhaps the western Saraswati, the floods filled numerous hollows (dhands), which for some months acted as reservoirs from which to draw water to irrigate the crops; many held water until December and some as late as February. The Indus people probably used lifting gear such as the shadoof to raise irrigation water from these and from streams and channels. One sherd of Indus pottery from Mohenjo-daro bears a scratched picture of such a device, a simple T-shaped arrangement of an upright and a horizontal pole, with a bucket on one side and a counterweight on the other.
More than seven hundred wells were sunk at Mohenjo-daro when the city was built. Over the centuries houses were rebuilt and street levels rose; new courses of bricks were therefore added to the wells to keep their tops at the same height with respect to the street. The removal of earth and debris during the excavation of the city has left many wells standing like towers high above the exposed remains of earlier streets. (J. M. Kenoyer, Courtesy Department of Archaeology and Museums, Government of Pakistan)
World History
