Tamed

by Alice Roberts

  When it comes to the question of exactly where rice domestication first germinated, it seems that archaeologists and geneticists broadly agree – but only broadly. Both genetic and archaeological paths lead back to southern China. But this is a pretty large place. The Chinese geneticists who first published that single-origin-with-later-interbreeding model identified the middle of the Pearl River Valley, in modern-day Guangxi Province, as the homeland of domesticated rice. The impression of timelessness, or at least, deep time, in those famous rice fields of Longsheng could be more than just a romantic notion. (Was Liao Jongpu, I wondered, a direct descendant of the earliest rice farmers? In fact, he’s very likely to be – such is the nature of our infolding family trees when you go back that many generations – along with everyone else in China.)

  The problem with this genetic identification of the Pearl River Valley as the homeland of domesticated rice is that it clashes with the archaeological evidence – the oldest traces of domesticated rice have been found around the Yangtze River, further north. Around the lower reaches of this river, going back between 10,000 and 12,000 years ago, there’s evidence of a growing focus on gathering wild rice, following on from a more sporadic use even earlier. Grinding slabs and wild rice husks have been found in caves and rock shelters in the Yangtze River Valley, dating to more than 10,000 years ago. And then the husks of what appear to be domesticated rice grains have been found in Neolithic pottery at a site in Shangshan, in Zhejiang Province – apparently mixed into the clay to temper it. The pottery dates to around 10,000 years ago. Rice spikelets from the nearby Huxi site, dating to 9,000 years ago, show clear signs of a non-shattering trait – a hallmark of domestication. Rice phytoliths, which vary between wild and domestic rice, have also been used to show a gradual transformation of rice into a domesticate, starting around 10,000 years ago. By 8,000 years ago, several archaeological sites in the Yangtze Valley contain evidence of domesticated rice – based on features of the grains themselves. Then, around 7,000 years ago, the balance starts to shift, with domesticated types starting to outnumber the wild types.

  There’s always the possibility, of course, that the earlier signal from the Yangtze is just an artefact – that researchers have simply been looking there harder, for longer, or have even been luckier in that area – and that earlier sites from the Pearl River Valley lie, as yet, undiscovered. So rather than relying on a handful of sites, one group of archaeologists took a more sophisticated approach and produced computer models of the spread of rice, using as much of the available archaeological evidence from across Asia as possible. These models also predicted an origin – or, even more likely, two, closely connected origins – of rice domestication in the region of the Middle to Lower Yangtze. If I had to make a bet right now, despite my romantic attachment to that wonderful landscape of Longsheng, I’d go for the Yangtze too.

  Winter is coming

  The timing of the beginning of rice domestication is important. At the same moment, right at the other end of Asia, people were also starting to cultivate the wild cereals that grew there – rye, barley, oats and wheat. Between 11,000 and 8,000 years ago, those cereals of the Fertile Crescent become staples – and were transformed from wild grasses into domesticates – just as millet and rice were in the Far East.

  It seems too much of a coincidence: two groups of hunter-gatherers at opposite ends of Asia developing a novel predilection for wild grasses, becoming increasingly dependent on them, and eventually cultivating them as crops. There’s surely something which links these identical changes in human behaviour – something which is at work in the Fertile Crescent, and in the Yangtze Valley, more than four thousand miles away. That ‘something’ is most likely to be climate change.

  During the cold, dry peak of the last Ice Age, wild rice would have been restricted to wetter refugia, in tropical areas of East Asia. As the climate warmed, from around 15,000 years ago, wild rice would have spread, getting an extra boost from the increasing level of carbon dioxide in the atmosphere. Dense stands of wild cereals, replete with grains, presented hunter-gatherers right across Asia with a dependable and easily harvestable food. Under these favourable climatic conditions, wild rice – and millet – may have been a more attractive prospect than it seems today. Perhaps, as we suspect with maize, there were some plants whose characteristics were closer to those which would be bred into whole populations during the process of domestication – plants with larger grains and fewer side-branches – that already looked like a good source of food, and which were easier to gather.

  But around 12,900 years ago, the Younger Dryas hit – the cold drought that lasted more than a thousand years. Faced with a dwindling supply of wild food, people may have been driven to control those resources, to start cultivating the wild grasses they’d become so dependent on. The population boom that occurred just before the Younger Dryas would have meant that resources would have been placed under even more pressure during the coming climatic downturn. For wheat in western Asia and rice in East Asia – and possibly even for maize in Mesoamerica – the Younger Dryas may have been the crucial factor pushing these species together with humans, forging an alliance which would resonate down the centuries and millennia. As a dependable resource, cereals became more important in diets – becoming a staple. Cultivation would be the next step.

  It’s a very different view of human history than perhaps we’re used to. Not a series of triumphant advances, driven by sheer ingenuity and inventiveness, but a story of misfortune and accidents, contingency and serendipity. People falling on hard times and being forced to change their lifestyles, to adapt and accommodate to changing environment. The circumstances which led to cereals becoming staples, and then being cultivated, at opposite ends of Asia, makes more sense if we see it as less of a choice – more of a necessity – driven by a climatic downturn.

  Yet even if the development of cereal cultivation in western Asia and East Asia was somehow motivated by climate change, the Neolithic evolved in very different ways at each end of this vast continent. In the west, the invention of farming preceded the invention of pottery, with a long ‘pre-pottery Neolithic’ lasting from nearly 12,000 years ago to around 8,000 years ago. In East Asia, pottery came first, appearing in the archaeological record much, much earlier than the earliest evidence of agriculture. Instead of a pre-pottery Neolithic, there’s pre-Neolithic pottery. The dates keep getting pushed back and back.

  Pottery from the sophisticated hunter-gatherer Jomon culture of Japan, dating back to nearly 13,000 years ago, was long thought to represent the earliest pottery anywhere in the world. But in the last ten years, evidence of even earlier pottery traditions has emerged in Asia. Sites in eastern Russia and in Siberia have been found to contain evidence of pottery going back an astonishing 14,000 to 16,000 years ago. Analysis of potsherds and associated remains from a cave in Daoxian County, southern China, emerged to suggest a startlingly early date – 15,000 to 18,000 years ago. That study was published in 2009. And then it got even earlier. In 2012, a paper in the journal Science announced the discovery of pieces of pottery from Xianrendong Cave in Jiangxi Province – dated to around 20,000 years ago, right at the last glacial maximum. In China, then, pottery was being used some 10,000 years before the development of agriculture. What were the pots used for? Bones from deer and wild boar were also found in the cave – as well as rice phytoliths. Even this far back, in the Ice Age, it seems that hunter-gatherers were eating a bit of wild rice alongside other plant foods and meat. No residue analysis from the potsherds has been reported yet, but black scorch-marks on the outer surface of the pot fragments suggest that they would have been used over a fire. Even if we don’t know what they were having for their dinner, it certainly looks like the foragers of Jiangxi were cooking something in their pots. The archaeologists reporting on these early pieces of pot talk about the energy gains to be had from cooking starchy foods and meat. But I think that sometimes we focus so much on quite abstract ideas like this that we mis
s a more obvious benefit. In the depths of the Ice Age, hot food must have been something wonderful to look forward to, after a hard, cold day of hunting and gathering.

  Other prehistoric pottery has been shown to be used for storage, food preparation (don’t forget the cheese-makers), and brewing up alcoholic drinks. This is a technology which, in China, predates the development of agriculture, and perhaps even helps to push society in that direction – towards complexity, stratification and a more settled way of life. The details of the story vary, but perhaps, once again, we’re seeing the old idea of agriculture driving the development of complex society being stood on its head. But we have to be careful – the arrival of settled societies and agriculture comes a long way after the first evidence of pottery use in China. Thousands and thousands of years stretch in between.

  Still, the old ‘Neolithic package’ of pottery, sedentism and farming is smashed into tiny pieces with the discovery of the Xianrendong sherds. By the time people are living in villages like that at the Shangshan site, tempering their pottery with rice, they’ve made that transition to a settled way of life, and they are cultivating as well as gathering. But the pot-makers of Xianrendong, twenty thousand years ago, were nomadic hunter-gatherers.

  The march of rice

  Shangshan and other early Neolithic sites in China – dating to around 9,000 years ago (7000 BCE) – provide us with glimpses of the new way of life that would transform people, the landscape, and rice itself. These ancient villages consisted of clusters of rectangular houses, some up to 14 metres long. People were still using old-fashioned Stone Age tools – mostly flakes knocked off stone cobbles – but they also had adzes for hoeing the ground, axes for chopping down trees and grinding stones for pulverising seeds. They were using pottery for storing, preparing and food. They were still, in the main, hunters, gatherers and fishers, but rice would become more and more important.

  By 6,000 years ago (4000 BCE), rice was being farmed, alongside millet, across a broad sweep of land between the Yangtze River to the south and the Yellow River to the north. Rice cultivation continued to spread southwards, with extensive cultivation evident in the Pearl River Valley between 5,000 and 4,000 years ago. Rice cultivation also spread north in China, and to Korea and Japan. Early rice cultivation took off in Japan from around 4,000 years ago – this is when rice-seed impressions appear in Jomon pottery. Rice was probably quite a minor crop at this time, being grown in relatively small quantities alongside more important crops like millet and beans – but of course we know that it would only increase in prominence. It’s now hard to imagine Japanese cuisine without it.

  There’s early evidence of rice use in northern India – which prompted archaeologists to suggest the existence of a separate centre of domestication there. Charred rice grains from the site of Lahuradeva in the Ganges region have been dated to around 8,000 years ago (6000 BCE) – but these now appear to be wild rice. It can be a tricky distinction to make, but wild grains tend to have a smooth-edged, circular abscission scar where they separated from the rachis, while domesticated rice grains tend to have a slightly ragged, kidney-shaped scar. By 4,000 years ago, there’s definitive evidence of firmly domesticated rice, in the form of rice spikelets from the Neolithic site of Mahagara in north-eastern India. These spikelets had clearly developed a non-shattering trait. This is also precisely when japonica was arriving from the east – bringing its domestication genes with it. Other East Asian crops, such as apricots, peaches and cannabis, and stone harvesting knives, similar to those found at older sites in China, also make their way into northern India at this time. Archaeologists suggest that these novelties arrived in India via a networks of exchange – precursors of the Silk Road – connecting the cultures of East and South Asia.

  It’s been argued that, when japonica arrived from the east, it was interbred with early Indian cultivars – which hadn’t quite developed the full suite of characteristics expected of a domesticated variety. then interbreeding between the immigrant domesticate variety and local protodo-mesticates could have produced crops which combined the beneficial traits of domestication with local adaptations to climate – becoming Oryza sativa indica.But Recent research at archaeological sites in north-west India has challenged this sequence of events. At the 4,500-year-old sites of Madsudpur I and VII, 10 per cent of rice grains appear to be domesticated. This seems to be too early for the domestication traits to have been bred in from japonica arriving from the east. Archaeologists have suggested that this raises the possibility that there was indeed an independent, though later, centre of rice domestication in northern India. But this doesn’t fit with the genetic data. Crucially, the domestication alleles in today’s indica rice, which are related to the non-shattering trait as well as to a white-coloured husk, and large grain size, have all come from japonica.

  There seem to be two possibilities here. Either an early variety of Oryza sativa indica had already evolved, with its own domestication alleles, that would later be entirely replaced by japonica alleles, producing the signature we see in rice today. Or – and perhaps more likely – Oryza sativa japonica arrived somewhat earlier than 4,000 years ago in northern India. The only way of settling this question would be to analyse ancient DNA preserved (if indeed there is any) in the grains of rice from Madsudpur.

  This debate also focuses on that important difference between cultivation and domestication. Cultivation is something people do to plants – sowing them, tending them, harvesting them. Domestication describes the genetic and phenotypic changes that occur in species when they are under particular selection pressures which are knowingly or unknowingly produced by humans interacting with that species. Even if the rice of northern India was not a true domesticate until it came into contact with incoming eastern varieties, northern India may yet have been, effectively, an independent centre of agriculture. This is certainly true for other crops: there’s good evidence that local plants such as mung bean and some small-seeded grasses were cultivated on the Ganges plain long before any crops were brought in from elsewhere. Whatever happened right at the start of rice cultivation in India, by the first millennium BCE the crop was being grown right across the subcontinent.

  There’s no such argument about the origin of domesticated rice in West Africa. There, an entirely separate centre of agriculture saw rice being domesticated around 3,000 years ago (1000 BCE) – from a completely different wild progenitor. The Neolithic in West Africa had begun with the introduction of cattle, sheep and goats, and gradually the herders had settled in the landscape, and started cultivating cereals such as rice, sorghum and pearl millet, as well as yams. Early farmers around the Niger River cultivated wild Oryza barthii, which evolved into the domesticated species, Oryza glaberrima, also known as African rice. Genome-wide analyses of African rice suggest that it derived from a single, fairly discrete origin, rather than many separate centres of domestication. These studies also revealed something fascinating about the process of domestication. Geneticists scrutinised the genome of wild and domesticated African rice to look for regions which had been influenced by artificial selection – presumably linked to phenotypic traits which were being selected for. They were interested to see how those traits and genes would compare with those selected for in Asian rice. They looked at homologous (or equivalent) genes in each species – these are genes which are very similar, inherited from a common ancestor of African and Asian rice long before domestication. And what they found were important changes in several genes relating to domestication traits – including husk colour, shattering and flowering. But the changes themselves varied between these domesticated species. For instance, in a certain gene controlling shattering, domesticated African rice showed up a missing section of DNA compared with its wild ancestor. In the equivalent gene in domesticated Asian rice, there was an extra length of DNA compared with wild Asian rice. The effect of these completely different changes to the genetic code – deleting a piece of the code in African rice and inserting a piece in As
ian rice – was the same. Both altered genes were linked to reduced shattering. So the Asian and African farmers had each selected for similar traits in the rice they’d grown, and that selection pressure had led to structurally different, but functionally similar, changes in homologous genes. This was good evidence, not only for similar traits being favoured by early rice farmers in both Africa and Asia, but also for the completely separate domestication of African rice. Unlike Oryza sativa indica, with its domestication alleles bred into it from Oryza sativa japonica, African rice – Oryza glaberrima – had its very own, quite distinct, domestication genes.

  Wet feet and dry fields

  While many plants abhor waterlogged soil, others thrive when fields are flooded. Rice happens to love it, and this secret was discovered back in the Neolithic. The first evidence of wet paddy fields comes from the lower Yangtze Valley, where ancient irrigation systems have been found, dating to the third millennium BCE. There are botanical clues too: archaeologists sifting through the ancient sediments from the Neolithic site of Baligang, situated on a tributary of the Yangtze, have found seeds from wetland weeds, as well sponge spicules and diatoms – tiny algae with silica cell walls – which all speak of well-watered fields, some 4000 to 5000 years ago. The practice spread, and many archaeologists believe that the beginning of wet-field rice farming in Korea and Japan, around 2,800 years ago, reflects an incoming migration of early farmers.