by Rud Istvan
A novel hypothesis on the role of CO2 in the technological transition from hunter/gatherers to sedentary agriculture.
This guest post has been formulating for several weeks. It summarizes a fascinating hypothesis in a guest post by Don Healy at WUWT some weeks ago, albeit posted for a different purpose. Credit where due; this is not my idea. It is merely my formal hypothesis formulation of his original insight, involving a lot of casual research outside my wheelhouse. Yet another Climate Etc. learning experience.
It has been known for decades from archeology that the incredibly important technological transition from small bands of hunter-gatherers to sedentary agriculture (wherever possible) took place about 10,000 years ago. This transition eventually enabled specialization of labor, towns, cities, and all of modern civilization. This technological transition happened all over the world, in many ecosystems, with people geographically isolated, so independently developed. It was spontaneous, similar, roughly simultaneous, and ‘finished’ by about 9000 years ago. There are many archeological hypotheses as to how, why, and most importantly for this post, when. None satisfactorily explain the rough global similarity and simultaneity. We know from ice cores that the atmospheric CO2 level was ~180-190ppm at the last glacial maximum (LGM). We know that it was ~280ppm preindustrial, and that that level had persisted for millennia in the present Holocene interglacial. The natural ocean mediated rise in CO2 from near plant starvation levels to pre-industrial levels fully explains the similar and roughly simultaneous emergence of sedentary agriculture globally, as a simple function of plant primary productivity, foodshed dimensions and natural human behaviors.
Hunter/gatherer to sedentary agriculture transition
There are many aspects to this crucial transition, backed by thousands of science papers. There is no simple, single narrative, although I develop an overarching one here at the risk of oversimplification.
Domesticated dogs (Canis familiaris) emerged from wolf progenitors (Canis lupus), probably twice. First domesticated in Europe about 16000 before the current era (BCE, and note this term itself conveys inherent timing uncertainty). And then, based on DNA evidence, dogs were domesticated a second time in Asia about 14000BCE. Both times likely as a result of scavenging cohabitation, with domestication many millennia before sedentary agriculture emerged.
It is also likely that at least some animal domestication occurred before true sedentary agriculture emerged. This is easy to envision with grazing animals. Cattle were domesticated from wild aurochs perhaps 11000BCE in Meospotamia. Sheep were domesticated from wild mouflons about 11000BCE—and new genetic studies suggest not once, but twice, once in Asia and once in Mesopotamia, both about the same time BCE. (Pig domestication follows as an example of animal domestication as a consequence of sedentary agriculture.) All that was necessary is that the ecological foodshed was within a reasonable distance of ‘home base’ for the shepherds tending the grazers. Domestication would have proceeded in two steps. First, simply taming wild animals so they weren’t dangerous to humans. This is easy to imagine by capture near birth while hunting the mother for meat. Second, selecting for desirable domesticate traits (like docile, else eaten). So animal domestication is a process. Jarod Diamond says it took place before (grazing) or in parallel with (pigs) the development of sedentary agriculture and domesticated crops.
The previous paragraph on grazing domestication introduced the idea of a foodshed. It is central to this climate/agriculture hypothesis, and borrowed from the idea of a woodshed, central to planning dimension lumber and plywood mill capacity in the forest products industry. A ‘shed’ is like a river catchment basin. The foodshed concept is quite simple. A hunter/gatherer band has to forage over an area that during a year (seasonality) provides enough food to sustain that band. A typical band might have been <30-50 individuals based on present day hunter/gathers. The poorer the productivity in the surrounding environment, the larger the band’s foodshed area was. Lower plant productivity means not just lower plant food availability for humans, it means less hunted animal meat from herbivores from other plants.
As a familiar foodshed example, the Neolithic Plains Indians were hunter/gatherers because the conditions for sedentary agriculture never emerged despite agriculture being practiced in the American Southwest, the Mississippi River valley, and the East Coast. The high North American plains blossom in spring from winter snowmelt. But by late summer they are brown, desiccated, dormant grasslands. The bison that grazed those plains had to keep moving to find sufficient food year round since it only grew in spring/early summer. So did the Plains Indians that depended primarily on bison for food, shelter (teepee hides) and dung fire fuel. Their foodshed was very large. Their nomadic ways evolved technical innovations such as the travois and the teepee. But not sedentary agriculture, despite almost certain cultural exposure to it.
Sedentary agriculture transition
About 11,000 year ago, things started changing all over the world. Sedentary agriculture began to to emerge. The most interesting thing is when the first wave of this fundamental technological revolution ‘ended’ with clearly domesticated suites of plants, animals, implements, and permanent settlements. Globally that was about 9000BCE. We will provide specifics in what follows.
The ‘end’ at about 9000 BCE is of course subject to many uncertainties. The definition of plant domestication isn’t precise (following paragraphs give examples). The archeological indicia of domestication are subject to interpretation. For example, is that Neolithic stone tool the cutting edge of a scythe? Is that a grain storage container or just another big pot? (In north China, the archeological evidence is incontrovertible. At Cishan, storage pits were unearthed containing 50,000kg of domesticated common millet radiocarbon dated to between 10,300 and 8700BCE.)
Taming/domestication of grazing animals is only a partial agricultural indicator, for reasons already discussed. The domestication of plants is a better indicator, and also one more archeologically certain. In general, like animals, neolithic plant domestication took two identifiable steps.
The first was indehiscense. That means formerly foraged wild plants lost their ability for natural seed dispersal. Taking a plant category as an example, it meant the grains derived from grasses like emmer (wheat) did not shed their seeds as readily to natural wind and rain when ripe. This indehiscent property would have emerged and strengthened naturally. Hunter/gatherers gleaning what was left of ripe wild emmer would only have gathered seeds from plants already inclined to be indehiscent; the rest was more readily already lost. And every year, human driven selection would have increased that indehiscent tendency.
The second was human selection for productivity. Larger seed size is the typical archeological example. The extreme is maize, human selected from wild progenitor teosinte. Some examples are illustrated below. An exception was the Papau New Guinea domestication of bananas, where larger fruit size was also human selected for smaller seed size.
At several places around the world, the evidence points to the initial transition to sedentary agriculture being ‘completed’ about 9000BCE—despite completely different ecosystems. In Mesopotamia, wheat and barley. In China, millet and rice. In the New Guinea Highlands, taro and bananas. In Mexico, maize from teosinte. In southern Peru/northern Bolivia, the potato.
Take the example of the crop ‘trinity’ of central and north America, maize (corn), beans, and squash. Nutritionally, these together provide all 20 essential amino acids. So, from a modern nutritional perspective, the ‘three sisters’ explain the huge population growth of Peruvian Incans, Yucatan Mayans, and Mexican Aztecs without an abundance of domesticated animals for meat.
Modern genetic analysis proves beyond doubt that maize originated from the Mexican highland C4 grass teosinte, the original domestication transition finishing about 9000BCE. This is the most famous phenotypical human selected change of a wild type precursor plant.
Modern genetic analysis of the common bean (P. vulgaris, another of the Amerindian ‘trinity’ of maize, beans, squash) shows it emerged at least two separate times (and possibly three) nearly simultaneously. Domesticated P. vulgaris began independently in Peru, (landrace kidney beans), in Mexican highlands (landraces pinto and red beans), and probably once more in lowland MesoAmerica (landraces black and navy beans), all around 9000BCE (depending on archeological site and landrace, anything between 10000BCE and 8000BCE). These landraces are all genetically similar, varying mostly in phenotype. How the phenotypes could vary so much without much underlying genetic differentiation is now understood thanks to growing knowledge of epigenetics.
The third of the Amerindian trinity, squash (C.papo, predecessor to the pumpkin) was also domesticated before 8000 BCE.
As a final near simultaneous timing example, hogs (pigs) are non-foraging so have to be fed domesticated plants. The pig was domesticated from the Eurasian wild boar (Sus scrofa) of which there are several subspecies that can be distinguished by variation in mitochondrial (maternal) DNA. Based on analysis of present day wild and domestic types, domestication happened independently in the Near East and in China from at least two separate wild boar subspecies at about the same time ~9000BCE. The dating is quite solid since based on mDNA mutation rates (a new standard ‘clock’ method). Darwin knew there were two basic domestic pig types, but not why. The Asian domestics were introduced into Europe for further cross breeding in the 18th and 19th centuries (the ‘subsequent introgression’ part of the linked paper title).
Similar domestication timing around the world despite very different ecosystems and agricultural crops cannot be a coincidence. Nor can it be from cultural diffusion of agricultural knowledge; at that time, these areas were geographically isolated.
Archeological explanations for the transition
There are several ‘explanations for this Neolithic revolution in technology. Wiki has a decent overview.
The ‘oasis’ theory has largely been discredited.
The ‘hilly flanks’ theory might explain Mesopotamia, but not Borneo.
The ‘stable Holocene climate’ theory is falsified by the Younger Dryas.
The ‘Younger Dryas’ theory might explain Mesopotamia. But not Bolivia or Borneo.
The ‘it wasn’t simultanous’ theory depends heavily on dogs and grazing animals while overlooking much of the solid evidence for near simultaneity in plants and foddered animals like hogs or chickens.
In short, it appears no one had provided a good explanation for simultaneity before Don Healy’s post.
CO2 explanation for Simultaneous Transitions
Ice cores suggest that the CO2 concentration at the last glacial maximum (LGM ~21,000BCE) was 180-190ppm. That compares to ~280 ppm pre-industrial and ~400 ppm now (with noticeable greening over the past 35 years). The change in CO2 concentration from the LGM to the Holocene is easily explained based on Henry’s Law and reduction in warming ocean dissolved CO2. The ice cores show that the CO2 rise lagged temperature rise by about 800 years, the period of the thermohaline circulation.
For C3 photosynthetic pathway plants, that LGM CO2 level is quite low, hindering their productivity. About 85% of all plant species are C3. All trees, fruits, vegetables, and most food crops are C3. The only C4 food crop exceptions are maise, millet, sorghum, and sugarcane. There is an excellent long review paper by Gerhard in New Phytologist (2010) on this general topic. The paper describes experiments with various C3 plants over a growing season at the pre-industrial CO2 280ppm and at 150ppm, below the LGM level. On average, at 150ppm the primary C3 plant productivity was reduced an average 92% as measured by dry weight biomass.
So at LGM CO2 levels, foodsheds would have necessarily been quite large, since plant photosynthetic productivity was low. Sedentary agriculture would have been impossible. As CO2 concentrations rose, foodsheds shrank. Eventually they would have shrunk to the point that permanent settlements with food storage became possible compared to nomadic temporary shelters, with food resources perhaps within just a couple days walk. At that point, the near simultaneous emergence of sedentary agriculture around the world–despite very different ecosystems–became inevitable.
That has proven to be very beneficial. Climate change increased natural atmospheric CO2, which in turn enabled agricultural technology development. Which in turn enabled modern civilization. Which enabled exploitation of fossile fuels. Which further raised beneficial atmospheric CO2. The current further greening documented by NASA is also beneficial, with crop yields rising while needing less water. The opposite of alarming.
Moderation note: As with all guest posts, please keep your comments civil and relevant.