The eastern fertile crescent returns

The recent paper in Science by Riehl et al. on  the evidence for Chogah Golan has rightly garnerd wide attention (e.g. Science news; commentary by Willcox). This is a highly significant paper, which shows that the beginnings of cultivation were indeed mutlicentric within the fertile ccrescent, and it suggests that there was an independent domestication process for emmer wheat in the eastern fertile crescent in addition to that in the western fertile crescent. 

Are there surprises? Yes. The big surprise here is the emmer wheat domestication, as many have argued on biogeographical and modern genetic grounds that there should have been and eastern and western barley domestication, but this has been little considered for wheat. This is mainly because the modern distribution of wild wheats does not extend that far east and south, and thus the data from the Chogha Golan, especially the lower levels indicates that the distribution of wild wheats at the start of the Holocene/end of the Pleistocene was indeed different and more extensive than modern wild wheats. This further implies that some starts to cultivation and domestication events could have drawn on wild population that are extirpated today and therefore are not reflected in modern wild germplasm collections. Modern collection used in genetic studies are only a fragmentary representation of the past, although geneticists often fall into the trap of assuming that good wild sampling in the modern day means they have captured the wild diversity from which domestication began.

What this what we/ I suspected? Yes. I am one of a number of scholars who have been arguing for a multicentric process of parallel starts to cultivation and parallel, and protracted, domestication processes around the Fertile Crescent, i.e. De-centering the Fertile Crescent. Mostly we have argued this on contrasts between the Southwest and the north/central fertile crescent and the contrasts between morphological diversity in archaeological samples and that in modern germplasm. As the authors here note with their triticoids, they are dealing with a wild wheat type not well represented in modern collections; this is equally true of early domesticated wheats in Syria/Anatolia and even in Neolithic Europe. In the Neolithic there are extinct genetic lineages, that are morphologically distinct, that are not found in modern landraces. In other words there are several lost crops of early agriculture. 

Were things really synchronous? This remains a little unclear. The lower levels seems to have pre-domestication cultivation of barley and lentil and lost wild wheat in the equivalent of PPNA/ EPPNB time periods-- this is indeed the same period that we see this in Jordan, Israel, north and south Syria. However in the Chogha Golan there is then a break and a large minority of domesticated type emmer appears. But this is mainly in the Late PPNB (ca. 7800 BC)! By this period domesticated crops are well established at higher frequencies (usually 60-70% non-shattering spikelet bases) in western fertile crescent assemblages (mainly of einkorn wheat or barley). Emmer at Tell Aswad in Syria is ~23% non-shattering at 8300 BC and at Tell el-Kherkh in NW Syria it is 44% at ca. 8400 BC. This evolution of non-shattering (a key domestication trait) appears  slightly ahead in the west. This could mean the that 25%-domesticated assemblage at upper Chogha Golan has spread from early cultivated population elsewhere that were undergoing the gradual selection for non-shattering, or it could indicate a local process, maybe not at Chogha Golan, but nearby that simply got started a bit later. 

I would note in passing, that there was one previous suggestion of eastern emmer domestication, many years ago by Hans Helbaek in the 1960s based on rather poor samples collected by the Braidwood expedition at Jarmo in Iraqi Kurdistan in the 1950s, in which Helbaek reported intermediate types and mixtures of wild and domesticated emmer. These data were never quantified nor fully published but would potentially fit with the Chogha Golan finds. So a return to Jarmo may yet have some important archaeobotanical contributions to make.

On the whole, however, these new data offer strong support from a new dataset and a different research group for what I have been championing as a paradigm shift in agricultural origins research. From the paradigm of a rapid and singular agricultural revolution to a paradigm of protraction and entanglement that was messy and non-centric. (See, e.g. refs 2 and 19 cited by Riehl). 

A genome map that is not a map of origins (Rice Genetics Watch returns)

Last week Nature ran an article (Huang et al) with the headline that " A map of rice genome variation reveals the origin of cultivated rice." I here to report that this paper does not do what is says. There is nothing obviously relevant to locating where rice was first brought into cultivation, and the claims in the article are misleading and misguided. This is apparently one of most read Nature papers at the moment, so no doubt we will have to face lots of additional confusion over rice domestication-- and I thought there was already enough confused and misguided info out there. I have had several queries on this over the past week, so below is my quick response.

 There is some important data here and details, and much for further critical analysis. BUT: This study changes nothing. Its stated conclusions are misleading, making false unstated assumptions and arriving at unreasonable and unbelievable conclusions.  In a way this mistake was inevitable and obvious. The authors have concluded the the closest wild ancestors to cultivated rice are living wild populations in the  Pearl River basin. The problem is that rice was domesticated not from living populations but from past populations almost certainly from regions where wild rice is now extinct (technically, we would say, extirpated). This study demonstrates that big science and lots of resources do not inevitably produce answers, but that nuanced analysis and critical thinking, and in this case some knowledge of Chinese history, are necessary to direct analyses.

It is clear that wild rice (O. rufipogon) formerly occurred much further north, through much of the Yangtze valley and even as far north as the Shandong peninsula and lower Yellow River basin. This is clearly attested from Chinese written sources of the Song Dynasty (i.e. about 1000 years ago). Even by that period it is likely that wild rice distribution was greatly reduced by the impact of China’s huge human population and agricultural expansion which took place between 6000 years ago and 1000 years ago. More so than anywhere else on earth central China (from the Yellow river  to the Yangtze) has supported massive human populations and suffered the corresponding habitat loss. In the late Bronze Age (Zhou dynasty), they were hunting elephants on the banks of the Yellow river (for a wonderful book on Chinese environmental history that takes this as representative of the broader sweep of Chinese history, see the Retreat of the The Elephants by Mark Elvin 2004). These would certainly not be represented in a genetic study of living elephant populations! (as blogged previously these may actually be an extinct elephant species with straight tusks)

Although reference 2 in the article is to a paper I co-authored (Fuller et al 2010), this study clearly did not take on part of the fundamental implications of the maps and discussion early in that paper about the past distribution of wild rice, which has been modified both by major climatic change since the wetter and warmer early Holocene and by the impact of habitat destruction by Chinese farmers since the Neolithic. Areas that could support wild rice made excellent areas for agricultural reclamation: domesticated rice replaced wild rice over much of its original range in central China, which had the highest human populations. Even clearer, I think, is the paper I published in the journal Rice in 2011, which includes maps and a phylogenetic diagram illustrating the fallacy of using modern extant wild rice to represent the full diversity of past wild rice. By making this assumption in pinpointing a pearl river origins for rice pretty much all the authors subsequent conclusions are inevitably problematic. The  only way oin which genetics is going to advance pinpointing the number and location of domestication events in rice is through the recovery of ancient DNA. The fallacy an approach that relies purely on the modern time-frame of sampling is well-illustrated with European pigs and boar genetics, in which only via recovery of ancient DNA is it possible to see that the first Neolithic pigs were derived from Near Eastern boar and pigs but were later replaced by genetics from European wild boar (see: Larson G, et al. (2007).

That modern populations of Oryza rufipogon are not the direct ancestors of japonica rice is implicit in the data in fact. The “obvious genetic distinction between japonica and Or-IIIa (Fig. 2a)”, implies that domestication rice and South Chinese rufipogon are in fact not really so close, just the closest available in linving populations. The intermediates found with Or-1 and indica are because Indian wild rice have been less decimated by the combination climatic changes and human impacts. Indeed this pattern is not new, but was already evidence some years ago, especially in the study of Cheng et al 2003. (Polyphyletic origin of cultivated rice: Based on the interspersionpatterns of SINEs). —this is discussed on the basis of the more detailed Ohtsubo et al paper or 2004 in my 2010 paper and various earlier articles in the archaeological literature). It is nice to see a much larger dataset in the this new paper re-affirm the results of the  p-Sine study, but there is not really anything new accept that the present authors have tries to grab a headline by claiming a Pearl River  origin for rice. It is the populations that bridge the gap between OR-IIIa and japonica which are crucial and these must be extinct populations of Oryza rufipogon that were brought into cultivation in the earlier Holocene. Geographically, this points back towards the north and the Yangtze.

The authors have found more extensive evidence that most domestication genes were selected in japonica and then entered indica through hybridization. Some geneticists, like the Japanese scholar Y-I Sato, or Susan McCouch at Cornell have been discussing this for years, and evidence for this has been mounting—you will also find discussions in the "rice consilience paper" or the "pathways to Asian civilizations" paper. It is misleading, however, to speak of this as “introgression” which implies that pollen flow from domesticated japonica into wild populations in India created indica. What is missing here, and clearly absent from this study, as it was from the Molina et al PNAS paper last year (see previous blog), is consideration of the chloroplast genome. This is older work, but really key, because chloroplasts are not carried in pollen. The Chloroplast (cpDNA) genome of indica and japonica are completely different. Thus introgression by pollen flow from japonica into wild rices is a very convoluted way to account for this hybridization as it would require domestication genes to then persist in wild population that were then re-domesticated. More reasobale in the model I have been promoting as the “proto-indica”model in which wild ancestors of indica (with indica chloroplasts) were under early cultivation or management and the were improved by hybridization with introduced japonica. This does not require domestication gene to somehow persist in wild population where they would be selected against (actually I would expect such introgression to lead to the evolution of weedy rices by "de-domestication": see this blog: ). It also implies a role for human agency in this hybridization process. This means that there were separate starts to cultivation (the human behaviour) for indica and japonica even if the domestication syndrome was shared and evolved one time. 

Does genetic evidence on its own trump fossil evidence? No. Archaeological evidence, which is a fossil record of past rice and past human activities, has once again been simply ignored! Archaeologically early farming societies, with sedentism and villages and evidence for rice cultivation and rice undergoing morphological changes of domestication are found only in the Yangtze valley, as you probably well know. There is no equivalent evidence from Guangdong/ Pearl River. In fact when rice in the Lower Yangtze is showing morphological evolution under cultivation, i.e. between 5000 and 4000 BC, in the Pearl River and South China there are only sparse populations of hunter-gatherer fishers, represented mainly by coastal shell midden sites. These sites provide the earliest evidence for ceramics in the coastal zone (more than 10,000 years later than pottery in the Yangtze!). The first agriculture, based on rice, was introduced between 5000 and 4000 BP, although finds remain few and focused on the southern mountain slopes and north of the Pearl River delta. By this time the Lower Yangtze support urban sites, such as Liangzhu, support by extensive paddy field systems and intensive cultivation of fully domesticated rice. It makes no sense for rice domestication to be placed in the Pearl River region

Weed evolution by de-domestication: the case of rice

The study of weed origins and evolutionary history is the poor cousin of the archaeobotany of crop domestication. Archaeobotanists can potentially do much more on this, and undoubtedly should. To provide some inspiration it is worth considering some recent insights from genetics, to do with weedy rice. While it is surely the case that rice's wild progenitors may act as weeds in the crop, it now appears that much weedy rice is descended from the crop and not directly from the wild progenitor. A recent paper in Genetic Resources and Crop Evolution by Zhang et al. explores the variation in weedy rices in southern China (Guangdong) and northeast China (Liaoning). In Liaoning there are no wild progenitor populations so it is cultivated fields or their margins which provide the only real habitat for spontaneous rices. In Guangdong by contrast there are populations of wild O. rufipogon. Based on microsatellite markers they show that weedy rice in each region clusters with crops, which represent indica and temperate japonica rices, and the weedy rices are distant from true wild populations. They take this to support the hypothesis that weedy rice is secondarily derived from the crop.

Last year this hypothesis also got support from an anatomical study of weedy rice in the USA, by Thurber, Kepler and Caicedo in BMC Plant Biology which shows that the abscission layer which leads to shattering is clearly distinct from non-shattering domesticated rice but also differs from shattering wild rices in terms of its timing in development: it breaks down sooner leading to earlier shattering than in wild rice. This presumably is an useful adapation for beating the farmer to it and getting into the seedbank before the rice harvest. Thurber et al conclude that this points to unidentified regulatory genes that allowed weedy rice, derived from the crop, to reacquire wild-type shattering. (Whether one might be able to tell weedy from wild rice on the physical remains of spikelet bases is another matter, but surely worthy of investigation by an archaeobotanist!). What is more,  genetic characterization (Thurber et al 2010 Molecular Ecology) found that these weedy rices all possessed the sh4 mutation that characterizes domesticated non-shattering rices! This points unambiguously to the acquisition of a different novel mutation that allows shattering. A few years ago Londo & Schaal (in Molecular Ecology) did some haplotyping of American weedy rices and found mutliple origins, with haplotypes from japonica, indica and aus rices (as well as some hybridization).

So rice has a proclivity to weediness, as with many other crops, and the wild progenitor per se may be less to blame. Contrast this with crops that have been domesticated from weeds (oat, rye, kodo millet) and we can begin to think about alternative pathways to and from being a weed.

Does starch evidence push back foxtail millet domestication?

A recent article in PNAS, by Yang Xiaoyan et al., on starch grains (and some phytoliths) from two early Holocene sites in northern China (approx. 9500-7500 BC), pushes back evidence for the early cultivation of foxtail millet. This builds on recently reported identification criteria for millet starch.  This paper is an important contribution to the earlier prehistory of grass and millet exploitation in Northern China and provides important new evidence for the discussions and debates about the timing, areas and processes of millet exploitation and initial cultivation. For on thing it highlights the importance of grasses, including a large proportion of millet grasses, in the subsistence of later hunter-gatherer, or cultivator transition, northern China. This hints at a contrast with nut-hungry and acquatic-focused south (Yangtze). These new data make foxtail millet exploitation, and possible cultivation, at least as early, if not earlier, than the claimed early Panicum from Cishan (although in the latter case, hard evidence for cultivation rather than gathering remained elusive). Phytolith data indicates the occurrence of Panicum miliaceum types only from late Donghulin. Taken with other recent finds this is suggestive that Panicum and Setaria were not brought into cultivation together, perhaps each more than once, and separately, in different parts on northern China. The bringing together of these two crops, fully domesticated, in an integrated system, would seem to be a key transition, yet to be identified, but which must have happened by the time of the emergence of the Early Yangshao tradition in the first half of the 5th Millennium BC.

The trends that Yang and colleagues have found, towards more of the larger Setaria italica like starches are suggestive a subsistence shift, and potentially as the authors argue, of changes evolving in foxtail millet, as part of the domestication process. More data from more sites and periods are needed, however, to confirm whether this a real trend. It would also be nice to see what these kinds of ratios look like on later sites with clear macro-remains of domesticated foxtail millet.

Nanzhuangtou and Donghulin as early Holocene/terminal Pleistocene transition sites with early pottery are often cited as the precursors of more typically Neolithic miller cultivators but have lacked much of any archaeobotanical evidence. These findings will take on an obvious significance to those interested in the early Holocene or early agriculture in Northern China. This study is also a good example of careful archaeological starch research, and is therefore wider methodological interest. Although starch grain research has become increasingly popular in China, some studies have been rather unconvincing, especially with regards to methodologies for identification and for being clear about uncertainties. That is not the case here, where a clear methodology for identification of millet-type starch grains has been employed which is in part qualitative and in part quantitative and it includes a clear recognition of some of the uncertainties in secure species level identifications, especially of smaller and less fissured starch grains. It is also clear that the reference collection of modern material that has been studied is the most extensive to date for Chinese grasses and this increases greatly the likelihood of their reported identifications. The inclusion of control samples of sediment and loess from Donghulin to check for contamination is also methodologically very important, since modern contamination will always be a concern in starch studies. It is also good to see consideration of the presence of immature starches, since immature millet grains are a significant component of charred assemblages, and significant present on immature/unfilled spikelets has been noted in Chinese agricultural texts throughout history, since records in the Han Dynasty. The recognition of carbonized immature Panicum grains has recently been bolstered by an experimental study blogged previously.

Emmer wheat synthesis of genetics & archaeology

Both genetics and archaeobotany provide vantage points of reconstructing the early history of crops, where and how they originates, and how they spread. There has been a growing recognition in both fields that some of our previous conclusions were false truths, based on simplistic assumptions. There has been a small, but growing trend, for some true interdisciplinary synthesis, of archaeobotanists teaming up with geneticists to write the history of crops, and a new example comes from the journal Genetic Resources and Crop Evolution, "Geographic distribution and domestication in wild emmer wheat (Triticum dicoccoides)" in which archaeobotanist George Willcox teams up with geneticists (including Ozkan, Salamini and Kilian) to provide an updated compilation of what is know about Emmer and what it might mean for multiple starts of cultivation, gradual domestication, but the possible predominance of one domesticated line at the end of the process. It generally points towards a much more complex picture genetically and biogeographically, and it recognizes the incompleteness of sampled datasets, including the range of modern emmer landraces and wild populations as representatives of what would have been there in the past. This paper hasn't solved all the mysteries of the origins of agriculture in the Near East, nor of emmer, but it represents serious progress through a conversation between archaeology and genetics, rather than just talking past each other.

Divergence and gene-flow between the wild rices

A important study has recently come out in the journal Molecular Ecology, on  "Ecological divergence in the  face of gene flow in two closely related Oryza species (Oryza rufipogon and O. nivara" by Zheng and Ge. In it they have sequenced a few chloroplast and nuclear DNA loci (7 in total) across 26 populations of these two wild rices. Nivara and rufipogon represent two ends of a adaptive spectrum in the wild genepool of Asian domesticated rice, with O. nivara being an annual adapted to seasonal water from the monsoon and lacking daylength seaonsality controls (so its life cycle is driven by water availability), while O. rufupogon sensu stricto is a perennial occurring in perennial wetlands, and is often highly structured in terms of the seasonality of seed set (especially important in its northern range in China). As I have discussed previously in a 2009 World Archaeology paper, these differing ecologies had important consequences for how these plants would have been utilized by hunter-gatherers, with the perennial rufipogon requiring more environmental manipulation to force it to produce more grain.

In this study Zheng and Ge show that these two wild rice ecotypes are well diverged overall, and they estimate a last common ancestor about 160,000 years ago, but they also show evidence for recurrent gene flow. This strongly suggests that these have diverged as ecological adaptations despite being in continued genetic contact, providing a nice case of the strength of natural selection in pushing divergence even when species ranges overlap (i.e. a case of sympatric speciation). It is interesting to note that the divergence time they have calculated is very similar that those calculated for the last common ancestor of indica and japonica domesticated rices, which have ranged in various studies between 86,000 and ~400,000 years ago, but most focus on 100,000-200,000, much like nivara and rufipogon. As many have argued (see for example the recent "Rice consilience" review article), indica and japonica appear to have different origins in cultivation from different maternal ancestors, one from a nivara-like annual and the other from Chinese rufipogon perennials-- although they also have undergone recurrent geneflow (see  McNally et al 2009 on SNPs and hybridization), which has introduced selected domestication traits amongst others. Thus the process of differentiation in cultural ecologies in the rice crop, despite continued geneflow, continues that ecological and genetics dynamics of the wild progenitors.