Saturday, June 30, 2012

Parallel Domestication of Food Crops

Qasr Ibrim, as photographed by the Allaby Research Group (Link)

This recent paper from the Allaby Research Group discusses the case for parallel domestication of food crops in diverse regions of the world.  Its focus is on agricultural crops, so it is not directly related to the process of cattle domestication.  However, the discussion in the paper of an independent domestication of cotton at Qasr Ibrim supports the notion of a localized agripastoral tradition on the Nile.  Related to this discussion is the previous post which mentions cotton domestication among Saharo-Sahelian speakers.

The blossoming of plant archaeogenetics
Sarah A. Palmer, Oliver Smith, Robin G. Allaby
Annals of Anatomy, Volume 194, Issue 1, 20 January 2012, Pages 146-156


"For the past two decades scientific drivers concerned with the origin and spread of domesticated crops have dominated the utility of ancient DNA from plants (Fig. 1). It is a current aspiration that genomic studies of plant remains could unlock critical evidence of the origins of domestication and strengthen the bridge between phylogenetic studies of modern plants and the archaeology of plant remains (archaeobotany). This study of plant domestication has itself undergone substantial change in recent years in response to new data leading to a different framework of understanding. The interpretation of phylogenetic data with regard to domestication is often seated on the framework of models of human behaviour in the shift from hunter-gathers to agriculturalists ( [Allaby et al., 2008] and [Allaby et al., 2010]). However, theses based on plant phylogeny, which uses extant allelic diversity to infer past genomic events, and archaeobotany, which examines archaeological plant remains by morphology, have sometimes appeared to contradict one another. Recently, mounting archaeological evidence has spurred a shift in thinking from a rapid transition paradigm for the shift from hunter-gatherers to agriculturalists in which a ‘Neolithic package’ of staple food crops were domesticated over a relatively short time period by a small farming group, to a protracted transition paradigm in which domestication of food crops occurred in parallel, over a long time, in diverse regions worldwide (Fuller et al., 2010). The interpretation of genetic diversity needs to be reviewed in light of the modified genetic expectation of the paradigm of protracted transition (Allaby, 2010)."

Reconstruction of the early history and movement of cultivated plant species:

"The arguments for the origin of crop species and movement of these species to where they can be found today are largely dependent on genetic diversity. With the classical underlying expectation of a rapid transition paradigm (where domestication events are discrete and quick), monophyly or polyphyly has been interpreted as evidence of single or multiple origins, respectively (Allaby et al., 2010). Archaeogenetics has been used in parallel with these studies to mine for past allelic diversity (Navascues et al., 2010). Phylogenetic analysis to determine centre of origin and movement has been undertaken of archaeobotanical remains of many of the crop species important to human societies, including members of the cucurbitaceae ( [Szabo et al., 2008], [Zoltan et al., 2007] and [Erickson et al., 2005]), the poaceae ( [Nasab et al., 2010], [Li et al., 2011], [Tanaka et al., 2010], [Lia et al., 2007], [Gyulai et al., 2006], [Lagler et al., 2005], [Lagler et al., 2006], [Freitas et al., 2003] and [Allaby et al., 1999]), grapes ( [Cappellini et al., 2010] and [Manen et al., 2003]), the Prunus genus (Pollmann et al., 2005) and olives (Elbaum et al., 2006)."

"The most extensively archaeogenetically characterised gene family is that of the high molecular weight glutenin gene (Glu) in wheats. The Glu genes have a conveniently short polymorphic region in the promoter that is ideally suited for the identification of alleles unique to hexaploid bread wheat ( [Allaby et al., 1999] and [Nasab et al., 2010]). Recently, Li et al. (2011) discovered the Glu D genome allele in Bronze Age wheat samples from China indicating hexaploid bread wheat. This is in contrast with archaeobotanical evidence of tetraploid emmer wheat predominating Western Eurasia. This finding echoes previous studies in which the D genome was similarly identified in an assemblage from Bronze Age Greece that was thought to be tetraploid wheat only (Allaby et al., 1999). Furthermore, an entirely extinct expansion of wheats into Europe containing the more unusual G genome was initially detected through ancient DNA (Brown et al., 1998), a finding that was later confirmed by archaeobotany (Jones et al., 2000). A similar pattern of replacement has been observed in recent studies of early Japanese rice agriculture which suggest that tropical rice varieties were grown which were later replaced by temperate varieties (Tanaka et al., 2010)."

"Ancient DNA as a marker of phylogeographic stability has also been used with maize ( [Freitas et al., 2003] and [Lia et al., 2007]). In this case the primitive landraces used by Native Americans in South America were used to establish a modern phylogeographic distribution of alleles of alcohol dehydrogenase and microsatellites the antiquity of which was confirmed with archaeobotanical samples, giving a picture of a distinct east west divide suggestive of two expansions of maize into South America from the meso-American homeland of maize. The establishment of 10,000 year old remains in the New World as bottle gourd (Lagenaria siceraria) sets an important biogeographic precedent in that it establishes that paleoindian populations transferred domesticated plants from the Old World, despite their hunter-gatherer life-style (Erickson et al., 2005). Until then evidence that they had been associated with any plant transportation had been scant."

"More recently, archaeoegenetic studies which have exploited herbarium samples of barley to expand the sample set to include extinct landraces and cultivars sampled where they are no longer grown today, facilitated resolution of genetic structuring on a fine geographical scale in Sweden that would have been unavailable from modern crops (Leino and Hagenblad, 2010). In a similar case to that of the maize, in this latter study the distribution is suggestive of two entries of barley into Sweden suggestive of two cultural influences."


"Archaeological plant remains are also studied for the integral part they play in human society. Often excavated in close proximity to remains of humans or their dwellings, plant remains can be examined in light of what plants were being produced, foraged, eaten, traded, or used in some other way. The marriage of archaeogenetics and ethnobotany is one of the most interdisciplinary of the sciences, and is characterised by the interpretation of archaeobotanical phylogenetics in light of theories formulated from archaeology. Archaeobotanical reconstructions are frequently hindered by ambiguous identification of samples (Schlumbaum et al., 2008). Archaeogenetics has been used as a tool for species identification where the morphology is ambiguous."

"A common use of archaeogenetics on ethnological remains has been to reconstruct diets of past peoples including the vegetal dietary elements ( [Rasmussen et al., 2009], [Rollo et al., 2002] and [Poinar et al., 2001]). Of archaeobotanical remains, the inferences of resolved identification have been used to confirm an indigenous independent domestication of the A genome cotton, Gossypium herbaceum in Africa prior to Asian influence (Palmer et al., unpublished) and several instances of hexaploid bread wheat during the Bronze age in China (Li et al., 2011) and Greece (Allaby et al., 1999), much earlier than previously estimated in the former case. Also, the use of Prunus spinosa and P. avium/cerasus by the Romans in the Northern alpine region has been established (Pollmann et al., 2005)."

"The mode in which archaeological remains were used and the evaluation of the importance of trade versus local agricultural diversity have also been areas of ethnobotany where archaeogenetics has made valuable contributions. In our lab, we have examined Barley and Cotton remains from Qasr Ibrim, a border settlement on the River Nile that was subject to numerous cultural transitions. Samples of barley from all six strata examined, spanning more than 3000 years, displayed the same allele at the vrs1 locus (Palmer et al., 2009). Given the settlement's situation on the trade route of the River Nile, it was considered likely that conquering peoples may have introduced foreign germplasm to the area. However, this local landrace appears to have persisted throughout the occupation of this settlement. The identification of the cotton remains at this site as G. herbaceum, suggests local production, rather than the opposing theory that cotton was generally imported from Asia and processed locally."


  1. I don't think anyone disputes that there are multiple independent sites of Neolithic era crop domestications (Ferile Crescent (wheat, flax); Sahel (pearl millet; sorghum); S. China (rice); N. China (millet); Papua New Guinea highlands; Mesoamerica (maize).

    Most would argue that within the regions there were also actually multiple distinct domestications that gelled into a local package that once complete enough started to expand from the core area (e.g. barley and wheat and lentils may have been domesticated in different places and combined into a Fertile Crescent package over a couple of thousand years; likewise maize and beans were probably domesticated in different places and then combined into a single package). Most packages probably eventually accumulated at least two to four food crops and at least one fiber crop.

    Most would argue that most secondary receipents of these Neolithic packages in turn had secondary domestications shortly after the local arrival of the Neolithic that added to the package received (donkeys in Egypt, Llamas and Guinea pigs in South America, bannanas in tropical Southeast Asia, coffee in Ethiopia; a few other crops Europe whose name escape me) that added onto these packages. Most would also argue that there were also some late bloomer domesticates that arrived long after the Neolithic took hold (e.g. camels and apples).

    The case for cotton as a secondary domestication that added to the Fertile Crescent package in Africa (and perhaps independently elsewhere) much as the donkey did is pretty strong. The case for some domesticates that fell out of favor where better domesticates came along (e.g. in the SE United States region pre-Maize) is also decent. But, the case for multiple independent domestication of the same major Neolithic plants is one made strongly in only a very small number of cases (perhaps a couple of types of rice, perhaps several varieties of cotton) and a somewhat larger number of animals (probably horses, probably pigs, possibly dogs).

    I'm less impressed for the evidence of some of the other local varieties being local independnet domestications as opposed to locally evolved or bred derivatives of received package member crops.

  2. Andrew,

    As with the Gifford-Gonzalez/Hanotte paper on the domestication of animals in Africa, the Allaby Research Group in the above paper is proposing an alternate way of viewing domestication as a long time base process, in parallel, in many regions. It is not arguing for a single domestication of crop A here and crop B there. It doesn't use the language "secondary domestication". What it is pointing out is that archaeogenetic evidence is beginning to mount to show that common crops such as wheat, maize, cotton and barley probably underwent processes of domestication in multiple regions over time scales much longer than those that are currently accepted. Additionally, the paper summarizes evidence such as the transport of the bottle gourd to the New World more than 10,000 years ago that shows that societies that have been viewed as hunter-gatherers did engage in limited forms of domestication.

    Regarding your statement:

    "But, the case for multiple independent domestication of the same major Neolithic plants is one made strongly in only a very small number of cases (perhaps a couple of types of rice, perhaps several varieties of cotton) and a somewhat larger number of animals (probably horses, probably pigs, possibly dogs)"

    Allable points out that the nascent field of archaeogenetics is uncovering the likelihood of multiple independent domestications of wheat, rice, cotton, barley, maize, and gourds. Your "only a very small number of cases" argument against multiple independent domestication processes doesn't hold up under the weight of current archaeogenetic research.

    Allaby presented an interesting paper last week at SMBE 2012 Dublin last week on Qasr Ibrim barley, which you might also find interesting. I haven't found an open source copy of this paper, so I haven't posted it.

  3. I read the maize paper (I don't recall how I got past the paywall, at the time, I think I was chasing down a footnote from something I read in one of Fuller's earlier papers and ended up wikiwalking until I found it), and I am pretty familiar with maize domestication as it integrated into Mesoamerican Neolithic from at least a few articles on the subject. It looks like all that paper cited on maize suggests is that farmers continued to refine the original domesticate over time in different places as they artificially selected for the traits that they wanted. Call it a "sequential refinement" model as oppposed to a true independent separate domestications model.

    The maize paper is cited in this paper not as an example of independent domestication but as an example of an early example (2003) of a study that used DNA analysis of ancient seeds to infer properties of the plants analyzed that are visible from the morphology of the seed itself.

  4. Andrew, if you have something to add to the discussion on *this* paper, that would be great.

    You've made a point for the last month or so of posting on my blog and generally being dismissive of the information that is posted: dismissive of Chris Ehret's dates for Nilo-Saharan cattle domestication, dismissive of Chris Ehret in general, and dismissive of this paper on the domestication picture that is now coming out with new archaeogenetics techniques. Your posts suggest that you do not even read the posted material.

    I would ask that you stick with the topic of the posts.


  5. This comment has been removed by a blog administrator.


Comments have temporarily been turned off. Because I currently have a heavy workload, I do not feel that I can do an acceptable job as moderator. Thanks for your understanding.

Note: Only a member of this blog may post a comment.