Wild about rice landraces

There’s been some interest in a new rice variety that grows better in soils deficient in phosphorus. The BBC touted Wild rice gene gives yield boost and said that

A gene from wild Indian rice plants can significantly raise the yield of common varieties in nutrient-poor soils.

Moments later, however, the report informs readers that

The gene came from a variety called Kasalath, native to nutrient-poor soils of eastern India.

I guess we all have a ways to go in raising media awareness about the subtleties of genetic resources. A wild plant would hardly be a variety that has a name now would it?

IRRI’s press release and the scientists’ paper in Nature are both clear that the gene in question came from a “traditional rice variety”. And the BBC’s report — despite later referring to “wild varieties” — picked that up. But someone, probably some poor put-upon sub, decided they knew better.

What does it matter? Partly for reasons of conservation. That’s of no interest to the BBC, but IRRI proudly “conserves more than 114,000 different types of rice in the International Rice Genebank”. If they are there, does it matter whether they are still in farmers’ fields? At least one person, however, is using the mistaken characterisation to ask an odd (rhetorical?) question:

[T]his research supports claims that wild crop relatives hold an inventory of genes, the value of which is huge. How do we protect more effectively this rich resource?

I’ll leave others to answer that one, if they must.

As for the gene in question, it seems to promote root growth, which is what enables the plant to scavenge more nutrients from poor soils. I may well have more to say on that in a day or two.

28 Replies to “Wild about rice landraces”

  1. As I posted on FB earlier today – where I first saw this rice story – I wonder if the accession number(s) of Kasalath (presumably from the International Rice Genebank Collection – IRGC?) is cited in the full article. I have often argued with IRRI scientists of the need to cite the actual accession ‘provenance’ of the germplasm used in research and breeding, rather than just referring to this variety or that. There is a myth that a variety with the same name is genetically the same. And from our own evaluation of IR36 accessions (I think that was the variety, or maybe IR64) in the IRGC made in the 1990s there were at least six different types – even though the breeders stated, with confidence, that they knew what was the variety and what was not. I also hope that the actual lines in which this gene was discovered have been pure-lined and a sample entered into the IRGC – under a new accession number, of course!

    1. Dare we be more optimistic? There is growing awareness of the need for better management of provenance and diversity within accessions, as evidenced by the various initiatives to pure-line accessions and maintain them as separate accessions, in order to improve the precision and repeatability of genetic analyses.

      But the genebank standards community is not keeping pace. How should we document that accession B is a pure line selection from accession A? There’s no place for such info in the MCPD. ICIS users can of course use genealogies: but even then should we also embed info on the selection history in the name so that users don’t have to trace through genealogies to find the origin? In the absence of any agreed guidelines, we’ve started our own standards. Look in IRIS for names containing “::”, and you’ll find the pure line copies of accessions.

      And then how should we deal with the additional cost of keeping two versions of an accession? We might want pure line copies of all accessions, but we surely don’t want to double the size of our collections. This needs to feature in genebank standards

      1. Ruaraidh – good to see that despite the lack of standards (smacks of performance targets, etc. – don’t do something until the necessary paperwork is in place) IRRI IS doing something positive and practical. Lead from the front and the others will follow. I agree there are some documentation issues but not in my opinion insurmountable problems – and definitely never a reason NOT to do something positive.

  2. I tell a lie (in my comment on the featured comment). This from the Nature paper:

    Rice accessions included in this study were: Kas group, Kasalath, AUS196, AUS257, Dular, IR84144-11-12, Lemont and Vandana; Nip group, Bala, CT6510-24-1-2, IR 42, IR64, IR66424-1-2-1-5, IR73678-6-9-B, IR 74, IR74371-46-1-1, K36-5-1-1BB, Nipponbare, PM-36 and Vary Lava 701.

    GenBank protein accession numbers for OsPupK04-1, OsPupK05-1, OsPupK20-2, OsPupK29-1 and PSTOL1/OsPupK46-2 are BAH79993, BAH79994, BAK26565, BAH80018 and BAK26566, respectively.

    Any help?

    1. Nope. These are variety names or line names, not accession numbers, and tell you nothing about the actual provenance. Many rice breeders think that varieties are ‘immutable’. I once had a discussion about multilocation trials and the need to include some controls. My colleagues were all for using a ‘known variety’, say IR64 as the control, but sourced from each country. I insisted that the control varieties had to be from the same seed source, even if local ‘versions’ of these varieties were also included. This was indeed news to these agronomists, physiologists, and some rice breeders. I know for a fact that there are several different lines of the same bred variety supplied from different locations in the IRGC (why they are there in the first place is another question). But they are clearly genetically different. I think that one of the reasons where there have been some discrepancies in the rice genomic work is that different labs used different sources of Nipponbare on the assumption that all Nipponbares were the same.

  3. Dear Dr. Mike: Thanks for your words , agree with you “There is a myth that a variety with the same name is genetically the same”. in Iraq now there is many variety with the same name is genetically the same this is from Iran ,I try to collect some of them and will be send to Dr.Luigi with best regards

  4. Mike – This brings up a generic problem. As many (most?) genebank accessions are mixtures, how can we rely on evaluation data on what are inevitably samples of the original accession to characterize the original accession in databases? The trial sample may have features not found in the remainder of the original accession (and the remainder of the original accession may have features not found in the trial sample). As you suggest, there is a need for any evaluation for useful characters to return seed of the actual samples evaluated under a new accession number to the parent genebank.
    And if samples after field selection are to be returned to genebanks, then why stop at individual accessions? Why not mix up seed of , say, 5000 accessions, bulk sow it, to be grown under high selection pressure for whatever feature required, and then return the bulk harvest to the genebank – now as a population with a higher level of the feature of interest. This would seem to have more going for it than the present hit-and-miss attempts at evolution by in-situ on farm conservation.

  5. Kasalat is actually a Bangladeshi variety not Indian. In Bangla it means Kacha Lota or green shoot. From time immemorial it has been grown in the eastern district of Sylhet of Bangladesh from where it might have gone to India. Indeed all Aus varieties are of Bangladeshi origin. I myself saw Kasalot being grown over several decades in bangladesh and it is just plain wrong to describe it as an Indian variety

  6. Language plays an important role in determining origin. The Aus rice is actually pronounced Aoosh, which is derived from Ashoo meaning “early”; so earliness is the main character of these lines and lines of diverse provenance were selected on the basis or earliness. This dsparate group were given additional names such as Dhumai (or Du-mahee meaning maturing in 2 months or sixty days) and Kasalot meaning green shoot. An unfortunate fact in modern plant science is to ignore such heritage-based information which is a clue to earlier selections. As early as as in 1925, the then mega varieties of eastern part of Indian subcontinent were sent to USA which included rice such as Kasalat, Dhumai, Baurosh, and intriguingly, Basmati which too carries a Bangla name and used to be a widely grown rice in the Sylhet-Assam area

  7. Dear Jeremi, to struggle for establishing Sovereign rights never been considered as “Ugly”. “Ugly” evolves from nasty politics. And according to your reference and logic, should all the natural resources of Bangladesh be considered as “Indian”? As Bangladesh was the part of Indian subcontinent while the resources were formed!!
    If the claim of Mr. Abed is true (we believe so), it must deserves more in depth discussion.
    I am taking the opportunity to discuss a little on our double kernel rice variety (or species!) Biram Sundori- that is the only grass on earth is being cultivated, which contains the unique character.
    Initially all of you including Mr. Ruaraidh supported us with various information and links. We were advancing well. But suddenly all were silent. May be in that time, IRRI had the sample from Dr. Mazid in Bangladesh. But the question here is, has the accession been documented at the Bank properly? Or we will have to wait another 50 years to struggle for the right on Biram Sundori?

      1. Yes I remember it too. But I also remember that Biram Sundori was not the only such grass on earth – we found one from Nepal as well, and references to another from China.
        Silence from me because I’ve heard no more about it. No seed has come to us. If it does come, I will of course take great care to document it properly.

  8. As far as I remember IRRI has the “twin seed” rice lines from China, way back in 1993 we screened those lines for signs of apomixis, during a workshop organised by Dr. Khush. The chinese materials had low “penetrance” for twinning; in contrast I myself have observed 70% twinning in Bangladesh material. It is important to ask if the additional embryo might be maternal or if both are zygotic

  9. Kasalot is available in BRRI Gene Bank (accession number is 3283) collected from Kamalgonj of Sylhet.

    In a publication in Bengali called “DESHI DHANER JAT ” from BRRI Aus accessions named ‘kasalot’/’kasalota’/’kasaloti’- is reported in Moulvi Bazar, Kamalganj, Rajnagar, Bishwanath, Chahatak, Gopalgonj, Kulaura of Sylhet region are reported.

    Is the above-mentioned accession from BRRI not available at IRRI since the collection took place in 1983 and BRRI sends everything to IRRI; would be also nice to know what happened to the other collections from all over Sylhet

    1. For the record, no this accession is not in the genebank at IRRI. The highest BRRI genebank accession number we have recorded of accessions duplicated in IRRI is 1971, sent in 1980. For material sent 1983 onwards, we know the collector’s sample ID but not the BRRI accession ID.

  10. Dear Abed and All,
    Thank you very much for the lively discussion on our paper and the origin of Kasalath. I learned a lot in the process and will follow up on this by genotyping the different Kasalath accessions that we have at IRRI and will also ask BRRI to do the same for accessions from Bangladesh.
    As you may know from our previous publications on Pup1 (Chin et al 2011 Plant Physiol 156: 1202–1216; Chin et al 2010 Theor Appl Genet 120(6): 1073–1086), we find the tolerant Pup1 haplotype in many stress-adapted varieties of various origins and also in IRRI breeding lines developed for rainfed environments. We mention this in the paper. Whether the Pup1 locus/PSTOL1 has the same origin in all these accessions and whether the gene that we cloned from the one specific Kasalath sample is the “original” gene is not known and might be difficult to determine.
    Given the high conservation of Pup1/PSTOL1 in stress-adapted accessions, it appears that breeders and farmers have been selecting this gene without knowing about it. We take this as further support of our data showing the importance and potential impact of this gene under P deficiency. We can now provide the gene-specific markers and technology to ensure that the gene is present in all breeding lines. The markers can also be used for targeted introgression of the locus into those varieties that do not yet have PSTOL1 to make them better adapted to stress.
    To initiate and out-scale our efforts to use the gene for breeding, we had a Pup1 training workshop last week with participants from six countries, including Bangladesh and India. The participants are now well trained and highly motivated to initiate the genotyping of their local germplasm and to select the best suitable donor – whether this will be Kasalath, one of our Pup1 breeding lines, or a locally adapted variety that has Pup1 might vary from country to country.
    We also encourage the participants to start their own screenings for P efficiency to help us identify additional tolerance mechanisms. PSTOL1 is not the sole solution to all problems as we all know and really, I think PSTOL1 is just a first step, though hopefully significant, into the development of nutrient efficient crops. Much more needs to be done and we need additional tolerance genes. Matthias Wissuwa at JIRCAS has already identified some additional tolerant genotypes and we are also working on this at IRRI. We should all join forces to address this important problem and get more funds to work on this with priority.


  11. Many thanks Sigrid. You have published a paper which is more than outstanding, it is uplifting and electrifying bringing the issue of conservation of traditional germplasm into the fore-front of plant research. As to why this particular gene has been selected by a number of ecotype is intriguing. I have the following observation.
    1. Could the actual selection be acidity tolerance as high-rain in Sylhet basin is supposed to make soil very acidic and selection for acid tolerance would also give tolerance to low P; in this scenario the ability to make powerful root is more a result of acidity tolerance
    2. The PSTOl1 gene seems to have been mobilised by retro-elemnt; thus elements are known to become active under stress
    3. Stresses often cross-feed so one stress might make plant resistant to another stress by activating a family of retro-element
    4. In the low lying areas of Sylhet there are many nivara/rufipogon lines; do they also have PSTOL1; when in evolutionary type this haplotype originated?

    Anyway this is an exciting area that is now feeding the imagination of many. Many thanks to you and congratulations to you and your group

  12. Dear Abed,

    Thank you very much for your kind and encouraging reply!
    Thanks also for bringing up the putative role of TEs- we are currently looking into this and will hopefully publish this soon.
    I am now very curious to look at the varieties from Sylhet in more detail- do you know someone there that can help us do that?
    We only now have enough seeds of our indica-Pup1 lines to do the agronomic studies and multilocation trials. This will help us to better define where Pup1 is most beneficial. For now, we can try to check the effect of Pup1 under different pH in hydroponics. All the soils we have tested so far are around pH4-5 .

    Best regards


  13. Hi Sigrid, as you know there is a BRRI collection of Kasalot from 1983; you can get it through BRRI; regarding other areas reported in the 1980′ BRRI book also BRRI might be able to shed light as they did this thorough study. I will use my Sylhet connections to obtain as many lines as possible. Wide-spread use of hybrids and HYV lines have led to precipitious decline in the use of heritage land-races and a concerted collection attempt will have to be launched to salvage these lines, Kasalot and others. You might also ask Susan Mccouch from Cornell as she has been very active with Bangladesh germplasm. What about greater Sylhet districts of India, for instance areas around karimganj , Hailakandi, Shilchor etc, you might try your Indian contacts for that. I also think that the numerous lines that you have in IRRI collection as Aus-1…. Aus200 might have putative Kasalot type sitting there. Aus rice always have a name and never just a number, so those need to be screened also. I have an ongoing plant production operation going in greater sylhet area so let me know what I could do for you and your research efforts

  14. yes transposable retro elements might bring them in horizontally even so it need not have happened vertically at all. In stress-prone condition sine and line elements might have mobilised key genes. I’d predict that PSTOl1 gene would also be present in other wild grasses that grow extremely well in hardy soil

  15. Please note we have unpublished the information about the origin of Kasalath pending corrections.

    Apologies to all concerned – we will work to resolve the matter and let you know the outcome.

  16. Hi everyone – sorry for the delay.

    We’ve re-worked the Kasalath article to capture real quotes from this very blog! Many thanks to everyone for their help in finalizing this version.

    We hope we have acknowledged everyone accurately but if not please let us know. Other comments are welcome too – we’ll post it on our facebook page so you can comment direct there.


  17. Dear All,

    Please find below the IRRI blog in which we clarify that the Pup1 donor used in our study is a Kasalath accession from India. Thanks to Abed, we are all aware now that Kasalath is an important variety in Bangladesh and we are working together to look at this in more detail. Hopefully we can share data on the origin and genetic diversity of this important variety some time soon.

    Best regards

    Sigrid Heuer


    Copy IRRI blog text:

    Origin of the rice variety Kasalath
    In a recent Nature paper and a related media release, the source of an important gene (PSTOL1) that helps rice take up phosphorous was reported as a rice variety from India called Kasalath. The discovery of the PSTOL1 gene is very exciting news as it has such potential to improve the food security of rice farmers with the lowest value phosphorus-deficient land.
    In August, the International Rice Research Institute (IRRI) held a training workshop for plant breeders across Asia to help them use molecular marker technology to breed the PSTOL1 gene into their rice varieties. This will mean that farmers will get access to new locally-adapted rice varieties with the PSTOL1 gene and the trait that helps the plants take up phosphorus that can boost yields by 20%.
    Putting this important discovery and work aside, there’s an interesting side story here that relates to both understanding genetic diversity of rice and conserving it.
    IRRI manages the International Rice Genebank where more than 115,000 types of rice are conserved. Conserving this rice genetic diversity is one of our most important roles. Even within one singular variety of rice, such as Kasalath, plants that grow in different regions will be genetically different from each other. For scientists studying and identifying genes in rice, it is important to know the exact heritage of the specific plant sample that they are investigating – because what they find for one individual plant sample may be different to another even if they are from the same variety of rice. The PSTOL1 gene was discovered in a specific plant sample (also known as an accession) of the rice variety Kasalath that was from India.
    “The samples of Kasalath used for the published research trace back to a sample collected from Karimganj, Assam in 1925, in what is present-day India,” said Dr. Ruaraidh Sackville Hamilton in his response to questions about the source of the material. “It is not clear if the sample used by the Japanese traces back to exactly the same collected sample, but I’m informed it is also from a very old collection in Assam. So, based on the CBD’s [Convention on Biological Diversity] definition of country of origin, India is the country of origin of the material studied.”
    However, according to Dr. Abed Chaudhury, a plant scientist from the Sylhet area of Bangladesh, the geographical range of the entire variety of Kasalath includes the Sylhet region, which encompasses Bangladesh’s Sylhet Division and neighboring parts of northeast India.
    “Sylhet farmers knew it [Kasalath] hundreds of years ago calling is Kasa Lota or “young green shoot” rice,” said Dr. Chaudhury in his comment on a blog post.
    Kasalath has been widely grown throughout Sylhet as a prized traditional variety since time immemorial. It has been recorded in Maulvi Bazar, Kamalganj, Rajnagar, Bishwanath, Chhatak, Gopalganj, Kulaura, and Karimganj. Now, Karimganj is in India, but most of the rest of the Sylhet region is in Bangladesh.
    Kasalath is also part of a broader group of rice called the Aus variety group. This group of rice is famously known to tolerate difficult growing conditions and was also the source of another important gene (Sub1) that IRRI has used to develop submergence tolerant rice varieties. Other members of the Aus group of rice are known to carry genes for tolerance to drought, heat, salinity, and other stresses.
    A figure used in the Nature paper illustrates that the Aus variety group was developed in India. It is an accurate depiction of the specific results for India however, the absence of evidence on the involvement of Bangladesh should not be misinterpreted as a denial of the role of Bangladesh in the development of Aus.
    “Language plays an important role in determining origin,” said Dr. Chaudhury. “The Aus rice is actually pronounced Aoosh, which is derived from [the Bangla word] Ashoo meaning “early”; so earliness is the main character of these lines and lines of diverse provenance were selected on the basis of earliness.”
    So while the Nature paper outlines specific evidence for domestication of Aus rice in India, it would be misleading to generalize this to conclude that Aus was not also domesticated in Bangladesh.
    “What is clear is that Aus varieties, and Kasalath, have been grown in India as well as Bangladesh for many years,” concluded Dr Ruaraidh Sackville Hamilton.
    Understanding the heritage of plant samples used in research and conserving the complex diversity of rice varieties is critical to rice research. Without it we would not have access to the immense genetic diversity of rice to develop new rice varieties that can help farmers face challenges like problem soils or harsh weather extremes. And of course there is intrinsic value in conserving the genetic diversity of rice too – which is done at the International Rice Genebank.
    Thanks to everyone who provided information to help fully explain this: Dr. Abed Chaudhury (Food Security Development Center, Mauritius), Dr. Tanvir Hossain (formerly of BRRI),Prof. Zeba Seraj (University of Dhaka), Md. Sazzadur Rahman (University of Dhaka and BRRI), Dr. Sigrid Heuer (IRRI), Dr. Ruaraidh Sackville Hamilton (IRRI), Dr. Matthias Wissuwa (JIRCAS), Prof. Susan McCouch (Cornell University), Dr. Pat Gonzales (IRRI), Dr. Ken McNally (IRRI) and all the bloggers over at the Agricultural Biodiversity Weblog.

  18. Many thanks to Sigrid, Ruaraidh and IRRI; needless to say this blog played an pivotal role in clarifying this important issue; we look forward to more exciting postings here

  19. Not the last word, I’m sure, but I would like to thank everyone involved for their contributions and willingness to share information and credit. It is truly marvellous what can be achieved with the help of easy, global communications.

  20. What would be the problem if the line was completed Like “So, based on the CBD’s [Convention on Biological Diversity] definition of country of origin, India is the country of origin of the material studied and both Bangladesh & India are the country of origin of the genetic resources.”

    1. Zakir: The problem remains that it is the FAO Seed Treaty and not the CBD that covers rice (as an Annex 1 crop). Both Bangladesh and India have accepted the Treaty (I think), which is multilateral, not recognizing the country of origin. There were arguments at the time that we couldn’t know where genetic resources originated (the CBD position for species – partly true for crops and their varieties) but this ignores that plant breeders and farmers work from samples: handfuls of seed from a registered collection that we generally know the origin of. To my mind it would be far more acceptable if the Treaty recognized country of origin of the sample (and eventually rewarded that country) if only to encourage countries to put samples in genebanks from whence they would be widely available – as for IRRI. Then there could easily be full recognition of the country of origin of the sample: in this case India, but if there were lots of samples from Bangladesh readily available, it could well have been Bangladeshi samples that were used and credited. Moving samples around is the best way to get credit as the country of origin.
      I suspect that even the big CGIAR genebanks still have holes in their collections – lacking samples from places where they should have samples from. For example, ICRISAT had nowhere like the range of sorghum samples from Ethiopia that we had in the Addis genebank 35 years ago. Ditto for sorghum from Yemen, where the USDA had a collector 40 years ago. The FAO Seed Treaty does not seem to be helping to fill these holes – needing little more than a transfer of existing samples.

Leave a Reply

Your email address will not be published. Required fields are marked *