Running a genebank for rice . . .

In March this year, I posted a story about the International Rice Genebank (IRG) at the International Rice Research Institute (IRRI) in Los Baños, the Philippines. Now, I thought it would be interesting to describe some of my early challenges when I joined IRRI as head of the Genetic Resources Center (GRC) in July 1991.

Running a genebank is not one of your run-of-the-mill endeavors even though the individual technical aspects that make up genebank operations are relatively straightforward – for rice, at least. It’s their integration into a seamless, smooth and efficient whole, to ensure long-term genetic conservation, that is so demanding.

This is how the genebank is running, more or less, today:

Before I joined IRRI I’d never actually managed a genebank, although I had trained in genetic conservation, worked in South America on potato genetic resources, and spent a decade teaching various aspects of genetic resources conservation and use at the University of Birmingham.

My predecessor at IRRI was Dr Te-Tzu Chang, known to everyone as ‘TT’. He joined IRRI in 1962 and over the years had built the germplasm collection to about 75,000 or so accessions by the time I joined the institute, as well as leading IRRI’s upland rice breeding efforts.

Following in the footsteps of such a renowned scientist was, to say the least, quite a challenge. I was also very conscious of the great loyalty that the genebank staff had to TT. But I had to look at the genebank through a fresh pair of eyes, and make changes I thought necessary and appropriate to what it did and how it was managed.

I spent several months learning about rice (since I’d never worked on this important crop until then), about the workings of the genebank  (in July 1991 it was still called the International Rice Germplasm Center), and assessing the genebank staff for possible new roles. I asked a lot of questions, and slowly formulated a plan of the changes I thought were necessary to significantly up-the-game, so to speak, of genetic resources conservation at IRRI.

From the outset, the local staff were rather wary of this assertive Brit who IRRI Management had brought in to deliver change. After all, most of them had only ever worked for TT. Here I was, asking lots of questions and expecting straight answers. But until I arrived on the scene – with rather a different management approach and style – they’d been used to a regime under which they were merely expected to follow instructions, and were given little if any individual responsibility.

Elaborating the best personnel structure with sufficient staff was a critical issue from the outset, just as important as upgrading genebank operations and the physical infrastructure. I was determined to eliminate duplication of effort across staff working in different (sometimes overlapping) areas of the genebank, who seemed to be treading on each other’s toes, with little or no accountability for their actions. In 1991, it was clear to me that making progress in areas such as seed viability testing, germplasm regeneration, data management, and curation of the wild rices would be hard going if we had to depend on just the existing staff. Furthermore, many of the genebank facilities were showing their age.

So I was fortunate to persuade IRRI Management that the genebank should be one of its priorities in the institute-wide plan for an infrastructure upgrade. I developed several initiatives to enhance the conservation of rice, eliminate geographical gaps in the collection, as far as possible, through a major collecting program, as well as begin research about on farm conservation, seed conservation, and the taxonomy of the wild rices. In November 1993, the Swiss Development Cooperation (SDC) approved a five year project, which eventually ran until early 2000, and provided a grant of more than USD 3.2 million. Click on the CD image to read the Final Report published in July 2000, just a few weeks after the project ended. We also released this on an interactive CD, with the Final, Annual and Interim Reports, copies of published papers, etc., all collecting trip reports, and those about the various training courses, as well as some 1000 images showing all aspects of the project.

I should add that starting research on rice genetic resources had been one of the conditions I made when accepting the headship of GRC.

Quite quickly I’d also come to the conclusion that I needed a focal person in the genebank who would in effect become the genebank manager, as well as other staff having responsibility for the different genebank operations, such as seed viability testing, regeneration, characterization, the wild rices, and data management. I just felt that I needed to be able to go to a single person to get information and answers rather than several staff each with only part of what I needed.

By the end of 1991 I’d named Flora ‘Pola’ de Guzman as the genebank manager. She had a background in seed technology, so seemed the right person to take on this important role. Pola is now a Senior Manager, the highest level among the national staff, although in 1991 she was only a Research Assistant.

I placed all field operations under Renato ‘Ato’ Reaño, who also took direct responsibility for germplasm multiplication and regeneration, while Tom Clemeno managed the characterization efforts of GRC.

Socorro ‘Soccie’ Almazan became the curator of the wild rice collection and manager of the special quarantine screenhouses where all the wild rices had to be grown – at a site about 4.5 km away across the IRRI experiment station.

Adelaida ‘Adel’ Alcantara became the lead database specialist (supported by Myrna Oliva, Evangeline ‘Vangie’ Guevarra, and Nelia Resurreccion).

And two staff, Amita ‘Amy’ Juliano (who sadly succumbed to cancer in 2004) and Ma. Elizabeth ‘Yvette’ Naredo (now Dr Naredo since 16 October 2012) moved over to full-time research activities related to rice taxonomy.

One of my staff concerns was what to do with Genoveva ‘Eves’ Loresto. I needed to find her a role that took her away from any direct supervision over the others. She helped me with the overall infrastructure changes, liaising with contractors, but once we had the SDC funding secured, I was able to ask Eves to take on a major project management role, as well having her lead the germplasm conservation training courses we organized in many of the 23 countries that were project partners. Eves eventually retired from IRRI in 2000.

A ‘new’ genebank
In terms of infrastructure, we had opportunity to make many changes. We remodelled the data management suite, giving each staff member proper workstations, and constantly upgrading when possible the computers they used. I made it clear to everyone that the database staff would have first access to any computer upgrades, and their machines would filter down to other staff whose work depended less on using a computer. And of course in 1991 (and for some years afterwards) the PC revolution was only just beginning to have an effect on everyone’s day-to-day activities.

Seed drying was one of my concerns. Before my arrival seed drying was done on batch driers immediately after harvest, with no precise temperature control but certainly above 40°C; or in ovens well over the same temperature. We designed and had installed a seed drying room with a capacity for 15 tonnes of seeds, at 15°C and 15% RH, and seeds dried slowly over about two weeks to reach equilibrium moisture content suitable for long-term conservation.

Incidentally, in recent research [1] supervised by Dr Fiona Hay, GRC’s resident seed physiologist, initial drying for up to four days in a batch drier before slower drying at 15°C and 15% RH seems to have a beneficial effect on viability.

We doubled the size of the wild rices screenhouses, and converted the large short-term storage room in the genebank to a seed cleaning and sorting laboratory for about 20 technicians. Previously they’d been squeezed into a small room not much more that 4m square. Another general purpose room was converted to a dedicated seed testing laboratory, and a bank of the latest spec incubators installed. We converted a couple of other rooms to cytology and tissue culture (for low viability seeds or for embryo rescue) laboratories. Finally, in the mid 90s we opened a molecular marker laboratory, initially studying RAPD and RFLP/AFLP markers, but it’s now taken off in a big way, and a whole range of markers are used [2, 3], led by Dr Ken McNally (who was my last appointment to GRC before I moved from there to become one of IRRI’s directors in 2001).

We were also fortunate in the mid 90s to have a very successful collaboration with the University of Birmingham (and the John Innes Centre in Norwich, UK) to explore the use of molecular markers to study rice germplasm, funded in the UK by the Department for International Development (DfID). One of the most significant achievements was to demonstrate – in one of the first studies of its kind – the predictive value of molecular markers (RAPD) for quantitative traits, the basis of what is now known as association genetics [4]

Today, the genebank has an Active Collection (using hermetically-sealed high quality aluminium foil packs, based on advice from seed physiology colleagues Roger Smith and Simon Linington at Kew’s Wakehurst Place) at about 2-3°C, and a Base Collection (a much smaller room, with two sealed aluminium cans, about 150 g, per accession) maintained at -18°C. In recent years a third cold room has been added.

The herbarium of the wild species was also expanded significantly, and provides an invaluable resource for both the conservation and taxonomy research of the wild rices.

The challenge of data management
There are two cultivated species of rice: Oryza sativa (commonly referred to as Asian rice) and O. glaberrima, found mainly in West Africa. There are also more 20 species of wild Oryza, and several genera in the same broad taxonomic group as rice, some of which have been looked by breeders as sources of useful genes; because of their genetic distance from rice, however, their use in breeding is both complex and complicated.

I discovered – to my great surprise – that, in effect, there were three rice collections in the genebank, all managed differently. One of the fundamental issues I grappled with immediately was the need for a functional database system encompassing all the germplasm, not three separate systems that could hardly communicate with each other. These had been developed on an Oracle platform (and an old version that we didn’t have the resources to upgrade). But more fundamentally, database structures and data coding were neither compatible nor consistent across the cultivated and wild species. Many database field names were not the same, nor were the field lengths. Let me give just one fundamental example – the accession number. For O. sativa and the wild species this was a numeric field (but not the same length) while for O. glaberrima, it was alphanumeric! Even the crop descriptors (now updated) were not the same across the collection. For example, the code value for ‘white’ was not consistent. As you can imagine making all the database conversions to achieve consistency and harmony was not without its pitfalls – without losing any data – but we did it. We also went on to develop a comprehensive genebank data management system, the IRGCIS, linking germplasm and genebank management modules with passport, characterization, and evaluation data.

Seed conservation
The FAO Genebank Standards provide guidelines to manage many different operations of a genebank, including seed drying. The drying of rice seeds to a low moisture content and storage at low temperature (as indicated earlier) presents few problems, as such. What is more of a challenge is the multiplication and regeneration of rice germplasm in a single environment at Los Baños in the Philippines, especially for less adapted lines like the japonica rices that are more temperate adapted. We began a collaboration with Professor Richard Ellis at the University of Reading and a leading expert in the whole area of seed conservation. To assist with this research looking at the seed production environment and its effect on seed quality and viability in storage, I hired a germplasm expert from ICRISAT in Hyderabad, Dr N Kameswara Rao, who had completed his PhD with Richard and Professor Eric Roberts a few years earlier. We had already decided to multiply or regenerate germplasm only during the Los Baños dry season (from December to May) when the nights are cooler in the first part of this growing season, and the days are generally bright and sunny. We had anecdotal evidence that seed quality was higher from rice grown at this time of the year than in the so-called wet season, from about July onwards (and the main rice growing season in the Philippines) which is characterized by overcast and wet days, often with a much higher pest and disease pressure. In parallel approaches at Reading (in more or less controlled environments) and in Los Baños, we looked at the response of different rice lines to the growing conditions, and their viability after seed ageing treatments, and confirmed the regeneration approach we had taken on pragmatic grounds. Incidentally, we also moved all field characterization to the wet season, which gave us the advantage of having the field technicians concentrating on only one major operation in each growing season, rather than being split between two or more per season and at different sites on the experiment station.

Germplasm collecting
In 1992 the Convention on Biological Diversity was agreed at the Rio Earth Summit, and is now the legal basis for the biodiversity activities of 193 parties (192 countries plus the European Union) that have ratified the convention, or formally agreed to accept its provisions. For many years, uncertainty over access to and use of biodiversity placed a major block on germplasm collecting activities – but not for rice. Through the SDC-funded project referred to above, we successfully sponsored collecting missions in most of the 23 countries, mainly for traditional varieties in the Asian countries and Madagascar, and for wild rices in these, several eastern and southern African countries, and Costa Rica. We based one staff member, Dr Seepana Appa Rao in the Lao People’s Democratic Republic (Lao PDR). Over more than four years, the various teams collected more than 25,000 samples of rice, and with other donations to the IRG, the collection now stands at more than 110,000 accessions.

Appa Rao and his Lao counterparts visited almost every part of that country, and collected more than 13,000 samples, and in the process learned a great deal about rice variety names and management approaches used by Lao farmers. Duplicates of this valuable germplasm were sent to IRRI, and Lao breeders immediately began to study these varieties with a view to using them to increase the productivity of rice varieties grown by Lao farmers. I believe this is one of the few good examples, within a national program, of an organic link between conservation and use. Regrettably in many national programs conservation and use efforts are often quite separated, so germplasm remains locked up in genebanks that some commentators refer to as ‘germplasm mausoleums’, fortunately not the case with IRRI nor the other CGIAR Consortium centers.

An active research program
In addition to the molecular marker research described earlier, our research focus was on the AA genome wild and cultivated rices, germination standards for wild rices, and on farm conservation.

In 1991, there was a British researcher in the IRGC, Dr Duncan Vaughan, who undertook collecting trips for wild rices, and made some preliminary taxonomic studies. When Duncan moved to the National Institute of Agrobiological Sciences in Tsukuba, Japan in 1993, I hired Dr Bao-Rong Lu, a Chinese national who had completed his PhD on wheat cytogenetics with Professor Roland von Bothmer at the Swedish University of Agricultural Sciences. Bao-Rong stayed at IRRI until 2000, when he moved to Shanghai to become Professor in Biology/Genetics, and Chairman of the Department of Ecology and Evolutionary Biology at Fudan University. He developed an active group working on the wild rices, and also made several collecting trips to Indonesia, Cambodia, and Australia, among other countries, to collect wild rice species.

In 1995 the genetic resources literature was full of papers advocating the virtues and necessity of both in situ conservation of wild species, and the on farm conservation or management of farmers’ varieties as a parallel to conservation, ex situ, in a genebank. While I was neither for or against on farm conservation, I was very concerned that this approach was being ‘pushed’ – at the expense of ex situ conservation, or so it seemed – without really having any empirical evidence to support the various ideas being put around. So I decided to do something about this, and hired a population geneticist and a social anthropologist to study the dynamics of farmer-managed systems in the Philippines, Vietnam, and eastern India. Geneticist Dr Jean-Louis Pham joined IRRI on secondment from IRD (Institut de recherche pour le développement, formerly ORSTOM) in Montpellier, France until 2000 when he returned to IRD.

There were two social anthropologists. Dr Mauricio Bellon, from Mexico, joined in 1995 and stayed for a couple of years before moving to CIMMYT in Mexico; he’s currently with Bioversity International in Rome. He was replaced by Dr Steve Morin from Nebraska in the USA. When the SDC-funded rice biodiversity project ended in 2000, Steve stayed on for a couple of years in IRRI’s Social Sciences Department, but is now with USAID in the Middle East.

Two important findings from this on farm research concern development of different cropping systems options to permit farmers to continue to grow their own ‘traditional’ varieties while increasing productivity; and responses of farmers to loss of diversity after natural disasters (such as typhoons in the case of the Philippines), and how different approaches are applicable for long-term conservation and adaptation.

Click here to see a full list of publications.

The new century
After I left GRC in May 2001 to become IRRI’s Director for Program Planning and Communications, my successor as head of GRC, Dr Ruaraidh Sackville Hamilton, joined IRRI in August 2002. An evolutionary biologist, Ruaraidh is a graduate of Cambridge University, and came to IRRI from the Institute for Grassland and Environmental Research (IGER), now part of the Institute of Biological, Environmental and Rural Sciences at Aberystwyth University.

Fiona Hay joined IRRI in 2009 from the Millennium Seed Bank at Kew, and Ken McNally, who originally joined IRRI in the 1990s as a post-doctoral fellow working on perennial rice, has taken GRC’s molecular research from strength to strength for over a decade, and this has been accelerated by the completion of the rice genome and identification of whole suites of molecular markers.

I am gratified to know that many of the changes I made in GRC are still in place today, even though Ruaraidh has made further improvements, such as the bar coding of all germplasm accessions, and a re-jigging of some of the laboratories to accommodate greater priority on seed physiology and molecular research. Ruaraidh has further championed links with the International Treaty on Plant Genetic Resources for Food and Agriculture, and securing long-term financial support.

A major step forward came about three to four years ago when the Global Crop Diversity Trust began to support the International Rice Genebank. When the Global Seed Vault at Svalbard was opened in 2008, the first samples placed inside were from the International Rice Genebank.



[1] Crisostomo, S., Hay, F.R., Reaño, R. and Borromeo, T. (2011) Are the standard conditions for genebank drying optimal for rice seed quality? Seed Science and Technology 39, 666-672.

[2] McCouch, S.R., McNally, K.L., Wang, W. and Sackville Hamilton, R. (2012) Genomics of gene banks: a case study in rice. American Journal of Botany 99, 407-423.

[3] McNally, K.L., Bruskiewich, R., Mackill, D., Buell, C.R., Leach, J.E. and Leung, H. (2006) Sequencing multiple and diverse rice varieties. Connecting whole-genome variation with phenotypes. Plant Physiology 141, 26–31.

[4] Virk, P.S., Ford-Lloyd, B.V., Jackson, M.T., Pooni, H.S., Clemeno, T.P. and Newbury, H.J. (1996) Predicting quantitative variation within rice using molecular markers. Heredity 76, 296-304.