Regular visitors to my blog will, by now, know that for many years from July 1991 I worked at the International Rice Research Institute (IRRI) in Los Baños in the Philippines, south of Manila. For the first 10 years, I was head of the Genetic Resources Center (GRC), having particular responsibility for the International Rice Genebank (now supported financially by the Global Crop Diversity Trust). Elsewhere on this blog I have written about the genebank and what it takes to ensure the long-term safety of all the germplasm samples (or accessions as they are known) of cultivated rices and related wild species of Oryza.
Well, consider my surprise, not to say a little perplexed, when I recently read a scientific paper¹ that had just been published in the journal Annals of Botany by my former colleagues Fiona Hay (IRRI) and Richard Ellis (University of Reading), with their PhD student Katherine Whitehouse, about the beneficial effect of high-temperature drying on the longevity of rice seeds in storage. Now this really is a big issue for curators of rice germplasm collections, let alone other crop species perhaps.
So why all the fuss, and why am I perplexed about this latest research? Building on a paper published in 2011 by Crisistomo et al. in Seed Science & Technology², this most recent research¹ provides significant evidence, for rice at least, that seed drying at a relatively low temperature and relative humidity, 15C and 15RH—the genebank standard for at least three decades—may not be the best option for some rice accessions, depending on the moisture content of seeds at the time of harvest. It’s counter-intuitive.
But also because germplasm regeneration and production of high quality seeds is one aspect of germplasm conservation most likely to be impacted by climate change, as Brian Ford-Lloyd, Mauricio Bellon and I emphasized in our chapter in Genetic Resources and Climate Change.
To explain further, it’s necessary to take you back 24 years to when I first joined IRRI.
The first six months or so
The Director General in 1991, Dr Klaus Lampe, encouraged me to take a broad view of seed management services at IRRI, specifically the operations and efficiency of the International Rice Genebank (IRG). It was also agreed that I should develop research on the germplasm collection and its conservation, something that had not been considered when the GRC Head position was advertised in September 1990. I should add that in negotiating and accepting the GRC position, I had insisted that GRC should have a research arm, so to speak. I guess I was in a fairly strong negotiating position.
Once at IRRI, I didn’t rush into things. After all, I had never run a genebank before let alone work on rice, although much of my career to that date had been involved in various aspects of germplasm conservation and use. But after about six months, I reckon I’d asked enough questions, looked at how the genebank was running on a day-to-day basis. I had developed a number of ideas that I thought should vastly enhance the long-term conservation of rice germplasm, but at the same time allow all the various operations of the genebank run smoothly and hopefully more efficiently. In one sense, managing the individual aspects or operations of a genebank are quite straight-forward. It’s bringing them all together that’s the tricky part.
There was another ‘delicate’ situation to address, however. All the Filipino staff had worked for only one person for many years, my predecessor as head of the genebank (then known as the International Rice Germplasm Center, or IRGC), Dr TT Chang. It’s not an understatement to say that many of these staff were fiercely loyal to Dr Chang (loyalty being one of their greatest virtues), firmly fixed in their ways, and didn’t feel—or maybe understand—that changes were desirable or even necessary. It was a classic change management situation that I was faced with. I needed to help them evaluate for themselves the current genebank management focus, and propose (with more than a little encouragement and suggestions from me) how we might do things differently, and better.
Some radical changes
But I don’t think anyone foresaw the radical changes to the management of the genebank that actually emerged. The genebank was ‘the jewel in IRRI’s crown’, the facility that every visitor to the institute just had to see. It seemed to run like clockwork—and it did, in its own way.
Staffing and responsibilities
Apart from several staffing issues, I was particularly concerned about how rice germplasm was being regenerated in the field, and how it was handled prior to medium-and long-term storage in the genebank. There were also some serious germplasm data issues that needed tackling—but that’s for another blog post, perhaps.
In terms of genebank operations, it was clear that none of the national staff had responsibility (or accountability) for their various activities. In fact, responsibilities for even the same set of tasks, such as germplasm regeneration or characterization, to name just two, were often divided between two or more staff. No-one had the final say. So very quickly I appointed two staff, Flora ‘Pola’ de Guzman and Renato ‘Ato’ Reaño to take charge of the day-today management of the seed collection (and genebank facilities per se) and germplasm regeneration, respectively. Another staff, Tom Clemeno, was given responsibility for all germplasm characterization.
Working in the field
But what seemed rather strange to me was the regeneration of rice germplasm at a site, in rented fields, some 10km east of the IRRI Experiment Station, at Dayap. This meant that everything—staff, field supplies, etc.—had to be transported there daily, or even several times a day. It made no sense to me especially as the institute sat in the middle of a 300 ha experiment station, right on the genebank’s doorstep. In fact, the screenhouse for the wild rice collection had been constructed on one part of the station known as the Upland Farm. To this day I still don’t understand the reasons why Dr Chang insisted on using the site at Dayap. What was the technical justification?
Also the staff were attempting to regenerate the germplasm accessions all year round, in both ‘Dry Season’ (approximately December to May) and the ‘Wet Season’ (June to November). Given that the IRRI experiment station has full irrigation backup, it seemed to me that we should aim to regenerate the rice accessions in the Dry Season when, under average conditions, the days are bright and sunny, and nights cooler, just right for a healthy rice crop, and when the best yields are seen. The Wet Season is characterized obviously by day after day of continuous rainfall, often heavy, with overcast skies, and poor light quality. Not to mention that Wet Season in the Philippines is also ‘typhoon season’. So we separated the regeneration (Dry Season) from the characterization (Wet Season) functions.
But could we do more, particularly with regard to ensuring that only seeds of the highest quality are conserved in the genebank? That is, to increase the longevity of seeds in storage—the primary objective of the genebank, after all, to preserve these rice varieties and wild species for future generations? And in the light of the latest research by Katherine Whitehouse, Fiona and Richard, did we make the right decisions and were we successful?
Seed environment and seed longevity
That’s where I should explain about the research collaboration with Richard Ellis at that time (Ellis et al. 1993; Ellis & Jackson 1995), and helpful advice we received from Roger Smith and Simon Linington, then at Kew’s Wakehurst Place (and associated with the founding of the Millennium Seed Bank).
I hired a post-doctoral fellow, Dr N Kameswara Rao, on a two-year assignment from sister center ICRISAT (based in Hyderabad). Kameswara Rao had completed his PhD at Reading under seed physiologist Professor Eric Roberts.
We set about studying the relationship between the seed production environment and seed longevity in storage, and the effect of sowing date and harvest time on seed longevity in different rice types, particularly hard-to-conserve temperate (or japonica) rice varieties (Kameswara Rao & Jackson 1996a; 1996b; 1996c; 1997). And these results supported the changes we had proposed (and some even implemented) to germplasm regeneration and seed drying.
In 1991, the IRG did not have specific protocols for germplasm generation such as the appropriate harvest dates, and seed drying appeared to me to be rather haphazard, hazardous even. Let me explain. Immediately after harvest, rice plants in bundles (stems, leave and grains) were dried on flat bed dryers before threshing, heated by kerosene flames, for several days. Following threshing, and before final cleaning and storage, seeds were dried in small laboratory ovens at ~50C. It seemed to me that rice seeds were being cooked. So much for the 15C/15RH genebank standard for seed drying!
During the renovation of institute infrastructure in the early 1990s we installed a dedicated drying room³, with a capacity for 9000 kg, in which seeds could be dried to an equilibrium 6% moisture content (MC) or thereabouts, after a week or so, under the 15/15 regime.
Now this approach has been apparently turned on its head. Or has it?
To read the headlines in some reports of the Whitehouse et al. paper, you would think that the 15/15 protocol had been abandoned altogether. This is not my reading of what they have to report. In fact, what they report is most encouraging, and serves as a pointer to others who are engaged in the important business of germplasm conservation.
In her experiments, Katherine compared seeds with different initial MC harvested at different dates that were then dried either under the 15/15 conditions, or put through up to six cycles of drying on a batch drier, each lasting eight hours, before placing them in the 15/15 seed drying room to complete the drying process, before different seed treatments to artificially age them and thereby be able to predict their longevity in storage before potential germination would drop to a dangerous level.
This is what Katherine and her co-authors conclude: Seeds harvested at a moisture content where . . . they could still be metabolically active (>16.2%) may be in the first stage of the post-mass maturity, desiccation phase of seed development and thus able to increase longevity in response to hot-air drying. The genebank standards regarding seed drying for rice and, perhaps, for other tropical species should therefore be reconsidered.
Clearly seeds that might have a higher moisture content at the time of harvest do benefit from a period of high temperature drying. Because of the comprehensive weather data compiled at IRRI over decades, Katherine was also able to infer some of the field conditions and seed status of the Kameswara Rao experiments. And although the latest results do seem to contradict our 1996 and 1997 papers, they provide very strong support for the need to investigate this phenomenon further. After all, Katherine studied only a small sample of rice accessions (compared to the 117,000+ accessions in the genebank).
The challenge will be, if these results are confirmed in independent rice studies—and even in other species, to translate them into a set of practical genebank standards for germplasm regeneration and drying and storage for rice. And it must be possible for genebank managers to apply these new standards easily and effectively. After all many are not so fortunate as GRC to enjoy the same range of facilities and staff support.
I’m really pleased to see the publication of this research. It’s just goes to demonstrate the importance and value of research on genebank collections, whatever the crop or species. Unfortunately, not many genebank are in this league, so it behoves the CGIAR centers to lead from the front; something I’m afraid that not all do, or are even able to do. Quite rightly they keep a focus on managing the collections. But I would argue that germplasm research is also a fundamental component of that management responsibility. Brownie points for IRRI for supporting this role for almost a quarter of a century. And for Fiona as well for ensuring that this important work got off the ground. Good luck to Katherine when she comes to defend her thesis shortly.
A recent seminar
On 12 November, Fiona gave a seminar at IRRI in the institute’s weekly series, titled How long can rice seeds stay alive for? In this seminar she explores changes that have been made to genebank operations over the years and the extent to which these did or did not affect the potential longevity of rice seeds in the genebank. She talks in some detail about the benefits of initial ‘high temperature’ drying that appears to increase potential longevity of seeds. As I queried with her in a series of emails afterwards, it’s important to stress that this high temperature drying does not replace drying in the 15/15 drying room. Furthermore, it will be necessary at some stage to translate these research findings into a protocol appropriate for the long term conservation of rice seeds at -18C.
Fiona has graciously permitted me to post her PowerPoint presentation in this blog, and the audio file that goes with it. You’ll have to open the PPT file and make the slide changes as you listen to Fiona speaking. I’ve done this and it’s actually quite straightforward to follow along and advances the slides and animations in her PPT. Click on the image below to download the PPT file. Just open it then set the audio file running.
Here’s the audio file.
I am also pleased to see that the CGIAR genebanks have also established a seed longevity initiative under the auspices of the Global Crop Diversity Trust. You can read more about it here.
Seed storage – an interesting anecdote
In 1992 we implemented the concept of Active (+3-4C) and Base (-18C) Collections in the IRG. Before then all rice seeds were stored in small (20g if I remember correctly) aluminium cans. We retained the cans for the Base Collection: once sealed we could expect that they would remain so for the next 50 years or more. But in the Active Collection there was no point having cans, if they had to be opened periodically to remove samples for distribution, and could not be re-sealed.
So we changed to laminated aluminium foil packs. Through my Roger Smith and Simon Linington we identified a manufacturer in the UK (from near Manchester I believe) who could make packs of different sizes, using a very high quality and tough laminate of Swedish manufacture (originally developed to mothball armaments). It had an extremely low, if not zero, permeability, and was ideal for seed storage. Unfortunately by the time we made contact, the company had gone into liquidation, but the former managing director was trying to establish an independent business. On the strength of a written commitment from IRRI to purchase at least 250,000 packs, and probably more in the future, this gentleman was able to secure a bank loan, and go into business once again. And IRRI received the seed storage packages that it ordered, and still uses as far as I know. The images below show genebank staff handling both aluminium cans in the Base Collection and the foil packs in the Active Collection. You can see the Active Collection in the video below at minute 1:09.
¹ KJ Whitehouse, FR Hay & RH Ellis, 2015. Increases in the longevity of desiccation-phase developing rice seeds: response to high-temperature drying depends on harvest moisture content. Annals of Botany doi:10.1093/aob/mcv091.
² S Crisostomo, FR Hay, R Reaño and T Borromeo, 2011. Are the standard conditions for genebank drying optimal for rice seed quality? Seed Science & Technology 39: 666-672.
³ If you would like to see what the seed drying room looks like, just go to minute 9:40 in the video below: