The road toward global food security is not without challenges. The population will balloon to 9 billion in 2050. The signs of climate change have never been so real—frequent floods, droughts, and storm surges. Storm surges make farmland in coastal areas too salty for most crops to grow. Also, pathogens and pests evolve. Therefore, a rice variety may lose its resistance to new strains of pathogens or insects.
"With so many challenges that we are facing now, we can't just continue with what we are doing," said Eero Nissilä, head of the Plant Breeding, Genetics, and Biotechnology Division (PBGB) at the International Rice Research Institute (IRRI). "There need to be changes in the way we do breeding at IRRI."
IRRI plant breeder Bert Collard agrees. "A revolution in rice breeding is what we need now," Dr. Collard said. "Not much has changed for the last 50 years. The methods used today in Asia are generally the same as the ones used in the 1960s-'70s. More importantly, the rate of yield increase or genetic gain for irrigated varieties is less than 1% per year."
Doubling genetic gains
Thus, Dr. Nissilä and his team are now restructuring IRRI’s entire breeding operations. Transforming Rice Breeding (TRB), a project funded by the Bill & Melinda Gates Foundation, is one important component of this new breeding factory, which focuses on irrigated rice. IRRI is aiming to double the rate of genetic gains—the increase in crop performance that is achieved through genetic improvement programs per unit time of breeding—or even make it higher (more than 2%).
"To make breeding more efficient, we need to change how we organize our breeding operations," Dr. Nissilä said. "We need to restructure the overall breeding pipeline."
A shorter cycle
Traditional breeding takes 8 to 9 years to develop a variety and even longer for the variety to reach farmers. Now, the breeding process has been shortened to about 6 years (See How a modern variety is bred ).
"In breeding, one year is significant," said Dr. Nissilä. "Investments in plant breeding with costs incurred during the phase of developing varieties are only realized when farmers grow the varieties. A longer breeding cycle means economic opportunity cost to farmers of losing the chance to grow better varieties earlier."
A 1999 study conducted by IRRI shows that reducing a breeding cycle by 2 years has an economic benefit of about US$18 million over the useful life of the variety.1
Breeding by demand
Taking a cue from breeding in the private sector, IRRI's breeding pipeline should run like a business operation—production by demand. For Dr. Nissilä, this should be the underlying philosophy of the new breeding pipeline.
"That's why the TRB team is working closely with IRRI’s Social Sciences Division to better understand the needs of farmers," said Dr. Collard. "This will help guide us in 'must have' traits as well as ‘good to have traits' in our breeding objectives."
The team recognizes that national programs have different needs. It could be a breeding line, a gene donor, or a gene marker. If a researcher needs only a gene marker and wants to do the breeding in a lab himself or herself, IRRI will provide some help on how to use that marker. If a national program is interested only in salinity tolerance, the team will use its trait pipeline to provide the best package of salinity tolerance for the national program’s breeding. On the other hand, if a national program needs a new rice variety, then the team will use its product development pipeline.
"We can help organize a national program's overall product development and product profiling when needed," Dr. Nissilä said. "We will not only give a trait, say, salinity tolerance, but, it can be combined with many traits that one needs in one's 'market,' so to speak. We can fit their needs.
"But, our aim in the future is that we won’t be doing the 'job,'" he added. "We want them to do it themselves. Our role will be more of a consultant and provider of training. Then, we exit when they are already able to do the work."
The role of hubs
"But, if national programs want us to do this in the first round, we will breed the variety itself," said Dr. Nissilä. "This is where our regional breeding hub system comes into the picture. Our hubs in India, for the South Asian region, and in Burundi, for East and Southern Africa, allow us to localize the breeding process— we select and produce the material in the region where it will be used."
IRRI now has many modern breeding options for a more efficient and cost-effective process than the conventional pedigree method. The conventional breeding way goes like this: choose plants with the desired trait, cross-pollinate them, wait for the offspring to reach maturity, select the best performers, and then repeat the process to the nth degree until one obtains a plant that fits farmers' need.
Today, these modern methods include marker technology, multilocation trials, and rapid generation advance (RGA). RGA produces fixed lines or plants that no longer segregate. Scientists call them "homozygous" or "genetically stable." "In RGA, plants are grown at high density with low nutrients in greenhouses or screenhouses so that they flower and mature earlier, thus shortening their life cycle," said Dr. Collard. "Therefore, several generations (e.g., three to four) can be advanced in one year." In addition to time savings, this method also saves labor and resources, and costs considerably less than other methods.
New breeding pipeline model
The new structure of IRRI’s breeding pipeline will hopefully be a model for the national programs to follow. Historically, the national partners have modeled their programs on IRRI’s old system.
"We hope that the new processes of rice breeding at IRRI will help catalyze new thinking in the national programs so they can also restructure their own factory," Dr. Nissilä said.
"There is a great interest in Africa and Asia to do the same," said Dr. Collard. "There may be some reluctance about specific aspects because it is a 'big change' generally. But with the criteria of time and efficiency, it wins hands-down."
To have an evidence-based comparison between the old and the new system, IRRI will conduct a benefit-cost analysis.
"IRRI has never put value to its costs until recently," Dr. Collard recalled.
In terms of full-cost recovery, IRRI’s breeding pipeline is working to improve efficiency in three areas. First is the use of knowledge capital. The pipeline should maximize the use of expertise in the Division.
Second is the use of a running budget. Will it be split into a hundred small activities or put in a strategically based operation?
Third is the use of investment. "This is important since it is a technology-driven operation and this will ensure an efficient use of investment," said Dr. Nissilä. "This is where our cross-cutting strategy comes in. For instance, instead of adding seven more molecular labs, we will have one world-class lab!" With one state-of-the-art facility servicing all breeding pipelines, the use of expertise will be more efficient because activities won’t overlap or be duplicated compared to having several scientists working on their own in their respective small labs.
Moreover, Dr. Collard mentioned that the team is implementing computerized systems for data collection and exploring using mechanization in field trials to save on time and labor.
So, with efficient use of knowledge capital, a running budget, and investment capital coming into play, Dr. Nissilä concluded that there is no better way of organizing breeding operations.
In any business operation, one is usually asked how to be sure that the money is used efficiently for its intended purpose. According to Dr. Nissilä, the new breeding pipeline has put in place some indicators of success.
The first is cost-efficiency. What costs what? The second is progress. Have we increased the genetic gain as promised? The third is impact. What products are used in the national programs? How many farmers' seeds were produced? How much income did these products make for farmers?
'And, of course, one more measure of success is making our national partners more independent and their breeding activities more sustainable," Dr. Collard said.
A never-ending story of service
Will there be a time when IRRI’s new breeding pipeline will no longer be needed in the future? “Breeding is a never-ending story of service, because the technology upstream is developing,“ Dr. Nissilä explained. Things that are not normal today will be routine tomorrow. To stay alive in this business, we need to have a very wide product portfolio and diverse expertise.
"Besides, the needs are very wide. For example, in Rwanda, they will need our lines while in India we need collaboration in genomics and biotechnology applications."
The story has started
But, even before the funding of the foundation arrived in November 2013, Dr. Nissilä and his team had started carrying out some changes in IRRI's rice breeding pipeline (See The pipeline grows stronger). They restructured the breeding activities from research-centric to a product-oriented pipeline. Now, the TRB project brings the resources as the fuel to propel the new factory of irrigated variety pipelines.
Many examples of change are brewing in IRRI's new breeding factory. One is the modernization of IRRI's data collection and analysis and the organizing of a load of information on rice breeding.
In fact, the Institute has come a long way in terms of its baseline. In IRRI’s classic book on breeding called Rice Improvement by Jennings, Coffman, and Kauffman,2 published in 1979, data or analysis was not even mentioned, Dr. Collard said. "During that time, plant breeders decided whether a plant was good or bad 'by eye assessment,'" he added.
Today, decision-making strongly depends on data collection and analysis. "After field work, we spend a significant time in our office analyzing data," said Dr. Collard. "Modern techniques such as the use of new statistical analysis, computerized data management, and molecular breeding approaches may be small things. But, the sum of which was all these spells a great difference in streamlining high-quality data for a more efficient way of breeding."
A significant improvement of IRRI’s breeding program in the last 2 years has been its testing of breeding materials at multilocations much earlier than before. With this multilocation trial system and regional hubs, the effectiveness of the development of new varieties will improve. Furthermore, decentralized breeding in hubs allows easy seed transfer of IRRI breeding material to the region.
With the TRB project, Dr. Nissilä and his team in the PBGB Division are looking forward to a future when farmers need not wait long to plant an improved variety apt for the challenges of the time.
"The TRB project aims to accelerate the current breeding pipeline in developing varieties, shorten the breeding cycle, and dramatically increase the efficiency of breeding operations," said Dr. Collard. "But, the big picture of this project is to help resource-poor farmers in Asia and Africa improve their food and income security."