Dreher costed out field and lab procedures in great detail. The result will be more efficient breeding methods.

A Close Look at Biotech Breeding Costs: The Details Make a Difference

With the rise of biotechnology, novel techniques that complement tried-and-true breeding methods allow plants to be screened in the laboratory rather than the field. Are the new techniques always cost-effective?

To choose wisely among the many tools of the breeding trade, CIMMYT researchers need to answer that question. A judicious choice could lower the cost and speed the pace of plant breeding, bringing millions of dollars in additional benefits to farmers and consumers.

Plant breeding at CIMMYT is well positioned to benefit from the wave of innovation brought by biotechnology, including marker-assisted selection (MAS) of breeding materials. As biotechnology comes into its own at CIMMYT, the time has come to take a fresh look at the costs of conventional and molecular-based breeding schemes, their relative advantages and disadvantages, and optimal strategies for achieving different breeding objectives. This is a formidable task, but it is integral to conducting research efficiently.

Selecting a Real Case for Analysis

Kate Dreher, a research associate, and Michael Morris, an economist (both with the CIMMYT Economics Program) organized a study that examined the costs of various breeding schemes designed to transfer a single gene/single trait into maize. The study relied on the expertise of Applied Biotechnology Center (ABC) staff Mireille Khairallah and Jean-Marcel Ribaut, and Maize Program staff Shivaji Pandey and Ganesan Srinivasan.

In setting up the study, Dreher and Morris met with Maize Program and ABC researchers to identify examples of breeding projects that incorporated disparate parameters, such as the transfer of a single gene versus multiple genes, or the detection of important traits earlier rather than later in the breeding process.

Dreher, Morris, and the other researchers were interested in breeding projects already underway at CIMMYT because they wished to obtain actual, rather than simulated, cost data. For two reasons, the group decided to focus on projects related to quality protein maize (QPM).* First, CIMMYT was already using molecular markers to introduce the quality protein trait into experimental maize lines. Second, the conversion of "normal" maize lines into QPM lines involved selecting for only a single trait, the quality protein trait, which made QPM a relatively straightforward example for an initial cost study.

The case study concentrated on:

• generating detailed information about the costs of conducting conventional breeding operations and of implementing MAS procedures at CIMMYT's facilities in Mexico;
• determining the cost-effectiveness of using MAS for particular breeding applications, specifically QPM; and
• providing insights into the potential cost-effectiveness of future applications of MAS.

The analysis proceeded in three stages. First, field and laboratory operations involved in conventional and MAS breeding were identified and costed out. Dreher observed and questioned researchers and lab technicians as they went through each stage of their work, from start to finish, to gather specific information on the costs involved.

Second, CIMMYT maize breeders and molecular geneticists were asked to design representative breeding schemes for QPM line conversion, and four hypothetical, stylized breeding schemes were selected. Two of the schemes relied solely on conventional breeding methods and phenotypic evaluation, and two incorporated MAS.

Third, the laboratory and field parameters put forth in each breeding scheme were used to calculate the total cost of implementing that particular scheme.

Although they are still analyzing the results of the study, Dreher and Morris have come to several preliminary conclusions, the most salient being that the relative cost-effectiveness of various conventional and MAS schemes depends on the detailed circumstances of each particular application. Decisions about whether to incorporate MAS into a breeding scheme are likely to require a case-by-case analysis.

When phenotypic screening is simple (in other words, when it is relatively easy to determine whether a given plant variety possesses a given trait, such as a certain grain color), conventional breeding is, and will continue to be, extremely cost-effective. Conversely, when phenotypic screening is expensive, technically difficult, or even impossible, MAS will often be advantageous, according to Dreher.

"Nematode resistance or tolerance is a case in point," she says. "Most cereal nematodes are insidious, below-ground pathogens that are difficult to diagnose and quantify. Determining a variety's resistance through conventional selection techniques requires sampling, extracting, and counting the nematodes, extrapolating the results, and diagnosing the effects on the variety, a process that is costly, labor intensive, and only moderately reliable. In this case, if we have molecular markers that are linked to the gene (or genes) that confer resistance or tolerance, MAS offers an alternative that is simple, direct, and very reliable.

"Or take the case of maize streak virus," Dreher continues. "CIMMYT can't screen for it here in Mexico, because strict quarantine regulations rightly prohibit the introduction of the disease into the country. Molecular markers are the only choice for conducting such work at headquarters."

Time Is Still Money

Aside from cost, another important factor affecting efficiency--time--is addressed in the study. Marker-assisted selection often allows breeders to cut down on the number of seasons needed to produce a desired product. For farmers and seed companies, the benefits of developing and releasing a new variety more rapidly can be significant, as indicated in a study by the International Rice Research Institute (IRRI) and Chum Phae Rice Experiment Station, Thailand.** The study concluded that additional benefits of more than US$ 18 million were realized over the life of a particular rice variety because the variety was released two years earlier than usual.

"Even if we assume a high-end MAS scheme that might run a few thousand dollars more than a conventional scheme," Morris notes, "the extra cost pales in comparison to the additional benefit to our clients when a variety becomes available sooner to farmers."

Although data analysis is continuing, the study has already produced concrete results. "Kate Dreher spent a huge amount of time around the labs and in the field costing out day-to-day procedures that usually don't get reviewed at that level of detail," says Morris. "For station and lab management, these data are very useful. In addition they provide the kind of solid information needed by our national program partners to put together biotech project proposals based on real figures."

Finally, Morris adds, the analytical tools (linked spreadsheets) developed by Dreher will allow researchers and managers to conduct additional detailed case studies aimed at increasing the efficiency of breeding schemes for CIMMYT and national program scientists alike.

For more information:

* Quality protein maize, which was developed through conventional breeding processes, has the nutritional advantage of containing twice the amount of two essential amino acids, tryptophan and lysine, as normal maize. See p. 6.

** S. Pandey and S. Rajatasereekul, 1999, "Economics of planting breeding: The value of shorter breeding cycles for rice in Northeast Thailand," Field Crops Research 64:187-197.