Soybean breeders are something like financial investors with most of their assets in just a few stocks. Both need more diversity - the investors in their portfolios and the breeders in their soybean germplasm.

Fortunately, there is promising new material becoming available for soybean breeders. But let's first look at why diversity is needed.

"There is a very limited gene pool being used today to develop commercial soybean varieties, and that represents a major threat to long-term yield improvement and profitability," explains Randall Nelson. He's curator of the USDA soybean germplasm collection at the National Soybean Research Laboratory (NSRL), located at the University of Illinois.

"Five ancestral varieties account for 55% of the genes in all the varieties currently being grown in the U.S." says Nelson. "Another 17 ancestors provide an additional 35% of the genes. Breeding programs are relying almost totally on these few ancestors for the genetic resources for future yield improvement."

In an effort to increase genetic diversity, the NSRL is working with exotic soybean lines from China and other areas. The goal is to raise yields, enhance oil and protein content and improve insect and disease resistance.

"In our program to increase yields, we have tested progeny from crosses involving more than 250 exotic germplasm parents," Nelson reports. "Currently, we have developed high-yielding experimental lines from about 90 of those parents, and that is significant progress.

"With the support of the Illinois Soybean Program Operating Board and the United Soybean Board, we have been able to evaluate the yield potential of these experimental lines and determine their genetic relationship to our current varieties. Biotechnology allows us to compare varieties at the DNA level."

That technology, says Nelson, lets researchers compare two varieties that look and yield similarly, and determine if they are different genetically. "Our ultimate goal," he notes, "is to ascertain which of the genetic differences are really important in determining yield."

Early yield results are encouraging. In two years of testing in central Illinois, the experimental line LG92-1255, at 50.6 bu/acre, yielded significantly more than the best-yielding commercial variety, Savoy, at 46.3 bu/acre.

"One of the LG92-1255 grandparents is a Chinese line that came to the U.S. in 1926," Nelson says. "Our DNA analysis demonstrated that this Chinese introduction is genetically distinct from our modern varieties."

Nelson and his colleagues also have identified experimental lines derived 100% from exotic germplasm. All four grandparents are genetically different from modern varieties. They have been tested against top commercial varieties.

"Last spring we released one of those lines, LG94-1128," Nelson says. "Its average yield was 94% of the best commercial variety and 98% of the average of the four commercial varieties in the test. It's among the highest-yielding lines we've developed, and we're very excited about the possibilities it will have in the hands of breeders."

Nelson and colleagues are developing new lines combining more exotic parents. In 1999 they tested lines that had six to eight exotic lines in their ancestry and no U.S. varieties as parents.

"Some of those lines are yielding as well as the best varieties in our tests," Nelson reports.

They crossed another experimental line, derived from 100% exotic germplasm, with Chamberlain. From that combination they developed an experimental line that yielded significantly higher than Chamberlain did. In two years of testing in central Illinois it yielded as well as Macon, the best variety in the tests.

That line also has been released to soybean breeders.

"Soybean breeders could cross these experimental lines with modern varieties to incorporate significant new genetic diversity into their breeding programs," Nelson points out. "Because this diversity is already in such high-yielding experimental lines, it could be the needed genetic stimulus to improve the yield of new varieties and increase the rate of yield improvement in the long term."

Nelson reports that two lines containing exotic germplasm were released to about 45 public and private soybean breeders in 1997, and eight additional lines were released in 1998 and 1999. "We have had a very positive response," he says.