Story and photos by Eve Daniels
Soybean farming was a guessing game in Minnesota, circa 1950s. Breeders based their decisions on how the plants looked, which involved a lot of trial and error. Back then, soybeans were a forage crop, planted on just a few acres and used for livestock feed. Most varieties were straight introductions from China, with limited success on Minnesota soil.
Today, the guesses are far more educated. Minnesota farmland is currently home to almost 7 million acres of soybeans. We’re the third largest soybean-producing state in the country. It’s a booming industry, and that progress has much to do with the University of Minnesota.
To see how far we’ve come, look no further than the University’s soybean breeding and genetics program. We’ve evolved from handwritten notes to digital spreadsheets, floppy disks to cloud storage. Breeding decisions are based on DNA, versus the naked eye. Trial-and-error is out—and genomic selection is in.
In the soybean breeding lab on the St. Paul campus, space that was once used for lab equipment is now filled with computers for analyzing big data. The data continues to grow in size and scope, so those lab computers are backed by supercomputers on the Minneapolis campus.
“Nowadays, you can go to your doctor and get your DNA sequenced for a few thousand dollars to find out your probability for diseases. The same thing is happening in plant breeding,” says Aaron Lorenz, incoming director of the University’s soybean breeding and genetics program. He’s one of four new faculty members hired as part of the Global Food Ventures initiative, part of the Minnesota’s Discovery, Research and InnoVation Economy (MnDRIVE) partnership between the U of M, the state legislature, and industry.
After gathering soybean samples from the field, University researchers extract DNA and send it to a genotyping core facility. Then, they analyze that data and compare it with historical results. Soon, they’ll use all of this information to make highly accurate predictions.
“At this point, we can genotype our plant breeding progenies with high-density markers, pretty easily, inexpensively, and efficiently. The challenge, then, is to use that DNA marker information for making selection decisions,” says Lorenz. “At the same time, how can we compile historical data, put it into an analyzable format, and use it to make predictions going forward?”
As he plans the future of the University’s soybean program, Lorenz is determined to answer those questions.
Coming full circle
As a University of Minnesota undergraduate, Lorenz’s favorite course was plant genetics and breeding (of course!). His instructors, who led various plant breeding programs at the University, were eager to share their real-world knowledge. Lorenz enjoyed their class so much that he decided to major in plant sciences.
Fifteen years later, with a few more degrees and faculty positions in his CV, Lorenz has come full circle. This past April, he returned to the University as an assistant professor in the Department of Agronomy and Plant Genetics. Next summer, he’ll take the helm as director of the soybean program, which his former undergraduate instructor, James Orf, has led since the early 1980s.
As he plans for retirement, Orf is excited to pass the torch to Lorenz: “Aaron’s interests and area of expertise are on the cutting edge of plant breeding,” he says.
Prior to his current position, Lorenz was a researcher in corn breeding at the University of Nebraska, but his agricultural experience dates back to childhood. He grew up on a corn and soybean farm near Worthington, Minn., which his parents still maintain. Even as a child, he preferred the study to the practice.
“Farming is a business,” says Lorenz, “and I’ve always been more interested in science than business.”
Six decades and counting
Lorenz will be the third director of the soybean breeding and genetics program. It started with Jean Lambert in the 1940s, who helped develop Minnesota’s first soybean variety, Renville, paving the way for nearly 100 new Minnesota varieties to date.
When Orf took over in the ‘80s, he shifted the focus from commodity crops to developing food-grade varieties for tofu, nattō and related products. He also played a lead role in expanding the growing range further north.
In the coming years, Lorenz plans to shift more toward breeding methodology research. “That means figuring out how to optimally use markers for breeding programs, and actually conducting those experiments, as opposed to focusing all efforts on developing new varieties,” he explains.
Used in combination with good field evaluation practices, this approach means greater efficiency for breeders, farmers, and the industry in general. “Every year, a breeder will generate between 10,000 and 20,000 candidates to go into the field,” says Lorenz. “I envision that 10 years from now, each of those candidates will be predicted for every single environment. Based on those predictions, the breeder will choose which to plant.”
To help in making those predictions, Lorenz and his fellow researchers also will compare field notes dating back to 1945. They’ve already scanned in all the old log books. Now they need to enter relevant information into a database. Next, they’ll add a decade of data stored on 3.5-inch floppy disks, and work their way up from there. Eventually, they’ll connect this historic data to current records to develop a statistical model for future varieties.
Why all this extra effort? Bottom line: “Genotyping on a per progeny basis is more cost-effective than phenotyping across multiple locations and years,” says Lorenz. “The less guesswork, the better off we are financially.”