Arguably the most famous release from the University of Minnesota’s apple breeding program is Honeycrisp. According to David Bedford, a senior research fellow and apple breeder at the university, the groundbreaking variety completely reset the bar for the industry.
“It changed our world,” Bedford says. “We had to really clean out some of our breeding program, some of the selections that were in the hopper that we thought were good but just didn’t live up to that.”
Honeycrisp isn’t the program’s only notable variety; other major releases include Zestar, Rave/First Kiss and SweeTango.
Whittling 10,000 Crosses Down to One
Fruit breeding is a long and arduous process. Bedford and Matt Clark, an associate professor in the Department of Horticultural Science and director of research at the Minnesota Landscape Arboretum, make roughly 10,000 crosses each year. Over several years, those thousands are whittled down to just one or two selections fit for commercial release.
Bedford likens fruit breeding to running a marathon, where success requires playing a long game. He notes that a common mistake early in a breeder’s career is believing every cross is merely misunderstood and deserves a home. Early on, he gave each cross three strikes before culling it. Eventually, he reduced that to two. Today, it’s one strike and out.
“Maybe they’re having a bad year, but they still should have good texture and good flavor,” he says.
From initial cross, transferring selected pollen from one variety to the flower of another, to harvesting seeds, rearing seedlings and waiting for the tree to fruit is a four-to-six-year timeline. Woody plants typically go through a lengthy juvenile period before bearing fruit, but the university speeds this up by grafting young seedlings onto dwarf rootstock rather than growing them on their own roots. This technique also maximizes orchard space, allowing roughly 2,000 trees per acre instead of 400.
“We cut that bearing time down. I’d say on their own roots they would take oh six, seven, eight, even 10 years sometimes,” he says. “Now we’ve cut it down closer to four years. And they still have to go through their juvenile period. It’s not quite the same as buying a tree from a nursery because … the scion on that tree is already in its adult phase, and we have to still grow through this juvenile phase. But we’ve cut that juvenile phase in half at least.”
Once the fruit is ripe, a cross has only two to three minutes to impress breeders with its texture and flavor. If it succeeds, it receives an official tracking number and advances to the next round of trials.
“Once we’ve determined that they’re ripe, that’s about how long we’re going to spend making that decision because there’s another thousand ahead of them,” Bedford says. “We have to make that decision quickly. That’s all part of the process is getting through. It’s really a numbers game. We’re loading the best things we can into that numbers game, but they don’t always come out the way we want.”
Passing that first selection takes four to five years, but it only unlocks additional rounds of rigorous testing.
“I do remember the study that we did on that showed that they spent an average of seven years in that second test before we decided they’re good enough to keep going forward with,” he says. “Something that’s on a good track forward is going to be probably 12 years before we can make a decision to go on to the next and final round. And yeah, not cheap.”
Striking Gold With SweeTango
SweeTango was one of those good surprises, he says, with both flavor and texture.
“There’s only a couple apples that I remember the first time I ate them and SweeTango, Minneiska, was one of those. It just stood out,” he says.
A cross between Honeycrisp and Zestar, SweeTango blends Zestar’s sweet-tart balance with Honeycrisp’s distinct crunch. Bedford notes that achieving this level of crispness and texture is exceptionally difficult.
“We’re making the best crosses we possibly can with good crisp parents … maybe 2% of all those crosses come out with the good texture,” he says. “Even with our best intentions, we don’t see it that often and then we have to satisfy all the other 20 criteria that we’re trying to find in an apple.”
With SweeTango, however, the program achieved what Bedford calls the “gold standard” profile of high sugar, high acid and visual appeal.
“That one not only got a number that day, but that’s one of the ones you start thinking about, ‘Boy, should we start thinking about a name?’” he says. “Of course, it’s another 10 years or so before that actually happens. But that was a good one right from the start.”
Accelerating the Breeding Pipeline With Genomics
While the timeline can seem overwhelming, Clark says the sheer volume of evaluation work at every stage keeps breeders grounded and focused.
“You don’t have to be worried about 10 years from now,” he says. “That stuff is going to get its chance next year or in five years. … The pipeline is always full. Every step of the pipeline has work that has to be done and so you have to stay focused on that too.”
To manage the pipeline, Clark and Bedford use DNA markers to whittle the 10,000 annual crosses down to roughly 5,000, filtering early for the seedlings most likely to succeed.
“Obviously we don’t get all the bad ones, and we’re not sophisticated enough to only pick out the good ones, but that means when we plant out those 5,000, they’re sort of partially vetted at least rather than just wild cards,” he says. “It’s better to throw that seedling away when it takes up 3 square inches in a greenhouse than 30 square feet in a field seven years later. It definitely helps us. What it doesn’t do — and it surprises people sometimes — it doesn’t speed up the process. It just gives us an enriched pool to draw from, and that’s great, but that part of it isn’t any faster.”
Clark notes that modern genomic sequencing, AI and advanced technology make this an exciting era for fruit breeding. Currently, a doctoral student in the program is evaluating flavor and aroma by linking consumer sensory feedback directly to specific chemical compounds in the fruit. The ultimate goal is connecting those compounds back to the genetic markers responsible for producing them.
“Getting the data takes a long time, but our aim is that hopefully we can associate what the human is sensing, what the computer is detecting through this mass spectrometry, and then where in the genome so that we can maybe develop some DNA markers,” he says. “That’s a huge process. We’ve done this in grapevine on a different project that I worked on and so as a proof of concept, we know it works and we can get there but it’s a huge lift.”
Once this screening model is fully established, if nothing else, Clark says it could help cull the initial crosses down sooner, without waiting until the cross bears inedible fruit.
“If there’s a proportion of those that we can discard seedlings before they ever leave the greenhouse, that saves us on the back end,” he says.
Chasing Uncharted Flavors and Ultracrisp Crunch
Beyond flavor and texture, the program targets traits like apple scab resistance for sustainable growing, alongside novelty characteristics like red flesh and distinct aromatics.
Experimental crosses have produced off-putting flavors like mushroom, chlorine, sweet corn and cut grass, all automatic rejections. But other unexpected notes are highly desirable. Bedford notes that an ancestor in Honeycrisp’s lineage regularly introduces intense, surprising flavors like star anise, clove and cinnamon. One profile Clark and Bedford continue to chase is “cherry Life Saver.”
“We’ve gotten it a number of times, but never in quite the right package,” he says. “It comes in an apple that doesn’t have the crispness or doesn’t store. We know it’s there. We keep reaching for that prize.”
Bedford says cracking the code on complex flavor profiles is the program’s next true frontier.
“There are so many flavors and so many genes, that it’s a much harder target where Honeycrisp — and not that not that crisp is simple by any means — but it’s a single trait. It’s multiple genes, but it’s easier to focus on that than, let’s say, 10 different flavors. But we’re pursuing all the viable options.”
Another challenge to breeding for flavor is that it’s subjective. Some consumers want sweet, some consumers want tart and some consumers want a mix of both.
“We can’t control exactly the amount of acidity or sweetness in each one, but we have a whole range,” he says. “When we’re going down making our selections, we try to select something that from sweet to tart and in between. In the end, the customer will decide what they like.”
Even when achieving the high-sugar, high-acid ideal, Bedford finds that consumer interpretations vary wildly.
“If you have both of those in a high enough level, everybody tastes a little bit of what they want,” he says. “In SweeTango, there’s high sugar and high acid. I’ve had people say, ‘Oh, it’s so sweet and nice,’ and other people say, ‘Oh, it’s so nice and tangy.’ That’s what the high sugar and high acid gives you — lots of room to find yourself.”
While flavor remains the complex frontier, the team isn’t finished with texture. Bedford says the program is simultaneously taking crispness to its physical limit with what he colloquially calls “ultracrisp” crosses — selections engineered for even more crunch than Honeycrisp.
“We have been working on texture crispness in particular long enough that we’ve developed sort of a little a scale of crispness,” he says. “It and it ranges from what we call tender crisp, which is a lower pressure crisp, more like a watermelon almost, up to what we call firm crisp, which can be 22, 23 pounds of pressure. It’s not hard to get a 23-pounds-of-pressure apple somewhere, but try to get one that’s crisp and juicy.”
While early in evaluation, this next tier of crunch shows immense promise.
“Most things in breeding, we make incremental steps,” Bedford says. “We just take these baby steps and because we’re in it for the long haul. But every now and then, you get this exciting leap forward and they’re not predictable, to be honest. That’s what’s exciting. So, that’s what Honeycrisp was and we’re looking and hoping for the next big leap forward.”
That element of unpredictability remains the central excitement of breeding for consumers who may not even know what flavor or crunch is possible yet.
“We have these ideas, but genetics surprises us in a good way many times,” he says.
