The modern supermarket produce aisle is full of visual illusions. The strawberries are plump and glistening; the tomatoes smooth-skinned and lustrous; the melons firm and brightly colored—yet all too often devoid of flavor. We have no one to blame for these bland beauties but ourselves. By selectively breeding crops to be as prodigious as possible and to survive weeks of shipping and storage in dark, cool conditions, we have sapped flavor, aroma and nutrition from our food.

Consider the dilemma that cantaloupes presented to plant breeders. To enjoy a cantaloupe's full flavor, you must pick and eat it at peak ripeness, before it goes too soft. Toward the later stages of a cantaloupe's development, a burst of the hormone ethylene causes the fruit to ripen and soften quite quickly. This speedy puberty made transporting cantaloupes across states or from one country to another problematic: even on ice, the melons turned to mush by the voyage's end. So plant breeders decreased the levels of ethylene in cantaloupes intended for long-distance shipping by cross-pollinating only melons that naturally produced the lowest amounts of the hormone. Without a strong spurt of ethylene, the melons stay firm on the trip from field to produce aisle, but the chemical reactions that produce a ripe melon's aroma and taste never happen.

Breeders have had some success in overcoming this predicament. In the 1990s Dominique Chambeyron—a plant breeder employed by the Dutch De Ruiter Seeds Group—managed to create a variety of small, striped cantaloupe that retained both its firmness and flavor for weeks after it was harvested. Known as Melorange, this cultivar is grown in Central America and shipped to Sam's Club and other select chain stores in the U.S. between December and April, when it is too cold to grow cantaloupes locally. I tried a slice of one this past March: it had a dense texture and a splutter of flavor and aroma so robust, it bordered on spicy. Unfortunately, the kind of traditional breeding responsible for the Melorange relies an awful lot on serendipity. It can take more than a decade to perfect a consistently impressive new cultivar. Breeders must cross-pollinate plants over and over, hoping that some of the offspring will inherit the right characteristics. And they generally have to wait for the plants to grow and produce ripe fruit to find out. Much of what they reap is way off the mark and entirely unusable.

Genetics has recently offered an alternative path. At Monsanto, which acquired De Ruiter in 2008, Jeff Mills and his colleagues are able to predict the quality of a cantaloupe plant's eventual fruit before they ever put a single seed in the ground. First, Mills and his team pinpointed the melon genes underlying Melorange's unique combination of flavor and firmness. They can look for these genetic “markers” in cantaloupe seeds with help from a group of cooperative and largely autonomous robots.

I saw many of these machines in action when I visited Monsanto's molecular breeding laboratory at its vegetable research and development headquarters in Woodland, Calif., last November. A seed chipper shaves off a sliver of a seed for DNA analysis, leaving the rest of the kernel unharmed and suitable for sowing in a greenhouse or field. Another robot extracts the DNA from that tiny bit of seed and adds the necessary molecules and enzymes to chemically glue fluorescent tags to the relevant genes, if they are there. Yet another machine amplifies the number of these glowing tags to measure the light they emit and determine whether a gene is present.

Such techniques, known as marker-assisted breeding, are not brand-new, but they have enabled unprecedented feats of produce perfectionism in the past 10 years because genetic sequencing has become so much faster and cheaper. Monsanto's seed chippers can run 24 hours a day, and the entire system can deliver results to breeders within two weeks. In the past 10 years breeders at both private companies and universities have managed to create a cornucopia of more flavorful, colorful, shapely and nutritious fruits and vegetables, some of which are already available at grocery stores and farmers' markets. In addition to increasingly appetizing melons, the bounty includes broccoli that brims with even more nutrients than usual, truly succulent strawberries, and tomatoes that please both the eye and tongue.

“The impact of genomics on plant breeding is almost beyond my comprehension,” says Shelley Jansky, a potato breeder who works for both the U.S. Department of Agriculture and the University of Wisconsin–Madison. “I had a grad student here five years ago who spent three years trying to identify DNA sequences associated with disease resistance. After hundreds of hours in the lab, he ended up with 18 genetic markers. Now I have grad students who can get 8,000 markers for each of 200 individual plants within a matter of weeks.”

All this talk of DNA analysis sounds suspiciously like genetic engineering—the gene-editing technique that creates genetically modified organisms (GMOs). But it is not. It is a completely non-GMO technology. In fact, that is one of the main reasons it is so attractive to researchers and seed companies such as Monsanto.

Sowing Change
People have been changing plants to suit their purposes for at least 9,000 years. Just about every fruit and vegetable we eat is a domesticated species that we have transformed through decades of artificial selection and breeding: saving seeds only from plants with the most desirable characteristics and deliberately mating one plant with another to create novel combinations of traits. In this way, our ancestors turned a scrawny grass named teosinte into tall, plump-eared corn and molded a single species of wild cabbage into broccoli, Brussels sprouts, cauliflower and kale.

By the 1980s scientists had devised a much more exacting way of changing a plant's DNA: genetic engineering, which involves adding, removing or otherwise directly altering genes in a plant using lab tools. GMOs first appeared on the market in the U.S. in the 1990s. Although more than 70 percent of processed foods in the U.S. contain ingredients made from GMO corn, soybeans and canola, very few of the fresh vegetables and fruits sold in supermarkets have been genetically engineered. Exceptions include virus-resistant papaya, plum and squash, as well as pest-resistant sweet corn.

One reason that so few fresh fruits and veggies are GMOs is that, on the whole, they are far less profitable and less widely grown than the country's biggest crops: corn, soybeans, hay, wheat, cotton, sorghum and rice. When it comes to fruits, veggies and other so-called specialty crops, seed companies are not as motivated to deal with the burdensome and costly safety tests and federal regulatory procedures required to approve a GMO for sale.

The other big hurdle to GMO fruits and vegetables is public opposition. Universities and seed companies know that introducing new GMOs to the produce aisle today could ignite a furor among the segment of the U.S. population opposed to what they believe are “Frankenfoods.” Most shoppers remain oblivious to the few genetically engineered fruits and veggies already in stores because they are usually not labeled as such.

In the past decade marker-assisted breeding has become an increasingly viable way to improve fruits and vegetables while circumventing this controversy, especially as genetic technologies have improved and scientists have continued to sequence the genomes of more and more crop plants. In particular, the marriage of traditional breeding and DNA analysis is helping breeders turn their attention toward qualities of food that are important to consumers. “Asking what the consumer wants sounds really obvious, but it's not,” says Harry Klee, a tomato breeder at the University of Florida. Instead, he says, you almost always see breeders prioritizing the needs of farmers or food distributors.

A perfect example is the classic supermarket tomato. For decades experts have regarded the balance of acids and sugars in a tomato as the primary factor that determines whether we enjoy its taste. In general, people like tomatoes on the sweet side. But most breeders have not been chiefly concerned with flavor. With large-scale commercial growers in mind, breeders have instead favored tomato plants producing lots of smooth, hardy fruit that remain plump on the often long journey to the grocery. The more tomatoes a plant makes, however, the fewer sugars it can give to each one. Typical supermarket tomatoes may look pretty, but they do not have enough sugar to satisfy our taste buds.

Klee is determined to rescue the industrial tomato from its current gustatory doldrums. Through a series of large taste tests, he has evaluated nearly 200 varieties of heirloom tomatoes—older cultivars preserved by small groups of farmers and gardeners and sold at some grocery stores and farmers' markets. Heirlooms are known for their vibrant colors and fantastic flavor, but their skin easily cracks and scars, they go soft quickly, and they come from plants that do not make enough fruit to meet the demands of large commercial farmers.

In his research, Klee has learned that many heirlooms are tastier than standard tomatoes not because they have more sugar but because they are chock-full of a much more complex component of flavor: pungent chemicals known as volatile organic compounds that waft off plants and into our nostrils (think: freshly cut grass or the alluring smell of citrus). In a 2012 study Klee and his colleagues discovered that people actually enjoy a tomato with moderate levels of sugar if it contains enough of an aromatic compound named geranial. Klee suspects that geranial and other volatiles not only give a tomato its scent but also magnify the fruit's innate sweetness. In follow-up studies, he created tomatoes that lacked geranial and other fragrant molecules. People did not like them. If a tomato had average to high sugar levels but no volatiles, volunteers did not perceive it as sweet.

Lately Klee has been trying to make hybrid plants that give growers and consumers the best of both tomato worlds, old and new. In the past three years he and his colleagues have mated the most delicious heirlooms they could find with modern conventional tomatoes to create crossbreeds that yield well, are firm and smooth-skinned, and taste great. Klee routinely stocks up on cheap electric toothbrushes, which he and his team use to gently but thoroughly rattle tomato flowers, gathering the pollen that falls off in test tubes so they can play matchmaker. All the while, the breeders have been using hole punches to collect bits of leaves and analyze the plants' DNA, looking for genetic patterns that correspond to high levels of volatiles, for instance, or flawless skin. “Genetic analysis has definitely informed crossing decisions,” Klee says. “Our work has really accelerated in the last couple of years, with the emergence of the tomato genome sequence.”

The University of Florida recently released two of these hybrids—Garden Gem and Garden Treasure—that it would like to license to a seed company for mass distribution. Although the hybrids do not yield quite as much as commercial tomatoes, they produce more than three times the number of fruit as their heirloom parents, they have tremendous flavor and they can survive a good deal of shipping. Klee's colleague Vance Whitaker is making good progress on a related project to enhance the pungency of supermarket strawberries, which have also gained size and durability at the expense of taste.

Another notable victim of distribution difficulties, along with cantaloupes and tomatoes, is broccoli. About 75 percent of broccoli harvested in the U.S. is grown in California. Broccoli adores cool weather and flourishes in the Salinas Valley's occasional fog blankets. When forced to endure hot, sticky summers in the Northeast, the vegetable produces gnarly heads with buds of mismatched sizes. Each of the small buds that together make up broccoli's treetop dome is a flower that has not yet blossomed. Thomas Björkman of Cornell University and his colleagues recently discovered that during a critical period of its development, broccoli tracks how many hours of cool temperatures it enjoys and produces a uniform flowering head only if the tally is high enough. That is why broccoli grown on the East Coast might end up with an unattractive mix of nice, plump flower buds and dinky, almost imperceptible ones.

Three and a half years ago Björkman, Mark Farnham of the usda and their many collaborators decided to breed a new kind of broccoli that would thrive in the eastern part of the country. In their lab's growth chamber Björkman and his team have been subjecting broccoli to East Coast levels of heat and humidity, keeping seeds only from the plants that grow the most attractive flowering heads under these conditions. Although they have a lot of work ahead of them, they have already bred broccoli that can deal with a few more weeks of summer heat than the cultivars currently grown in the East. Meanwhile the researchers are searching the genomes of the various plants they grow, looking for genes that explain why some broccoli fares better than others. Finding them could shave years off the journey toward their ideal plant.

Breeding broccoli that stays beautiful in the heat is not just an exercise in aesthetics—it is also about getting tastier and more nutritious broccoli to farmers' markets and grocery stores. Fresh broccoli consumed the same day it has been harvested is different from typical supermarket fare, Björkman says—it is tender, with a mellow vegetative flavor, a hint of honeysuckle and no sharp aftertaste. Trucking broccoli from California to other parts of the country requires storing the vegetable on ice in the dark for days. With no light, photosynthesis halts, which means that cells stop making sugars. Rapidly dropping temperatures rupture cell walls, irrevocably weakening the plant's structure and diminishing its firmness. When the broccoli is thawed, various enzymes and molecules that escaped their cells bump into one another and trigger a sequence of chemical reactions, some of which degrade both nutritional and flavorful compounds. Giving farmers in the East broccoli they can grow and sell locally solves all these problems.

In a separate effort to boost the nutritional value of broccoli, Richard Mithen of the Institute of Food Research in England and his colleagues used marker-assisted breeding to raise levels of glucoraphanin, a compound that some evidence indicates may help fight bacteria and cancer. They have since licensed the resulting broccoli, called Beneforté, to Monsanto; you can find the florets at some Whole Foods Market and Stater Bros. stores.

Seeding Innovation
To get the grant that kicked off their initiative, Björkman and Farnham had to assure the usda that seed companies were genuinely interested in this potential new regional market for broccoli by securing funding from the private sector. Monsanto, Syngenta and Bejo Seeds are contributing even though they are competitors. In theory, both seed companies and university researchers can benefit from such collaborations. During the research and development phase, they all share information and exchange seeds. Eventually, however, it is time for negotiations. As is the case for Klee and his tasty tomatoes, Björkman hopes that once he and his colleagues get closer to their beauteous broccoli, a private company will license those seeds and produce them on a massive scale for commercial growers. Björkman and his team do not have the capital to do so themselves. Looking for genetic markers in individual plants may be getting cheaper all the time, but generating enormous quantities of seed and marketing it to farmers is still costly.

Some plant breeders worry that because giant seed companies have far more financial and technological resources than smaller firms and universities, true innovation will wither. “There's been a rather large decline in public sector breeding programs as the technology has transferred into the private sector,” says Irwin Goldman of the University of Wisconsin–Madison, who recently debuted flame-orange table beets with concentric gold stripes. “Some people argue that this transfer is a success for this country, but public breeding will do things that the private sector won't do—things that take too long or are too high risk.”

Jack Juvik, who directs the plant-breeding center at the University of Illinois at Urbana-Champaign and first got into breeding in the 1970s, remembers when big companies were not nearly as dominant as they are today. “When I first started, there were a lot of smaller companies selling a lot of seeds, but they have all been basically bought out or driven out of the market by the mega companies. That has changed the whole texture of the industry,” Juvik says. “Instead of having people at public institutions developing finished varieties, most of us design germplasm [seeds and cultivars] for big companies to work with. These large companies have the resources to do some good testing and make some really good varieties, but they end up controlling most of the germplasm and technology used to make it.”

Goldman and his University of Wisconsin–Madison colleague Jack Kloppenburg belong to a group of 20 breeders and farmers from around the country who are interested in creating the equivalent of open-source software for seeds—nonpatentable varieties that anyone can use. There is no precedent for such an arrangement in the 21st-century commercial seedscape, though. One potentially expensive option is for plant breeders to hire lawyers and obtain standard patents or copyrights on their seeds with the intent of letting just about anyone use them (excluding mammoth private companies, of course). Alternatively, they could try to create a kind of open-source license that allows people to use seeds only if they, too, agree to freely share them and anything they make with them. Goldman has also proposed a compromise in which breeders license some seeds to the private sector to make a profit but give others away.

Klee wonders whether a certain degree of conciliation is the best way forward. “The reality is that we in academia cannot compete with the Monsantos or other big seed companies,” he says. “Breeders at universities are pushed out of big crops and into niche crops. In my department, we have a peach breeder, a blueberry breeder and a strawberry breeder. I know a lot of people at Monsanto who have dropped these kinds of crops that are marginal for them.” Such dichotomy, he hopes, can be complementary, with the public and private sector relying on each other for distinct specialties.

Ultimately what Klee cares most about is the same prospect tantalizing more and more plant breeders: bridging the gulf between what growers need to make a living and what consumers want on their plates. “Marker-assisted breeding makes it possible to go back and fix things like flavor and texture,” Klee says. “In the end, it's really very simple: let's give people stuff they like.”