In July 2013, Adam Austin placed a bet on a red-and-white heifer.

Austin and his father run Lincoln Hill Farm, a dairy operation on the border of New York and Vermont. Lincoln Hill earns most if its revenue by selling its milk at Stewart’s Shops, a locally ubiquitous convenience-store chain, but milk isn’t the Austins’ only cash cow.

Adam Austin leads Delicacy, a heifer that he bought for $10,000 because of the animal’s promising genetic profile, at Lincoln Hill Farm in Eagle Bridge, N.Y. (The white mass behind Austin is an enormous bag of cattle feed.) (Luke Yoquinto)

Cheap, powerful genomic analysis technologies have revolutionized the way that dairy cattle are bred. With a little genomic know-how and a lot of luck, it’s possible for even small-scale farmers to produce some of the world’s most robust, fertile and productive dairy cattle — and then sell them at a profit. And that’s what Austin set out to do.

Austin’s plans centered on that red-and-white heifer, named Delicacy, whom he first saw at an auction not far from his farm. He had gone to the auction to sell a cow, not to buy one, but Delicacy immediately caught the eye of his wife, Meggi, and Austin ended up purchasing the 15-month-old heifer for $10,000, one of the higher prices fetched that day.

Delicacy’s promise lay in her genes. A handful of companies do for cattle what the genetic testing company 23andMe became famous for offering humans: the identification of tiny portions of DNA that correspond to important traits.

The data set connecting the genomes and physical attributes of dairy cattle is enormous, says Alison Van Eenennaam, a cattle genomicist at University of California at Davis.

Austin pats Delicacy. She’s milked in this stall at Lincoln Hill Farm in Eagle Bridge, NY. (Luke Yoquinto)

As a result, within a small margin of error, breeders can tell from a DNA sample taken from a young cow — or even an embryo — what it will be like as an adult. This portrait includes whether it will grow horns, the fat and protein composition of its milk, and the size and shape of its udder — an important consideration, because standardized milking machines require standard-size teats.

Once these and other values are known, it’s possible to assign an animal an overall score. The main scoring index for Holstein cattle, the breed that accounts for 90 percent of milk cows in the United States, is the Genomic Total Performance Index, or GTPI.

Austin knew that Delicacy had a fairly high GTPI number, and he hoped that with careful breeding she would give rise to an even higher-scoring calf. But her overall score belied a troubling weakness: Though she stood out in some characteristics, such as those for overall health, fertility and production longevity, her projected annual milk production was below average.

So Austin went shopping for bull semen with something specific in mind.

With deep-enough pockets, it’s possible to order semen from the best bull in the world. Currently that title is held by a black-and-white bull named Jedi, said Charles Will, the Holstein sire program manager at Select Sires, the largest bull semen provider in North America.

The cost of high-grade bull semen is usually capped at $100 per unit, but, as with scalped concert tickets, prices can rise dramatically on the secondary market — up to $1,000, in this case.

On a tight budget, Austin ordered the best he could justify buying: 10 units at $25 apiece. He prioritized the genetic characteristics that, he hoped, would complement Delicacy’s. “I wanted to increase milk [production]” in her offspring, he said.

Adam, his wife Meggi, and their sons Carter and Ashton with Delicacy. (Luke Yoquinto)

Cows at Lincoln Hill Farm in Eagle Bridge, NY. (Luke Yoquinto)

But in every mating of two animals, even where the full genome of both parents is known, there is always an element of chance, due to what Van Eenennaam calls the “roulette wheel of meiosis” — meiosis being the type of cell division involved in sexual reproduction. For the same reason that human siblings tend to look different from each other, calf siblings are almost always born with different GTPI scores.

Austin undertook the mating the old-fashioned way: He injected the mail-order bull semen into Delicacy himself; seven days later, he had a veterinarian transfer the three resulting embryos from Delicacy into three different surrogates, a common process used to maximize the number of offspring a mother can have. Only one embryo survived the procedure.

Larger-scale breeding operations are more likely to use in vitro techniques: inseminating many egg cells from a given cow in a Petri dish, creating dozens of embryos instead of a handful. Some breeders then test the DNA of each embryo, bringing only the most desirable ones to term. The effect is rapid genetic changes in cattle populations that once would have taken decades.

“There are all sorts of things like that that people are working on — sort of newer, novel traits that we haven’t been able to do in the past,” says Kent Weigel, chair of dairy science at the University of Wisconsin at Madison. “The big one that we’re looking at is feed efficiency,” the amount of milk a cow produces per pound of food consumed. Others include decreased methane production — cow burps are a significant source of this greenhouse gas — and increased resistance to postpartum health disorders.

Between Delicacy’s good traits and the DNA in the bull semen, Austin expected decent GTPI results for her calf, which he decided to name Dreamer. When Dreamer was born in June 2015, Austin sent a hair sample from her tail to a cattle genomics firm. The results came back via e-mail.

(Luke Yoquinto)

Compared with other red-and-white hornless Holsteins, Dreamer had the highest GTPI in the world. Her genes foretold a long, healthy life of impressive milk production.

“For whatever reason, that mating worked out just right,” he said. Dreamer’s score of 180 points was significantly higher than the average of her parents. “She did that on her own.”

This fall, hoping that this genetic triumph would pay off, Austin and his wife traveled to Madison to auction off Dreamer at the World Dairy Expo, the largest annual gathering of dairy farmers in North America. (This year’s theme: “Dairy in Our DNA.”) Because Dreamer was the top animal in her category, the Austins were told to expect a windfall far beyond the value of an everyday high-production milk cow.

Compared with an average cow, Adam Austin said, “a high-producing cow is going to make you five or six hundred bucks more per lactation.” (That is a cow’s annual period for milk production, roughly 305 days.) A young, otherwise unremarkable high-production cow might sell for $2,500. The Austins were told to expect at least 10 times that much for Dreamer.

She was worth so much, in theory, because in addition to being strong, fertile and a rapid producer of milk, she was the top animal in her niche category of red, hornless cattle — which actually means very little in terms of utility on a farm. Hornlessness saves a farmer only the moderately annoying task of horn removal, and the difference between red and black cows is “like a yellow Lab versus a black Lab,” Weigel said.

Nevertheless, owning the top animal in any category can bring a degree of cachet to a farm. More important, the possibility of using that cow to breed and sell the next generation of top bovines adds an element of speculative fervor to the market. Consequently, buyers often pay top dollar for cows and bulls with notable genes. The University of Missouri College of Agriculture discovered this year that a cow in its campus herd was the country’s second-highest-scoring individual in its category: carriers of the gene for red-and-white hair. It sold for $25,000.

The World Dairy Expo’s auctions took place in a 10,000-seat coliseum, filled with bidders and their proxies from around the world. Attendants led cattle around on a raised enclosure, announcers read out the genomic strengths and pedigree of each animal, and a fast-talking auctioneer coaxed bids from the audience. To the Austins’ dismay, the bidding for Dreamer stopped soon after it started, stalling at $8,700. It was a high price for a prepubescent calf, but still less than expected. After subtracting transportation costs and the auctioneer’s commission, Austin said he probably made $5,000.

What the caprices of genetics had provided, those of economics had taken away. It was hard to pinpoint exactly why. Hornlessness “was really hot a couple years ago,” but perhaps demand for it had waned, the owner of the auction company said. And although Dreamer’s milk production metrics were above average, the loftiness of her overall GTPI scores was mainly due to other traits, such as overall robustness and fertility.

The Austins weren’t entirely disappointed. They had made money and they still own Delicacy. Austin intends to go through the same process to breed her again. And in nine months, he said, “who knows — maybe I’ll make another number-one calf.”

Yoquinto is a research associate at the MIT AgeLab and a freelance science writer.