Doing What Comes Artificially

ALMOST 60 PERCENT OF ALL CALVES BORN ON AMERICAN dairy farms are offspring of parents that have never met. Artificial insemination (AI), once something that would “never work,” is close to being standard practice, selectively or across herds, wherever cows are milked. AI is the biggest reason why today’s dairy farms are producing at a rate that was unthinkable 50 years ago. Acceptance by owners of beef cows, while not yet as impressive, is slowly on the upswing as well.

AI makes it possible for a genetically superior bull to sire 10,000 to 20,000 calves and often more—every year. In natural service, by contrast, 100 calves by the same bull would reflect a year of very good fortune. The “often more” record is held by a Holstein bull in the Netherlands. Over his 12-year lifetime he serviced more than two million cows. At his peak he had a calf being born somewhere in the world every two minutes. That’s a record, but bulls breeding upward of 30,000 cows a year are quite common.

Since most young bulls now being proven in AI are the results of mating one popular bull with a daughter of another, inbreeding is a problem in the making. It has yet to become serious, but it is definitely a matter of concern. To avoid it, herd owners keep track of their cattle’s ancestors, and geneticists calculate degrees of inbreeding in advance of mating. At the same time, breeders are searching for the increasingly rare “outcross” bulls that have little genetic contribution from the most popular sires. So far, at least, increases in milk per cow resulting from AI have greatly outweighed losses in production and calf viability traceable to inbreeding.

In theory, one mature bull could breed every cow in the world in a single day. It all comes down to simple arithmetic. There can be 10 or 12 billion live sperm in a single collection. Pregnancy requires just one.

Things aren’t that finely tuned in practice. Allowance must be made for the millions of sperm that perish in the cow’s reproductive tract in AI and natural breeding alike. Even so, bulls that are harvested every other day can produce enough sperm to AI 1,200 to 1,800 cows a week. And the dairy industry needs to keep impregnating its she-stock, because a cow’s milk output drops to unprofitable levels if she doesn’t freshen—have a calf—and start milking with renewed vigor every year or so.

Obviously, all this begins with collecting semen from a bull. An artificial vagina with warm water between its double walls is used, along with a “mount” or “jump” cow or steer to provide inspiration. Freshly collected semen is sped to a laboratory, where live sperm cells are counted through a microscope or with a spectrophotometer.

The key to the process is the extender—egg yolk or milk plus antibiotics—which dilutes the semen in an amount determined by the sperm count. This step is what allows a genetically superior bull to produce hundreds or thousands of times the offspring he could have done by natural means. Extended semen is packaged in plastic straws containing half a cubic centimeter, enough to breed one cow. Strawed semen is then frozen under rigorously controlled conditions.

After freezing, semen is shipped in small liquid-nitrogen refrigerators to customers or distributors. Most AI organizations have fleets of delivery trucks, but air shipments are becoming more frequent. Twenty-five years ago most AI distributors also provided technician service. Nowadays the so-called direct herd approach, with herd owners breeding their own beef and dairy cows, is becoming more popular.

There are those who will tell you that all this began with quality-conscious horse thieves. Folklore traces AI back to long-ago Arabs “borrowing” the services of rich sheiks’ heavily guarded stallions under cover of desert night. That was in the fourteenth century. In Massachusetts we didn’t get around to applying the idea to dairy cattle until 1946. Thanks to World War II we were years behind several other states, including New Jersey, where this country’s first organized AI program had been launched in 1938.

AI was already long-established practice in much of Europe by that time. The ever-hungry Soviets, for example, set up an extensive program in 1928, and by 1938 they were AIing well over a million cows a year. And the trailblazing effort in New Jersey grew out of visits to Denmark by Enos J. Perry, an extension dairyman at Rutgers University. Virtually all Danish cows were being AIed even then.

Back in Massachusetts, helping a new AI cooperative get off the ground was among my principal duties as a young, freshfrom-the-Army county agricultural agent. Artificial breeding was the term we used; insemination was years away from being a mixed-company word in New England, even among farmers.

The co-op was a branch of a statewide organization. Walter Hurlburt, a Holstein breeder, probably knew more about AI than anyone else in the county, me included. Years earlier he had mated his best cow to a bull in New Jersey through what amounted to a typically pioneering effort.

Cow and bull were 150 miles apart. AI was in its preorganized infancy. Interstate highways had yet to happen. So Walter made seven round trips in his family car at illegal speeds, pausing just long enough in New Jersey each time to pick up a test tube of liquid semen.

Liquid semen doesn’t stay usable long, and cows stay in estrus (or heat) for just a few hours. All this contributed to a number of unsuccessful attempts before it finally took. But Walter hoped the mating would produce a good bull, and it did. Hurlwood Conqueror, the end result of all that frantic driving, sired some of the best cows Walter ever owned. Then, fittingly, he was sold to a bull stud, as AI organizations came to be known.

Conqueror was a “plus-proven” sire by that time. That meant that his daughters had outmilked their mothers (or dams). Methods for rating sires have been overhauled and improved many times since, but in the years surrounding World War II, bulls with “big plus” daughter-dam comparisons were considered top of the line.

Dr. C. Y. Cannon had sold me on this concept when I was a student at Iowa State. Why didn’t everybody use proven bulls? I was certainly going to when I got some cows. “You’ll have to wait in line,” Cannon told me. “There aren’t enough proven bulls to go around. Only one for every four hundred or so dairy farms, as closely as I can figure.” Dr. Cannon was a friend of the aforementioned E. J. Perry, then knee-deep in starting the first AI co-op in New Jersey. Perhaps that’s why he went on to say, “But maybe that’ll change if EJ. can make this artificial thing work back East.”

No maybes about it, “this artificial thing” worked and eventually put proven bulls within easy reach of all who wanted to use them. Stanley Gaunt, the Perry protégé who spearheaded AI in Massachusetts, saw daughters of such bulls as “genetically capable of taking maximum advantage of good feed and management practices.” In other words, AI by itself would not boost milk production greatly, but in combination with other improvements in dairy herd management, the results could be phenomenal.

In the 1940s Gaunt and other visionaries foresaw cows producing 9,000 pounds of milk a year each (a gallon of milk weighs 8.62 pounds). You’re dreaming, many farmers thought. Nine thousand pounds was almost twice the national average. And that average had increased by only 165 pounds from 1940 to 1945, despite heavy wartime emphasis on producing more milk per cow. Even the cows in the USDA’s Dairy Herd Improvement test—then, as now, looked upon as the best-fed and best-managed segment of the dairy cattle population—averaged just under 8,600 pounds of milk in 1945.

The need for genetic improvement was obvious, and AI offered an unprecedented opportunity to do something about it. Our co-op took on a slogan: “Better Bulls Beget Better Bossies” ( bossy means cow). That line would prove to be as prophetic as it was alliterative, and not just in Massachusetts. The average U.S. dairy cow’s output would jump from 4,622 pounds of milk in 1940, the last year without sizable numbers of AI daughters involved, to 16,915 pounds in 1997.

AI was a major reason for the improvement. But it didn’t happen easily. The application of science to agriculture in general has been described by scholars as a “resisted revolution,” and AI was certainly something to resist in this most traditional of pursuits. Many of the dairy farmers who pioneered it, for example, still used four-legged horsepower for at least part of their plowing, haying, manure spreading, and other cropping operations. Many intelligent, reasonably well-educated people, like the clergyman who advised one of our technicians to “get a job you won’t be ashamed of,” saw AI as something akin to black magic, even sacrilege.

AI amounted to a radical change. And in the circles in which I traveled, a change had ways of generating bucolic humor. Take the poultryman who called to have his family cow bred on the co-op’s first day of operation. Our secretary, following instructions, advised him of the technician’s estimated time of arrival and said, “If you’re not going to be there, tie a string around the cow’s tail so he’ll know which one to breed.”

“If he’s that dumb, don’t bother to send him!” said the poultryman. “We’re talking about the only damn cow on the place.”

Ward McCarthy, our most conscientious technician, once bred a Guernsey cow that did have a string around her tail. The whole herd was stanchioned in the barn. Their owner and his wife had gone Christmas shopping. On his way out Ward met a man leading another cow to her annual “date” with the dairyman’s Guernsey bull. Ward explained that the dairyman no longer had a bull. He was breeding his herd artificially instead.

The stranger, a neighbor with a family cow, asked how AI worked. “Tie her to that fence post and give me six bucks and I’ll show you,” said Ward. And that’s just what happened. From a technician’s point of view, a cow was a cow.

Another early convert had been milking 40-odd cows for years but had never learned to detect estrus. There was no need to in pre-AI days. Like his father before him, he had operated in what he liked to call “Junior Prom style,” simply pasturing a young bull with his cows. Eager for the superior genetics AI had to offer, he adapted his system to fit. He kept the bull in the barn and led him past the cows when they were tied there side by side for milking twice a day. If the bull showed interest, his owner put him back in his pen and called our technician. When the bull eventually grew too strong to be led away, he was replaced with a less libidinous, easier-to-handle animal.

In the early days many farmers relied on such ad hoc methods because there was no backlog of experience and research to draw on. The word extend , for instance, had yet to be applied to semen. Dilute understandably worried the uninitiated. We made sure to point out that it took only one sperm to get a cow in calf. “Maybe,” snorted an old-timer whose dairyfarming success was reflected in his shiny new Cadillac, “but that calf’ll be too puny to raise. All that sperm you say goes to waste when my bull over there breeds a cow the old-fashioned way is a tonic for the unborn calf. I’m surprised you didn’t learn that in college. It’s common knowledge.”

Bill Robinson, another dairyman, wasn’t interested in folklore, but low conception rates in his herd had him worried. Convinced that diluted semen was the cause, he brought down the house at one of our meetings by countering our only-takes-one talk with “Okay! Send me that one sperm tomorrow morning!”

Bill and those who laughed with him weren’t being negative. Anything but. They did see AI as a way to use better bulls than they could own, and they knew that genetic improvement takes time. But at least in our case, AI simply wasn’t resulting in enough expectant cows.

A calf we had built a fair booth around went to another barn for a bucket of milk just before a cynic we knew happened along. He spotted the empty pen, read our TYPICAL ARTIFICIAL CALF sign aloud, and muttered, “About the way I had it figured!”

It really wasn’t funny, though. Owners of cows that should have been pregnant but weren’t saw smaller milk checks ahead. At the time, we were breeding cows with liquid semen collected from bulls 100 miles away. Semen, extended and packed in ice, was mailed special delivery to our two technicians six times a week. They picked it up the same day, then carried it on their farm-to-farm rounds in back-seat ice chests. Even if they did manage to keep the semen at the recommended 40 degrees Fahrenheit, its usefulness was very much in doubt after 24 hours.

Bob Boese, an AI pioneer with whom I later worked in Illinois, said this problem had plagued him over the nearly 10 years he had shipped liquid semen coast to coast from an outfit near Chicago. At one particularly desperate point he had hired a barnstorming aviator to expedite deliveries in Indiana. Technicians spread sheets in their yards as targets, and the flier swooped down and dropped his packages.

Each package was rigged with a mini-parachute that would, in theory, float it to earth unharmed. It might even land on the sheet. The problem was that it often didn’t come close. In one bad winter storm the plane wound up skip-bombing Pennsylvania with semen packages while technicians and cows waited less than patiently in Indiana. Bob Boese went looking for less spectacular methodology after that.

Back-to-basics thinking was resurging in Massachusetts too. Maybe our conception problems could be solved by providing dairy farmers with a deeper understanding of the reproductive process. So we held meetings at which 10 or 12 dairymen sat around a table while veterinarians dissected and discussed bovine reproductive tracts. Real ones.

The tracts were blood-dripping messes, and one night we held forth in a courtroom without a garbage pail. I tried dumping the organs in a snow-filled gully en route home. One of them wound up on the floor of my car’s cluttered trunk. As I picked it up, a car pulled up behind me. “Need some help, buddy?” the driver yelled.

Bathed in his headlights, I turned, entrails in hand, and started to say, “No thanks.” Before I could finish, the woman riding with him screamed. The car took off at high speed.

One of our technicians, Norman Everett, had the lowest conception rate in the state. We couldn’t figure out why. Chet Putney, who managed the statewide co-op, checked Norm’s technique step by step at least 10 times. So did Stan Gaunt. So did I. We came away baffled—until somebody gave Norm’s preschool daughter two goldfish.

Both fish died shortly after their first water change. Their new home, it turned out, had copper water pipes. Norm used tap water to sterilize the glass syringes and tubes that we used to AI cows in the days before plastic became cheap enough to replace it.

A USDA farm-pond specialist told us that just a few parts of copper per million parts of water would kill fish. Would it kill sperm too? Nobody knew, though we asked every expert we could find.

So we took no chances. Norm started borrowing water from a lead-piped neighbor. With the help of a papermill chemist we knew, he boiled colloidal copper off his glassware with acid. His conception rate jumped 20 percentage points.

Resistance to AI had slackened enormously by the time I moved on to a job with what is now the Holstein Association USA in 1955. Gretchen, a Holstein owned by Doug Phelps, was the cow that put AI over the top in our area. She milked close to 20,000 pounds in 305 days. This was big news in the 1950s, when the average cow was still in the 5,000-to-6,000-pound range. Gretchen was just the first in what turned out to be a long line of prodigious producers.

We didn’t have a monopoly on success stories. AI daughters were revolutionizing dairy farming coast to coast. More than half the calves registered by the Holstein Association were sons and daughters of AI bulls. Even more important, frozen semen was entering the picture. It could be stored almost forever in liquid nitrogen at 320 degrees below zero Fahrenheit. Freezing ended the problems of short shelf life and the follow-up practice of throwing away unused semen.

At Holstein my job touched but did not center on all this. AI seemed to be over the hump. If I didn’t figure on selling the idea again, it was because I didn’t foresee applying “Better Bulls Beget Better Bossies” to beef cattle. But as things turned out, beef cows were front-and-center potential when I moved on to Curtiss Breeding Service in 1963. A big reason why, said my new coworker Dale Maples, was that dairy-cow numbers were dropping “like anvils down a silo.” Americans had been milking close to 22 million cows when Curtiss broke into AI in 1949. The figure was down to 17 million in 1963, and no end was in sight. AI was mass-producing cows that made better use than ever before of good feeding and management. And feeding and management, like semen processing and sire selection, were constantly improving as well. From our point of view the AI industry was a victim of its own success.

But as meat consumption climbed, beef cattle were getting more and more numerous. AIing them seemed a heaven-sent opportunity to keep sales volume growing. It looked logical, even easy. But out where the cows were, we heard replays of “It ain’t natural” and “Calves’ll be puny,” plus new ways to say no, like “AI is for dairy cows.” Of course, there is no physiological reason why beef cows can’t be bred artificially. The problems involve fitting AI—and the estrus detection vital to its success—into beef-herd management.

It’s fairly easy to check dairy cows for signs of heat, since they get close enough to people to be milked two or three times a day. Beef cows on range, by contrast, often go unobserved for months. They also don’t stay in heat as long as dairy cows do, and they aren’t as demonstrative about it. We knew going in that natural breeding —turning bulls out with the cows- was the most convenient solution for herd owners.

So once again we used the idea of having better bulls than the farmer could afford to own as our sales pitch. This time around, “better” hinged on performance and progeny testing—impressive rates of weight gain by bulls and their offspring alike. There weren’t many truly superior bulls around.

Putting them within reach drew a lot of ranchers to AI. On the whole they were disenchanted with the slowgaining, show-type bulls that the breeders of all three beef brands then numerically important—Angus, Hereford, and Shorthorn—had to offer. Ranchers were our main target because in the farm states where we had technicians, beef-cow herds were mostly sidelines. Their owners were too busy farming corn, cotton, and such to check cows for heat, so they stuck to natural breeding. But beef cows took top priority on Western ranches. Once AI was adopted, economics demanded tailoring herd management to fit.

The first step is to build a corral and a chute to hold skittish range cows during the AI process. Since 320 or more acres of range can be required to support one cow and her calf, it helps to concentrate the herd near these facilities during AI season. Back in the 1970s our man in the Dakotas, Stan Pozorski, and I visited a customer who had built his breeding chute and corral on lush river-bottom grazing land. His cows were kept there for two estrus cycles—42 days—before moving on to more extensive range. Since well-fed cows are thought to be most likely to conceive, their grass diet was supplemented with a small daily grain ration. The rancher and his cowhands rode through the herd twice a day, looking for signs of estrus.

“Hot cows,” as the cowboys called them, were corralled near the chute. Those spotted in heat in the morning were AIed that same afternoon. Those spotted in the afternoon were AIed the following morning. This timing tied into the ovulation cycle. After two AI go-rounds, the herd went out on range with a few “pick-up” bulls to breed the small number of cows not yet in calf. The procedure involved a lot of extra fuss and bother, but it paid off in extra profits.

The rancher’s 600 Angus cows and their calves all were the offspring of our progeny-proven bulls. The herd was several notches above average—so good, in fact, that the best of the bull calves hadn’t been castrated; they were going to be sold as breeding stock. Neighboring ranchers all but stood in line to buy sons of AI bulls because they didn’t want to “mess”—their word—with AI itself. Once again AI was working too well for its own good.

We finally found a way around the bull-meets-cow convenience factor with Parisien, the first Simmental bull to set hoof in North America in 75 years. A syndicate of Canadian ranchers brought him from France to Alberta in 1967. We had exclusive U.S. semen distribution rights. When Parisien calves from Hereford and Angus cows started arriving, one thing became clear: If messing with AI was the only way to get more like them, ranchers would mess with AI.

Parisien calves grew like inspired weeds, sometimes weighing 75 to 100 pounds more than their straight Hereford or Angus herdmates at weaning. The heifers among them became the milkiest mother cows ranchers had ever seen. Many of those calves were big enough to have birthing problems. But the cash value of halfSimmental heifers at weaning was more than twice that of their Angus or Hereford mothers. We were soon selling semen from a lot of Simmental bulls.

So were our AI competitors. And after three crosses to AI bulls, seven-eighths-blood Simmental bulls in natural service were competitors too. Bull-meets-cow convenience was back. But before that happened, other new-tous “exotic” European breeds stampeded out of the woodwork and into AI: Chianina, GeIbvieh, Limousin, MaineAnjou, Pinzgauer, Tarantaise, and more. Oldline Angus and Hereford breeders were more than a little concerned and resentful.

We had bulls of 30 beef breeds at Curtiss on the day a Hereford man, who looked ready to sock somebody, burst into our booth at the Denver stock show with “Why are you trying to louse things up with all these neurotic exotic cattle?”

His finger was pointing at me. So instead of backing away, I told him, truthfully, that my personal reasons were two kids I had to put through college. “Hey!” he said as he shook hands. “I never thought of that! Whataya say I buy you a drink?”

His kind offer was as close as I ever came to a happy ending with beef AI. It worked. It grew. But it has yet to take off and fly. Experience has convinced me that the convenience of natural breeding is the big roadblock. However, estrus-synchronization products now available could wipe out the need for labor-intensive heat detection and unlock the potential of AIing beef cows.

That potential is enormous, if dairy cows are anything to go by. There were 23,215,000 of them on U.S. farms when the first AI co-op began operations in New Jersey in 1938. They produced a total of 105,807 million pounds of milk that year. Parallel figures for 1997 are 9,258,000 cows producing 156,602 pounds of milk. That’s almost 50 percent more milk from 60 percent fewer cows. Average production per cow has more than tripled.

Artificial insemination certainly isn’t the only reason. Modern cows lap up more grain than most of their grandmothers ever saw. Quality, as in milk-making value, of pasture, silage, and hay has improved enormously. So has knowledge of ruminant nutrition and the aggressive use of growth hormones. Average herd size has almost quadrupled since 1970. All manner of better management practices have resulted. In one of many examples, milk per cow jumped six pounds a day across one Wisconsin herd after sand replaced straw as bedding.

But it takes nothing away from any of that to add that 1999-style feeding and management would have made 1938-model dairy cows fatter, not milkier. To adapt a line from Stan Gaunt, they weren’t genetically capable of putting it to maximum use.

As a footnote, I’m still in Christmascard touch with a few of those who, back in the 1940s, said it would never work. Now they’re blaming daughters of AI bulls for flooding the milk markets with surplus. I’m willing to bet that I’ll be getting that kind of mail from more than a few beef cowmen someday.