There is a lot of talk today about the pros and cons of artificial intelligence in agriculture and the potential impact of AI on dairy farming and cattle reproduction.
Will AI help dairy farming and reproduction? Or is it just a bunch of bull (hype)? AI is an acronym for artificial intelligence or the capability of computer systems to perform tasks typically associated with human-intelligence-like learning, problem-solving and decision-making.
Dairy farmers use the term AI differently. AI in dairy farming is the acronym for artificial insemination, where dairy cows are bred with bulls without having bulls on the farm. This article will briefly review artificial insemination history and evaluate the potential impact artificial intelligence will have on dairy farming and cattle reproduction in the future.
Artificial insemination
Artificial insemination for dairy cattle became commercially available in Ohio when the Northeast Ohio Trumbull-Ashtabula Breeders Association was incorporated in 1941 with 260 farms enrolled. The first reported AI calf from the association’s six bulls was born Dec. 25, 1941. This original association later became the Northeast Ohio Breeders Association, with a consolidation in the mid-40s to the Central Ohio Breeders Association and in 1965 with Select Sires of Plain City, Ohio.
Prior to artificial insemination, each dairy farm had a bull on their farm to breed the cows. As dairy reproduction developed, bulls have become less prevalent on farms, which has improved farm safety. In the early days of artificial insemination, my father-in-law’s post-milking routine required him to stand by the pasture gate for 30 minutes each day and observe cows for standing estrus. After observing the cows, he would phone in to the breeder association an order for bull semen that had previously been collected and stored frozen in glass ampules. This bull semen had a short shelf life, and the ampules would be boxed up with ice and air-dropped by a small parachute from a small single-engine plane into the pasture field — an early form of Amazon delivery trucks! Upon receiving the glass ampule, he would siphon the semen into a breeding sheath and inseminate the cows observed in standing estrus earlier that morning. Each day, they’d repeat the process.
Today, artificial insemination bull semen is purchased from dairy bull associations and stored in small plastic-like straws that are stored for months in on-farm liquid nitrogen tanks. No more last-minute phone calls or airplane deliveries. The farmer or the trained AI technician daily receives on their cell phone a list of dairy cows in estrus. This AI list is generated from the on-farm computer from individually cow-mounted activity monitors. Each cow’s activity is correlated to cow estrus activity. The AI technician uses a pre-selected bull (selective mating) to inseminate the estrus cow. Each day, repeat the process.
The selective mating process matches the AI bull to the specific cow. The selective mating is done with the experience of a trained AI bull association expert, the gut instinct of the farmer with an experienced eye, and specialized DNA genomic traits from each cow and bull. It’s a process that requires coordination of information. The overall goal is to enhance the genotype of the next-generation milking cow.
Artificial intelligence
How will artificial intelligence impact future dairy farms? Here’s a brief background on artificial intelligence. Artificial intelligence started in 1956 as an academic computer-based discipline. Computer speed, the internet, communication with text and on-farm data have expanded rapidly in recent years. Dairy farms today have on-farm computers that collect individual cow hourly steps, ruminations, lying time, rumen pH and body temperature from cow wearable devices, like a Fitbit or Apple Watch for cows. Milk production, SCC and milk composition are also measured at each milking. In former days, the milk house breeding wheel, along with the DHIA once a month, gave the farmer what cows to breed, dry off and calve. A much slower pace was evident. Today, how does a farmer evaluate the massive amount of cow data to make important decisions that affect the profitability of the farm? Here is “artificial intelligence.”
Don’t think of AI as “man against machine” but rather “man plus machine.” Artificial intelligence does not replace the AI breeder nor the final decision of the farmer. The complexity of selecting the best bull for each cow in the herd has become more complex with a multitude of 20 or more traits to balance for each mating.
To effectively use the large amount of on-farm cow data along with advanced genomic information, the complexity of a selective mating decision becomes more complicated.
Then, to increase the complexity, what should the future dairy cows be bred for? In recent years, milk fat percentage has increased dramatically, but milk protein demand is driving the need to breed and feed for higher milk protein. So, what should the future cow be bred to produce? It is like putting together a 1000-piece puzzle, and the picture on the box keeps changing. Artificial intelligence can be used to assess this future on a herd basis and provide relevant breeding options and risk assessment for the farmer. The farmer is still in control.
A key fundamental farm management statement is, “If you can’t measure it, you cannot manage it.” This applies to future dairy farms that will collect data for improvement in dairy cow feed efficiency. Feed efficiency has a moderate heritability, and bull associations have focused on feed efficiency. Feed efficiency has high potential to improve farm profitability.
Systems are available today (www.vytello.com) to measure feed intake of each cow to provide additional information on refining feed efficiency on each farm. As more data is collected, artificial intelligence programs will be used to improve performance, breeding and reproductive selections on farms. For further information, listen to Dairytalk podcast “AI in dairy farming: from hype to practical support.”
Summary
Artificial intelligence is a tool to organize dairy farm data and dairy cattle breeding information that the human brain may miss. As advances in dairy reproduction have developed over the past decades from on-farm bull breeding to artificial insemination to cattle genomics, the complexity of the selective mating process will continue to increase. Artificial intelligence will not replace the breeder or the dairy farmer. But AI will provide information to the dairy farmer to make better-informed decisions about the breeding and reproduction of their farm’s dairy cattle. No bull about it, AI is here to stay and help.
