It was time for a change. After the passing of my grandfather in the fall of 2013, it became clear to my uncle, Marty Costello, the 70-cow dairy needed a major update from the facilities that had been built before he was born – or he would have to stop dairying altogether.

Not willing to give up on the dairy lifestyle, a partnership between himself and several of his siblings, including my father, was formed to facilitate the update and expansion of Cost Farms. The discussion quickly turned to a new freestall barn with robotic milkers – as no one had the desire to hire or manage outside employees.

Ground broke in the fall of 2014 and after a long and bumpy road, the new facility became operational in September of 2015.

Expanding the herd, inside and out

Once the decision was made to expand the herd to approximately 120 milking animals and two robots, a strategy was outlined to grow the herd from both the inside and out. Knowing we wouldn’t be able to achieve internal growth by fall of 2015, the hunt for a group of 40 to 50 cows to purchase began, ending with a purchase of an outstanding herd of cows from 140 miles away.

Adding that many cows to an existing herd at one time came with its challenges, but the two groups eventually became acclimated to each other and the new facilities.

Advertisement

To grow internally, a strategy using a combination of sexed semen and flushing of our top cows was used. Embryos were implanted into our grade heifers, and almost every high-producing, registered animal was bred to sexed semen.

Doing all this would allow us to expand upon only the animals we wanted to keep working with, or that we felt would excel in a robot environment.

Heat detection and synchronization

Once the robots were up and running, our heat detection and breeding program changed dramatically – from a completely visual, natural heat detection system, to a detection system based almost completely on activity monitoring and estrus synchronization on non-expressive cows. As any robotic dairyman will tell you, the amount of data a robot collects on an individual cow basis is astronomical.

Daily measurements of activity, rumination, temperature, weight, plus individual quarter measurements for milk output, component percentages and conductivity, can easily overwhelm a person without having a system to analyze and monitor the data in an organized fashion.

Fortunately, the software packages that can be purchased with the robots do just that. For breeding based off of heat probability, or the combination of increased activity and reduced rumination, the system creates timeframe windows for optimal time of insemination. This allows for the highest percentage chance of conception, whether you are using conventional or sexed semen.

In addition to increasing the odds of detecting natural heats, we now incorporate a sophisticated program to monitor cows that were not expressing heat. By using the information collected by the robot, a cause can often be diagnosed.

Using the information, you can see underlying events that may not have raised a red flag at the time, but can have lasting effects on a cow’s reproductive health. To overcome these issues, other than preventative care, synchronization protocols can be implemented, with which the robot assists. This will improve herd reproductive performance.

Now a cow that needs to enter a shot protocol and be sorted from the main herd can be easily and automatically identified. This reduces the amount of labor needed to ensure the success of the protocol.

Change in philosophy

Once we began production in the robot facility, it became clear a change in breeding philosophy was needed – and that change needed to be done through sire selection. Historically having been a herd focused on conformation and production, cows having sound udders, correct teat placement and functional feet and legs was not something with which we needed to be concerned.

A very small portion of the culling within the last 12 months has been for conformation issues. There are always extreme cases, however, and those happen in some circumstances. Robots are very capable of making adjustments to accommodate cows you would not think will work in such a system.

The areas that needed improvement were those we had not previously thought to look at. In our old flat stanchion barn, we came in close contact with every cow, twice a day, which allowed us to very closely see changes in health. Now with the cows free to come and go as they please, we don’t have the opportunity to spend time with each cow daily, but we still need to know cows are productive and healthy.

Robots do an outstanding job of illness detection; however, we preferred to rely on preventive actions and hearty breeding to promote cow health. Traits such as milk-out efficiency, desire to come to the robot, mastitis resistance and claw health immediately came to the front of our minds.

Because of my experience in the A.I. industry, I have an understanding of the information available to producers, and was able to begin using trait heritability to our advantage. Understanding the heritability of traits is critical within any breeding program.

Once problems are identified in a herd, knowing how many generations it will take to correct the issue can assist in steering sire selection and trait emphasis to capitalize upon the genetics you have available.

We quickly recognized hoof health, hardness and soundness of hoof, combined with the ability to fight infection, was a concern. Constant concrete exposure is extremely hard on a cow’s feet. A strict, twice-a-week footbath protocol was implemented, which is a solution that adds additional costs to the operation.

Using hoof health breeding values available on bulls, we look for bulls that are above average for this value in our selection criteria. Knowing the heritability of hoof health is 12 percent, similar to that of somatic cell score, we can expect within two to three generations we will see a natural reduction in claw afflictions.

Box time, or time spent at the robot, also became a key performance indicator for our cows. A cow that gives 120 pounds per day is great, but if she spends 15 minutes in the box three times per day to do it, she prevents other cows from visiting the robot and reduces the amount of cows we can milk.

Being able to use robot efficiency, or milk produced/letdown per minute in the box, in our sire selection criteria with a heritability of almost 30 percent, allows us to mate slow milk-out cows to the right bull and make significant improvement within just one or two generations.

Similar to robot efficiency, we look at a bull’s value for robot interval, or how frequently his daughters visit the robot, and we only use bulls above average for this trait. We have even started genomic testing calves born from these “corrective matings” to monitor changes in genetic level from generation to generation and can clearly see the positive results.

Like any major change, time is a critical factor. Whether adjustments need to be made quickly, or are long processes, monitoring and measuring are essential to know you are headed in the right direction. With an understanding of where our herd was at the beginning of this process and where we wanted to be in the future, we were able to create a roadmap for getting us there, with reproduction as a pivotal factor to achieve our goals.

Pairing the tremendous amount of daily data collected on a single cow with sire selection based both on trait criteria and the needs of our herd, both phenotypically and genetically, we are able to create a herd of cows we are sure will thrive in the robot environment.  end mark

Matthew Costello