EPDs 101: Use Information to Improve Your Herd

Jared E. Decker
Associate Professor, Division of Animal Sciences, University of Missouri
http://blog.steakgenomics.org/2019/03/epds-101-use-information-to-improve.html

Can we be frank for a minute? It is quite simple: EPDs work. When we use EPDs to make selection decisions (which bulls to buy, which females to keep and cull), the performance of our herd improves. Let’s discuss why EPDs work, how they can be used, and pitfalls to avoid.

Defining EPD

EPD stands for Expected Progeny Difference. “Expected” in this context is a loaded word. We use it here the way a statistician would use it. Expected means we are describing a prediction of the future. Expected also means we are discussing an average, not a single observation. What is the average that we are predicting with EPDs? We are predicting the average progeny, or the average of an animal’s calf crop. Finally, when we are discussing EPDs we are discussing differences. Either the difference between two animals or the difference between an animal and the breed average.

The Key to EPDs

What makes EPDs special? Genetic predictions, i.e. EPDs, separate genetic variation from the total variation in a trait. There is a key piece of information that is essential to do this. We need to have some measure of genetic similarity. In the past, we used pedigree information to estimate this genetic similarity. Now we use a combination of DNA data and pedigree data to measure genetic similarity. With measures of genetic similarity, we can separated the bell curve for the variation in the trait into two parts: first, a genetic variation and second, other sources of variation. Now that we have the genetic variation isolated, we can use that informationto make genetic decisions.

Contemporary Groups

Contemporary groups are another important piece of genetic evaluations. Contemporary groups are groups of animals from the same farm or ranch that were managed the same, are the same sex, and are similar in age. By accounting for contemporary groups in genetic evaluations, we remove sources of similarity or differences that are not due to genetics. This makes the genetic evaluation more accurate. By using contemporary groups, we can perform a national genetic evaluation.

The Purpose of a Bull Sale

What is the purpose of a bull sale? What are we buying at a bull sale? What is the purpose of a bull? And, what distinguishes a great bull from a good bull?

We do not go to a bull sale to buy the environment. We do not go to a bull sale to buy management practices or feed rations. When we purchase a bull, we are buying genetics. A bull is simply a delivery mechanism for the genetic potential of our next calf crop. Bulls should be measured on the performance of their calves. The ultimate measure of a bull is not how he looks, but how his calves perform and affect our profitability.

Far too often, bull buyers try to buy management, nutrition and environment when they purchase a bull. While a bull needs to be fertile and sound to do his job, his main purpose is to provide genetics for the next generation. When we make genetic decisions based on actual performance or adjusted performance, we are trying to purchase the management of the bull. Management is not passed on to future generations. Only by using EPDs to make genetic decisions do we focus our decisions on genetic merit. Genetic evaluations take raw measures and create information we can use when purchasing a bull.

How EPDs Are Used

We use EPDs to rank animals from the most favorable to the least favorable. For example, suppose we want to increase the growth potential of our herd. We can rank potential artificial insemination (AI) sires from those with the largest Weaning Weight (WW) EPD to those with the smallest WW EPD. We would then purchase semen on the bulls with the largest WW EPD to increase the weaning weights in our herd. But, we need to discuss two pitfalls with this hypothetical example.

First, we need to avoid single trait selection. We do not want to focus on one trait. By focusing on one trait, we often take one step forward for that trait but two steps back on other economically important traits. Focusing on multiple traits can be difficult, but economic selection indexes make this easier. Economic selection indexes combine multiple traits into a single number based on each trait’s economic importance. Commercial producers should identify an index that matches their production system and marketing endpoint.

Second, extremes are not always better. For some traits, a middle or optimal value is best. We want birth weights that are small enough to avoid calving difficulties. But, we do not want calves born so small that they do not thrive as newborns. We also want Milk or Maternal (MAT) EPDs that match our forage resources. We want cows that provide adequate nutrition to their calves. However, cows with high genetic potential for Milk or MAT have high maintenance energy requirements and cannot fully express their milk production potential. These cows waste forage resources and in some environments may have trouble breeding back. EPDs can be used to select cattle that are at or near breed average; EPDs do not require selection for extremes.

Avoiding Pitfalls

EPDs can be imprecise, e.g. miss the mark, for individual animals. This is especially true if it is a young animal with very little data. There are a couple of strategies to counteract this.

First, we can turn in more data on an animal. Seedstock producers can collect as much data as possible and turn complete, clean data into their breed association. Commercial producers can work with seedstock producers who are passionate about data reporting. An easy way to get more data is made possible through technology. DNA testing a bull to produce genomic-enhanced EPDs is equal to reporting about 20 progeny records for all published traits. Genomic testing improves the accuracy of EPDs as it allows us to better measure the genetic similarity of animals in the genetic evaluation.

However, we should not double- or triple-count information. One of the ways we commonly see people do this is when selecting a calving ease bull. They often look at the actual birth weight, the Birth Weight EPD, and the Calving Ease Direct EPD to make this decision. But, this practice makes these decisions less accurate, because instead of adding information they are adding together uncertainty. Further, the Calving Ease Direct EPD contains the information in Birth Weight EPD and Birth Weight EPD contains the information in the actual birth weight. So, the best practice is to simply look at the Calve Ease Direct EPD that contains all of the available information.

Second, we can hedge our bets. Too often in beef breeding, we are looking for the “One” great bull. But, in reality, the “One” does not exists. Instead of using one bull, we hedge against changing EPDs by using a group of bulls. Because EPDs are unbiased, some of the bulls in the group will have their EPDs go up, others will have their EPDs go down, and others will stay the same as EPDs are updated with more data. The average of this group of bulls will remain the same before and after the EPD updates. However, we are at the mercy of randomness if we use a single bull. His EPDs might stay the same, go up, or go down. Because EPDs are predictions of the average, we can use this property to protect ourselves against uncertainty.

How EPDs Work in Practice

Suppose we have two bulls, Black Bull and Gold Bull. Black Bull has a WW EPD of 2 and Gold Bull has a WW EPD of 22. There is a 20 pound difference in the EPDs of these two bulls. At the same ranch under the same management and environment, we mate each of these bulls to 100 cows apiece. Fifteen months later, we wean the resulting calf crops. The Black Bull’s calves average 495 pounds at weaning, with the majority of the calves weighing between 445 to 545 pounds. The Gold Bull’s calves average 515 pounds at weaning. The 20-pound difference in the EPDs of the bulls is reflected in a 20-pound difference in the average of their calf crop. However, due to the randomness of inheritance and genetics, some of the Gold Bull’s calves under perform the Black Bull’s average and some of the Black Bull’s calves outperform the Gold Bull’s average. We cannot do much to change the shape of the bell curves. However, with the use of EPDs, we can move the bell curve in the direction we want it to go.

Conclusion

Cattle producers can use EPDs to use all the data available boiled down to information they can use. EPDs predict genetic differences and inform selection decisions. EPDs produce the desired results when used consistently and properly. Whether in good times or bad, EPDs help us accomplish our goals.

http://blog.steakgenomics.org/2019/03/epds-101-use-information-to-improve.html