Have you ever wondered why some sires’ daughters perform better in some herds or environments than they do in others? I have. The current sire indexing system may rank two sires as being of equal genetic merit, yet their daughters may perform differently in the individual tie stall barns of cold Minnesota compared to the 400+ cow groupings in the heat and humidity of a Florida cow shed. The system assumes that there are not performance expression differences due to environment.
Geneticists do not know enough about what happens on farm
It is a known fact that our geneticists do not have enough details about the animals’ health events, ability to perform in large groups, differing nutritional programs within a herd, calf-heifer disease and many other matters when processing the genetic evaluations to produce genetic indexes. Without the details, geneticists can only assume all animals in a herd are treated equally. We all know that this not the case.
Other Livestock have similar Challenges
Recently I read an interesting presentation (EPDs only one part of the genetic selection formula, 2018 Canadian Beef Breeds Council’s Technical Forum) by P J Budler of Modern Ova Trends on beef cattle genetic indexing. He cautioned about using EPDs (Estimated Predicted Differences aka genetic indexes) without also considering nutrition, herd management, animal health, forage program, animal marketing program, record keeping, human capital and farm finances. His article also made mention about breed performance differences that depend on environment. His example was fertile Black Angus cows that are great at raising calves in the sometimes harsh cold of the Upper Plains of the United States and Western Canadian Provinces but put them in a hot semi-tropical environment and they do not graze, stand in ponds and they do not breed back. My summation of Budler’s presentation is – a) environment, management and nutrition play a role in an animal’s expression of its genetic make-up and b) sires need to be proven in the environment in which their future daughters will perform.
Plant scientists in genetically evaluating varieties of corn, need to know the length of the growing season, heat units, soil type, tillage program, nutrient program, plant population, spray program and more in order to make accurate predictions on a variety’s ability to perform. The extent of the data captured from corn test plots is huge.
Likewise, it is a fact that livestock genetics do not work independent of nutrition, animal health, animal care, animal management and the environment.
Assuming can lead to Errors
Budler’s presentation got me thinking. Does the dairy cattle breeding industry make too many assumptions about animal treatment equality, when we do our genetic evaluations?
We have super super computers and very advanced methods to statistically analyze data, but we have not expanded the data forwarded to genetic evaluation labs.
Every Bullvine reader can think of a long list of factors beyond genetics that can affect an animal’s performance and for which geneticists do not have data available for inclusion when they do their analysis. This list includes all the things that happen from birth to removal from the herd. Some things like calf morbidity, calf growth, hoof trimming, disease occurrence and animal grouping are not known. And yes, each one on its own may be minor in its affect but in total they lead to errors being made, when it comes to genetically ranking animals in the population.
More Data Can Help
I often hear dairy people say – but that trait has a low heritability so we should not pay much attention to an index until the reliability of prediction is over 90%.
We need to ask – if we could have more data for the animals could the prediction accuracies be increased?
Feet, as currently scored by classifiers, has a low heritability. Could the heritability for feet be increased if the geneticists knew details about calf hoof growth, housing environment of calves, heifers and cows, how recent was the last hoof trimming, have the feet ever been trimmed and has the animal ever been lame?
For more and more milking cows we electronically have observations from every milking (90 data points per month), the nearest weather station can provide the weather for the each day, in-barn monitors capture extensive information, … yet, the dairy cattle improvement industry (breeders and organizations) persist in using one milking or one day’s observations per month to calculate milk yields and ignore data from in-barn monitoring systems. In addition, animal performance beyond milk cows is non-existent in our central data bases.
There are never too many known facts when it comes to making accurate genetic index predictions and information available for managing a dairy herd.
The Goal in Genetic Evaluations
The goal in genetic evaluations is to accurately predict an animal’s ability to transmit a trait relative to other animals in the population. Of course, ability can be both positive and negative.
Every breeder’s goal is to have the perfect animal for a trait and for that animal to transmit that perfection to the next generation. Perfection is not achieved by making decisions based on averages.
More Data Points affect all Aspects of a Dairy Herd
- As mentioned above having more animal, herd and farm data will enhance herd nutrition and management. In fact, those two disciplines will determine 75% of herd profit.
- Bullvine readers continuously learn about new on-farm monitoring devices. The data they supply should be included in the national data base if it can assist in improving herd profit.
- Dairy farmers will experience even tighter financial margins in the future. Data points that contribute to increased profit are a “must have” in the national data base.
- With more and more cloud or on-farm animal / herd management softwares in use, some farmers are talking about discontinuing to use DHI and breed services. If that is done it stops data from being available for benchmarking and for enhancing improvement services.
- It is highly unlikely that sires will ever be sampled and proven randomly across all herd environments scenarios. So, having more data points will assist in genetic index accuracy, especially for low heritability traits.
- More data especially feed efficiency, animal health, animal fertility, calves and heifers will assist in increasing the reliabilities of genomic indexes. Even to 90+% REL within the next decade.
Something to think about
Determining an animal’s lifetime profit is a marathon that starts at birth and ends when the animal leaves the herd. The performance and events focus in the past has been the lactations of the milking cows, thereby the industry has been missing the data from significant parts of each animal’s life.
The Bullvine Bottom Line
It is time for breeders and their representatives on committees and boards to think to the future and the need to use more on-farm data.
The accuracy and number of traits included in genetic evaluations and on-farm performance reporting can be significantly increased by having more on-farm data reach the central national data bases. Use it, not waste it!