March 3, 2020
the april 2020 genetic base change
A 5-year report card on our genetic progress
The US genetic evaluation system will have a genetic base change in April 2020. This is in accordance with global standards set by the International Committee for Animal Recording (ICAR). This ensures that everyone evaluates the genetics of their animals to an up-to-date average population. Every base change brings a considerable shift in PTAs even though the genetics of our breeding stock obviously don’t change overnight. The more genetic progress we make, the bigger the base changes are.
Why do we do this, what does it mean exactly and what will be the changes in April?
Why a base change?
When we say: a bull has a PTAM of +1,000lbs, we expect the daughters of this bull to produce 1,000lbs more milk than the average cow. But what is the average cow? In order to express PTAs, we need to determine what our average is. The population of cows that we call the ‘average’ is referred to as the ‘base population’. All PTAs are expressed relative to this population. For most traits, animals with positive PTAs are population improvers and those with negative PTAs move the genetic level of the population back.
Our dairy population keeps improving and its average performance keeps changing. If we would compare our bulls to the average cow population in 1950, all bulls would have extremely high positive PTA values and we wouldn’t be able to pick out the ones that bring progress. We therefore want to keep comparing the PTAs of our bulls to a current milking population. We do that by updating our base population every 5 years. All the breeding values are then re-calibrated and expressed relative to the genetic level of this new group of animals which we will call the ‘average’ for the next 5 years.
What happens at a base change?
Since December 2014, our base population consisted of all cows born in 2010. In April 2020 we will update that to all cows born in 2015. We do so by evaluating the average PTA of cows born in 2015 (that have completed at least one lactation) and equating that number to zero. This happens within breed and for each separate trait. All PTAs will then be expressed relative to this new genetic base for the next five years which makes it clear if a PTA is favorable in relation to the current milking population.
Because we make genetic progress, the genetics of the population born in 2015 are better than the animals born in 2010. This is expressed in their performance, along with the environmental effect.
U.S. Holstein cows born in 2010 produced on average 26,994 pounds of milk (mature equivalent (ME)) whereas U.S. Holsteins born in 2015 produced 28,071 pounds (ME), an improvement of 1,077 pounds. Of this 1,077-pound gain in performance, CDCB reports 984 pounds to be due to genetic improvement (Norman et al. 2020, April 2020: Genetic Base Change). The rest is caused by environmental developments. A PTA reflects what the animal passes on to the next generation, which is 50% of her genetic merit. The average PTA for milk of Holstein cows born in 2015 is thereby +492lbs (see Table 1). To make this average genetic level our new zero, we will subtract 492lbs from all the PTAs for milk of our Holstein bulls and cows in April 2020.
Whereas his genetics won’t change, a bull with a current PTAM of +1000lbs for milk will therefore have a PTA of 1000-492=+508lbs (plus or minus any changes that come with a normal evaluation) after April 2020 as we will now compare the daughters of this bull to an up-to-date milking population with a much higher average genetic merit. Because the genetic base changes equally for all bulls and all traits within breed, the base change itself will not affect the rankings between bulls. All Holstein bulls will have 492lbs subtracted from their PTAM in April 2020. For Jersey, the change will be a subtraction of 524lbs and Brown Swiss animals will see a decrease of 214lbs etc. There are some traits where a lower PTA is desirable, such as SCS, or calving ease and stillbirth. We want to keep moving the genetic level of the population for these traits downwards. The base change for these traits will therefore not be a subtraction but an addition. A good example is SCS: the cow population born in 2015 had a lower average SCS than the population in 2010. To set this average to zero, we have to add (in case of Holstein) +0.08 to all PTAs for SCS.
Please refer to Table 1 to see the base changes for all the breeds and traits.
By setting the average PTA of cows born in 2015 to zero, we must mathematically subtract 50% of the genetic gain accumulated since the last base change in 2010. After all, for milk, the average PTA for Holstein cows born in 2015 was +492lbs while the genetic gain for the period 2010-2015 was 984 pounds. The amount of change is thereby reflective of the genetic progress that was made in the period between 2010-2015. You could say that the values of the base change make up the report card of how well we have done with moving our population forward genetically.
When evaluating the genetic base change, you can see how much progress was made for each trait based on the size of the genetic base change. Some key progress points, highlighted by the Council of Dairy Cattle Breeding (CDCB), are:
- Favorable gains were shown for 81 out of 102 traits (excluding type)
- Genetic improvement was made in production traits for all breeds, with Holstein and Jersey showing large progress for milk lbs.
- All breeds have increased their genetic capacity for a longer productive life
- All breeds show genetic improvement for the lifetime merit indices. Largest gains were for Holstein, Jersey and Ayrshire
The effect of genomics
Because of technological developments, the genetic progress we make keeps accelerating. The more genetic progress we make in the 5-year period between base changes, the larger our base changes will be.
Genomic selection was introduced in 2009 and its uptake has been remarkable. Figure 1 demonstrates that in 2009, 20% of marketed AI bulls were young genomic bulls, and 80% were still daughter proven. Hence, the population born in 2010 was likely sired by predominantly proven bulls. Fast forward four years and we have a totally different situation. In 2014, 61% of marketed bulls are young genomic bulls and only 39% are daughter proven. A much larger proportion of our new base population, born in 2015, are thereby daughters of young genomic bulls. It therefore comes as no surprise that the magnitude of change that the April 2020 changes bring is much higher than those in 2014 which is proof of how well we are improving our US dairy population. The period of 2015-2020 has added genomic selection in commercial females. A growing proportion of the female commercial cows in the United States are genotyped and genomic PTAs are playing an increasingly large role in the selection and mating decision on farm. With genetic progress increasing on both the male and female side of the equation, we can expect the base change in 2025 to be even larger.
Base changes for type evaluations
Production traits are intuitive where a higher number means better. Type evaluations, however, include some traits with an intermediate optimum. For these traits, it’s important to remember that the zero level is simply the base population average. This doesn’t necessarily mean that a 0 PTA represents the genetics for the ideal phenotype. We always aim to keep moving the population to a phenotype that we consider desirable whether that is healthier, more fertile or higher producing animal. For type traits, the ideal is less defined and differs between breeds.
Table 2. presents the base changes for the type traits per breed. Most apparent are the adjustments in Final Score and Udder Composite for Holstein, in addition to the +0.27 adjustment for teat length. The latter makes it appear that teats have gotten considerably shorter over the last 5 years. However, if we look at the underlying phenotypic scale the +0.27 genetic change equates to only 0.019 inches phenotypically as each point in PTA for teat length represents 0.2 inches. So, between 2010 and 2015, the average teat length of our Holstein cows became 0.019 inches shorter. For reference, that is the thickness of 5 sheets of stacked normal paper and less than 1/3 the thickness of a quarter. Similarly, for udder depth, which shows a drop of 0.84 points this base change, the seemingly large change on a genetic level represents a much humbler change of 0.23 inches on a phenotypic scale.
Across all breeds, most PTAs increased between 2010-2015 indicating that selection has been for the higher scores.
What if a bull has a negative PTA after the base change?
Because the majority of the PTAs will be lower after April 2020, it is possible that a bull that you currently use drops under the zero level for a trait and receives a negative PTA. These zero thresholds are often given more credit than they deserve. A below zero PTA does not mean that the bull will suddenly decrease the genetic average of your herd. Remember that a negative PTA simply means that the bull is lower for that particular trait than the cow population born in 2015. If the drop is minor, that could equate to tenths of an inch for traits such as udder composite and final score or tenths of a percentage point for fertility traits such as DPR or health traits such as mastitis.
So, when a bull does drop under zero for one or multiple traits in April 2020, don’t kick him out of your breeding program too quickly. Keep comparing the bull to your own herd average. It would be a shame to miss out on a good bull for a tenths of an inch adjustment that will likely go unnoticed.
In order to keep improving our US dairy population using PTAs as our genetic evaluation system we need to adjust the population we refer to as the ‘average’ every 5 years. In April 2020, we will therefore update our base population to cows born in 2015 that have completed one full lactation. Because we make fast genetic progress, many of our PTAs will change significantly and most will decrease in level. This doesn’t mean that the genetics of the animal change after April, but rather that we compare those genetics to a more current population of higher genetic value. Due to genomic selection, the 2020 base changes that we will observe are higher than those we had in 2015. Similarly, we can expect even greater changes at the next base change in 2025. Although those larger changes can be temporarily disruptive, they are nothing to be concerned about. In fact, they are indicative of the excellent job the US does in genetically improving its population and that should be a compliment to all involved.
Table 1. Average PTAs of the new base population per breed. PTAs will decrease by these amounts at the 2020 genetic base change
Source: Norman et al. 2020, April 2020: Genetic Base Change
Table 2. Average PTAs of the new base population for the type traits, per breed. PTAs will decrease by these amounts at the 2020 genetic base change
Source: Norman et al. 2020, April 2020: Genetic Base Change (https://www.uscdcb.com/wp-content/uploads/2020/02/Norman-et-al-Genetic-Base-Change-April-2020-FINAL_new.pdf)
- Figure 1. Genomic vs Proven bull availability 2009-2019
Source: National Association of Animal Breeders (naab-css.org)