Traits that have a higher heritability can be quicker to make changes in than traits with a lower heritability. The heritability of a trait tells us whatproportion of the variation we observe is due to genetics. The key word here is variation.

Variation is the key to genetic improvement

Variation is defined as a measure of spread in the data, and some traits have more variation than others. For instance, the more than 12 million adjusted weaning weights (WW) recorded in the American Angus database vary greatly from the lightest to heaviest weaned calf, while functional longevity (FL) records vary from one to nine calves. This variation itself does not tell us much, and the more important question is, where does the variation come from?

As geneticists, we would like to claim all the observed variation is because of genetics; but we recognize this is not true. There are many sources of variation that influence traits in different ways. For instance, if you feed your cows appropriately and expose them earlier in the season, you most likely improve your herd’s conception rate. The better conception rate is because of nutrition and management, and not related to genetics.

When the goal is to make genetic selection, we need to identify genetic variation, because that is what can be passed on through the generations. To do that, look at phenotype variation and separate what is causing it and how much of the variation is due to genetics.

The “how much” portion is what we call heritability. For instance, the heritability of WW is 0.28, which means of all the WW variation we observe, 28% is due to genetics and 72% is due to environment and management. For fertility-related traits like functional longevity, the heritability is much smaller, 0.1, meaning of the variation we observe in the number of calves cows have, 10% is due to genetics, while 90% is due to unknown environment and management effects.

What kind of variation?

When we look at what causes variation, it helps us better understand what factors are contributing to the variation we observe. We can break down the phenotypic variation we observe (P) into two components: genetic (G) and the environment (E). The E component can be further divided into known and unknown environmental effects (for instance, the cotemporary group is a source of known environmental variation). This gives us the classic P = G + E formula.

In FL for example, 10% of the variation is due to genetics and the rest is nongenetic factors. The nongenetic portion contains several components, many of which are captured by the contemporary group; such as the year, season, farm and management style the cows were raised in, which captures climatic variation across years and farms. Some of these are accounted for by the contemporary group, while others fall in that unknown environmental effects contributing to the remaining nongenetic variation.

Another important component of the variation we observe in FL records is the variation over time. FL is a longitudinal trait, which simply means the performance records are measured during a cow’s lifetime. This is different from traits measured only once in a lifetime, such as WW.

For FL, the variation over time is important because what happens in one year’s breeding season may affect next year’s reproductive performance. These cumulative factors, most of which are notobserved or recorded, cause the unknown environmental variation to be a much higher fraction of the phenotypic variation than other traits, such as WW, making the heritability lower.

For example, if a cow calves earlier in the calving season, she will have more time to recover her body condition and return to estrus sooner, potentially improving her chances of getting pregnant. Being bred earlier or later in the season is a nongenetic component, but still affects her lifetime performance.

One could argue there is a genetic component to breeding earlier as well, but for this example we are focusing on the nongenetic part only. Other examples of nongenetic factors include nutrition and management during early growth, disease exposure when young, management practices year over year, and bull/semen interactions.

Modeling all sources of variation

The FL genetic evaluation model is specifically designed to handle longitudinal data and longlasting nongenetic effects with a component called the permanent environment effect. Permanent environment accounts for factors that affect the cow’s performance consistently across time for all records. This component and modeling strategy is used for any traits in which one animal can have multiple records (e.g., foot scores, teat and udder scores, FL, maternal component of WW).

For FL the permanent environment accounts for 84% of the variation we observe. If you are doing the math, 10% is genetics (FL heritability of 0.1), 84% is the permanent environment (consistent environmental variation across time), and the remaining 6% is due to environmental variation that is unique to a specific breeding season that does not affect subsequent performance.

What does it mean for such a large proportion of the variation we observe in FL to be due to the permanent environmental component? The environment, nutrition and management, even early in the cow’s lifetime, play major roles in whether a cow has a calf or not. So, for cows with exceptional performance, much of that is a result of your good management practices. This is not different in other traits, even for WW, with a
heritability of 0.28. That means 72% of the variation we observe is due to your management and stewardship practices preweaning.

Is the hope lost for genetics?

At a superficial level, it may seem like genetic selection is not worth the effort since for some traits, it’s responsible for only a small proportion of the variation you see. However, it is quite the opposite.

Genetic selection is the type of improvement that in many cases, while slower/smaller, is cumulative and permanent. As the population improves genetically, all herds benefit from it. Do not be discouraged by traits with lower heritabilities and EPDs with a smaller spread, as long as there is variation to select from (e.g., better vs. worse cows).

A disciplined and consistent selection program allows you to make progress, even in more challenging traits, and those can arguably be some of the most important.

As you work on your plan for genetic improvement, think about it as a long-term investment where time in the market is more important
than timing the market, and each generation builds incrementally on the last, compounding gains over time.

Decades of selection for weaning weight and carcass traits have proven that selection results in success; and with time, selection for newer traits such as foot scores, functional longevity and others will also improve the population and move the productivity needle.