In the last article I suggested that reading Dr. Enns’ Producer Report would be a good place to start for looking at how to use AOA’s calculated EPDs. If you have not done so already you can find the current report www.AlpacaInfo.com/report.
After giving us a summary of the data submitted, Dr. Enns goes on to write about Analysis Procedures, in which he tells us about heritabilities and accuracies. I would like to look at the subject of heritability today for two reasons. First, it is the basis for calculating EPDs, and it is the main reason why we need to use mathematically calculated expected progeny differences as one of our selection tools. Second, because it is one of the most easily misunderstood foundational genetic concepts. When we are mating two alpacas the goal of course is to produce an offspring. We would like that offspring ideally to have certain characteristics that are important to us based on whatever our own personal goals are for our breeding program. Each of the two parents we have selected to mate with each other will pass on one half of their genes to the offspring. This is all they can pass on through inheritance. They cannot pass on any characteristics for the traits that they exhibit that come from environmental influence. The problem lies in the fact that we cannot actually see or measure exactly what those genes are that each parent’s cell contain. What we can observe or measure in some way is the characteristic for each trait that the parent exhibits to us on a particular day. For most of the traits that are important to us as alpaca breeders the characteristics which we can actually see or measure, which is what we call the animal’s phenotype, are a result of the combination of two kinds of influence. From conception until the end of the animal’s life, its inherited genetic information guides and influences the growth, development, and production of the traits we are interested in and environmental influences of all kinds, either permanent or temporary, also affect the growth, development, and production of those same traits. All we can actually see or measure is the result of all of those effects.
To simplify a bit, let us choose a particular trait to talk about, staple length for instance. We can measure the length of staple in the fleece which an animal produces over a period of time. Part of the staple production is due to purely genetic influences. The other part of that staple length produced is based on the environmental influences that affected the animal during that time period. This could possibly include many environmental influences. The animal’s nutrition or general health or exposure to toxins or a myriad of other factors could possibly have an effect. In order to evaluate this animal’s genetic value as a parent, his or her breeding value, we would like to know exactly what part of this staple length production is from genetic influences that the parent might pass along to the next generation. Heritability is simply a term used to describe just how strong, or weak, the relationship is between what we can actually see, the phenotypic value, and what the animal can pass along, the genetic breeding value. If this relationship is very strong, then heritability is high. If this relationship is weak, then heritability is low.
If heritability is low, it does not mean that the parent does not pass on genes to the offspring that influence this trait. It just means that what we can actually see in the parent is more a result of the environment than a result of those genes it will pass along. The genetic breeding value is more hidden from us in other words, but none the less it is still there. It is just harder to select for accurately with phenotypic selection.
The table of heritabilities in Dr. Enns report shows us the heritability for fiber diameter traits, for instance, are higher than those for fleece weight or staple length. That says to us that looking at the animal’s staple length tells us less about the genetic influences that affected that staple length than looking at the animal’s fiber diameter tells us about the genetic influences which contributed to that particular fiber diameter measurement.
Because of the mathematics involved, lower heritablities mean that we will need more data to calculate EPDs of the same accuracy for that trait than we will need for an equivalent accuracy in the calculation of a trait with higher heritabilty. Ultimately high accuracies can be obtained for traits with lower heritablities as well. That means that if we want to make accurate genetic selection choices, using EPDs for these traits of lower heritability is even more useful compared to the standard phenotypic way of making selection and mating decisions than it is for traits of higher heritability.
Dr. Enns next writes about accuracy. We all have an understanding of what the concept of accuracy is and often I hear this being a major concern that breeders have about using EPDs. In the next article we will look at what the accuracy numbers themselves mean and we will explore why Dr. Enns writes “No matter the accuracy, however, the EPD for a trait is a consistently better prediction of an animal’s genetic merit than its own performance alone because EPD are based on considerably more data than that from just the individual.” This is a very important statement and will do much to understanding of the importance of using EPDs as well as the confidence level we have in using them.