keepThe “off-the-cuff” answer is when it comes to animal genetics, what you see is not necessarily what you get. Animals with tremendous individual phenotypes (e.g., heavy weaning weights, stellar reproductive records) can certainly have lackluster genetic value for those traits and vice versa. The reason for this is because an animal’s phenotype is the result of both genetic and environmental factors–and for most traits we measure, the environment has a much stronger impact on an animal’s phenotype than does genetics. How do we know this? Through knowledge of a trait’s heritability.
Heritability measures the proportion, on average, of the differences between animals within in a contemporary group due to genetics. For most of the traits we deal with, heritability is less than 50%–meanno ping that less than 50% of the differences between animals within a contemporary group are due to genetics for those traits. For example, the heritability of reproductive traits (e.g., pregnancy status) is quite low (~10%). This means 90% of the differences we see within a contemporary group for pregnancy status are due to factors that aren’t heritable (unaccounted environmental influences or hybrid vigor for instance). In other words, whether or not a cow is pregnant at the end of the breeding season has very little to do with her genetic potential to conceive and maintain a pregnancy. The growth traits we measure are typically in the 20-30% range for heritability. Though genetics certainly have a larger impact on growth traits than the paltry 10% it accounts for in reproductive traits, the environment still plays the major role. With heritability estimates around 50%, carcass traits are the only area where genetics plays an equal role to the environment. Even then, the environment has a sizable impact on an animal’s phenotype–meaning an animal’s genetic level for carcass traits can be quite different than its phenotype.
Because an animal’s genetic level can be very different than its phenotype for most any trait we are interested in, we must avoid assuming that a shiny phenotype translates to superior genetic value and vice versa. Because what we see isn’t necessarily what we get, we must have a means of addressing that fact–which is what EPDs do. Rather than relying on one phenotype (the animal’s own record), EPDs leverage lots of phenotypes along with knowledge of a trait’s heritability in the calculation of an animal’s genetic merit. In fact, EPDs use all available information (phenotypes on the trait and related traits, pedigrees, DNA) in the prediction of an animal’s genetic level. Because EPDs leverage all available information and do it in a statistically unbiased manner, we can be assured that we have the best possible prediction of an animal’s genetic value–even if it doesn’t match its phenotype.