“Excellence is doing ordinary things extraordinarily well.”

Dr. Wade Shafer, Ph.D wrote this article back in 2011 and when we discussed the Milk EPD again last summer, he forwarded this to me and stated the logic still applies today. If you have questions on the Milk EPD, this does a great job breaking it down for us.

An Examination of Milk EPDs
A discussion of the validity of EPDs for milk.
By Wade Shafer, Ph.D., ASA Director of Performance Programs
Published in March 2011

Over the last several years I have fielded too many calls to count from breeders feeling that Milk EPDs do not do justice in gauging their herd’s genetic milking ability. Though you may not be one of the many breeders who have voiced reservations about Milk EPDs to me directly, I suspect
you may have had qualms with them from time to time. In talking with people of other breeds, this sentiment is not unique to Simmental breeders. Though many breeders may question the validity of milk EPDs, they are without a doubt the best predictor we have
of animals’ genetic level for milk. That said, when we scrutinize the situation, breeders’ skepticism about Milk EPDs is understandable. In the following paragraphs I will explain why EPDs gauge genetic level for
milk as accurately as possible and offer up my opinions as to why breeders often question their validity. To understand a widely held belief, it usually helps to get at the root of it. What is at the core of breeders’ misgivings about Milk EPDs? From my experience, breeders seldom question milk EPDs on sires — rather, doubt tends to permeate when cows in their herd wean big calves yet have lower-percentile
Milk EPDs. On the surface, their skepticism seems justified. It just seems logical that high-producing cows should have upper-end milk EPDs. It turns out there are solid reasons why a top producing cow does not
necessarily possess genes for high levels of milk production — and, therefore, should not have high-level milk EPDs. To gain some perspective, let’s take a look at factors affecting the measurement
we use to gauge milking ability — weaning weights. (Obviously, we do not have the latitude to measure milk directly, as does the dairy industry. Even if we could, the trait we are concerned with in the beef business is the pounds of weaned calf expected from a cow due to the genes she
possesses for milk and mothering ability — which we can compute using weaning weights.) From a cow’s genetic perspective, the weaning weight of her calf is affected by her own milk genes and half of her weaning weight genes (the calf carries half of her genes). (This fact is the reason
Maternal Weaning Weight EPD is calculated as Milk EPD+ ½ Weaning Weight EPD.) Beyond that, there are non-genetic factors affecting the calf’s weaning weight (age of its dam, stocking rate, weather, creep feed, and many other non genetic effects we are not able to distinguish).
Due to the myriad of factors affecting weaning weight, using actual and/or adjusted weaning weights and/or ratios will not serve us well in predicting an animal’s genetic value for milk production. One glaring shortcoming of these statistics is that they do not partition weaning weight into separate growth and milk components — an obvious requisite for assessing milk genetics. Besides that monumental failing, weights and ratios do little to nothing when it comes to sifting through the non-genetic influences on weaning weight and they ignore information on relatives. Because the methodology used to calculate EPDs (Best Linear Unbiased Prediction; BLUP) allows us to parse weaning weight by the factors that impact it, EPDs permit us to zero in on an animal’s genetics for milk — independent of all the other factors. This is accomplished by leveraging the sophisticated statistical techniques of BLUP on the 6 million+ weaning weights in our database. These features make EPDs vastly superior to other methods of gauging milk genetics.
After extolling the virtues of EPDs, how is it that we can have cows that wean heavy calves with poor milk EPDs (and vice versa)? One obvious reason is that these cows may be weaning heavy calves via exceptional
growth genes they pass to their calves. (Just look through ASA’s database and you will find plenty of animals that have low Milk EPDs but high Maternal Weaning Weight EPDs due to their extreme growth. Remember, Maternal Weaning Weight EPD = Milk EPD+ ½ Weaning Weight.) A typical retort to this assertion is “how could the calf express exceptional growth if the cow did not give loads of milk?” Though high levels of milk certainly help a calf express its growth potential, a calf gets a significant amount of its nutrients from sources other than milk
(particularly after the first few months). Another plausible cause for a heavy weaning calf out of a low milk EPD cow is simply that she produced lots of milk. In these circumstances, our genetic evaluation is telling us that the reason for the souped-up milk production is likely non-
genetic in nature. This is a big pill to swallow, as we tend to assume
that a heavy milking cow must be genetically outstanding for milk
production; however, since milk production is only low to moderately
heritable we should not be surprised to have heavy milking cows that
are below average genetically, and vice versa. From a genetic evaluation standpoint, milk’s meager heritability means that it takes more data to move Milk EPDs and improve accuracies than it does with more heritable traits (e.g., growth and carcass). Since a cow is quite limited in the number of calves she can have naturally, unless she has numerous daughters (typically only possible for donor cows) and they have many calves or her sire or maternal grandsire change dramatically for milk, a colossal shift in her milk EPD from where she starts is not in the cards. A big jump in accuracy is just as unlikely. These limitations do not in any way imply fault with EPDs, the limitations are simply a function of biology — low heritability and small numbers of offspring are not conducive to accurate estimation. Always keep in mind, however, even though a cow’s Milk EPDs may not be highly accurate, research has clearly shown that they are far and away the most accurate estimate of her genetic merit for milk. In our population we have estimated the heritability of milk to be 16%. This means 84% (100% – 16%) of the differences in milk production between cows during a lactation are due to non-genetic factors — and that is after we remove differences due to age and contemporary group (i.e., herd, pasture). Given these circumstances, it iseasy to see that a cow’s milk production ability may be very different than the genes she possesses for the trait.