New Evaluations Offer a Genetic Approach to Improving Cow Fertility

Dr. Kent Weigel, Dairy Science Department, University of WisconsinMadison
Dr. Paul VanRaden, Animal Improvement Programs Laboratory,
Agricultural Research Service, USDA, Beltsville, MD

What can I do to help my high-producing cows get pregnant? That question has probably received more attention from veterinarians, consultants, nutritionists, and reproductive specialists than any other during the past few years.  Now geneticists are taking a crack at it as well. Why the attention from so many different groups?  Because reproduction is a complex process, that's influenced by almost everything in a cow's environment. An easy catch-all phrase for those factors is "cow stress". Many herds now average more than 30,000 lbs. of milk per cow per year, and the physiological challenges are easy to imagine. The list of contributing factors is long. Let's start with housing – cows are spending more time on concrete, and that doesn't encourage expression of visible estrus symptoms.  How about labor?  Farms are getting larger, with more cows per worker and more hired employees. Nutrition also makes the list: cows are eating a high-octane diet in huge quantities. What about the weather? The dairy industry has certainly grown in the South and West, hot climates that are not conducive to reproduction.  Management practices, such as the use of BST, also contribute: not necessarily to a decline in conception rate, but rather to a decrease in the cost of keeping cows open. On the other hand, reproduction specialists have developed an array of heat synchronization programs that are designed to get more cows pregnant by inseminating them sooner, more frequently, and with less reliance on estrus detection.

Where does genetics come into play?  Genetic recessives, such as CVM or DUMPS, can affect fertility by causing embryonic deaths or abortions, and inbreeding levels are on the rise in all major dairy breeds.  However, the main culprit seems to be selection for high milk production. We've created high-producing cows that can mobilize body reserves and divert feed resources toward putting money in the bulk tank. However, the antagonism between production and reproduction is well known: the genetic correlation between milk yield and days open is about +0.35. We've nearly doubled milk production over the past 40 years (partly because of genetic selection for production), and over the same period we've also increased days open by about 40 days (also partly because of genetic selection for production). We haven't tried direct selection for improved fertility yet, mainly because the heritability of female fertility is only about 4% (remember all the environmental factors listed above). But we have been indirectly selecting for fertility for the past eight years via genetic evaluations for length of productive life.  The genetic correlation between productive life and days open is a whopping -0.60, about twice that between productive life and other important traits. It's simple, if cows don't get pregnant, they get culled.

Genetic selection for cow fertility will rely on several different sources of data. The most reliable will be days open as reported through the DHI milk recording system and as verified by a subsequent calving date. However, some cows don't calve again. Their reported days open can still be used, but they can't be checked against calving dates. On the other hand, some producers don't routinely report insemination or pregnancy data. For cows in those herds, we can simply subtract 280 days (a typical gestation length) from the calving date to calculate days open in the previous lactation. Lastly, some cows are culled for infertility, and we can assign them a large (albeit arbitrary) value of 250 days open.  Genetic evaluations then proceed using the same animal model system as for milk, type, and other traits and adjustments for known environmental factors, such as contemporary group, and age of cow.

Genetic evaluations will be published as daughter pregnancy rate (DPR). Many producers are already familiar with 21-day pregnancy rates, because they are routinely used by consultants, veterinarians, and herd management software programs. For those who aren't yet familiar with pregnancy rates, their calculation is relatively simple: choose a recent 21-day time period (the length of a typical estrus cycle), and then divide the number of cows that became pregnant during this period by the number of cows that were eligible for breeding. For national genetic evaluations, all open cows that have passed the voluntary waiting period but have not yet reached 250 days in milk are considered to be eligible for breeding. Therefore, although the original data are recorded as days open, 21-day DPR can be obtained using a simple approximation. A 1% increase in DPR corresponds to a decrease of 4 days open, and vice-versa. DPR encompasses a cow's ability to return to normal reproductive status after calving, to display visible signs of estrus, to conceive when inseminated, and to maintain the pregnancy.

Daughter Pregnancy Rate (DPR)

The amount by which daughters of a particular sire are expected
to exceed or to fall short of your herd's average 21-day pregnancy rate.

+1% DPR = 4 fewer days open

Let's look at some example bulls; we'll use "oldies, but goodies" that are well known among producers:

Bull's Name



Days Open


+ 0.9 %

- 3.6 days


+ 0.2 %

-0.8 days


- 2.6 %

+ 10.4 days


+ 1.2 %

- 4.8 days


+ 2.4 %

- 9.6 days


+ 3.2 %

- 12.8 days


- 1.6 %

+ 6.4 days


- 2.0 %

+ 8.0 days


+ 2.0 %

- 8.0 days

For example, if your herd's 21-day pregnancy rate is typically around 20%, your Duster daughters would be expected to have an average 21-day pregnancy rate of 22.4%. Likewise, if your herd averages 153 days open, you can expect your Duster daughters to average 143.4 days open. On the other hand, you can expect your Melwood daughters to have a 21-day pregnancy rate of about 18% and an average of 161 days open. You can see that cows from the best daughter pregnancy rate bulls will have pregnancy rates 2-3% higher than the other cows in your herd, whereas the poorest daughter pregnancy rate bulls will sire cows that fall 2-3% below your average 21-day pregnancy rate. When expressed as days open, differences between the best and worst fertility bulls span about three weeks. 

What's the downside, you ask? Well, you should remember several key points when adding DPR to your sire selection toolbox:

¨ Reliability values will be low for AI bulls that have only first-crop daughters.  The heritability of female fertility is only 4%.  Therefore, most recently-released bulls will have reliabilities for DPR of 45% to 60%, and pedigree influence will be a very important factor.

¨ Evaluations for some bulls may change significantly over time, as new information arrives.  Although this is true whenever reliability is low, it is more important for DPR, because additional types of data (reported breedings, calving intervals, and reproductive culls) will arrive at different times during a sire's life.

¨ A delay will occur between a bull's first genetic evaluation for milk production and his first evaluation for DPR.  This delay may be three or more sire summaries, because we must wait for cows to have the opportunity to become pregnant before contributing to their sire's proof or to have a subsequent calving if they are in a herd that does not report breedings.

¨ Many high-ranking bulls for milk, fat, and protein will have negative evaluations for DPR.  Remember that production and reproduction have a negative genetic correlation and that relatively few bulls will rank highly for both traits.  You shouldn't ignore all bulls that are average or slightly below average for fertility, however, if those bulls are superior for other economically important traits.

¨ DPR is not the same as estimated relative conception rate (ERCR).  DPR is a measure of female fertility – a cow's ability to cycle, to show estrus, and to conceive in a timely manner.  On the other hand, ERCR is a measure of male fertility – a bull's ability to produce fertile semen that will result in a pregnancy.  Male and female fertility are independent traits, and both are useful in predicting reproductive success.

¨ DPR is a new trait, and we're all still learning how to use it.  Semen salesmen, AI technicians, veterinarians, and (maybe even some) extension geneticists won't have all the answers right away.  Don't rush in – take some time to study the DPR values of your favorite bulls, and watch how they change over time.  Most importantly, try to view DPR as the first step toward genetic improvement of cow fertility.  Official DPR evaluations are scheduled for February 2003, and further details regarding their calculation will be announced in upcoming months.  As we refine the statistical procedures for analyzing fertility data, and as producers supply better data for inseminations, veterinary palpations, ultrasound examinations, and so on, we can continue to improve the evaluation system over time.