Heifer Livability: The new trait was released in December, but 99% of the attention went to the simultaneously released Feed Saved trait. Of course, we have been getting used to Cow Livability for a while, but it is still worthwhile to reflect on Heifer Livability for several reasons.
First, the trait is the first to target a period that is otherwise largely neglected by the genetics industry. Calving and stillbirth are traits developed to genetically improve the first few hours of life, and Heifer Conception Rate improves that first breeding. But we didn’t have a national trait that genetically aims to increase the chance of survival from calf to breeding.
The average loss of a heifer is estimated at $500. The latest 2021 dairy statistics published by Progressive Dairy, however, show that the number of replacements is the lowest since 2014, and there are now 48.8 heifers per 100 cows, the smallest ratio since 2009. Sexed semen and genomic testing have allowed a precise and strategic management of replacement heifers of considerable genetic quality. Now more than ever, the loss of a replacement animal means the loss of significant economic and emotional investment. You can state that Heifer Livability comes at the right time.
So why does it feel like “just another health trait”? Where is the excitement?
Let us look at the details of the trait first before answering that question.
The details of Heifer Livability
After Cow Livability was introduced in August 2016, Heifer Livability seemed like a logical follow-up. Thankfully, data was already available through heifer records being coded with so-called disposal codes, or termination codes. Disposal codes were studied from 3.4 million heifer records of all breeds with birth dates between 2009 and 2015. Only data from herds with death losses between 1% and 25% was used. Although data for all breeds was evaluated, only Holstein and Jersey proved to have a large enough dataset for publication of the trait. Figures 1 and 2 present updated numbers of data records per year of recording by breed. Please note that records aren’t included until three years after birth so that the live status of contemporary heifers can be confirmed.
Heifer Livability looks at death between day 2 and when the heifer leaves the herd, or until 18 months of age. Calves dying between birth and 48 hours after birth are included in the stillbirth evaluations.
The predicted transmitting ability (PTA) for Heifer Livability thus represents the expected livability percentage of an animal’s female offspring from 2 days after birth up to 18 months of age in a herd with average management conditions. Larger, positive values are more favorable.
The trait is expressed in percentage points with a mean of 0 and a standard deviation of 0.5% for Holsteins and 0.2% for Jerseys. This effectively means that approximately 68% of Holstein bulls will have a PTA between -0.5% and +0.5%, while 95% of Holstein bulls are expected to fall within the PTA range of -1.0% to +1.0%. Similarly, 68% of Jersey bulls are expected to have a PTA between -0.2% and +0.2%, while 95% of Jersey bulls will range from -0.4% to +0.4%.
The average Heifer Livability was found to be 96% for both Holstein and Jerseys (Figure 2). Therefore, a Holstein bull with a PTA of +1.0% for Heifer Livability would be expected to average 97% of his heifer daughters surviving. Daughters of a Holstein bull with a PTA of -1.0% are expected to have an average Heifer Livability of 95%.
The heritability of Heifer Livability was originally estimated at 0.4% but has recently been updated to 0.7% following some additional research. Young genomic bulls are expected to have reliabilities for the prediction of the trait averaging 46% in Holsteins and 30% in Jerseys. Progeny-tested bulls are expected to have genomic reliabilities averaging 55% in Holsteins and 46% in Jerseys.
There has been a favorable genetic trend for Heifer Livability in recent years, likely because of the selection of correlated traits. Heifer Livability is favorably correlated to yield, Productive Life, Calving Trait Dollars, as well as Early First Calving. Selection on any of these traits will have yielded an indirect selection for higher Heifer Livability.
The buzzkill
So we now have a national trait that producers can use to improve the overall resistance of heifers to causes leading to mortality. Attention for health and well-being of growing calves is important; no one will contest that. Then why isn’t this trait getting more attention?
The honest truth is that the trait has a heritability of 0.7% and a reliability of approximately 40%. In other words, yes, progress is definitely possible, but it will be slow. Very slow. And to put it bluntly: That is just difficult to get excited about.
Now there are other traits of low heritability that we made considerable genetic progress in such as Calving Ease, stillbirth and Daughter Pregnancy Rate, but these have heritability ranging from 3%-8%, and the reliability of these traits are higher because there is more data available.
But for Heifer Livability and other health traits, data collection is a struggle.
And that is the crux of the matter.
We need to make it better
It is extremely likely that the majority of new traits to come will be of low to very low heritability. We have simply run out of traits that are easily measured and on which there is plenty of data. We are now looking at important traits that lead to more sustainability, better health and more efficiency, but those are not easy to record.
The heritability of Heifer Livability does not have to be 0.7%.
A heritability is the proportion of phenotypic variation that can be explained by genetics. But our heritabilities are estimates, they are not static. A heritability definitely has bounds, not every trait is equally regulated by genetics, and many are mainly affected by management. But the heritability estimate is subject to how well we can estimate that phenotypic variation and how well we can subsequently mathematically derive the genetic proportion. And for that, we need good-quality data. Good data leads to better estimates for heritabilities and higher reliabilities for PTAs. And that, in turn, leads to faster genetic progress.
While Heifer Livability is new to the U.S. national evaluations, it is certainly not new to the world. Other countries have launched calf wellness traits before the U.S., and the heritabilities of these traits vary. However, it is interesting to note that countries with government-mandated recording of calf deaths or countries that have historically high emphasis on animal health (such as Germany, Scandinavia and The Netherlands) have higher heritabilities and reliabilities for Heifer Livability. Heritabilities vary between 1%-4%. While the trait definitions vary slightly, the common denominator is that their livability data is more precise.
With 3.4 million records going into heifer livability, it is safe to say that quantity of data is not the issue. It’s the quality of data where the U.S. can still improve.
Ninety percent of usable records for Heifer Livability came from Dairy Records Management Systems (DRMS). This is because of the availability of disposal codes that categorized heifers for the Council on Dairy Cattle Breeding’s (CDCB) analyses. Still, the trait is based on a binary recording of dead or alive. This shows how subject trait development is to the type of data recorded by the U.S. dairy producer and by the availability of herd management software that allows detailed recording.
If we want health traits that are impactful, we must realize that it starts with good data recording on the farm. Heifer Livability could be much more impactful when heifer cause of death is recorded, when she left the herd, why she left the herd and where she went. And this needs to be recorded in a preferably similar and easy way by all major herd management software packages that feed data to the Dairy Herd Information Association (DHIA) and the CDCB.
Yes, we can develop a trait with the bare minimum information, but it should not be what we want as a community.
CDCB is doing their best to continue to improve Heifer Livability through improving data quality. A collaboration between their Pursuing Data Quality committee with the Dairy Records Processing Centers and the USDA Animal and Plant Health Inspection Service (APHIS) led to revised disposal codes that now include more reasons for termination. Each party will work on implementing these codes in programming and available software to make the data flow. But it will still be up to the producer to record it and realize this data is valuable beyond their herd.
The bottom line
While Heifer Livability is important and this newest trait is more than worth selecting for, its current genetic parameters are less than exciting.
Adding another health trait to the national trait portfolio just because it looks good isn’t good enough. It just confuses marketing, selection and mating algorithms. These traits are too important not to have an impact, but for impact, we need better data.
So when you know that all heifers’ lives matter but you are frustrated about the low reliabilities, then think about how we can make that better. You are the customer of both the national evaluations and your herd software provider – and you own your data. Demand that you can record important parameters, start recording your data, allow data release to the CDCB and then request for traits to either be developed or improved. Whether we will get a Heifer Livability or Hoof Health or Birthweight or Milking Speed that we can get excited about depends on exactly that. Good data.
And a bit of buzz.
References omitted but are available upon request by sending an email to the editor.
