In the last several years, the production of a Stayability EPD has been given an increasing amount of attention by the beef industry. Today most breeds are either publishing the genetic predictions or are in the process of developing a genetic prediction for sustained fertility. Despite the flurry of recent activity around the trait, Stayability has actually been around for 25 years, having first been published in 1995 by Red Angus.
The trait Stayability is defined as the probability of a cow that calves as a 2-year-old will remain in the herd calving at 6 years of age or older. This assures that cows in a commercial herd last long enough to be profitable. Its importance is demonstrated in indexes, which have a maternal component, where Stayability is routinely identified as the most economically important trait, and given the most weight in an index.
Stayability’s economic importance was elaborated upon in 2006 by Dr. Dorian Garrick. He listed the advantages of high Stayability cattle including: “Cows that are able to stay and produce longer, affect beef herd productivity through a decreased need and cost for replacement, less calving difficulty, more sale offspring with heavier weights, and greater average weaning weight.” Dr. Matt Spangler of the University of Nebraska also points out that the real costs of cow herd depreciation, and culling cows before they are fully depreciated represents an economical loss to a commercial producer.
However, the problem with the Stayability genetic prediction has always been its low accuracy until a sire is relatively old and generally past its productive lifespan. This is because by the time a sire’s first set of daughters reach 6, the bull will be at least 8 years old. Therefore, the model has been under continuous improvement since first released in 1995.
The first Stayability model designed by Dr. Warren Snelling only needed two data points, which were having a calf prior to 6 years old and then at 6 or older. It was designed this way to account for selective reporting before Whole Herd Reporting (WHR) systems were put in place by breed associations. The heritability of this trait was also low, ranging from 0.07 to 0.20.
The first major improvement in the model occurred in 2007 when Colorado State University designed a multi-threshold model for ages 3 to 6. This took advantage of WHR systems, and helped build accuracy at a younger age. The prediction from each age’s threshold was then indexed to represent a 6-year-old prediction. When genomics became available, the prediction was further enhanced by indexing the prediction with its molecular breeding value. This multi-step process was an improvement over the original model, but multiple indexing processes still had much to be desired.
To solve the shortcoming of these early threshold models, Dr. J. Jamrozik and his colleagues at the University of Guelph proposed using a random regression model. Using random regression that is based on the data from each calving for 3-6 years of age resulted in higher estimates of heritability, which ranged from 0.35 for Stayability to second calving to 0.13 for Stayability to the eighth calving. The correlation between earlier measures of Stayability to 6-year-olds ranged from 0.89 to 0.98, which was a significant improvement over the index model where correlations ranged from 0.18 to 0.47.
Other major advantages of the random regression model are that it utilizes more data including data on ages when there was no variation in a contemporary group, as well as data from cows that missed a calving opportunity but remain in the herd. With random regression, even contemporary groups without variation are still informative because it provides information on the trajectory of a regressions slope and intercept. It also combines data from all ages, and provides the best fit in terms of the correlation between observed and predicted values.
Another major advantage of a regression model is it is now being run as a single-step process where all ages are utilized to project the 6-year-old prediction, as well as incorporating the genomics in the same calculation. The genomics provide the same amount of information as a sire having 15 to 20 daughters.
Stayability will continue to evolve by adding information to the model that will increase the accuracy of the prediction. This includes utilizing disposal codes, as well as data from commercial herds. Spangler points out that commercial herds are generally run in a tougher environment than purebred cattle, and could serve as a valuable source of information to a Stayability prediction.
Having been calculated for the last 25 years, the trait Stayability is well understood by most commercial producers. However, a problem in recent years is various breed associations each coming up with their own prediction of sustained fertility. This includes Gelbvieh’s 30-month pregnancy EPD, and Hereford’s new sustained fertility to 12 years of age.
In the case of Gelbvieh, all the information contained in the 30-month pregnancy prediction is also contained in the breed’s Stayability EPD calculated by International Genetic Solutions. This means utilizing both predictions will actually lower the power of selection.
In terms of Herefords’ sustained fertility EPD approach using random regression, in the 1995 Journal of Animal Science paper unveiling Stayability, the authors warned: “For Stayability evaluations to be most useful, the trait considered should reflect the age that has the greatest economic impact, weighted by the amount of information available to make the predictions. Although extremely old cows may be the most profitable, genetic predictions for the probability of surviving to an extremely old age may be meaningless without sufficient information to make reliable predictions.”
This means the later data in the sustained fertility analysis will likely be low in heritability and contained in small contemporary groups on sires that are 14 years of age or older, which brings into question the value of setting the threshold that late.
With the new Beef Improvement Federation Guidelines being released in the WIKI format, it is the hope that all the breeds will stick with one definition of Stayability. This will produce predictions more easily understood by commercial producers, and provide more useful information in an across-breed analysis.
Stayability has long been the most economically important trait in a suite of EPDs. Luckily, the science of the prediction has continued to improve over time from the very rudimentary prediction developed at Colorado State University in the 1990s to the sophisticated genomically-enhanced random regression model used today. Because of its economic relevance, selection for improved Stayability is having a very positive impact on the industry. — Dr. Bob Hough
(Dr. Bob Hough is the retired executive vice president of the Red Angus Association of America and a freelance writer.)
