Since you can't change what you don't measure, the seedstock industry is seeking new ways to measure valuable traits fot commercial cattlemen.
Genetic predictions in the form of EPDs (expected progeny difference) have been fully implemented for 30 years. Centered on growth traits, these predictions have led to tremendous genetic improvements not imagined when the only selection tool was visual appraisal. The industry went on to tackle carcass traits with EPDs, and for the last 20 years, seedstock producers have enjoyed similar genetic progress.
However, many areas of harder-to-measure traits such as reproduction have lagged far behind despite their economic importance, and objectively describing traits like disease resistance and soundness have also been sorely lacking. Genomics have also become extremely important tools to the industry, but have strained our EPD models based on 30-year-old technology, which has not allowed genomics to fully reach the potential that they can play in objective selection. This article will cover the advances in objectively describing fertility and novel traits as well as the new EPD models, which will finally fully leverage the information genomics have to offer. Fertility traits Reproductive traits have been shown to be the most economically-important traits to commercial producers, and the seedstock industry is just beginning to objectively describe reproduction. This is because in order to describe female reproduction, breed associations need to have an inventory-based breed registry which annually accounts for the production of every female in every calving season. This has long been talked about, but implemented slowly by many associations.
Red Angus has been the notable exception going to mandatory total herd reporting in 1995, which featured a spring and fall calving inventory. This allowed the Red Angus Association to release the industry’s first reproductive sire summary in 2002, which among other genetic predictions, featured EPDs for heifer pregnancy and Stayability. Even with this history, Dr. Mark Enns of Colorado State University says, “We have just scratched the surface on reproduction and fitness traits.”
The first female reproductive EPD was Stayability, which is defined as a female already in the cow herd producing a calf at six years or older. The first model was very simplistic with only two meaningful data points; having a calf before six years of age, and then again at six or older. This meant with the Stayability EPD, sires were quite old before high accuracy was achieved, past the sire’s normal productive lifespan. This model has been incrementally advanced in recent years, which has allowed accuracy to build earlier in an animal’s life. However, new models for producing Stayability or longevity EPDs have been sorely needed.
The research community has answered this need by introducing a new statistical model using a procedure called random regression, which will greatly advance the well-understood Stayability EPD or allow the calculation of longevity EPDs. This statistical model tests the reproduction of a cow each year during her productive lifetime and allows for unknown values when inventory-based reporting is not fully implemented. A number of breeds are on the cusp of implementing this new procedure, which will greatly advance the industry’s genetic description of sustained fertility.
International Genetic Solutions (IGS), which calculates multibreed EPDs for 12 breed associations, plans on implementing a multibreed Stayability EPD on the same base and scale in the coming year. Included in IGS’ effort are Simmental, Red Angus, Shorthorn, Gelbvieh, Limousin, Chinana and Maine-Anjou, as well as their Canadian counterparts. Hereford and Angus are looking at calculating longevity EPDs, and Brangus and Beefmaster are considering a Stayability EPD in the future.
Where Angus and Hereford plan on a longevity EPD, and the breeds associated with IGS will continue with the well-understood Stayability EPD, both methods will be based on an economicallyimportant trait that accounts for any reason a cow leaves the herd and/or fails to report a calf. The primary reason will always be reproduction, but can also include production or soundness issues, so they are currently investigating feet and leg scoring and other fitness traits, which they hope to add as correlated traits.
Another relevant fertility trait being pursued is heifer pregnancy. This is described as a heifer that is exposed to become pregnant successfully entering the herd. With the implementation of inventorybased systems, heifer pregnancy EPDs have been implemented by Angus and Gelbvieh following Red Angus’ lead, and it is on the drawing board for other breeds like Hereford. IGS is also working on a multibreed heifer pregnancy EPD and expects to release it in the near future.
As breeds like Angus gather more phenotypes, genomics will play an increasing role in adding information to the accuracy of the trait. For instance, due to the larger phenotypic database in Angus’ latest genomic recalibration, the correlation between genomic information and phenotypic data jumped from .45 to .62, which means genomics will add significantly more accuracy to their heifer pregnancy EPD.
Novel traits
A number of breeds are currently looking at a host of novel traits including soundness and disease resistance. Perhaps none is more economically important than the disease resistance work being conducted by a group of universities.
According to Enns at Colorado State University, bovine respiratory disease (BRD) causes over $1 billion in losses to our industry annually. Luckily, BRD susceptibility is a heritable trait (.18 to .21), so genetic progress can be made on lowering BRD incidence. The genomic companies are also working with the universities to identify gene markers to add to their seedstock panels, which will increase the accuracy of a BRD EPD. The methodology for producing BRD EPDs is also straightforward, so the key to producing an EPD is the collection of phenotypes. A whitepaper has been presented to Beef Improvement Federation (BIF), and proposed guidelines for collecting data will be approved this year. Once the breed associations have the guidelines for collecting data, an EPD and genomic enhancement can be achieved in the next couple of years. To get a jump on the collection of field data many of the breed associations are collecting BRD data in their structured progeny tests on many of their high-use sires.
Feet and leg scoring has been receiving increased interest by many breed associations. For example, Gelbvieh, Simmental, and Red Angus have an active multi-year research project going on with Kansas State University to determine the heritability of the trait and candidate gene markers. Angus is leading the industry in collecting field data and hopes to come out with a genetic prediction as soon as this fall. All of their goals are to add it to their Stayability/longevity models to help add accuracy to the genetic predictions. Other associations are looking at a longer horizon before they produce feet and leg predictions.
Using genetic predictions, teat and udder quality has long been objectively selected for in the dairy industry, so the methodology is well understood. In beef breeds, Hereford has led the industry in the production of these EPDs, and they have found the traits to
be moderately heritable and teat and udder scores to be highly correlated. Red Angus is also among the many breeds currently working on the traits and is in the process of calculating genetic parameters at Colorado State University needed to produce a genetic prediction. Simmental and Gelbvieh are also active in this effort. Breeds hope to incorporate the genetic predictions as correlated traits to add accuracy to their Stayability model.
In the West, certain breeds run at altitude are susceptible to high altitude disease, more commonly known as brisket disease, which can be fatal and may also be associated with feedlot death in cattle nearing market readiness. Colorado State University has been actively engaged in working with Angus cattle on producing an EPD on the indicator trait pulmonary arterial pressure (PAP) because cattle with higher PAP scores are more likely to suffer from brisket disease. PAP is a heritable trait (.25 to .40), so genetic advancement can be made through a genetic prediction. The technology to produce a PAP EPD has been fully developed and as soon as the affected breeds collect field data, EPDs should become available in a timely manner.
Genomics and EPDs
The seedstock industry is in the process of making a quantum leap in methodology on how it incorporates genomics into genetic predictions. Currently the breeds with genomically-enhanced EPDs (GE- EPDs) include Angus, Beefmaster, Brangus, Charolais, Gelbvieh, Hereford, Limousin, Red Angus, Santa Gertrudis and Simmental. The primary method used to achieve GE- EPDs is a two-step method.
Other breeds and universities are working on environmental adaptability traits. Hair shedding is necessary for productive cattle especially in hot environments and for cattle grazing fescue. Angus Genetics Incorporated (AGI) is currently researching hair shedding and has already found it to be very heritable (.42). This means genetic improvement can be made on the trait and EPDs can easily be produced once field data is gathered. Other universities are looking to see if there is any value in regional EPD calculation, and Mississippi State University is investigating the effect of hair color on productivity in hot environments. This whole area of environmental adaptability has significant potential to improve the objective description of cattle.
First a molecular breeding value (MBV) is calculated and then the MBVs are used to enhance the EPDs calculated from the existing EPD models. This is because genomics have been treated as an add-on to how EPDs have traditionally been calculated. This has many drawbacks, but has been the most expedient way to achieve GE-EPDs.
Currently there are four major companies calculating EPDs: Agriculture Business Research Institute (ABRI) in Australia calculates EPDs for several U.S. breeds including Hereford, Salers, and South Devon; AGI, which does Angus, Charolais, and Senepol as well as Canadian Black Angus and Charolais; IGS, which does the genetic predictions for 12 American and Canadian breed associations; and Livestock Genetic Services, LLC (LGS), which calculates genetic predictions for Brangus, Beefmaster, Santa Gertrudis, Braunvieh, and Akauski.
IGS and ABRI achieve GE-EPDs by blending the MBVs into the EPDs post analysis. The major drawback of this method is that it does not allow the genomic information to flow up and down the pedigree to parents and progeny. AGI uses the MBV as a correlated trait much as birth weight is a correlated trait with weaning weight, which allows genomic information to flow up and down the pedigree. This is a step forward, but does not make full use of all the information genomics have to offer.
The answer to these drawbacks is to design completely new models with genomics as a main component. This has already been achieved in dairy, swine, and poultry, and has been implemented in beef by LGS with American (Bos Indicus influenced) breeds. This new “One Step” process is very computationally intensive. With One Step, all the markers from a genomic panel are added to an EPD equation, significantly increasing the size of the mathematical problem to be solved. For each animal you still have all the phenotypic and pedigree data, but add approximately 50,000 markers per animal that is genotyped, exponentially increasing the size of the mathematical problem to produce EPDs.
The One Step method described above was first pioneered in beef by John Genho’s company, LGS. According to Genho, the major advantage of One Step is that through genotypes the actual pedigree relationship becomes known. For instance, theoretically full siblings can be 0-100 percent related according to the way genes are randomly passed from a sire and dam to their progeny. With genomics, the amount of relationship is known, which according to Genho, makes single step GE-EPDs much more precise.
The concept of EPD precision can be demonstrated by a genetic prediction calculated from two contemporary groups of equal size, but one containing biased data and the other with unbiased data. Each will come up with an EPD for the trait in question with the same accuracy since the same amount of information went into the calculation. However, with Charolais.
Hardware hardships
One of the challenges of using these new ways of calculating EPDs is that the calculations are massive, but with new computer hardware this is now possible.
IGS, which calculates GE- EPDs for more breeds than any other company, and ABRI are taking a whole new approach to One Step genomically-enhanced EPDs using a model called BOLT (Biometry Open Language Tools) developed by THETA Solutions, LLC, founded by geneticists Drs. Bruce Golden and Dorian Garrick. Their method is much more computationally intensive, utilizing equations that were only thought to be practical with small data sets and nothing as large as a breed database. Interestingly enough, they solved this problem using processing units designed for computer games—called graphic processing units or GPUs—which have huge processing capabilities. When they stack multiple GPUs, it exponentially increases computing power, but the key is to have software that can use the multiple processors simultaneously to solve an equation. The power of this hardware is demonstrated by being able to calculate traditional EPD procedures in less than 24 minutes instead of the 24 hours IGS now takes. With this processing power these much more intensive computations associated with BOLT can be achieved.
Like the Georgia methodology, the BOLT software takes advantage of the actual genotypic relationship between animals, but also allows markers to be weighted or discarded depending on their effect on a trait, which further improves a genetic prediction’s precision as well as accuracy. In addition, the BOLT software computationally gives a more reliable estimate of accuracy and can even produce accuracies for indexes.
ABRI and IGS plan on implementing the BOLT software this fall. Although it does not have the track record in production like the Georgia model, BOLT is fully developed and tested, and due to the speed of calculation will easily allow for more frequent evaluations, daily if desired. Angus already does weekly analysis, so implementing One Step will not be a concern from a data management perspective. However, IGS plans on going to a more frequent analysis when it implements BOLT, as does Hereford, making database management the step they are working on now. This is especially challenging for IGS, which deals with 12 breed associations running a multibreed breed analysis producing EPDs on the same base and scale for all breeds in order to facilitate planned crossbreeding systems.
Merging and cross referencing these divergent breeds’—and their hybrids’—databases from 12 different breed associations is a tremendous logistical problem. IGS is currently testing a web portal system and universal animal identification system. The EPD analysis will be done weekly, and various breeds can update their EPDs as frequently as they desire. Red Angus and Simmental have committed to weekly runs starting sometime this year while others might choose to have their EPDs calculated less frequently. Hereford at ABRI will also be using BOLT and in the next year expects to move from a monthly to a weekly analysis.
Summary
The future of objective selection is exciting and foretells heights previously unimagined. Traits of significant economic importance related to soundness and fertility are on the cusp of being fully objectively described with EPDs. A genetic prediction for something like bovine respiratory disease will have a huge economic impact on the industry and is just one of a number of novel traits on the threshold of having EPDs. New models will allow the industry to fully leverage all the information genomics have to offer. This will lead to a more profitable and sustainable industry.
Entering A New Era In Objective Selection:
http://www.thebeefblog.com/1112-2222