Originally Published: https://www.agproud.com/articles/63351-when-it-comes-to-genetic-improvement-phenotypes-are-king

When we look back at the evolution of genetic improvement in the beef industry most all of our major improvements have come relatively recently. Each of the steps in this evolution improved the accuracy with which we make selection decisions. This increase in selection accuracy led to jumps in rates of genetic improvement. From the visual appraisal that drove selection decisions to the current genomic-enhanced EPD calculations, individual animal phenotypes have been absolutely essential. As the industry looks to the next frontier of genetic improvement, our ability to capture important phenotypes will be as critical as ever. The development of new technologies are making the capture of phenotype easier, more precise, and widely applicable across the beef industry.

Historical developments in genetic improvement programs

The history of selectively breeding beef cattle looked largely the same from the dawn of domestication until the 1950s. Only then did widespread performance testing programs begin that allowed us to more precisely measure differences between animals at a level beyond visual appraisal.  These programs aimed to standardize measurements and use contemporary groupings to identify genetic outliers within common management groups. The next major leap in improving selection accuracy came with the advent of expected progeny difference (EPD) calculations from this performance data. The EPD calculations leveraged both the phenotypes collected by performance programs and information from the deep pedigrees kept by breed associations to make more accurate estimates of animal genetic potentials. The subsequent five decades contained incremental and large improvements to EPD accuracy, including the addition of genomic information.

While we continue to make incremental improvements in how we calculate EPDs, I would argue that the next great leaps in beef genetics will come from measuring new phenotypes for things that drive profitability across the industry.

Phenotypes: The currency of genetic evaluations

EPD calculations rely on the widespread collection of phenotypes that capture the traits driving profitability to breeding programs. This can range from calving ease scores to weaning weights to fertility outcomes to carcass quality.  EPDs can be calculated on any phenotype that can be measured on a very large number of animals. More records result in a more accurate EPD, but there is often an inverse relationship between the number of phenotypes collected and the difficulty of measurement. It’s worth noting that data quality matters just as much as data quantity. Inconsistent scoring or measurement error across operations can erode EPD accuracy just as quickly as a shortage of records. Often, we rely on “indicator phenotypes” to ramp up the number of records captured. For example, capturing real carcass records on registered animals is quite difficult, motivating the use of live animal carcass ultrasound as a useful indicator trait for building greater numbers of phenotypic records. It is also important to remember that phenotype collection must remain consistent, even after the number of records is sufficient to calculate an accurate EPD. The models that produce EPDs need to be constantly trained on new phenotypes to maintain their accuracy.

New phenotypes

Genetic evaluations are always looking to develop new EPDs that fill gaps in our genetic understanding of profitability. In recent years, multiple breed associations have added EPDs for structural soundness (e.g., Angus’ CLAW and ANGLE), udder structure (e.g., Hereford’s TEAT and UDDR), and adaptive traits (e.g., Angus’ Hair Shedding and PAP). In each case, standardized scoring systems allowed widespread capture of these phenotypes by individual breeders.

Numerous traits are of interest to phenotyping initiatives moving forward, including more precise methods for capturing the red meat yield of carcasses, understanding bull and cow fertility, capturing health outcomes from calves, and more directly measuring grazing efficiency. All of these sets of traits are heritable and have major impacts on profitability across our industry. Genetic tools that allow for these to supplement our selection decisions would have an enormous impact.

Selection is a complex process that almost always involves multiple traits of interest. The continued addition of new EPDs to genetic evaluations presents challenges with “information overload” as we attempt to balance an already large number of profit-driving traits. Optimally, we do this using selection indexes that weight traits by their economic importance to a particular breeding goal. This means that the addition of EPDs to a genetic evaluation should be accompanied with an understanding of the trait’s economic impact and its integration into appropriate indexes.

Technology and AI in phenotyping

The rapid expansion of technology and AI tools puts our industry on the cusp of major breakthroughs in how we capture phenotypes for hard-to-measure traits. The advent of wearable sensors, computer vision, and herd management technologies now allow us to passively collect animal phenotypes and mobilize them for genetic evaluation without a producer ever having to touch a piece of data. Most of these technologies are designed to aid in management decisions and action, making the passive phenotype collection a valuable by-product. Indicator phenotypes related to animal health, fertility, and grazing behavior that used to be impossible to measure at scale can now be captured automatically on large groups of animals.

Phenotyping in new contexts

One major challenge in genetic evaluations is the sheer diversity of environments in which we observe phenotypes. Beyond natural environmental variation (e.g., rainfall, temperature, elevation), differences in management can mean that different genetic optima exist even between neighboring operations. When EPDs are calculated, the result is a prediction of genetic differences between animals placed into an “average” environment. This “average” environment likely looks quite different from the commercial settings where our seedstock genetics are asked to function. This makes phenotype collection in commercial herds incredibly valuable both to evaluate the efficacy of seedstock EPDs and to add valuable records to EPD calculations.  Additionally, mobilizing phenotypes from feedlots and packing plants could help make tools that fill major gaps in our understanding of post-weaning performance and carcass merit. These traits are difficult to capture in seedstock settings but matter enormously to the commercial producer’s bottom line.

The bottom line is this: the beef industry’s ability to make meaningful genetic progress depends on our collective willingness to measure things. Every score recorded, every weight captured, every sensor reading mobilized into a genetic evaluation moves us closer to tools that help producers make better decisions. Phenotypes have always been the foundation of genetic improvement, and as we look ahead, they remain our most powerful currency.