Sheep Industry News August 2021

ARS Reference Flocks: The Path to Genetic Improvement

TOM MURPHY, BRAD FREKING, JOAN BURKE & J. BRET TAYLOR USDA Agricultural Research Service B reed evaluations and the development and dissemination of composite breeds have perhaps been the most recognized sheep breeding and genetics experiments led by the U.S. De partment of Agriculture's Agricultural Research Service scientists. A key component of both efforts focuses on evaluating aver age trait performance. Averages are certainly important for many aspects of sheep husbandry, but an often overlooked descriptor of performance is variation or how “spread out” individual animals are around the average. Figure 1 (at right) displays “bell curves” for number of lambs born per ewe lambing, ewe breeding weight, and lamb 90-day adjusted weight for purebred Polypay and Suffolk reared at the U.S. Sheep Experiment Station in Dubois, Idaho, and purebred Katah din and Romanov reared at the U.S. Meat Animal Research Center in Clay Center, Neb. The middle of each breed’s curve represents the average and the width represents the phenotypic variation of each trait. To be clear, this is not a formal breed comparison, but it illustrates an impor tant point: while there are clear differences in average performance between breeds, there is substantial phenotypic variation in perfor mance within all breeds. Are there USSES Suffolk ewes that consis tently lamb triplets and USMARC Romanov ewes that consistently lamb singles? It’s rare, but yes. Likewise, only a couple of USMARC Romanov lambs have weighed 70 pounds at 90 days, while this is about par for the course for USSES Suffolk. This might be considered heresy to some, but no breed can excel in everything and each has their strengths and weaknesses. Although combining relative strengths of breeds through cross breeding is extremely valuable for commercial sheep production and will continue to be an important component of USDA ARS genetics research, this article will focus on improving purebred sheep populations. Part of the within-breed phenotypic variation seen in these graphs is due to genetic differences between animals. This is what ultimately matters when making selection and mating decisions in our flocks. The problem is all the non-genetic effects that impact trait performance cloud the ability to see an animal’s genetic merit. Estimated breeding values – which are provided to American sheep producers through the National Sheep Improve ment Program – remove most of this non-genetic haze and enable us to select individuals for their genetic merits. Standard NSIP traits for which EBVs are available are associ

Figure 1

ated with body weight and composition, wool, reproduction and parasite resistance. While these are economically important, they are limited to those that can be easily measured on the farm or ranch (e.g., number of lambs weaned, post-weaning weight, etc.) or inexpensively by experienced technicians (e.g., ultrasound loin muscle depth, fecal egg count, etc.). Furthermore, as one ardent NSIP breeder put it, “we have EBVs for traits that make us money, but we don’t have EBVs on enough traits that lose us money." What about traits like survival, feed efficiency and longevity? Are these not economically important? They are, but they’re just too labor intensive, difficult or expensive to measure at the producer level. Enter the genetic reference flock. Simply put, a genetic reference flock is one that is enrolled in NSIP, acquires representative rams from other NSIP breeders, and quantifies standard and novel traits in their GRF progeny. With strong genetic connectedness between GRF and industry NSIP flocks, new trait EBVs are available for use in selection by all NSIP

20 • Sheep Industry News • sheepusa.org