By: Kristin Lee, Jasper Munro, Ricardo Ventura, Flavio Schenkel, Gordon Vander Voort, Angela Cánovas
Over the past several decades, the global consumption of animal protein has doubled, and is expected to double again by the year 2050, when the global population reaches at least 9 billion (FAO, 2021). During this time, the demand for animal protein is expected to increase by 70% (FAO, 2021). This presents unique challenges as well as opportunities for beef production. Producers will have to contend with limiting resources (i.e. land, feed), and also be conscious of the environmental impact of their herd, to ensure the sustainability of the industry for the years to come. In other words, producers will need to raise more animals as efficiently as possible. Fortunately, innovative solutions are available to help resolve these challenges.
Traditional phenotyping is based on manual data collection of measurable characteristics such as average daily gain, feed to gain ratio, etc. The traditional phenotyping process is slow, limits the number of attributes and the size of the population that can assessed. This in turn limits the efficiency of matching phenotypic information with genomic data. New technologies, such as high-throughput phenotyping and genotyping, continue to emerge for the rapid and inexpensive collection of genetic data on large populations of beef cattle. For example, live imaging, wearable sensors, and fixed technologies can each be used to quickly measure a variety of economically relevant traits in both confined and grazing environments (Tedeschi et al., 2021), including animal location, movement, and behavior (González et al., 2014; Greenwood et al., 2014), body weight (González et al., 2014; Wang et al., 2006), feed intake (Greenwood et al., 2017, 2014; Wang et al., 2006), temperature (Schaefer, 2010), and methane emissions (Hammond et al., 2016). In addition, high-throughput genotyping methodologies are available at continually reducing costs.
Genotyping usually relies on high-density SNPs (single nucleotide polymorphisms) panels from thousands of performance tested animals (i.e. the reference population) and the selection candidates. To reduce the cost of genotyping for routine application in breeding programs , low-density SNP panels can be used for which per-sample genotyping costs can be a fraction of the cost of full high-density SNP panels, but this reduced genotype density comes at the expense of reduced prediction accuracy in a breeding program. Entire populations of beef cattle can now be genotyped using low (3K to 10K), medium (50K to 80K), and high (300K to 800K) density SNP marker panels, up to whole genome sequencing (WGS). Currently, commercial companies offer genotyping for an average price of $25 per individual depending on the number of SNP markers tested. Genotype costs can be further reduced using imputation, a process where a core subgroup of the population can be genotyped using a high-density panel, then unobserved genotypes can be inferred with high accuracy for animals that have been genotyped with only an inexpensive low-density panel (Berry & Kearney, 2011; Habier, Fernando, & Dekkers, 2009; Li, Sargolzaei, & Schenkel, 2014). Overall, these new technologies provide producers with the tools to improve the genetics of their herds. High-throughput phenotypes and genotypes will not only increase the size of the datasets that will be collected and managed, but it will also increase the computational demands required to analyze the data. Flexible genetic evaluations systems (GES) are needed to convert data into comprehensive results that can be utilized by producers to generate high accuracy breeding value predictions.
The University of Guelph and AgSights have started a collaboration on a major three-year project led by Dr. Angela Cánovas and funded by MITACS (Mathematics of Information Technology And Complex Systems) where AgSights GES will be transformed into a GES that is prepared for the future needs of the Canadian beef industry. MITACS Accelerate is a popular program that partners Universities and Colleges with businesses to solve research challenges with direct application to the industry. Therefore, a collaboration between the University of Guelph and AgSights will aim to update the GES:
1) to be flexible and adaptive in terms of its ability to efficiently evaluate novel and high-throughput phenotypes and genotypes, and
2) to assess the feasibility of including genotype information from purebred, crossbred, or multi-breed beef cattle in a single-step genomic evaluation procedure that combines information from traditional genetic selection (pedigree, phenotypes), with genomic data.
The resulting GES will be integrated into AgSights Go360|bioTrack tool, allowing for the efficient selection of novel and high-throughput phenotypes, in purebred, crossbred, or multi-breed beef cattle. This will provide producers with the tools to improve the genetics of their own herds with the intension of increasing the economic value of Canadian beef and facilitating the uptake of technology by the industry in the future.
The major benefit of implementing new genetic selection technologies into beef cattle production is that it will improve the accuracy of genetic selection, especially when genotypes are used. However, implementation of new technologies has proven difficult, as the population sizes for beef cattle breeds are small in comparison to dairy cattle and other livestock species. Common cattle breeds, such as Angus, Hereford, and Simmental make up 60% of the beef industry in North America, and the remaining 40% is made up of over 80 different breeds (Strauss, 2010). Large and high-quality datasets are needed to generate accurate estimated breeding values. Therefore, the small population sizes that exist in the beef cattle industry often lead to small datasets and low accuracy genetic prediction, which provides little incentive for producers to invest in recording of phenotypes and genotypes. This problem is exacerbated when considering crossbred or multi-breed populations, as animals are more diverse and variable in their genetics, and even larger populations, including all breed types and crosses, are needed for accurate breeding value predictions, compared to highly inbred, purebred lines. Therefore, an accurate, efficient, and easy to use technology could provide beef producers with the ability to begin widespread use of new technologies into their herds. Once implemented, more phenotypes and genotypes will be recorded, resulting in a large dataset including multiple breeds as well as crossbred animals. Over time, continuous feed back will generate larger dataset and provide further improvements to the accuracy of genetic selection and the genetic gain observed in Canadian beef cattle herds.
Kristin Lee, Flavio Schenkel, and Angela Canovas are from the Centre for Genetic Improvement of Livestock (Department of Animal Biosciences, University of Guelph), Jasper Munro and Gordon Vander Voort are from AgSights (Elora, ON), and Ricardo Venture is from the School of Veterinary Medicine and Animal Science, University of São Paulo.
References
Berry, D. P., & Kearney, J. F. (2011). Imputation of genotypes from low-to high-density genotyping platforms and implications for genomic selection. Animal, 5(8), 1162–1169.
FAO. (2021). Meat & Meat Products. Retrieved August 18, 2021.
González, L. A., Bishop-Hurley, G., Henry, D., & Charmley, E. (2014). Wireless sensor networks to study, monitor and manage cattle in grazing systems. Animal Production Science, 54(10), 1687–1693.
Greenwood, P. L., Paull, D. R., McNally, J., Kalinowski, T., Ebert, D., Little, B., … Bishop-Hurley, G. J. (2017). Use of sensor-determined behaviours to develop algorithms for pasture intake by individual grazing cattle. Crop and Pasture Science, 68(12), 1091–1099.
Greenwood, P. L., Valencia, P., Overs, L., Paull, D. R., & Purvis, I. W. (2014). New ways of measuring intake, efficiency and behaviour of grazing livestock. Animal Production Science, 54(10), 1796–1804.
Habier, D., Fernando, R. L., & Dekkers, J. C. M. (2009). Genomic selection using low-density marker panels. Genetics, 182(1), 343–353.
Hammond, K. J., Waghorn, G. C., & Hegarty, R. S. (2016). The GreenFeed system for measurement of enteric methane emission from cattle. Animal Production Science, 56(3), 181–189.
Kriaridou, C., Tsairidou, S., Houston, R.D. & Robledo, D. (2020) Genomic Prediction Using Low Density Marker Panels in Aquaculture: Performance Across Species, Traits, and Genotyping Platforms. Front. Genet. 11:124, https://doi.org/10.3389/fgene.2020.00124
Li, H., Sargolzaei, M., & Schenkel, F. (2014). Accuracy of Whole-genome Sequence Genotype Imputation in Cattle Breed. Proceedings, 10th World Congress of Genetics Applied to Livestock Production.
Schaefer, A. L. (2010). Identifying Efficient Cattle Using Infrared Technology.
Strauss, S. (2010). Biotech breeding goes bovine. Nature Biotechnology, 28(6), 540–543.
Tedeschi, L. O., Greenwood, P. L., & Halachmi, I. (2021). Advancements in sensor technology and decision support intelligent tools to assist smart livestock farming. Journal of Animal Science, 99(2), 1–11.
Wang, Z., Nkrumah, J. D., Li, C., Basarab, J. A., Goonewardene, L. A., Okine, E. K., … Moore, S. S. (2006). Test duration for growth, feed intake, and feed efficiency in beef cattle using the GrowSafe System. Journal of Animal Science, 84(9), 2289–2298.
Dairy and Beef Cattle Production 