Abstract Abstract Pig robustness has been an increasing concern in the global swine industry over the last 10 yr. This trend of increased mortality across all phases of production continues to worsen over time despite greater focus on research, management, and other production practices to control it. During this time, dramatic improvements in genetic progress have been realized through the selection of economically important traits. Historically, swine genetic companies have relied on data from their purebred animals reared in high-health and well-managed environments which has proven difficult to provide meaningful insight on variation in robustness traits. More recently, leading genetic companies have invested in large commercial test herds where crossbred pigs and breeding animals are individually identified with pedigree and genomic information. These pigs are reared under challenging and varying commercial conditions and mortality, removals, and reproductive efficiency data are gathered and used to generate crossbred breeding values. This type of data increases the accuracy of genetic estimates as well as creates new traits aimed at improving robustness and ease of management. Further investments have been made in precision livestock farming technologies such as computer vision which is quickly being developed due to the low cost of cameras and the ability to use edge computing to process images in real-time or to transfer images and data to the cloud. These developments are particularly compelling in the genetic improvement space when they can be paired with individual animal identification enabling novel traits to be measured in large populations more consistently and at a reduced cost compared with previous labor-intensive or cost-prohibitive methods. Early examples of this are the use of cameras to score feet and leg soundness in replacement gilts as lameness and associated feet and leg injuries are still one of the largest contributors to sow culls and mortality. This new digital feet and leg soundness scoring method has initially resulted in dramatically greater accuracy and heritability compared with conventional subjective human scoring methods. Other camera-based algorithms have been developed to track individual pigs and assess their daily behaviors and interactions. These developments are key to collecting new phenotypes for the selection of positive behavior traits for genetic selection compared with what was previously possible with human observer or scan sampling methods. Gene editing is another promising technology that has been used in human health and plant breeding but is now being widely researched in livestock. This technology has been used to create the first PRRS-resistant pigs to address one of the most costly diseases in the industry. New sources of data and technologies are key to creating and capturing existing and novel phenotypes on large populations of pigs in commercial environments to aid in the genetic improvement of non-conventional traits.
Read full abstract