AgBioResearch

EAST LANSING, Mich. – Michigan State University (MSU) animal geneticist Juan Steibel is leading a group of researchers to develop methods to breed pigs that are less aggressive. He has received a $980,000, three-year grant from the U.S. Department of Agriculture’s National Institute of Food and Agriculture (NIFA) for this project.

For centuries, pork producers have bred pigs for desirable physical characteristics such as faster growth and improved meat quality. Breeding the animals for more desirable social behavior has proved much more difficult. Aggressive behaviors such as fighting significantly and routinely compromise the welfare and productivity of pigs on farms by causing injuries that have negative impacts on both the health and the growth of the animals.

As Michigan’s pork industry voluntarily transitions from separating pregnant female pigs – a practice that eliminated fighting but has been linked to other health and welfare concerns – to keeping them in larger groups by 2020, reducing aggressive behavior through breeding has become an important priority.

“Keeping pigs in larger social groups solves some problems but creates others,” said Steibel, associate professor in the MSU Department of Animal Science. “Fortunately, we now have the technology to address those new problems.”

Previous research conducted by Steibel and MSU animal behavioral scientist Janice Siegford showed a connection between pig genetics and the tendency toward aggressive behavior. Now, Steibel and his colleagues plan to build on that knowledge by developing techniques using cutting-edge video technology to monitor large numbers of pigs at once and determine which exhibit less aggression.

One of the most challenging aspects in determining which pigs are more inclined toward aggression has been the time-consuming task of observing behavior. In the previous project, a team of trained scientists and students reviewed footage of approximately 1,000 pigs, making observations and identifying which animals were more prone to aggression than others. Though this approach is effective under highly controlled, small-scale experimental conditions, it is too time-consuming to be practical at the scale of commercial pork farms, which average more than 8,000 animals.

To remedy this challenge, the research team will test automated video software capable of observing large numbers of pigs at once and flagging instances of fighting. Though not as accurate as direct observation by a trained scientist, the video will help monitor commercial-scale pig populations.

“The 1,000 pigs we used in our previous study is one of the largest university groups available, and reviewing the footage from it has taken our students the better part of a year,” Steibel said. “If this technology proves reliable, it could allow the pork industry to analyze the behavior of thousands of pigs with minimal supervision, leading to major breakthroughs in pig breeding.”

The research team is working to program existing video software to recognize individual pigs within a group by subtle differences such as the size and shape of their heads. They are also looking at programming the system to identify instances of aggression, such as pigs standing alongside one another and pressing against and biting one another. Researchers at the University of Leuven in Belgium are working on software capable of noting such aggressive interactions.

Another research team, based out of Scotland Royal University College, is integrating thermal imaging into the video technology. When pigs fight and their blood flow increases, their body temperatures rise. This can be captured by infrared cameras. Incorporating this capability into the system will provide yet another way to identify violent encounters.

The MSU researchers, along with colleagues at the University of Buenos Aires, aim to use the data they collect to build new genetic models that will allow breeders to select pigs with characteristics for less aggression and increased performance.

“So far, because data collection has been limited to small numbers of pigs, our breeding models have not been well-developed to incorporate information on their social behaviors,” Steibel said. “Aggression has a substantial impact on pig group health, and we want to incorporate that into breeding models alongside traditional performance measures like weight gain.”

This project will give pig breeders their first opportunity to select pigs for beneficial social behaviors at an industrial scale.

“For the first time, we can add aggression to the suite of breeding characteristics,” Steibel said. “Getting these systems to work well together is very challenging, but now it’s feasible. Ten years ago, I would have called it science fiction.”

This project is also supported by the Michigan Alliance for Animal Agriculture and the National Pork Board.