“It was meant to be a very quick auditing,” said Shawna Weimer, a graduate student in animal physiology at Iowa State University.

But the plan to assess behavior by approaching piglets was eventually rejected because people observing piglets could not get consistent results.

“There is a need for these measures to be objective, repeatable and reliable,” said Ms Weimer.

Ms Weimer discussed the need for objective behavior measurement in a presentation titled “Willingness to approach (WTA): Live human observation and digital image” at the ADSA-ASAS Midwest meeting in Des Moines, Iowa. In a study of piglets in a commercial nursery, Weimer found that digital imaging can be a more reliable way of assessing welfare and willingness of pigs to approach a human.

To compare human observation with the reliability of digital imaging, Ms Weimer designed an experiment where she assessed pig behavior at the exact same time as a camera took a photo. She entered pens, and then crouched on the ground. She looked down for 15 seconds—a non-threatening posture for pigs—and then looked up to visually assess where the piglets were in the pen. At the moment when she looked up, Ms Weimer also used a remote control to take a photo with a camera fixed several feet above her.

Crouching in the pen, she recorded the number of pigs making physical contact with her, making eye contact with her or doing neither. Of the pigs not making physical or eye contact, Ms Weimer recorded whether they were sitting, standing, laying down, eating, drinking water, or “piling” on each other. In her presentation, Ms Weimer showed a photo of the “piling” behavior, which looked like the pigs climbing each other, perhaps in an effort to avoid her.

“Only piling would be considered a fearful behavior,” Ms Weimer said.

After her live observations, she analyzed the images of the pens for the same behaviors.

Ms Weimer used this method to assess 79 pens of about 23 six-week-old piglets, and she found that the digital image provided a more accurate measurement of pig behavior than human observation. On average, Weimer did not notice two of the pigs that made eye contact. The digital image did catch that behavior.

“Either methodology could be used on farms,” said Ms Weimer. “However, the digital image was more accurate.”

The study showed that the limited field of human vision and time constraint in the pen can affect observations of pig reactions to humans. Ms Weimer said consistent measurements of behavior can help producers reduce pig stress and improve welfare.