Exciting news! The pandas’ new home in Atlanta, Arthur M. Blank Panda Grove, is now open!  

Generic filters
Exact matches only
clock
Today
9:00 am - 5:00 pm
LAST ADMISSION 03:30 pm

Gaze analysis and eye tracking with lions

In our previous research blog on our lions (Lions, construction, and cognitive bias, oh my! (Part 1) – Zoo Atlanta), we shared how our cognitive bias study with Hondo, Hatari, and Azizi required some quick pivoting from scent cues to visual ones, and how we relied on the lions’ gaze to show us which object they were choosing. A clear look toward one shape counted as a choice, even if the rest of the body was angled away. That approach got us through training, but it also left us with a question we haven’t stopped asking: How do we know that what we think we saw is what actually happened?

When you work with animals every day, you develop a feel for them. A keeper can often tell from across the habitat that something is different, and that kind of experience is a big part of why studies like ours are possible at all. But in rigorous science, “the lion looked at me” is an impression, not a measurement. It tells us what an observer believed in the moment, and that is not quite the same as what the lion actually did. A look in the correct direction doesn’t always mean a solid choice. A lion could be looking past the target, watching for our reactions instead of the shape, or responding to a pattern he has learned, like which side could possibly come next. If we want to say something meaningful about how a lion feels during a period of change, we have to know that he is making a definitive and clear choice and not a choice we think he is making. That means we need measurable ways of determining their choice. To study behavior scientifically, we first have to define it in terms anyone could observe and measure the same way.

This is where gaze analysis and eye tracking come in. Instead of relying on a quick read of where a lion seemed to look, these approaches let us measure where his attention goes and how it moves. That gives us details a simple “he looked at the pyramid” would miss, like how quickly he made up his mind or whether he hesitated along the way. Just as important, a measurement can be shared. When we define exactly what counts as a look, someone else can review the same footage and see whether they reach the same conclusion. We are learning how to do this, and we are learning it by hand. Right now, that means going through our session videos slowly, one frame at a time, and placing dots on key points of the lion’s body in each frame. Those dots turn a moving image into positions we can measure, so we can see exactly where each mark was placed instead of relying on a general impression of where he was looking.

Luckily, we aren’t the first to ask this question about animals. Researchers have used eye-tracking equipment with dogs to record exactly where they look (Somppi et al., 2012; Williams et al., 2011). That kind of equipment isn’t always practical, so in many other species, such as fierce lions, scientists use the direction an animal’s head is pointed at as a stand-in for where it is looking. That is an important lesson for us: a head pointed at an object doesn’t automatically mean the animal is looking at it, so we have to test that idea rather than assume it. Marking points on the body as well as the head helps us do that, because we can check whether the two agree. It is slow and humbling, but we are still early in this process and have only begun working through our footage, so we are taking our time and checking ourselves as we go.

While working on learning eye-tracking, one of our favorite parts of this project has been sharing it with high school students who were learning about the scientific method. Our lion study turned out to be a great example of how real research works. We started with a question, designed a method to answer it, hit an obstacle when the scents were simply too exciting, adapted, and are now asking how to measure our results more carefully. The students got to see that science can sometimes completely avoid the plan, and that changing course is part of the process rather than a sign that something went wrong. We hope they better understand that science is about asking good questions, building ways to answer them, and then, most importantly, asking whether those ways can be validated.

We are still early in this process, and there is a lot left to learn, which is part of what makes the work exciting. Our goal hasn’t changed: to understand how Hondo, Hatari, and Azizi experience the changes in their world so we can keep making that world better for them. Thank you for following along, and we look forward to sharing what we learn next.

References

1. Somppi, S., Törnqvist, H., Hänninen, L., Krause, C., & Vainio, O. (2012). Dogs do look at images: Eye tracking in canine cognition research. Animal Cognition, 15, 163-174.

2. Williams, F. J., Mills, D. S., & Guo, K. (2011). Development of a head-mounted, eye-tracking system for dogs. Journal of Neuroscience Methods, 194, 259-265.

 

Alexz A.
Research Associate

Connect With Your Wild Side #onlyzooatl