South Africa’s samango monkeys live on what ecologists call a “vertical axis of fear.” They live and eat in trees, and if they climb too high looking for food, they’re liable to get attacked and eaten by an eagle. If they venture too low, it’s leopards and desert lynxes they have to worry about. With every trip for food and every climb up and down the tree, a monkey needs to assess risk and reward—the benefits of eating versus the risk of being eaten. But a new study suggests that it’s not just food and predators that the monkeys take into account when deciding where to eat. They also think about us. 

When scientists study animal behavior, they rely on the idea that they’ll become a neutral presence by not being too intrusive and habituating the animals to the fact that there are humans gawking at them. Once an animal is used to them, they hope, it will act naturally when they’re around. But they also know that that’s not always the case. In Tanzania, colobus monkeys fled whenever researchers studying their chimpanzee neighbors came around. The chimps eventually figured out that they could take advantage of that, and let the humans flush out their prey like hunting dogs for easy capture.

Even human infrastructure is enough to change the way that animals act. When bears recolonized Grand Teton National Park, moose shifted their birthing sites closer to the park’s paved roads, using them as a safe zone where the traffic-averse predators would leave them alone. 

A team of researchers from South Africa, the UK, and the Netherlands have now found that the samango monkeys likewise feel safer when humans are around and adjust their behavior accordingly. The scientists did this by looking at what’s called a “giving-up density” (GUD), the amount of food that a foraging animal will abandon in a given spot. Animals will probably feed more where and when they feel safe, so the GUD should correspond to the perceived riskiness of a feeding site and be lower where it’s safe and higher where there’s a threat. 

The researchers hung plastic tubs full of peanuts at different heights in the trees at a site in South Africa’s Soutpansberg Mountains. The lowest tubs were around four inches off the ground and the highest were 24 feet up, just below the forest canopy. Then they looked at how much of this food two groups of monkeys ate and how much they left behind on days when the scientists followed them versus days when there were no humans around, to get an idea of how their presence impacted the monkeys’ risk assessment and behavior. 

The team found that GUDs differed greatly between “follow days” and non-follow days. Both groups of monkeys ate more from the higher bins when the researchers weren’t around, a sign that ground-level food is a riskier option and big cats pose more of a threat than big birds. On days that the team hung around the feeding stations, though, the monkeys ate more from all the bins, with the biggest difference in the ones closest to the ground. With humans around, the monkeys seemed to feel safer everywhere up and down the trees, but especially near the forest floor. That suggests, the team writes, that they see us as “shields” against predators, particularly terrestrial ones. That makes sense because humans are usually on the ground and not up in the air, but also because leopards in the Soutpansberg area are often targets for poachers and ranchers who see them as a threat to livestock, and are generally wary of humans there. 

That the monkeys can assess how one animal’s presence impacts another, and use scientists and other people traipsing through the forest as human shields, shows how observant and clever they can be. It also made the researchers wonder just how neutral a presence humans are and how naturally animals act when they’re around. Both of the monkey groups they looked at have been regularly studied by scientists for the last few years and, in theory, should be pretty used to humans being around. But even these habituated animals responded to human presence by changing their behavior and feeding habits, implying that some of the observations of animal behavior that scientists have made could be skewed or complicated simply because there was someone there to make them.