A puma in Patagonia shortens its nightly patrol. A wild boar in Poland sticks closer to the forest edge. An elephant in Kenya reroutes around a road it crossed freely a decade ago. None of these animals has lost its habitat. The forest is still standing, the savanna still open. What changed is simpler and harder to fix: people showed up.
A growing body of peer-reviewed research, anchored by a landmark GPS tracking study published in Science in 2018 and reinforced by two major analyses released since, shows that wild mammals across the globe are compressing, rerouting, or timing their movements around human activity. The pattern holds on every inhabited continent and across dozens of species, from large carnivores to small herbivores. And it emerges not only where cities sprawl or forests fall, but in landscapes that still look wild on a satellite image.
What three major studies found
The foundational dataset comes from a global GPS tracking synthesis led by Marlee Tucker of Radboud University and published in Science. Tucker’s team compiled movement data from 803 individual mammals spanning 57 species across multiple continents, using GPS collar records gathered over several years. Their central finding was stark: terrestrial mammals living in areas with a higher human footprint traveled roughly one-half to one-third the distance of those in low-footprint zones. The reduction held after controlling for body size and diet, suggesting it was not simply a quirk of which species happened to live near people.
A 2021 meta-analysis in Nature Ecology & Evolution, led by Tim Doherty of the University of Sydney, approached the question from the disturbance side. Doherty’s team reviewed 208 studies covering 167 species to assess how animals responded to specific types of human activity, including recreational hiking, vehicle traffic, hunting, and infrastructure. In the majority of recorded cases, movement changed substantially. Both lethal threats like hunting and non-lethal encounters like a jogger on a trail triggered measurable behavioral shifts. The takeaway was that animals react to people, not just to bulldozers or chainsaws.
A third study, published in Science in 2025 by Mark Wilber and colleagues, added a critical layer of precision. Using fine-grained proxies for human presence, including anonymized mobile-device signals and vehicle-activity data measured at small spatial scales, the team tracked thousands of individual birds and mammals across dozens of species in the United States. Their key contribution was separating the effects of direct human presence from broader landscape modification. Both exerted independent pressure on animal movement. Even within otherwise intact habitat, a spike in human foot traffic correlated with animals shifting when and where they moved.
Why reduced movement matters
A mammal that moves less is not necessarily a mammal at peace. Daily and seasonal movements serve essential functions: finding food, locating mates, reaching water, avoiding predators, and migrating to seasonal habitat. When those movements shrink, the consequences can cascade.
Tucker’s GPS synthesis found that compressed ranges were most pronounced in landscapes with roads, agriculture, and urban edges, places where animals may be avoiding perceived danger rather than thriving in a smaller area. A deer that no longer crosses a road to reach a richer foraging patch may survive the season but enter winter in poorer condition. A carnivore that avoids a valley during daylight because of hikers may burn more energy hunting at night, when prey detection is harder.
Doherty’s meta-analysis documented the flip side: animals that increase movement in response to disturbance. Fleeing farther, more often, or at unusual times carries its own energy cost. Repeated flight responses can reduce time spent feeding, disrupt parental care, and push animals into unfamiliar terrain where predation risk is higher. Whether an animal compresses its range or expands its flight distance, the common thread is an energy budget thrown off balance by the presence of people.
The Wilber et al. study underscored that these effects are not limited to heavily developed landscapes. Animals in national forests and protected areas also adjusted their behavior in response to fluctuations in visitor numbers. That finding has direct implications for conservation strategies that rely on protected-area boundaries as a primary tool: if the boundary keeps out bulldozers but not hikers, the animals inside may still be paying a behavioral tax.
What researchers still do not know
The three studies converge on a clear pattern, but several questions remain open. The most consequential is whether movement changes translate into population-level declines. A species that shifts its activity to nighttime or avoids a trail corridor may persist for years before reduced reproduction or survival becomes detectable. The existing research documents behavioral change, not demographic collapse, and bridging that gap requires long-term monitoring that most study sites lack.
Thresholds are another blind spot. The data show that animals change their movements as human influence rises, but they do not clearly define how much activity triggers a shift, or whether there is a level of use that wildlife can absorb without measurable cost. For land managers weighing trail access against wildlife protection, that missing threshold is the most practically important unknown.
Species-level variation also needs more resolution. The global GPS synthesis and the meta-analysis both report aggregate patterns, but the degree to which individual species drive those patterns is less clear from published summaries. Large carnivores, which require vast ranges and are sensitive to disturbance, may account for a disproportionate share of the movement reduction. If so, the management response would differ from a scenario in which moderate behavioral shifts are spread evenly across dozens of species.
Finally, the welfare question is unresolved. Reduced movement could reflect efficient adaptation to a constrained space, or it could reflect an animal that has effectively given up access to resources it needs. Distinguishing between the two requires pairing movement data with physiological measures like stress hormones, body condition, and reproductive output, work that is underway in some systems but far from comprehensive.
What this means for trails, parks, and planning
For anyone involved in land management, the practical signal from these studies is hard to ignore. Protecting habitat from development is necessary but, based on the available evidence, probably not sufficient. The physical act of people moving through wild places carries its own biological cost, one that is now documented across continents and taxonomic groups.
Some parks and reserves have already begun responding. Seasonal trail closures during calving or denning periods, visitor caps in sensitive watersheds, and time-of-day restrictions on backcountry access are all tools that align with what the research suggests. As of June 2026, however, most protected areas worldwide still manage for habitat integrity without explicitly accounting for the behavioral effects of visitor traffic.
The Wilber et al. study’s use of mobile-device data points toward a future in which real-time human-activity monitoring could inform adaptive management: opening or closing trails based on actual visitor density rather than fixed calendars. That approach would require both technological infrastructure and political will, but the scientific foundation for it is now in place.
What the research makes clear is that the line between “wild” and “disturbed” is not drawn by fences or land-use maps. It is drawn, daily, by the movements of people. And the animals have already noticed.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
