When hikers hit the trails in Rocky Mountain National Park on a busy summer Saturday, the elk do not simply stand there and watch. Many pull back into tree cover earlier than usual, cutting short their time in open meadows where they feed. Others drift farther from established paths, burning extra calories to find quieter ground. These are not random reactions. They are part of a pattern that now appears to stretch across at least 50 countries and dozens of species, according to one of the largest wildlife-tracking studies ever conducted.
Published in Science in May 2025 (DOI: 10.1126/science.adq3396), the study was led by Roland Kays of North Carolina State University and the North Carolina Museum of Natural Sciences, drawing on the Movebank animal-tracking database and a collaboration of more than 50 research organizations worldwide. Over six years, the team collected roughly 11.8 million GPS location points from more than 4,500 individual birds and mammals spanning 37 species. Their central finding: more than 65 percent of those species shifted their behavior when the number of people nearby increased. Some animals shrank their active hours and hid. Others roamed farther, expanding their home ranges as if searching for somewhere humans were not.
What makes this study different
Wildlife biologists have long known that roads, fences, and buildings fragment habitat. What this study isolates is something subtler: the effect of people simply showing up. The research team paired GPS animal tracks with a novel metric of human presence, using counts of mobile devices and vehicles recorded per U.S. census block on a given day. That approach let them measure fluctuating, real-time human activity rather than relying on static proxies like road density or the footprint of a parking lot.
The distinction matters. A trailhead exists 365 days a year, but the crowd that fills it on a July weekend is absent on a rainy Tuesday in November. By controlling for permanent infrastructure, the researchers could ask a sharper question: does the physical presence of people, independent of what they have built, change how wild animals move?
The answer, across most species in the dataset, was yes. The behavioral responses fell into two broad categories. Some species, particularly those that rely on cover and are sensitive to disturbance, reduced their activity or retreated on days when human counts climbed. Others expanded their daily travel distances, as though compensating for lost access to preferred feeding or resting areas. Both strategies carry costs. Hiding means less foraging time. Roaming farther burns energy and can push animals into unfamiliar territory where predation risk or vehicle collisions increase.
Supporting evidence from other research
The GPS findings do not exist in isolation. Camera-trap studies have documented similar patterns at finer scales. Research in urban-adjacent wildlands has shown that mammal detection rates can drop sharply on days with higher visitor counts, following a dose-response curve: more people, fewer animals detected. A separate global camera-trap analysis covering 163 mammal species, published in Nature Ecology & Evolution, found that the dramatic reduction in human movement during COVID-19 lockdowns in 2020 corresponded with measurable shifts in when and where mammals were active, a phenomenon researchers dubbed the “anthropause.”
Experimental work adds another layer. U.S. Forest Service researchers have demonstrated that recreation-related human presence alone alters bird behavior even when the surrounding habitat structure stays the same. Birds in those studies displayed anti-predator responses, flushing earlier, calling less, abandoning nests more frequently, that closely resembled their reactions to actual predators. The implication is that many wild animals process a passing jogger or a chatting family the same way they process a coyote: as a threat that demands an immediate energy expenditure.
Where the picture is still incomplete
The study’s scope is impressive, but several questions remain open. The 65 percent figure applies to the 37 GPS-tracked species, and the published summaries have not yet broken down response rates by taxonomic group or body size. Whether small songbirds reacted at the same rate as large ungulates, or whether carnivores proved more sensitive than herbivores, is not fully detailed in the materials available as of June 2026.
There is also a geographic nuance worth noting. The GPS tracking drew on data from sites across 50 countries through the Movebank collaboration, but the mobile-device metric for measuring human presence was calibrated using U.S. census blocks. How human activity was quantified in other countries, and whether the 65 percent figure holds uniformly outside North America, deserves closer examination as the full dataset is analyzed.
Causal interpretation requires care, too. The statistical models controlled for built infrastructure, but in practice, trails and the hikers who use them are hard to separate. Animals living near a popular path experience the physical trail, the noise, the scent, and the visual stimulus of moving people all at once. Landscape-scale analyses of human disturbance have noted that built features and active use co-occur so tightly that even sophisticated models can struggle to assign effects cleanly to one variable or the other.
Perhaps the biggest gap is fitness data. A behavioral shift is not automatically a population decline. Some species may tolerate or even benefit from human proximity if it suppresses predators or opens new food sources. Research on flight initiation distances has found that wildlife in heavily visited areas sometimes develops greater tolerance, allowing closer human approach before fleeing. Whether the species flagged in this study are adapting successfully or slowly losing reproductive ground is a question the current data cannot resolve, because it lacks direct measures of survival, birth rates, and long-term population trajectories.
What land managers and visitors can do with this
For anyone who manages public lands or simply enjoys them, the practical message is clear: the volume and timing of human visits matter independently of whether a trail or road exists. Seasonal closures, time-of-day restrictions, and visitor caps are tools that address the presence variable directly, giving animals predictable windows of lower disturbance. Trail design and development footprint still matter, but they are not the whole story.
The study also highlights a monitoring gap. Most protected areas track visitor numbers at entry gates, if at all. Few pair those counts with real-time wildlife movement data. Integrating GPS tracking, camera traps, and acoustic sensors with visitor-use records would let managers identify which species are most sensitive, pinpoint the times of day when disturbance hits hardest, and set evidence-based thresholds rather than relying on rough precautionary estimates.
Threshold data is exactly what is missing right now. The study describes animals responding along a gradient of human presence, but it does not yet specify tipping points. Can a trail handle 50 visitors a day without measurable wildlife displacement but not 200? That kind of number would transform how agencies like the National Park Service set permit limits, and it is the logical next step for this line of research.
Why presence alone is a conservation variable
National parks in the United States set visitation records in recent years, with the National Park Service reporting more than 300 million recreation visits annually. Global nature-based tourism has followed a similar upward curve. Against that backdrop, the finding that simply being near wildlife, not building anything, not poaching, not clearing forest, reshapes animal behavior across most studied species is a significant addition to conservation science.
It reframes a familiar debate. Conservation policy has historically focused on habitat loss, poaching, and pollution as the primary threats to wildlife. This research suggests that even well-intentioned, low-impact recreation exerts a measurable biological cost. The animals are not waiting for a bulldozer. They are already responding to the footsteps.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
