Somewhere in the northern Rockies, a collared gray wolf veers off a ridgeline trail it has used for weeks. Nothing visible has changed in the forest. But a mile downslope, a pair of hikers are working their way up a switchback, and the wolf, according to its GPS collar, is already moving faster and burning daylight on a longer detour. Multiply that scene across 37 species and six continents, and you have the central finding of a landmark study published in Science in early 2025: the mere proximity of people reshapes how wild birds and mammals move, even in landscapes that look untouched from space.
The research team, led by movement ecologists working across dozens of institutions, tracked more than 4,500 individual animals (22 bird species and 15 mammal species) using roughly 11.8 million GPS location points collected over six years. According to the paper’s results, more than 65 percent of those species altered their speed, daily range, or habitat use the moment human presence ticked upward nearby. The effect held in intact forests, agricultural edges, and the fringes of towns alike.
How the researchers separated people from pavement
Previous work established that roads, buildings, and cleared land push wildlife around. What made this study different was its ability to separate the effect of people themselves from the infrastructure they leave behind. The team paired animal telemetry with two independent layers of human-activity data: real-time mobile-device pings and vehicle traffic counts at the census-block level, plus satellite-derived indices of permanent landscape modification. That dual-layer design let them isolate the immediate behavioral jolt animals experience when humans draw close, distinct from the slower drag of habitat loss.
The approach also captured human presence in places traditional maps miss. Busy hiking trails, seasonal logging roads, and popular fishing access points all registered as pulses of activity in the mobile-device data, even when satellite imagery showed unbroken canopy. That helps explain a pattern wildlife biologists have long suspected: animals can be disturbed in places that look perfectly wild, because the human footprint, as experienced by a wolf or a golden eagle, is defined as much by where people walk as by what they build.
The researchers also leveraged the sharp drop in human mobility during the early months of the COVID-19 pandemic as a natural experiment, comparing animal movement under reduced and normal activity levels. Institutional summaries of the methods describe this comparison, though the detailed pre- and post-2020 telemetry subsets have not yet been released publicly for independent replication.
Gray wolves feel the squeeze
Wolves drew particular scrutiny because they are large-bodied carnivores with enormous home ranges, exactly the kind of animal most exposed to a moving halo of human disturbance. A complementary study by Dickie et al. (2022), published in the Proceedings of the National Academy of Sciences, used long-term telemetry of wolves, grizzly bears, cougars, and other large carnivores to document measurable changes in carnivore movement near towns, trails, and high-traffic roads. Wolves and other species in that study traveled faster through those zones, a pattern consistent with active avoidance rather than simple habitat loss.
Faster travel over longer detours carries a real metabolic cost. Calories spent dodging hikers or looping around a trailhead are calories not spent hunting, resting, or raising pups. Over seasons and years, that added energy burden can erode reproductive success and shrink the effective habitat available to a pack, even when the physical land remains intact. The research has not yet pinned a precise caloric penalty to the behavior, but the direction of the effect is consistent across multiple studies and species.
Not every pack responds the same way. Some wolves living near settled areas appear to tolerate closer human proximity, possibly because they have learned that certain zones offer reliable food or pose little direct threat. Others show extreme avoidance. The Science dataset aggregates responses across individuals and regions, which means local variation is partly hidden inside global averages. Without finer breakdowns by pack, region, and season, the universality of the pattern for any single population remains an open question.
A signal across birds and mammals, not just carnivores
The 65 percent species-response rate, as reported in the Science paper’s results, is what elevates this from a wolf story to a broad ecological finding. Some species moved less when people were nearby, hunkering down in cover. Others shifted their routes or became more nocturnal. A prior global synthesis in Nature Ecology & Evolution had already documented movement disruptions of 20 percent or more across a wide range of taxa and concluded that direct human activities like recreation and hunting can produce stronger impacts than habitat modification alone. The new Science paper builds on that foundation with a dataset large enough to confirm the pattern at a global scale.
Three tiers of evidence now support the core claim. The strongest is the Science paper itself, with its six-year, 37-species, 11.8-million-fix dataset and the statistical framework linking direct human presence to movement changes. The second is the PNAS carnivore telemetry study (Dickie et al., 2022) supplying wolf-specific behavioral data from a multi-species carnivore tracking effort. The third is the Nature Ecology & Evolution meta-analysis establishing the broader 20-percent-plus disruption baseline. The studies complement each other without being identical in scope or method, and readers evaluating the wolf claim specifically should note that the wolf data originates primarily from the PNAS work.
What the data cannot yet answer
Movement shifts are only one axis of disturbance. The studies do not yet provide robust estimates of how altered travel patterns affect survival, reproduction, or population trajectories over decades. Stress physiology, disease dynamics, and predator-prey interactions can all shift in ways invisible in GPS data alone.
There are also sampling limits. GPS collars tend to go on animals large enough to carry them, living in regions with active research programs. Smaller birds, cryptic mammals, and species in under-studied parts of the world may respond differently. The census-block-level mobile-device counts used to gauge human presence have not been publicly released, so outside researchers cannot yet verify the thresholds that defined “high” versus “low” human activity or test how sensitive the results are to alternative cutoffs.
What this means for trails, timing, and land management
As of June 2026, outdoor recreation continues to push deeper into previously remote terrain across North America, Europe, and parts of East Africa and South America. The research suggests that controlling where and when people move through sensitive habitat may matter as much as preventing physical development. Seasonal trail closures during denning or nesting periods, rerouting paths away from core habitat, and maintaining large refuges where wildlife can move and feed with minimal disturbance all emerge as practical tools supported by the data.
The findings also reframe a common assumption in conservation planning: that if a landscape looks intact on a satellite image, it is functioning as habitat. The evidence says otherwise. A forest crisscrossed by popular trails can impose the same behavioral costs on wildlife as one fragmented by roads, because the animals respond to the people, not just the pavement. For land managers weighing new trail proposals or recreation permits, that distinction is no longer theoretical. It is backed by 11.8 million data points and six years of global tracking.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
