Science Not yet confirmed elsewhere 1 publisher2 min readPublished Aug 28, 2026, 3:21 AM PDT

Cellphone data puts human presence alongside roads in explaining where wildlife goes

Weekly GPS from thousands of tracked animals, matched against anonymized phone traces from 2019 and 2020, ties human presence to shifts in range and niche for roughly two-thirds of mammal and bird species. That points management toward access as much as acreage.

The Scientist · Science desk

How we use AI Send a correction

Photograph accompanying Cellphone data puts human presence alongside roads in explaining where wildlife goes — Photo: yale.edu

What happened

UC Santa Barbara, the Smithsonian's National Zoo and Conservation Biology Institute and Yale combined GPS tracking from 37 animal species with US cellphone location data in a paper published in Science.
The analysis rested on weekly GPS records for 4,581 individual mammals and birds across the continental United States, compared across matching periods in 2019 and 2020.
Lockdown policies changed where and how people moved, which let the team separate the influence of people being physically nearby from the influence of permanent changes to the land.
Fifty-seven percent of the species studied were affected by both human presence and landscape modification rather than by modification alone.
Human presence alone was associated with a change in either territory occupied or environmental niche size for 67% of mammal species and 68% of bird species.

Why it matters

constraint A protected-area boundary governs what may be done to a parcel, not how many people walk across it, so for the roughly two-thirds of mammal and bird species whose range or niche tracked presence, buying land leaves that driver in place.
decision Anyone hoping for one disturbance rule per landscape now has to weigh the study team's own conclusion that plans need to get narrower and more species-specific, which means more survey effort behind every acre designated.
constraint According to co-lead author Scott Yanco, the mobility data was opened to researchers during the pandemic and normally stays with private companies, so no agency can yet treat presence monitoring as a standing, budgetable input.

The headline results count species rather than individual animals, and that limits what they can bear. Across 37 species the team tracked 4,581 individuals ^[1]^[2], an average near 124 per species ^[15], but the reported fractions are fractions of species, so each species is worth about 2.7 points of any whole-sample percentage ^[16]. The 57% affected by both presence and landscape modification ^[6] is therefore about 21 species ^[17]. The mammal and bird figures sit on smaller denominators that the release does not state ^[12], which is why I read 67% and 68% ^[7] as a direction of evidence rather than a precise share.

What makes the design worth attention is that roads stayed put in 2020 while people did not ^[3]. Comparing matched weeks of 2019 and 2020 ^[2] holds season roughly constant while human presence swings, which no cross-sectional map of development can do. Earlier work generally leaned on proxies, such as how built up or farmed a place is, or whether it sat under lockdown orders ^[20]. A lockdown flag records that a policy existed; the anonymized phone traces record where people were, at neighborhood resolution, which the team says is the first use of such data on animal movement ^[19].

The interaction is where the ecology gets interesting. The authors report that the effect of people depended partly on how modified the surroundings already were, and that the effect of modification depended partly on how many people were present ^[8]. Mammals were likelier than birds to respond to the two influences together, 67% against 41% ^[9], a gap of 26 points ^[18]. Ruth Oliver of UCSB's Bren School, a co-lead author, frames it as needing information on both presence and habitat change to see the full human effect ^[10]. The university's summary goes further, saying physical presence can reshape where wildlife moves and lives sometimes as strongly as roads, cities and other development ^[21].

The thing this doesn't tell you is whether the animals are worse off. The quantities measured are the area an animal uses and the breadth of its environmental niche ^[5], and both are geometry. A contracted range can mean displacement from good foraging, or an animal that found what it needed closer in, and a polygon does not adjudicate between them. The August 28, 2026 release from UC Santa Barbara ^[13] reports shares of species affected without giving, species by species, whether presence went with wider or narrower use ^[14]. That direction is the condition on any management reading: a seasonal closure aimed at an animal that in fact ranges further when people are around buys nothing.

What to watch

Whether any mobility vendor licenses neighborhood-resolution location data for conservation research outside a public-health emergency.
Per-species direction of effect in the full Science paper: which of the 37 species expanded and which contracted as human presence rose.
The first agency management plan that cites presence effects to set visitor timing rather than parcel boundaries.

Clarity's read

What the record supports and how the coverage leans. The claims behind it follow.

Reality

Evidence58
Adoption: Insufficient
Hype gap+26
Incentives68
Confidence54

Why these scores

Claim ledger

Ranked by verification strength, evidence, and original report placement.

[1]
Researchers from UC Santa Barbara, the Smithsonian's National Zoo and Conservation Biology Institute, and Yale University combined GPS tracking information from 37 animal species with cellphone location data collected across the United States, in a study published in Science.
ReportedSupportedView cited source
[2]
The team examined weekly GPS records from 4,581 individual mammals and birds across the continental U.S. during matching periods in 2019 and 2020.
ReportedSupportedView cited source
[3]
COVID-19 lockdown policies dramatically altered where and how people moved, and that shift gave researchers a chance to separate the influence of human presence from the effects of permanent changes to the landscape.
ReportedSupportedView cited source
[4]
The UCSB release states that wildlife responses to humans differ substantially depending on the species and on how heavily a landscape has already been altered, and that conservation efforts may need to become much more targeted.
ReportedSupportedView cited source
[5]
The researchers examined how human activity affected both the physical area used by each animal and its environmental niche, which describes how an animal uses habitats and resources.
ReportedSupportedView cited source
[6]
Overall, 57% of the species studied were affected by both human presence and landscape modification.
ReportedSupportedView cited source
[7]
Human presence was associated with changes in either the amount of territory occupied or the size of the environmental niche for 67% of mammal species and 68% of bird species.
ReportedSupportedView cited source
[8]
The effect of people on an animal depended partly on how much the surrounding landscape had already been modified, and the influence of landscape modification depended partly on how many people were present.
ReportedSupportedView cited source
[9]
About 67% of mammal species and 41% of bird species responded to the combination of the two human influences.
ReportedSupportedView cited source
[10]
Co-lead author Ruth Oliver, an assistant professor in UCSB's Bren School of Environmental Science & Management, said humans have complicated effects on wildlife, from physical presence to how habitats are reshaped, and that the full impact cannot be understood without information on both.
ReportedSupportedView cited source
[11]
Co-lead author Scott Yanco, a research ecologist at the Smithsonian's National Zoo, said the cellphone data was made available to researchers during the pandemic to help reveal the impacts of COVID-19 shutdowns, and that such data are typically held by private companies, which made this a rare opportunity.
ReportedSupportedView cited source
[12]
The release reports percentages separately for mammal species and bird species but does not state how many of the 37 species were mammals and how many were birds.
ReportedSupportedView cited source
[13]
The release is dated August 28, 2026 and is sourced to the University of California - Santa Barbara.
ReportedSupportedView cited source
[14]
The release reports the share of species affected by human presence but does not state, species by species, whether presence was associated with larger or smaller areas used and niches.
ReportedSupportedView cited source
[15]
The tracked sample averages about 124 individual animals per species.
DerivedSupported1 source— create a free account to open them
[16]
With 37 species in the sample, one species accounts for about 2.7 percentage points of any whole-sample percentage.
DerivedSupported1 source— create a free account to open them
[17]
The 57% of species affected by both human presence and landscape modification corresponds to about 21 species.
DerivedSupported1 source— create a free account to open them
[18]
Mammal species exceeded bird species by 26 percentage points in responding to the combination of human presence and landscape modification.
DerivedSupported1 source— create a free account to open them
[19]
The researchers used anonymized cellphone geolocation data with neighborhood-level resolution; according to the team, this marks the first time such data have been used to examine how human presence affects animal movement.
ReportedInsufficientView cited source
[20]
Studies of interactions between people and wildlife often rely on indirect measures such as urban development, agriculture or, during the pandemic, whether an area was subject to lockdown restrictions, which give clues about human activity but do not directly show how people are moving.
ReportedInsufficientView cited source
[21]
The UCSB release states that human physical presence can reshape where wildlife moves and lives, sometimes as strongly as roads, cities and other human development.
ReportedInsufficientView cited source

Sources

1 independent publisher whose own reporting we read for this story.

sciencedaily.com
1 article · August 27, 2026
Scientists tracked 4,500 animals and found a surprising human effect

NextConfirmed2 publishersDoubly heavy nitrogen molecules expose the nitrogen gas microbes add to water

Ask Clarity

Have a question this piece left open? The answer comes from its reporting and claim ledger — nothing else.

Try

Clarity analysis

How the coverage leans and what the record actually supports. Read the methodology

Perspective Coverage

How much of the coverage speaks to builders, operators and investors. It measures emphasis, not whether the story is right.

Builder

For anyone who builds with movement data, the interesting part of the UCSB account is the join itself: weekly GPS fixes from 4,581 collared and tagged animals lined up against anonymized phone traces at neighborhood resolution, with 2019 as the control year. Yanco is candid that the human side of that join existed only because private holders opened it during the shutdowns. So the method is reproducible in principle and unreproducible in practice, and the release says nothing about vendor, sampling density or how presence was aggregated to weekly units.

Reality

Evidence, adoption, hype gap, incentives and confidence, each read independently from the current record on the scale shown beneath it. How these are measured.

Evidence58: 0100
Strong study, thin retelling
The underlying work is substantial and peer-reviewed — 37 species, 4,581 tracked animals, a matched pre-pandemic year, published in Science. But the version available to readers is a university summary that reports shares of species affected and nothing about how much, in which direction, or with what uncertainty, and it withholds the mammal-versus-bird denominators behind its own headline percentages.
AdoptionInsufficient: Nothing to count yet
Our coverage records no agency adopting the approach, no dataset or code released, no replication and no management decision citing it. The authors recommend species-specific planning; recommending is not uptake, and we decline to score an uptake that has not been reported.
Hype gap+26: −100 understated0 aligned+100 overstated
The headline outruns the numbers offered
'Sometimes as strongly as roads, cities and other human development' is the line doing the persuading, and it is the one number-free sentence in the piece — the statistics report which species were affected, never presence weighed against built development. Add a self-attributed 'first time' and percentages where one species is worth 2.7 points, and the framing sits ahead of what is shown. The gap is stretch, not invention: the sample and the interaction result are real.
Incentives68: 0100
The university narrating its own paper
The sole account is promotional by construction: a Bren School release about a Bren School professor's Science paper, quoting only the two co-lead authors, reprinted without added reporting. Both quotes point toward the same institutional interest in a follow-on research agenda and in the value of the data access. The data holder, meanwhile, is described only as a private company — an omission that serves whoever supplied the traces.
Confidence54: 0100
Credible source, no second look
We are confident about what was claimed and by whom — the release is specific and internally consistent — and much less confident about the parts a reader would act on. One publisher, one institutional origin, no outside ecologist, and a peer-reviewed paper we can point to but not read. Confidence would move on the paper itself or on any independent assessment.

Topics and entities

Follow any of these and your For You feed starts watching them — no settings page required.

Topics

Species-specific conservation planningFollow
Anthropause and pandemic natural experimentsFollow
GPS biologging and high-resolution telemetryFollow
Human presence and wildlife movementFollow
Commercial mobility and location dataFollow

Entities

University of California, Santa Barbara Follow
Smithsonian's National Zoo and Conservation Biology InstituteFollow
Yale University Follow
Science Follow
Ruth OliverFollow
Scott YancoFollow
COVID-19 Bio-Logging InitiativeFollow