GPS Tags Show Barred Owls Commandeer the Old Forest Spotted Owls Need

The work begins with a net that can rise ten meters into the trees. After local sunset, field crews broadcast barred owl calls across steep Klamath and Coast Range forest, wait for a territorial bird to answer, and fit the captured owl with a backpack harness carrying a GPS tag of about 20 grams. The tag is programmed to record one fix every four hours for 61 days. Then the schedule ends, and the same coordinated regional program that supplied the tracking data usually removes the tagged owl. In this study, measurement and management were never separate for long.

That pairing matters because barred owls are no longer a rumor at the edge of the northern spotted owl’s range. Native to eastern North America, they spread west during the twentieth century and now occupy forest across Washington, Oregon, and California. A 2019 synthesis concluded that barred owls exert an “overwhelmingly negative influence” on spotted owls where the species co-occur, a review for managers warned. What remained unresolved was spatial: not whether barred owls hurt spotted owls, but exactly which trees each invader chooses, and how much ground one bird effectively takes out of circulation.

The Tags Turned a Hunch Into a Map

Earlier radio telemetry could follow only coarse fixes, and acoustic monitoring can say a barred owl is present without saying what it uses. High-resolution GPS changed the evidentiary scale across animal ecology by tying movement to remotely mapped habitat, a 2015 Science review noted, and this study puts that capacity inside old-growth structure. The team analyzed 56 owls that met a ten-day minimum, after two of 58 tagged birds dropped out, and the cleaned tracks averaged 294.7 locations per owl across a mean of 56 days. Six stationary calibration tags, including one placed inside a tree cavity to mimic signal conditions for cavity-nesting owls, measured mean horizontal error at 47.2 meters, large enough that the researchers modeled it explicitly rather than pretending the forest floor was a clean grid.

The habitat layers underneath those tracks were built from gradient nearest neighbor maps and spotted owl models long used in Northwest Forest Plan monitoring. The owls did not spread evenly across the options. Their use loaded onto one principal gradient the authors call an old-forest axis, a composite of canopy cover, diameter diversity, stand height, large-conifer density, mean tree diameter, and dominant stand age. Six of those nine structural variables were positive with confidence intervals excluding zero. The dead-wood measures, snag biomass and down wood, pointed the same way but were weaker, and topographic wetness showed no selection at all in this dissected terrain.

The result is not simply that barred owls like big trees. It is that they concentrate where the old-forest community is richest. The same attributes that define late-seral forest also support flying squirrels, tree voles, fishers, Humboldt martens, marbled murrelets, amphibians, and the spotted owl itself. Globally, invasive predators are implicated in 87 bird, 45 mammal, and 10 reptile extinctions, 58 percent of recent extinctions in those groups, a 2016 meta-analysis found, and alien species are the most common threat associated with vertebrate extinctions overall in a companion Red List analysis. The California case is a local version of that larger pattern, sharpened by the fact that the invader is selecting the very structure conservation maps were built to protect.

Selection Peaked Where Spotted Owls Nest

The spotted owl layers made the overlap legible. Barred owls selected mapped spotted owl habitat with an intensity ratio of 1.30, meaning about 30 percent heavier use than non-habitat. Across the four nesting and roosting classes, selection climbed from unsuitable to marginal to suitable to highly suitable forest, where the intensity ratio reached 1.52. Inside an average 6.7-square-kilometer home range sat 3.83 square kilometers of spotted owl habitat and 2.35 square kilometers of highly suitable nesting and roosting forest. Sixty-four percent of the typical territory was spotted owl habitat on paper. Thirty-eight percent was the best of it.

“Every stand of that forest still shows up as spotted owl habitat on our maps,” said Dr. Vitek Jirinec, research scientist at Integral Ecology Research Center and lead author of the study. “But habitat isn’t just trees, it’s trees a spotted owl can actually live in, and barred owls have taken that off the table across much of the landscape. Managing the invader is proving just as critical as protecting the trees themselves.”

The quote draws the study’s central line: structure is necessary, but occupancy is the test. Under the ecological definition the authors use, habitat is the set of resources and conditions that produce occupancy. A barred owl does not have to cut a tree down to subtract the stand from a spotted owl’s usable world. It only has to hold the territory, hunt the shared prey, and dominate the encounter. That is functional habitat loss, a biotic subtraction that a vegetation layer cannot register.

The per-bird footprint makes the subtraction expensive. The same continuous-time movement framework that produced a 6.7-square-kilometer mean in California produced 0.24 square kilometers for barred owls in Louisiana, so each invader here commands roughly 28 times the space it would at home. The comparison is unusually clean because both estimates come from the same analytical approach. It also turns removal math into habitat math. If one resident pair vacates a territory, the forest structure is still standing, and the study estimates that several square kilometers can function again for spotted owls.

Removal Works Like Reopening, Not Like Banking

That hopeful arithmetic already has a policy frame around it. In 2024 the U.S. Fish and Wildlife Service finalized a barred owl management strategy after considering non-lethal alternatives, calling lethal removal in identified management areas the only population reduction method proven to work for reducing barred owls and improving spotted owl populations. The new paper supplies a finer-grained reason such a strategy can reclaim habitat: because barred owls select the same high-quality stands, taking out established birds is not a random subtraction from the landscape. It is targeted release of the most contested forest.

The estimate is deliberately conservative. Of 1,171 barred owls removed during the study interval, the authors scaled habitat release only from the 489 removed males old enough to be likely territorial residents, even though females and younger birds may also reduce pressure. Multiplying by the mean habitat inside a home range yields 1,873 square kilometers of spotted owl habitat made available, with 628 square kilometers marginal, 621 suitable, and 1,149 highly suitable nesting and roosting forest. But removals were clustered and recolonization is quick, so the team also drew equal-area circles around removal sites and dissolved the overlaps. The unique footprint came to 1,851 square kilometers, about 57 percent of the naive additive total. Gains are real, but they are not banked forever; they have to be maintained.

That maintenance problem is the caveat that changes the story. This is not a study in which protecting acreage finishes the job. It suggests a specialist can become conservation-reliant, dependent on continued control of the competitor even inside intact forest. The mechanism could also reach past the owl-owl contest. Barred owls are generalist predators whose diets include salamanders, reptiles, small carnivores, and the small mammals that move fungal spores and nutrients through old forest. Concentrated predation by a new apex consumer is exactly the sort of perturbation the trophic-cascade literature warns can propagate beyond the target species. The study does not measure those cascades. Its habitat results point where to look.

“Spotted owls are the reason this study exists, but they’re part of a much larger old-growth food web, one where every species plays a role in keeping the system intact,” said Dr. Greta Wengert, Co-Founder and Co-Director of IERC and principal investigator on the study. “This means protection of the old-growth biological community requires staying committed to managing this invader for the long haul. This is exactly the kind of science IERC specializes in: rigorous enough for a peer-reviewed journal, specific enough for a manager to rely on for years to come.”

The limit is built into the same data that make the finding precise. GPS locations and derived home-range polygons are not public because they contain sensitive spatial information and could enable inference about illicit trespass cannabis cultivation sites, posing risks to conservation and enforcement work. The forest that made the signal legible also made the raw map dangerous to publish. Summary data sit in the supplement, and de-identified data may move under request and agreement, but the exact tracks stay guarded. For a study about what an invader can find in old growth, that is a fitting edge: even the evidence has to be kept from becoming a map for the wrong kind of use.

FAQ