A Hidden Climate Threat Is Putting Range-Shifting Species at Greater Risk of Extinction
Species are moving as the climate changes, but the tools used to judge whether they are threatened may be failing to follow them. A study by ecologists Richard Keuth, Stefan A. Fritz and Dana Zurell warns that models supporting Red List assessments can underestimate the climate-related extinction risk of species whose geographic ranges are shifting. The problem is not simply that warming makes habitats disappear. It is that many species must track suitable conditions across landscapes that may be fragmented, occupied by competitors, or changing faster than populations can respond. When these complications are simplified or omitted, a species may appear safer on paper than it is in the real world.
The International Union for Conservation of Nature’s Red List is one of the most influential systems for evaluating extinction risk. Its assessments help governments, conservation organizations and researchers decide which species require urgent protection. A central source of evidence is the species distribution model, a computational tool that links observations of a species to environmental variables such as temperature, rainfall, elevation and vegetation. The model then estimates where conditions are suitable now and how much suitable area may remain under future climate scenarios. These projections are powerful, but they depend on assumptions about how species move, adapt and interact with the environments they are entering.
The new research focuses on “range-shifting species,” animals, plants and other organisms whose distributions are changing as climate conditions move across the landscape. In a warming world, many species are expected to shift toward higher latitudes or elevations, following cooler conditions. A mountain species may climb upslope, while a temperate species may expand toward the poles. At first glance, this movement can look like a natural escape route from climate change. However, a projected gain in suitable climate does not necessarily mean a viable population will survive there. A location can have the right temperature but lack food, nesting sites, pollinators, shelter or the ecological relationships that allow a species to reproduce.
This distinction is critical because conventional distribution models often treat climate suitability as a close approximation of a species’ future habitat. They may calculate whether temperature and precipitation fall within the range currently associated with the species, then estimate how much of that climate space will remain available. Such models can identify broad patterns of exposure, but they may not capture the biological costs of movement. Colonization takes time, dispersal routes can be blocked by roads or farms, and newly suitable areas may be separated from existing populations. A species can therefore possess a large amount of theoretically suitable future habitat while lacking a realistic pathway to reach it.
The researchers argue that these limitations can produce a systematic bias in Red List assessments. If a model allows a species to occupy every future location with favorable climate conditions, it may predict range expansion or only modest decline. The assessment could then assign a lower level of concern than would be justified by the species’ actual prospects. This is particularly important for organisms with limited dispersal, specialized habitat requirements or fragmented distributions. The danger is greatest when climate suitability moves faster than populations can track it, creating a widening gap between where conditions appear favorable and where the species is actually present.
Technical details inside the models can make that gap difficult to see. Distribution projections commonly use climate data averaged across relatively large grid cells, while populations experience conditions at much finer scales. A shaded forest, a wet depression or a north-facing slope may remain cool even when the surrounding landscape becomes warmer. Conversely, a model may classify a broad region as suitable even though it lacks the microhabitats required by a species. Models also differ in whether they include dispersal limits, habitat connectivity, land-use change and uncertainty in future climate trajectories. Each decision can alter the projected range and, ultimately, the apparent level of extinction risk.
The study highlights another challenge: the future is not only a map of temperatures. Species do not respond to climate variables independently of one another. Predators, prey, parasites, competitors, pathogens and mutualistic partners are also shifting their ranges, sometimes at different speeds. A plant may reach a newly suitable climate zone but fail to establish because its pollinator has not arrived. A bird may move northward only to encounter unfamiliar competitors. An insect may gain climatic space while losing the host plants on which its larvae depend. These ecological interactions can transform a seemingly favorable destination into a demographic dead end, yet they are rarely represented in standard Red List modeling frameworks.
The consequences extend beyond individual species assessments. If climate-driven risk is consistently underestimated, conservation planning may prioritize places that are already too late to rescue populations while overlooking climate refuges, migration corridors and transition zones. It may also encourage a false sense of security around species whose modeled ranges are projected to remain large. The authors’ message is not that species distribution models should be abandoned. Rather, models should be interpreted as structured estimates of potential habitat, not direct forecasts of population survival. Assessments become more informative when they distinguish between climatic suitability, accessible habitat and locations where populations can maintain positive growth.
Improving those assessments will require more biologically realistic modeling and better monitoring. Future approaches can combine climate projections with dispersal distances, demographic rates, habitat fragmentation, land-use scenarios and data on ecological interactions. Dynamic models that track population growth and movement may reveal risks hidden by purely correlative methods. Repeated field surveys, genetic studies and automated observations can help determine whether species are actually colonizing newly suitable areas or merely disappearing from their historical ones. Conservation strategies may then shift from protecting isolated patches to maintaining connected networks through which species can move, while also preserving refuges where local conditions remain stable.
The broader warning is urgent because climate change is converting extinction risk into a moving target. A species can appear resilient when judged by the amount of future climate space available, yet remain vulnerable if it cannot reach that space, reproduce there or rebuild its ecological relationships. Red List categories influence action, funding and public attention, so underestimating risk can delay intervention during the narrow period when recovery is still possible. By showing how range shifts can expose weaknesses in current assessment models, Keuth, Fritz and Zurell call for extinction-risk evaluations that treat movement not as an automatic escape from climate change, but as a difficult biological process with limits, costs and failure points.
Subject of Research: Climate-related extinction risk in range-shifting species and the limitations of models used for Red List assessments.
Article Title: Models used for Red List assessments underestimate climate-related extinction risk of range-shifting species.
Article References: Keuth, R., Fritz, S.A. & Zurell, D. Models used for Red List assessments underestimate climate-related extinction risk of range-shifting species. Nat Ecol Evol 10, 1501–1510 (2026). https://doi.org/10.1038/s41559-026-03125-y
Image Credits: AI Generated
DOI: https://doi.org/10.1038/s41559-026-03125-y
Keywords: climate change, extinction risk, Red List assessments, species distribution models, range shifts, biodiversity loss, conservation biology, habitat connectivity, ecological forecasting, climate refugia
Tags: climate change and extinction risk assessmentClimate Change Impactconservation assessment challengesecological modeling shortcomingshabitat fragmentation and climate changeimpact of landscape changes on species survivalrange-shifting species extinction riskRed List assessment accuracyspecies distribution modeling limitationsspecies movement and climate adaptationthreats to species tracking climate shiftsunderestimating climate-related threats
