When wildfires swept through Brazil’s Pantanal wetland in 2020, the flames killed an estimated 17 million vertebrates in a single season. The Pantanal had already been parched by drought, leaving animals with nowhere to retreat. That collision of fire and drought in one landscape is exactly the kind of scenario a new study warns could become routine across more than a third of the world’s land animal habitats within decades.
Published in May 2026 in Nature Ecology & Evolution, the research projects that roughly 36 percent of the land area currently occupied by amphibians, birds, mammals, and reptiles could be struck by at least two types of extreme climate events within the same decade by 2085. The estimate carries a confidence interval of 26 to 45 percent, meaning the true figure could range from about a quarter to nearly half of all terrestrial vertebrate habitat.
Four hazards, mapped together for the first time
The research team at the Potsdam Institute for Climate Impact Research did something most biodiversity studies have not: it layered heatwaves, wildfires, droughts, and river floods into a single global analysis. Previous assessments typically evaluated these threats one at a time, which made it difficult to see where multiple dangers converge on the same species in the same place.
To build the map, the researchers combined species range data for thousands of amphibians, birds, and mammals with the Global Assessment of Reptile Distributions (GARD), a peer-reviewed dataset covering more than 10,000 reptile species. They then overlaid downscaled climate projections from multiple model ensembles. For each grid cell within a species’ current range, the team counted how many of the four hazard types were projected to exceed historical thresholds within a given decade. Any cell where at least two hazards crossed those thresholds was flagged as a zone of compound exposure.
Adding wildfire to the mix proved especially consequential. The Potsdam team identified fire projections as a gap that earlier biodiversity-climate models had largely overlooked. Once fire was included alongside heat, drought, and flooding, tropical and subtropical regions climbed sharply on the exposure scale, reshaping the geography of risk in ways that single-hazard assessments had missed.
The emissions pathway matters
The 36 percent headline figure is tied to a specific future: SSP3-7.0, a scenario used by the Intergovernmental Panel on Climate Change that assumes continued heavy reliance on fossil fuels and limited international cooperation on cutting emissions. It is a high-end trajectory, well above the temperature targets set by the Paris Agreement, though not the most extreme pathway modelers consider.
If the world follows a more aggressive decarbonization path, the picture improves. The Potsdam Institute’s summary of the research states that fast emissions cuts could meaningfully reduce multi-event exposure by 2085, though it does not specify the exact magnitude of that reduction for individual regions or species groups. That gap leaves an important question partly unanswered: how much relief would ambitious climate policy actually deliver for the most vulnerable habitats?
The difference between the low and high ends of the confidence band, 26 percent versus 45 percent, also reflects how climate models handle precipitation extremes and fire weather, two variables with notoriously large regional error margins. In dryland regions and monsoon systems, small shifts in rainfall timing can dramatically alter drought and flood statistics, which feed directly into the compound-risk calculations.
What the models do not capture
Several important dimensions fall outside the study’s scope, and they cut in both directions.
First, the projections use current species ranges as a fixed baseline. In reality, many vertebrates will shift their distributions as temperatures rise and human land use continues to fragment habitat. Whether those shifts will carry animals toward safer ground or into new compound-hazard zones is a question the study leaves open. Dynamic range modeling that accounts for dispersal barriers, habitat corridors, and microclimate refuges would be needed to test whether 36 percent is conservative or optimistic.
Second, the results are aggregated across all four vertebrate classes. Amphibians, which depend on moisture gradients far more than birds or large mammals, may face disproportionate compound risk from overlapping droughts and floods, but the primary paper’s summary statistics do not break out taxon-level exposure. Supplementary materials may contain finer detail, and forthcoming analyses could sharpen the picture for the most sensitive groups.
Third, the models treat overlapping hazards as co-occurring physical stressors but do not simulate ecological chain reactions: vegetation dieback after fire, altered predator-prey dynamics, or disease outbreaks following floods. Those feedbacks could amplify the damage beyond what exposure maps alone suggest. In rare cases, they could also buffer it, as when post-fire regrowth temporarily boosts food supply for certain grazers. The compound-risk percentages should therefore be read as indicators of exposure, not direct predictions of population collapse.
Why compound risk changes conservation math
The practical implication is blunt. Conservation planning that treats heatwaves, fires, droughts, and floods as separate, low-probability shocks is likely underestimating risk for a large share of the world’s land vertebrates. The Pantanal fires offered a preview: when drought and fire hit the same ecosystem in rapid succession, the toll on wildlife far exceeded what either hazard would have caused alone, because surviving animals faced degraded habitat, reduced food, and limited escape routes all at once.
The Potsdam study puts a global number on that pattern. Under a high-emissions future, more than a third of current habitat could face overlapping extremes within the same decade, compounding stress on ecosystems that are already fragmented by agriculture, roads, and urban expansion. Integrating compound-event metrics into protected area design, species recovery plans, and climate adaptation funding could help direct resources toward the regions where multiple hazards are on track to collide.
The full paper is available through Nature Ecology & Evolution. Readers evaluating the 36 percent figure should treat it as a modeled projection built on established climate ensembles and species databases, not an observed outcome. Its strength lies in the peer-reviewed methodology and the multi-hazard framework, which marks a genuine step forward in how scientists assess climate risk to biodiversity. Its limits lie in the uncertainties that all long-range projections carry, and in the ecological complexity that no single model can yet fully represent.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
