Tropical trees are dying faster than they ever have on record, and the culprit scientists identified in a 2025 study isn't heat or drought — it's ordinary thunderstorms, now fiercer with climate change, responsible for up to 60% of tree deaths in some regions

The surprising part of the 2025 paper is not that storms can kill trees. Anyone who has stood in a forest after a hard squall has seen broken branches, split trunks and fresh gaps in the canopy.

The surprising part is the scale that Evan Gora and colleagues argue may have been missed. In a 2025 Ecology Letters perspective, the team writes that research into changing tropical forests has focused heavily on drought, heat, fire and rising carbon dioxide, while convective storms have often been left out of the explanation.

Those are not hurricanes or cyclones. They are the ordinary tropical thunderstorms that build quickly, throw down fierce wind and lightning, and then move on. The paper’s argument is that these storms may be doing far more work inside intact tropical forests than climate models and forest carbon studies have usually counted.

A different suspect in tree death

Tropical forests are not simply being lost to chainsaws and fire at their edges. Even intact forests have shown rising tree mortality in some long-term records. That matters because tropical forests store huge amounts of carbon, shape rainfall and support dense webs of life.

The usual suspects have been heat and drought. That focus makes sense. A hotter atmosphere can draw more moisture from leaves and soils, and severe droughts can push trees past hydraulic limits. But Gora’s team argues that storms should be treated as a major direct driver, not just background weather.

According to the PubMed abstract of the Ecology Letters paper, the authors reanalysed the largest plot-based study of tropical forest carbon dynamics and found that lightning frequency, used as an indicator of storm activity, strongly predicted forest carbon storage and residence time. Adding storm activity also weakened the apparent evidence for high-temperature effects in that analysis.

That does not mean heat is irrelevant. It means the answer may be less tidy than a single climate variable. A forest can be hot, dry, storm-battered, or all three.

What a thunderstorm does to a forest

A convective storm kills trees mechanically. Wind can uproot trees or snap stems. Lightning can kill a directly struck tree and damage many neighbours. A falling giant can crush smaller trees beneath it, turning one death into a small local disturbance.

The paper describes storms as killing groups of trees through both lightning and wind. That group effect matters because forest plots do not only lose isolated individuals. They lose canopy structure, carbon and future growth when large trees fall and take surrounding biomass with them.

Earlier work has shown how large the effect can be in a single event. In a 2010 study in Geophysical Research Letters, Robinson Negrón-Juárez and colleagues reported that one Amazon squall line in January 2005 may have killed hundreds of millions of trees across the basin. That was one storm system, not a slow drought unfolding across a season.

Lightning can be similarly concentrated. In a separate 2025 New Phytologist paper, Gora and colleagues found that lightning killed 56 percent of directly struck trees in their Panama dataset, while some unusually resistant trees benefited when lightning removed competitors and lianas around them. The PubMed summary of that paper gives the useful caution: lightning is not only a source of death, but also a force that can shape competition and forest structure.

The 30 to 60 percent estimate

The most attention-grabbing number comes from the team’s estimate of how much mortality storms may already explain. A Cary Institute release on the Ecology Letters paper quotes Gora as saying their estimates suggest storms were responsible for 30 to 60 percent of tree mortality in the past, with that contribution likely rising as storm activity increases.

That should be read carefully. The paper is a perspective and reanalysis, not a direct census of all dead trees across the tropics. It argues that storms are an overlooked driver and that existing evidence is strong enough to require a change in how tropical forest change is studied.

The Ecology Letters abstract says convective storm activity has increased by 5 to 25 percent per decade over the past half century. Extrapolating from historical trends, the authors estimate that storms likely account for about half of the reported increases in biomass mortality across Amazonia, with a wide possible range.

Wide ranges are not a weakness to hide. They are part of the point. Storm damage is hard to detect, highly local and often missed by the monitoring systems that do a better job with drought, temperature and deforestation.

Why storms are easy to overlook

Heat and drought can be measured continuously by weather stations, satellites and climate datasets. Storm damage is messier. A small storm cell can flatten part of a forest while leaving nearby plots untouched. Lightning can kill trees without causing fire. Wind damage may be visible only after researchers enter the forest, and by then the exact storm may be hard to identify.

Remote sensing helps, but not perfectly. Broken crowns, fallen trunks and small gaps under a dense canopy can be difficult to attribute. A forest plot might record a dead tree months later without showing whether the immediate cause was wind, lightning, drought stress, disease, competition or a combination.

That is why the Cary Institute’s Gigante project, described in the same release, is important. It combines lightning detection, drones and field teams to connect tree deaths to the events that caused them. The aim is not just to count dead trees, but to understand how they died.

Why it matters for carbon

Tropical forests have been treated as major carbon stores and, in some periods and regions, carbon sinks. If large trees die faster, that storage changes. Carbon held in wood can return to the atmosphere as dead material decomposes, and the forest’s future capacity to accumulate biomass can shift.

The storm argument therefore matters beyond ecology. If models attribute too much forest change to heat alone, or too little to mechanical storm damage, they may misread which forests are most vulnerable and which tree species are most likely to persist.

That is especially important because storm damage does not necessarily track drought. The Ecology Letters abstract notes that spatial variation in storm activity has weak relationships with drought, meaning a forest can face intense drought stress, high storm activity, or both. Those combinations imply different risks.

They may also imply different management choices. A tree species that tolerates heat and dry spells may still be vulnerable to windthrow or lightning. A forest restoration plan based only on drought tolerance could miss the mature-tree risk that appears decades later, when the canopy is tall enough to take the force of storms.

The ordinary storm becomes less ordinary

The unsettling thing about the study is that it does not point to an exotic new threat. It points to weather that has always existed in tropical forests, but may now be growing more intense or frequent under a changing climate.

That is a different kind of climate signal. Not only hotter days. Not only drier soils. Also harder gusts, more lightning, more broken crowns and more fallen giants after storms that pass quickly enough to vanish from the public record.

The paper’s argument is not that thunderstorms explain everything. It is that leaving them out may distort the picture. Tropical trees are dying faster in many records, and the cause may not be a single stress arriving slowly. It may be a combination of chronic climate pressure and sudden mechanical violence.

Seen that way, an ordinary tropical thunderstorm is not just weather overhead. It is a force moving through the carbon balance of a forest, deciding which trees survive, which gaps open, and how much of the tropical canopy remains standing long enough to store carbon for the future.