Rainfall and Soil Phosphorus Control Tropical Forest Recovery After Drought

Tropical forests have long been portrayed as the planet’s most resilient green infrastructure, capable of shrugging off disturbances that would devastate ecosystems elsewhere. But a new study published in Nature Geoscience delivers a sobering and remarkably detailed account of what happens to these forests after severe drought, and it identifies two factors that appear to decide their fate: how much rain a region normally receives, and how much phosphorus its soils hold. Drawing on more than 142,000 individual drought events recorded across the world’s tropical forest belt between 2003 and 2022, an international team of researchers has produced one of the most comprehensive assessments to date of post-drought ecosystem resilience, and the results carry an urgent message for forest managers and climate policymakers alike.

The research, led by Yang Chu of Beijing Normal University together with colleagues from institutions across China, Hong Kong, Australia, Sweden and beyond, tackled a deceptively simple question with sophisticated tools. When a tropical forest is hit by a severe drought, does it return to its previous state, and what determines whether it does? Resilience in this context is not simply a matter of how green the canopy looks. The team quantified it using a metric drawn from complexity science: the temporal autocorrelation of satellite-derived vegetation greenness. In practical terms, they measured how strongly the greenness of a forest patch at one point in time resembles its own greenness in the preceding period. When a forest is healthy and self-regulating, its greenness fluctuates in a relatively independent, noise-like way. When it is stressed and approaching a critical threshold, the system tends to recover more slowly from small perturbations, causing fluctuations to become more correlated with one another. This phenomenon, known as critical slowing down, serves as an early warning signal that an ecosystem is losing its capacity to withstand and rebound from disturbance.

Using the MODIS Enhanced Vegetation Index, a two-band satellite measure of canopy greenness, the researchers computed resilience before and after each of the 142,444 severe drought events in their catalogue. The verdict was stark. In 68.8 percent of the affected areas, resilience declined after drought, and the losses were concentrated in drier tropical environments. Only 20.3 percent of areas showed increased resilience, and these gains were located almost exclusively in moist tropical forests, where abundant rainfall appears to buffer ecosystems against lasting damage. The remaining fraction showed no detectable change. The asymmetry is telling: droughts leave most tropical forests measurably less stable than they were before, and the forests best positioned to gain resilience are those that were already water-rich.

The study also quantified how the character of the drought itself shapes the outcome. More intense droughts, in which water deficits push vegetation closer to its physiological limits, produced pronounced declines in resilience. So did prolonged droughts, which deplete deep soil moisture reserves and force trees to sustain hydraulic stress over months or even years. This dose-response relationship matters because climate projections consistently indicate that tropical droughts will become both more intense and longer-lasting as global temperatures rise. If resilience losses compound across successive events, forests may be pushed incrementally toward thresholds from which recovery becomes slow, partial, or in the worst case impossible. Previous work by some of the same researchers and by other groups has documented declining resilience in the Amazon and rising vulnerability of Southeast Asian forests to environmental stressors, and the new findings extend that picture to the entire pantropical domain with event-level resolution.

What elevates the study beyond a catalogue of damage is its systematic search for the regulators of resilience change. The team assembled an unusually rich set of candidate variables: mean annual precipitation from WorldClim, vapor pressure deficit from the TerraClimate database, terrestrial water storage anomalies, soil phosphorus and nitrogen content, soil texture, species richness, canopy height, wood density, fragmentation and forest structural complexity, among others. Using random forest analysis, a machine learning technique capable of ranking the explanatory power of many interacting predictors, they asked which factors best explained why resilience fell in some places and held steady or rose in others. Two variables rose above the rest, and they operate in complementary ways.

Mean annual precipitation emerged as the single most important regulator of post-drought resilience change. Forests embedded in wetter baseline climates consistently fared better, confirming that long-term water availability acts as a first-order control on how much punishment a tropical forest can absorb before its internal dynamics destabilize. This finding aligns with a growing body of evidence that water availability and its variability shape vegetation resilience at global scales, and it helps explain why the resilience gains observed in the study clustered in moist forest biomes. In practical terms, the wettest tropical forests still possess a hydrological cushion, in the form of deep rooting zones, reliable rainfall recycling and favorable microclimates, that drier forests simply lack.

The second key regulator was more unexpected and, in some respects, more consequential: soil phosphorus. Across forest biomes, soil phosphorus was the most consistent factor mitigating resilience loss, and its protective effect grew stronger along gradients of decreasing precipitation. In other words, in the drier tropical forests where drought damage is most severe, phosphorus-rich soils offered the greatest buffer against the erosion of resilience. The mechanism is rooted in plant physiology. Phosphorus is a critical component of ATP, nucleic acids and the enzymes that drive photosynthesis and carbon metabolism, and many tropical soils are notoriously phosphorus-poor because centuries of weathering and leaching have depleted available pools. Trees growing on phosphorus-rich soils can maintain photosynthetic machinery, repair drought-damaged tissues and rebuild non-structural carbohydrate reserves more effectively after water stress eases, whereas phosphorus-starved forests face a compounded limitation: even when water returns, the biochemical capacity to capitalize on it is lacking. Recent work has shown that phosphorus and potassium availability mediate tropical forest productivity responses to seasonal drought, and the new study scales that insight to the level of whole-ecosystem resilience.

The biome-specific analysis added further nuance. The relative importance of individual regulators shifted across tropical forest types, from humid rainforests to seasonal and dry formations, but phosphorus stood out for the breadth and consistency of its mitigating influence. This has direct implications for how scientists represent tropical forests in Earth system models, most of which still treat nutrient limitation crudely or omit phosphorus dynamics altogether. If soil phosphorus genuinely modulates drought recovery at pantropical scales, then models that ignore it may systematically misjudge how quickly tropical carbon sinks will recover from climate-driven disturbances, and by extension how much carbon the atmosphere will retain in the coming decades.

The authors are careful to frame their findings as guidance rather than prophecy. Because the datasets underpinning the analysis are publicly available, including MODIS vegetation indices, global soil phosphorus maps, canopy height models and wood density databases, and because the analysis code has been released on GitHub, the results can be scrutinized, extended and translated into operational tools. The most immediate application is targeted forest management. In phosphorus-poor, low-precipitation regions, interventions such as protecting soil organic layers, promoting species with efficient phosphorus-acquisition strategies, and reducing additional stressors like fragmentation and fire could help preserve whatever resilience remains. In wetter forests, the priority shifts to preventing the repeated, compounding droughts that could erode even their substantial buffers. The study’s event-level atlas of resilience change effectively hands managers a map of where tropical forests are closest to the edge.

The broader stakes are difficult to overstate. Tropical forests store decades’ worth of global emissions in their biomass and soils, and their capacity to keep absorbing carbon depends on their ability to recover from the droughts, fires and heatwaves that a warming climate is intensifying. By showing that resilience after drought is not a fixed property but a regulated one, shaped by rainfall regimes and soil chemistry, the study transforms an abstract ecological concern into a set of identifiable, potentially manageable levers. It also adds a caution: as droughts grow more intense and prolonged, the mitigating power of phosphorus-rich soils may be overwhelmed, and the 68.8 percent of drought-affected areas already losing resilience could grow. Whether the world’s tropical forests bend without breaking in the coming decades may depend, in large measure, on the rain that falls on them and the ancient, weathered soils beneath their roots.

Subject of Research: Post-drought resilience of tropical forests and its regulation by precipitation and soil phosphorus

Article Title: Precipitation and soil phosphorus regulate post-drought resilience in tropical forests

Article References: Precipitation and soil phosphorus regulate post-drought resilience in tropical forests. (n.d.). https://doi.org/10.1038/s41561-026-02095-x

Image Credits: AI Generated

DOI: 10.1038/s41561-026-02095-x

Keywords: tropical forests, drought, ecosystem resilience, soil phosphorus, precipitation, satellite remote sensing, critical slowing down, climate change, forest management, Nature Geoscience, vegetation greenness, biogeochemistry