A New Test for Tropical Forest Resilience

By: Onrizal, S.Hut., M.Si., Ph.D., Lecturer at the Faculty of Forestry, Universitas Sumatera Utara

A pantropical study has shown that tree species richness at the local scale is strongly influenced by water availability during the driest month. This finding offers practical policy guidance for climate adaptation and disaster risk reduction.

In Indonesia, society is already familiar with the two faces of extreme climate: heavy rainfall that triggers flash floods and landslides, and prolonged droughts that increase the risk of forest fires and water crises. What is often overlooked, however, is a crucial question that should become the starting point of policy-making: what indicator can most quickly signal that a landscape is becoming fragile before disaster occurs?

A recent collaborative intercontinental study published in the prestigious peer-reviewed journal National Science Review provides a clearer compass. Researchers trained a spatial random forest model using a network of 429 permanent one-hectare plots across three major tropical regions in Asia, the Americas, and Africa, with 24 environmental variables covering climate, soil, and topography. The results revealed that the combination of environmental variables explained approximately 86 percent of variations in local tree species richness (alpha diversity) across plots.

This finding is important because it demonstrates that tree diversity at the site level is not merely determined by “geographical fate.” Instead, it is strongly influenced by environmental conditions that can be mapped and, to some extent, managed. In other words, climate change and landscape fragmentation are altering the ecological foundations that shape tree diversity, with consequences for ecosystem functions such as carbon storage, water regulation, soil stability, and post-disturbance recovery.

The study’s strongest finding was the role of water availability during peak drought periods. In the non-spatial random forest analysis, precipitation during the driest month emerged as the single most important variable for predicting tree species richness, followed by soil pH. Researchers also found that the interaction between precipitation in the driest month and the proportion of silt particles in fine soil fractions significantly contributed to species diversity. When spatial autocorrelation was considered, indicators of water deficit and humidity volatility became even more dominant, including annual relative humidity range, precipitation seasonality (bio15), potential evapotranspiration (PET), and climate moisture index.

In simpler terms, the scientific message is straightforward: tropical forests with the highest species richness are generally located in areas that maintain sufficient water supply during the driest month and avoid extreme humidity fluctuations. Conversely, when dry seasons become harsher and humidity conditions become increasingly unstable, local tree diversity tends to decline.

For the public, these findings translate into one simple reality: planting trees alone is not enough. Law enforcement alone is not enough. Managing only one component without addressing others often produces superficial improvements. Biodiversity resilience—and landscape resilience against disasters—requires integrated interventions that simultaneously consider water, soil, topography, and disturbance management. This serves as an important note for Indonesia’s forestry policies related to forest and degraded land restoration.

So, what is the important implication of this study for climate change policy and disaster resilience in Indonesia?

First, the driest month should be treated as a sharper risk indicator than annual averages. Many planning systems still rely on average figures, even though ecosystems and disasters are often determined by extremes. The driest month represents the harshest test for trees: water stress intensifies, fire risks increase, regeneration weakens, and the effects can extend into the following rainy season through higher surface runoff in degraded landscapes. Ignoring the driest month means failing to recognize the ecosystem’s most vulnerable point.

Second, moist refugia must be protected as natural infrastructure. At the landscape level, humidity and topography-related variables play major roles. This means riverbanks, foothill forests, functioning peatlands and wetlands, and topographic mosaics that maintain microclimates should be treated as dual strongholds: strongholds for biodiversity and for disaster prevention. Protecting refugia is not only about conserving species, but also about reducing wildfire risks during extreme droughts and accelerating recovery after disturbances.

Third, restoration strategies must shift from simply “planting trees” toward restoring water and soil functions. Soil conditions—especially pH and physical properties—strongly influence restoration success at the site level. Effective restoration should begin with soil diagnosis and hydrological recovery through improved drainage systems, riparian rehabilitation, and erosion control. Success should be measured through resilience indicators such as stable site moisture, reduced fire hotspots, improved tree survival and growth, and strengthened natural regeneration, rather than solely by the number of trees planted.

Fourth, maps of “dry season risk and moist refugia” should become official planning layers. Central and local governments can develop risk maps based on indicators such as the driest month and water deficit, then integrate them with land cover, accessibility, and disturbance history. These layers should become part of spatial planning, licensing systems, and field preparedness strategies during peak drought periods.

Finally, policies must include measurable targets within clear timelines. Within 12 months, operational risk maps should be released, drought preparedness SOPs implemented, and pilot hydrology-based restoration projects initiated in high-risk landscapes. Within 24 months, risk layers should be integrated into spatial planning and licensing systems, while habitat connectivity and riparian corridors are expanded. Within 36 months, impact audits should evaluate whether repeated fires have decreased, whether restored sites show greater moisture stability, and whether biodiversity indicators demonstrate consistent recovery.

The most important lesson from this pantropical study is simple: tropical forests provide early warning signals, and that signal is the driest month. If Indonesia seeks disaster resilience that truly benefits society, infrastructure, and the economy, then policies must strengthen ecosystems precisely when they are most vulnerable—before fires ignite, before floods arrive, and before losses escalate.

This article was previously published on Kompas.id
Photo source: Dinos Grow