Water conservation across the Cerrado biome

Context

The Cerrado is the most biodiverse savanna on Earth and the headwater of three of South America’s largest river systems — São Francisco, Paraná, and Araguaia. Its hydrology is governed by a sharp seasonal regime: a concentrated wet season followed by a long dry period in which rivers depend almost entirely on groundwater discharge to keep flowing. That mechanism hinges on how much rainfall actually infiltrates into the soil rather than running off. Land use is the control valve of the system.

Instituto Cerrados brought me in precisely because this wasn’t a typical ecological study — it required a hydrological framework grounded in topographic controls on infiltration, which sits at the core of my doctoral research on TOPMODEL. My role was to lead the project rationale, coordinate the technical reports, and integrate the geospatial data pipeline, working with more colleagues (prof. Guilherme Marques and hydrologist Rafael Barbedo) to ensure the outputs were both scientifically sound and policy-ready.

Approach

The study was structured in three parts. The first established the hydrological theory specific to the Cerrado: how infiltration operates across hillslopes and plateaus, how hydrological connectivity accelerates or dampens runoff, and why the Forest Code — while protective of riparian zones — misses the upland areas where infiltration actually happens. The conceptual framing drew on the HAND metric (Height Above Nearest Drainage) as a proxy for saturation susceptibility and infiltration potential at the micro-scale.

The second part tested those ideas empirically using eight reference catchments in the Federal District spanning a conservation gradient — from nearly intact Cerrado to fully urbanized — under essentially the same precipitation regime. From the event-scale analysis of the Bananal catchment, a HAND threshold of approximately 11.5 m was identified as a practical separator between saturation-prone zones and areas with genuine infiltration potential.

The third part scaled the analysis to the entire biome. The APCAC classification system (Priority Areas for Water Conservation) ranked sub-watersheds across up to 16 classes based on three dimensions: natural hydrological importance, current conservation state, and risk of service loss. The analysis was run on the high-resolution Brazilian Hydrographic Framework (BHO 5K) to preserve local-scale signals that coarser meshes systematically suppress.

Outcomes

Preserved catchments produced two to four times more streamflow during the dry season than their degraded counterparts, even under identical precipitation — a direct, empirical case for native vegetation as water infrastructure. The HAND threshold provided a reproducible criterion for mapping infiltration-priority zones at the property and watershed scale.

At the biome scale, geology emerged as the dominant macro-scale driver of hydrological importance, with the Urucuia and Parecis aquifer systems ranking highest for infiltration and storage capacity. Two distinct risk frontiers were identified: the Matopiba region in the northeast, where large natural areas face high conversion pressure, and the southeastern Cerrado, where degraded landscapes combine steep slopes, erodible soils, and low cover into a compounding erosion risk.

The APCAC maps were designed to plug directly into existing instruments — from Environmental Economic Zoning and water use permits to Payment for Ecosystem Services programs such as ANA’s Produtor de Água.

In April 2026, the methodology was validated at a workshop held at the Ministry headquarters in Brasília, bringing together specialists from ANA, ICMBio, IBAMA, and federal universities. The discussion focused on whether the classification logic was robust enough to inform real policy instruments — and the consensus was that it was.

Resources

Project Info