Rainfall and Fuel-Mediated Fire Dynamics in the African Savannas: A Process-Based Model

Savanna ecosystems are shaped by intricate interactions among rainfall, vegetation productivity, and fire regimes, with these dynamics varying along aridity–humidity gradients. We developed a stochastic, process-based simulation model to investigate how rainfall, fire frequency, and fire intensity jointly regulate grass and shrub biomass in semi-arid Namibian savannas over a 100-year period. The model operates at monthly time steps, linking rainfall-driven biomass accumulation to fuel availability and probabilistic fire occurrence. Simulation results reveal marked contrasts across rainfall regimes. In drier savannas ( 200 mm/year), low and relatively stable grass biomass dominates, supporting frequent but low-intensity fires that exert limited influence on vegetation structure. Mesic savannas ( 300–400 mm/year) exhibit recurrent, moderate- to high-intensity fires that produce pronounced oscillations in grass and shrub biomass, suppressing woody encroachment while maintaining dynamic coexistence with grasses. In humid savannas ( 500 mm/year), high biomass accumulates and fuels are abundant, but fires occur infrequently due to high humidity. When fire events do occur, they are extremely intense, shaping long-term vegetation composition. Rainfall variability amplifies temporal fluctuations in biomass and fire intensity but does not fundamentally shift equilibrium states. Across all regimes, fire characteristics—especially frequency and intensity—emerge as the dominant regulators of savanna structure, while rainfall acts primarily as a stochastic modulator. These findings highlight the need for rainfall-specific fire management strategies: fire suppression in arid systems to protect fragile vegetation, prescribed burns in mesic regions to prevent woody encroachment, and moderate, regular fires in humid savannas to sustain grass–tree coexistence and reduce the risk of extreme wildfires.