Miombo woodlands are the largest savanna in the world and dominate southern Africa. They are strongly influenced by anthropogenic fires and support the livelihoods of over 100 million people. Managing the fire regime of these flammable systems is difficult, but crucial for sustaining biodiversity, ecosystem services, and carbon stocks. Fire intensity is more easily manipulated than fire frequency, because suppression is expensive and ineffective. However, there are important issues relating fire intensity to impacts on woody vegetation that need to be understood to inform management approaches. Such impacts include the links between fire intensity, tree top-kill, resprouting, and regrowth rates. Here we present results from a fire experiment in Mozambican miombo; the results of a 50-year fire experiment in Zimbabwean miombo; and observations of forest structure at a dry-forest site in Mozambique. We synthesize these data with a process-based gap model of stem growth, regeneration, and mortality; this model explicitly considers the effect of different frequencies and intensities of fire. We use the model, tested against the field data, to explore the sensitivity of woodland tree populations and biomass to fire intensity and frequency.
The fire experiments show that large (>5 cm dbh) stems are vulnerable to fire, with top-kill rates of up to 12% in intense fires. In contrast to idealized physical representations of tree mortality, stems of >10 cm dbh did not gain further protection from fire with increasing dbh. Resprouting was very common and not obviously linked to fire intensity. The modeling showed that miombo tree populations and biomass are very sensitive to fire intensity, offering opportunities for effective management. At any achievable fire return interval (<5 years), low-intensity fires are required to maintain observed biomass. Model predictions and field experiments show that no tree biomass can be sustained under annual fires.