Variability in fire‐induced change to vegetation physiognomy and biomass in semi‐arid savanna. Issue 12 (17th December 2018)

• Record Type: Journal Article
• Authors: Singh, Jenia; Levick, Shaun R.; Guderle, Marcus; Schmullius, Christiane; Trumbore, Susan E.
• Is Part Of: Ecosphere. Volume 9:Issue 12(2018)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2018-12-17
• Subjects: fire frequency -- Kruger National Park -- savanna -- terrestrial laser scanning -- vegetation structure; Ecology -- Periodicals; Ecology; Periodicals; 577.05
• DOI: 10.1002/ecs2.2514
• Languages: English
• ISSNs: 2150-8925

Abstract

Abstract: Fire plays an intrinsic role in shaping the biophysical attributes of savanna ecosystems. Savanna fires limit vegetation biomass below their climatically determined potential, but the magnitude of this effect and how it varies across heterogeneous landscapes are poorly understood. In this study, we explore woody tree structure and canopy characteristics across a fire manipulation experiment that has been maintained for 63 yr in South Africa's Kruger National Park. Our study design assessed three late dry‐season fire regimes (biennial, triennial, and unburnt) across a precipitation gradient (737–496 mm/yr) spanning four different landscapes with a mixture of sandy and clay soils. We used terrestrial laser scanning (TLS) to quantify tree height, canopy cover, and aboveground carbon storage across the experimental treatments. Vegetation physiognomy was influenced by the interaction between landscape and fire frequency. In the absence of fire, woody height, cover, and biomass increased with increasing rainfall. The presence of fire acted to reduce structure and biomass as expected, but the magnitude of this effect increased with increasing rainfall. We found minimal difference between the effects of biennial or triennial burning—except at the wettest site where the triennial fire plots had half the biomass of those burnt biennially. The rainfall dependent fire–vegetation relationships shown here provide empirical quantification of top‐down constraint by fire and highlight the challenges of predicting responses to disturbances in these inherently heterogeneous ecosystems. Robust quantification of 3D structure and dynamics through TLS will be useful for constraining carbon stock models and predicting trajectories of change under future climate and land‐use conditions.