Continent-level drivers of African pyrodiversity
2 July 2017Hempson, Gareth; Parr, Catherine; Archibald, Sally; Anderson, T. Michael; Courtney Mustaphi, Colin; Dobson, Andrew; Donaldson, Jason; Morrison, Thomas; Probert, James; Beale, Colin
Pyrodiversity, which describes fire variability over space and time, is believed to increase habitat heterogeneity and thereby promote biodiversity. However, to date there is no standardised metric for quantifying pyrodiversity, and so broad geographic patterns and drivers of pyrodiversity remain unexplored. We present the first generalizable method to quantify pyrodiversity, and use it to address the fundamental questions of what drives pyrodiversity, which fire attributes constrain pyrodiversity under different conditions, and whether pyrodiversity is spatial grain-dependent. We linked the MODIS burned area and active fire products to measure fire size, seasonal timing, return interval, and intensity for 2.2 million individual fires in sub-Saharan Africa from 2000–2015. We then quantified pyrodiversity as a four-dimensional hypervolume described by fire attributes within a grid cell, for any spatial grain of analysis. Environmental (rainfall, vegetation, soils, and topography) and human-associated (cattle biomass, cropland area, and human population density) variables were assessed as potential drivers of pyrodiversity. Rainfall was the main environmental driver of pyrodiversity, with higher pyrodiversity in drier regions (-1). Pyrodiversity was not strongly associated with human-associated variables across Africa. Rainfall and a human influence index had clear but contrasting effects on the variability of fire size, seasonal timing, return interval, and intensity. Our analyses show that fire size and seasonal timing constrain pyrodiversity in wetter regions, whereas none of the fire attributes pose a strong constraint in drier regions. We found no evidence that pyrodiversity was spatial grain-dependent when recalculated at 5-minute grain increments from 15 to 120 minutes. We hypothesise that the strongest positive effect of pyrodiversity on biodiversity in all its forms will occur at intermediate precipitation (650–1300 mm yr-1), where fire plays an important role in shaping vegetation structure and where pyrodiversity is still quite high.