Size fractionation as a tool for separating charcoal of different fuel source and recalcitrance in the wildfire ash layer

Charcoal is a heterogeneous material exhibiting a diverse range of properties. This variability represents a serious challenge in studies that use the properties of natural charcoal for reconstructing wildfires history in terrestrial ecosystems. In this study, we tested the hypothesis that particle size is a sufficiently robust indicator for separating forest wildfire combustion products into fractions with distinct properties. For this purpose, we examined two different forest environments affected by contrasting wildfires in terms of severity: an eucalypt forest in Australia, which experienced an extremely severe wildfire, and a Mediterranean pine forest in Italy, which burned to moderate severity. We fractionated the ash/charcoal layers collected on the ground into four size fractions (> 2, 2–1, 1–0.5, < 0.5 mm) and analysed them for mineral ash content, elemental composition, chemical structure (by IR spectroscopy), fuel source and charcoal reflectance (by reflected-light microscopy), and chemical/thermal recalcitrance (by chemical and thermal oxidation). At both sites, the finest fraction (< 0.5 mm) had, by far, the greatest mass. The C concentration and C/N ratio decreased with decreasing size fraction, while pH and the mineral ash content followed the opposite trend. The coarser fractions showed higher contribution of amorphous carbon and stronger recalcitrance. We also observed that certain fuel types were preferentially represented by particular size fractions. We conclude that the differences between ash/charcoal size fractions were most likely primarily imposed by fuel source and secondarily by burning conditions. Size fractionation can therefore serve as a valuable tool to characterise the forest wildfire combustion products, as each fraction displays a narrower range of properties than the whole sample. We propose the mineral ash content of the fractions as criterion for selecting the appropriate number of fractions to analyse.