Linking leaf traits and litter flammability using a novel framework, tested with Brazilian Cerrado trees

Abstract In tree‐dominated ecosystems, the behaviour of surface fires is largely controlled by the structure of fine fuels. Because leaves make up most of the fine fuels, the traits of those leaves—their size, shape and leaf area for a given mass—help determine the impacts of vegetation on fire regimes. In this study, we developed several methods to better quantify the linkage between leaf traits and fire behaviour and improve our understanding of the role of tree species traits in a Brazilian forest/savanna mosaic fire regime. We developed a novel framework to partition leaf volume into a linear combination of orthogonal traits and separate their influence on flammability, using experimental burns of constructed fuel beds. The key strengths of the framework are the use of specific volume rather than bulk density, and the representation of specific volume as a linear combination of uncorrelated variables: leaf size, curl and specific leaf area. We collected intensive data on leaf traits, including foliar aluminium content and leaf volume measured with a 3D scanner, from 90 woody species of savannas and forests in São Paulo, Brazil, to assess the importance of various measures of leaf traits in determining surface fire behaviour. Greater fireline intensity was primarily driven by greater leaf size, with leaf curl having a large but secondary effect. The effect of leaf size and shape was almost entirely mediated by their effects on litter specific volume. Specific leaf area increased flammability both directly and indirectly, while foliar aluminium concentration directly reduced flammability. Forest and savanna species did not differ in flammability, but generalist species were slightly more flammable, due to a combination of leaf size and specific leaf area. Simple measures of leaf traits performed as effectively as time‐consuming 3D scanning methods. Using uncorrelated measures of leaf traits allowed us to separate the independent contributions of leaf traits to experimental fire behaviour. The framework we present highlights the importance of leaf structure and litter volume while remaining flexible enough to incorporate leaf chemistry or other traits. Read the free Plain Language Summary for this article on the Journal blog.