Leaf traits predict global patterns in the structure and flammability of forest litter beds

Fallen plant material such as leaves, needles and branches form litter beds which strongly influence fire ignition and spread. Traits of the dominant species influence litter flammability directly by determining how individual leaves burn and indirectly through the structure of the litter bed. However, we are yet to determine the relative importance of these different drivers across a range of plant species from different biomes. We undertook a meta-analysis, combining leaf trait, litter structure and flammability data for 106 species from North America, South America, Europe, Asia and Australia. The dataset encompassed broad-leaved and coniferous species from seven different experimental studies. Relationships between leaf traits, litter structure and key flammability metrics—sustainability, combustibility and consumability—were analysed using bivariate and piecewise structural equation modelling (SEM). Traits which characterise the three-dimensional nature of the leaf and how much space a leaf occupies showed much stronger associations to litter structure and flammability than other morphological traits. Leaf curl, surface area to volume ratio (SAV) and SLA predominately influence litter flammability indirectly via litter structure with SLA being the only leaf trait which had a negative direct effect on flame duration. Packing ratio and bulk density were influenced by different combinations of leaf traits and, in turn, they aligned with different flammability metrics. Bulk density predicted flame spread rate and flame duration whereas packing ratio predicted consumption. Synthesis. We identified key leaf and litter traits which influence different components of litter bed flammability. Importantly, we show that the effects of these leaf and litter traits are consistent across a wide range of taxa and biomes. Our study represents a significant step towards developing trait-based models for predicting surface wildfire behaviour. Such models will more flexibly accommodate future shifts in the composition of plant species triggered by altered fire regimes and climate change.