COS 83-1 - Plant functional types can predict the influence of fire on species relative abundance

Abstract: Fire is an environmental filter that influences plant survival, reproduction, and establishment. Consequently, plants exhibit traits such as resprouting and fire-stimulated seed release to cope with recurrent fire. Plants with similar combinations of traits can be grouped into Plant Functional Types (PFTs) and used to form predictions of population change. However, if PFTs and associated predictions are to assist ecological understanding and management, we must know if these predictions are robust. To explore this issue, we developed a functional type approach that predicts how the relative abundance of plants changes as a function of time since the last fire, and tested it using empirical data collected in Mediterranean-type shrublands of southeastern Australia. First, we completed surveys of vascular plants at 57 sites, stratified by an 81-year post-fire chronosequence. Second, we used trait databases to assign recorded species to PFTs, based on the timing of life stages and their ability to persist, arrive and establish at a site. Third, for each PFT, we developed predictions of changes in relative abundance by time since fire and expressed these predictions graphically. Finally, we tested predictions using non-linear regression models that quantified relationships between plant species and time since fire. Results demonstrate reasonable capacity for predicting fire-driven plant relative abundance change. Predictions of the direction (increase or decrease) of changes in plant abundance from 0 to 81 years’ post-fire were correct for 18 of 24 species modelled. Predictions of the shape of changes in relative abundance across the 81-year post-fire gradient were not as accurate, but still useful: 13 out of 24 species showed ‘excellent’ conformity with predictions, 7 ‘good’ conformity, and 4 ‘poor’. Data and models indicated that going beyond broad functional groupings, such as resprouter versus non-resprouter, were important, as we observed different responses within these broad groupings. Importantly, this empirical test demonstrates that PFTs can be used to predict long-term changes in plant communities in fire-prone ecosystems. Because the approach we explored focuses on the mechanisms that shape plant populations, we expect that it will be applicable to many fire-prone environments globally.