Discussion 
Fire has played an important role in shaping tropical biodiversity for millennia (Kelly et al., 2020), with studies reporting contrasting impacts on biodiversity (Kelly et al., 2020; He et al., 2019). To explain this heterogeneity, we analysed a systematic compilation of data quantifying species richness and community composition responses to fire in tropical communities of major plant life forms. Our analyses account for variation across biomes and quantify responses to time since fire and fire type (prescribed or non-prescribed burns) and assess if protected area status modifies plant community responses.
Despite conducting a comprehensive literature search, we only found 28 studies that met our data analysis requirements. There is thus a clear need for additional empirical fieldwork that assesses plant community responses to fire; such studies should follow the open science principles of making underlying datasets freely available to facilitate meta-analyses. Our focal studies included ones that assessed biodiversity recovery up to twenty-nine years following fire events, but most studies were conducted within ten years of a fire event. Given that we find plant community composition can remain impacted by fires ten years after they occur (see below discussion) there is a particular need for long-term (> ten years) longitudinal studies. Our results indicate that changes in species richness and recovery of community composition following fire events vary across plant life forms. This heterogeneity across life forms suggests that apparent contradictions in the published literature regarding the impacts of fire on plant communities may arise from variations in which life forms dominate the focal plant communities. We thus encourage future studies to take this into account in their study design and interpretation. Despite the limitations of data availability and duration of studies our analyses provide important novel preliminary insights regarding biodiversity responses and recovery from fire events.