Ecological succession has been traditionally approached as a niche-driven process, where a deterministic shift in species composition is predicted to be driven by changes in environmental conditions as succession unfolds. In the last decades, advances in community ecology have shown how plant community dynamics are affected by a wide range of mechanisms involving local biotic and abiotic interactions but also strongly by dispersal-based processes at the landscape scale. In addition, empirical studies have repeatedly shown that even nearby abandoned fields with the same abiotic conditions do not necessarily follow a single and predictable trajectory, but can follow multiple successional trajectories. Together, this indicates how the re-assembly of plant communities during succession is affected by a myriad of ultimate and proximate factors operating simultaneously across multiple spatial scales and interacting among them.
We show that all such mechanisms can be grouped in a relatively limited number of series of ‘higher-order’ processes that are hierarchically linked, creating minimal causal pathways that each drive predictable successional dynamics of plant communities. Specifically, we identify these “high-order” causal pathways and integrate them in a graphical framework that provides a synthetic understanding of the major underlying causes and mechanisms leading to both predictable successional shifts in species composition and variation in species dynamics at multiple spatial scales. We further identify and discuss, within some of the main causal pathways, examples of specific ‘lower-order’ mechanisms that affect successional species dynamics at local scales.
Until now, studies on tropical forest succession have mostly been pattern-driven, based on observational studies aiming to detect recurring trends to infer a process behind the observed patterns. However, our framework illustrates how the same successional trajectories and dynamics observed in field studies can be explained by alternative causal pathways, or by different lower-order processes within those pathways. We therefore argue that a more process-driven research program is needed to advance our theoretical understanding of ecological succession, and to improve our ability to assist in the design of landscape restoration based on natural regeneration.
ecological succession; community ecology; species dynamics; conceptual framework; restoration