Primary plant succession provides an excellent natural experiment to test ecological questions about community assembly following major disturbances. Temporal phylogenetic and functional trait dispersion patterns can give insight into the relative importance of stochastic and deterministic processes, as well as the potential identity of deterministic (biotic vs. abiotic) drivers (e.g. dispersal, growth, nutrient acquisition, and herbivore resistance ability). We used 28 years of plant composition data across four primary succession sites at Mount St. Helens (collected since the 1980 volcanic eruption) to examine phylogenetic dispersion and trait patterns over time. We expected to find more evidence of clustering or random phylogenetic patterns early in succession (where environmental filtering and stochasticity are thought to dominate), with a switch to overdispersion via processes such as niche differentiation later in succession. Contrary to expectations, phylogenetic relatedness and trait dispersion patterns were idiosyncratic. We found evidence of deterministic community assembly even early in succession, suggesting that both local-scale abiotic filtering and biotic interactions start playing a role shortly after the initiating disturbance. Traits were less predictive of successional patterns, with few consistent changes in trait distribution across all sites, even though the traits we examined are linked to the processes thought to be important during succession. Together, these results suggest that the drivers of community assembly during succession may be less generalizable and more complex than previously thought. We suggest that combining experiments and these analytical tools with long-term monitoring could be a useful step forward.
