Inverse priority effects: The order and timing of removal of invasive species influence community reassembly

Abstract

Abstract An ongoing restoration challenge is to recover native communities after the removal of invasive species. Because priority effects (i.e. the order and timing of species arrival) can strongly determine the trajectory of community assembly, their intentional manipulation is gaining attention to manage invasive plants and achieve restoration goals. Yet, ecologists and conservationists rarely consider how the order and timing of species removal—inverse priority effect—may impact future plant communities. Here, we evaluated the dependence of community reassembly on inverse priority effects by experimentally removing the target invasives Sweetbriar rose (Rosa rubiginosa) and Scotch broom (Cytisus scoparius) in field and mesocosm communities. We manipulated removal order (rose before broom vs. broom before rose) and timing (simultaneously early vs. simultaneously late in the season). We performed a hierarchical modeling of species sommunity to assess differences in community structure in response to order and timing of removal, and to evaluate whether species origin and leaf and seed traits were associated with species responses. We found that the order of removal was as important as timing driving community reassembly. Simultaneous removal favoured nonnatives, more so when performed early. Sequential removals led to contrasting communities. Rose before broom removal also favoured nonnative grasses at expense of native species, whereas the inverse order produced small changes in communities. In general, species with high specific leaf area were boosted, regardless of their seed size. Synthesis and applications. Inverse priority effects are neglected mechanisms that can drive variability in the reassembly of plant communities and can potentially upgrade invasive species management. These historical contingencies suggest the existence of an optimal order of removal that facilitates the recovery of the native community. We found that simultaneous removal promoted secondary invasions to a greater extent than sequential removals. Furthermore, removal order affected post‐removal community structure. In our system, we suggest removing the rose before the broom to hinder nonnatives and pave the way for restoration of native communities. Our results show that manipulation of the order and timing of removal can help to achieve restoration goals.