Trait-based ecology has revealed potential processes behind the patterns of community responses to the environment. Changing leaf traits in vegetative communities often track responses to shifting resource availability or competition, and root traits are increasingly being applied in similar ways. However, undetected functional diversity may also be stored in the seedbank, and seedbank trait bias (differences between trait composition of seedbank and vegetation communities) could limit or expand the potential responses of communities to environmental change. Furthermore, if seedbank biases shift across environmental gradients, this could reveal more about the filtering processes that act on seed production or recruitment -- two potential drivers of community turnover.
We compared community-level trait distributions (means and variances) in paired vegetation and seedbank communities at 12 grassland sites across a soil age gradient. We ask: Q1) Do seedbanks contribute unique species and functional composition (relative to the vegetation) across an edaphic gradient, and Q2) What does seedbank bias (i.e. the difference in functional composition between seedbank and vegetation) reveal about potential assembly and filtering processes across the gradient?
Results/Conclusions
In vegetative communities, older soils were associated with a suite of community-level attributes tied to resource acquisition belowground (e.g., thin roots with low dry matter content) and conservation aboveground (e.g., high leaf dry matter content). Across the gradient, however, seedbanks stored distinct pools of species that added to site-level richness. Although contributions to functional diversity were less clear, seedbanks displayed unique functional signatures; dominant seedbank species tended to be smaller-seeded and more resource-acquisitive than those in the vegetation, possibly reflecting a bias towards ruderality. However, for some traits, seedbank bias shifted across the gradient. For example, species with higher specific leaf area (tied to rapid resource acquisition) dominated the vegetation on younger soils, but were more common in the seedbank on older soils. This could reflect a shift in the processes shaping recruitment from the seedbank, favoring recruits with efficient leaves on younger soils and tougher leaves on older soils.
Thus, while older soils supported a vegetation primed for accessing belowground resources and conserving them in aboveground tissues, seedbanks expanded our assessment of community composition and thus, adaptability. Although seedbanks tended to be dominated by more ruderal traits, shifting seedbank biases across the gradient demonstrate that seedbank-related processes in community assembly - including input, storage, and recruitment - may also change systematically with the environment.