A long‐term experiment reveals no trade‐off between seed persistence and seedling emergence

Abstract

Summary Information on seed persistence and seedling emergence from the soil seed bank is critical for understanding species coexistence and predicting community dynamics. However, quantifying seed persistence in the soil is challenging; thus, its association with other life‐history traits is poorly known on a broad scale. Using germination phenology for 349 species in a 42‐yr experiment, we quantified the persistence–emergence correlations and their associations with intrinsic regeneration traits using Bayesian phylogenetic multilevel models. We showed no trade‐off between seed persistence and seedling emergence. Physically dormant seeds were more persistent but exhibited lower emergence than nondormant seeds. Monocarpic species had both higher persistence and emergence than polycarpic species. Seed mass posed a marginal proxy for persistence, while emergence almost doubled from the smallest to the largest seeds. This study challenges the traditional assumption and is the first demonstration of noncorrelation between persistence and emergence, probably owing to the complexity of regenerative strategies. Species with short persistence and low emergence would be the most vulnerable for in situ conservation. Our analyses of this unique, long‐term dataset provide a strong incentive for further experimental studies and a rich data resource for future syntheses.