D3: The Dispersal and Diaspore Database – Baseline data and statistics on seed dispersal

Seed dispersal is hard to measure, and there is still a lack of knowledge about dispersal-related traits of plant species. Therefore, we developed D3, the Dispersal and Diaspore Database (available at www.seed-dispersal.info), which aims at simplifying ecological and evolutionary analyses by providing and integrating various items related to seed dispersal: empirical studies, functional traits, image analyses and ranking indices (quantifying the adaptation to dispersal modes). Currently, the database includes data for more than 5000 taxa and 33 items as well as digital images of diaspores (i.e. the dispersal units), seeds, fruits and infructescences. The included items cover common traits like diaspore mass, size, shape, terminal velocity and seed number per diaspore. Furthermore, we present newly or further developed items like ecomorphological categorizations of the diaspore and fruit as well as information from literature on prevailing dispersal modes. Finally, we introduce several items which are not covered in other databases yet: surface structure and form of the diaspore, the exposure of the diaspores in the infructescence and dispersal rankings. Dispersal rankings allow estimations of how well certain species are adapted to a specific dispersal mode in comparison to a larger species set. They are calculated as the percentile rank of an indicator of species’ dispersal potential in relation to a larger species set. Especially for the new and further developed items we outline the basic concepts in detail, describe the measurement and categorization methods and show how to interpret and integrate these data for single species as well as for larger species sets. Thereby, we calculate baseline statistics of seed dispersal of the Central European flora. We found that diaspores of 72% of the taxa show specializations related to long-distance dispersal, i.e. most often elongated appendages or nutrient-rich tissues. Diaspore masses, sizes and terminal velocities vary over several orders of magnitude and can be approximated by lognormal distributions.