Eliot Ruiz, Marion Thémèze-Leroy, Franck Ferraton, Jacques Panfili, Jean-Dominique Durand, Michel Kulbicki, Juliette Silhol, Austin Bernard, Raphaël Thomas, Théo Morisson, Yannis Djouldem, Elvin Baptiste, Pierre-Yves Pascal, Lucie Vanalderweireldt, Sébastien Cordonnier, Amélia Chatagnon, Pierre-Louis Rault, Prune Chareyre, Leila Leon, Iris Leone, Ylana Fouchan, Stéphanie Manel, Laureline Boulanger, Benjamin Victor, Séverine Albouy-Boyer, Thomas Le Berre, Charlotte R Dromard, Loïc Pellissier, Camille Albouy, Fabien Leprieur
Although larval stages ensure most connectivity between demersal fish populations, dispersal traits are increasingly less measured, despite the importance of correctly parametrizing biophysical models for spatial conservation planning. Moreover, existing data are scattered across a vast multidisciplinary literature. To address this mismatch between need and availability, we compiled data from 3,201 references into multiple datasets covering 35 key pelagic dispersal traits across fish early-life stages. We specifically compiled data on swimming speed, vertical position, and delayed settlement in demersal fishes (rafting or pelagic juveniles), as these traits exert a major yet understudied influence on dispersal scale. We systematically reported all descriptive statistics, enabling robust fitting of probability density functions to introduce particle-level variability in Lagrangian models. Overall, datasets comprise 32,988 species-level (1,709 for higher-taxon-level) records for 6,841 marine fish species worldwide, including 1,155 unpublished records (other teams), and 313 records newly collected during two fieldworks (Maldives & Guadeloupe). We also digitized 12,322 individual-level age-at-length/weight records from rearing experiments, enabling temperature-dependent growth model fitting to forecast global warming impacts on dispersal.