Modeling seed germination data to meet biodiversity conservation needs in the Mediterranean.

Seed-based conservation and restoration programs rely on sound knowledge in fundamental seed biology. However, the adequacy of current data on seed germination to meet biodiversity conservation needs remains largely unexplored. We assessed germination data availability and quality for 1495 Mediterranean species of conservation and restoration interest to determine how much germination information is currently available; whether existing data meet the requirements for thermal germination risk assessment; and which species or ecological features help explain current data gaps. To this end, we used phylogenetic logistic regressions and assessed the adequacy of the available data by modeling the species' seed germination thermal responses. Germination data were available for 31% of species but adequate for thermal modeling for only 9% of them. Species with data were mainly ecologically widely distributed, produced physiologically dormant seeds, including species suitable for restoration, and had traditional uses. Thermal risk was not phylogenetically clustered, but it was linked to habitat, with species from forests and scrubland exhibiting higher thermal risk. Based on our results, we devised a framework to support conservation and restoration efforts that includes steps to improve data availability (e.g., data gap analysis) and to classify species according to their vulnerability to climate warming. These steps will help distinguish between species that are potentially suitable for restoration programs and those that should be prioritized for conservation and demonstrate how germination information can be modeled and translated into thermal risk assessments to support seed-based conservation and restoration programs.