The seed germination spectrum of 486 plant species: A global meta-regression and phylogenetic pattern in relation to temperature and water potential.

• A new conceptual framework on germination niche shaped by hydro-thermal traits. • Key inter-relations between quantitative germination traits for almost 500 species. • Patterns and co-correlants in the cardinal values for species' germination niche. • A database to determine the vegetation boundaries of wild or cultivated species. The germination niche of plant species depends primarily on the seeds' responsiveness to temperature and water potential. However, to appreciate future climate risks to natural regeneration through germination, a global level synthesis across species is needed. We performed a meta-regression and phylogenetic patterning of primary data from 377 studies on 486 species, including trees, grasses, crops and wild species, to determine patterns and co-correlants in the cardinal values that define species' germination niches. We found positive correlations between base temperature and other germination traits related to cardinal temperatures suggesting that plant species alter base temperature values in harmony with other thermal traits as a highly efficient adaptation strategy to coping with harsh conditions. A negative correlation was found between thermal time and base temperature, and positive correlations between other cardinal temperatures and base temperature. Mean values of thermal time indicate that annual crops germinate more rapidly compared to wild species, potentially as a consequence of domestication, and tropical tree seeds the slowest. Dryland species (Cactaceae and Agavaceae) have the widest upper thermal and lower moisture niches, indicative of abilities to grow under harsh conditions, while forages have the narrowest thermal and moisture niches, suggesting higher sensitivity to frost or drought. We propose a new conceptual framework for future research on germination niche as shaped by thermal and moisture traits. Our database represents a unique source of information to further determine the vegetation boundaries of wild or cultivated species, including within simulation studies on plant species adaptations under changing land-use and climate. [ABSTRACT FROM AUTHOR]