2.4. Seed dormancy classes

To assign each species under a seed dormancy class (sensu Baskin & Baskin, 2014), information related to seed coat permeability and embryo type was obtained from the literature (Baskin & Baskin, 2007; Martin, 1946). To determine physiological and morphophysiological dormancy level, generalized lineal mixed models (GLMM) with binomial error structure and logit link function were built for each species. In these models, seed germination proportion (i.e., number of germinated seeds out of number of viable seeds of each species) was the response variable, whereas dormancy‐breaking treatment [four‐level categorical variable including 0 CS, 3 CS, 5 CS, and GA₃], temperature (two‐level categorical variable, including 15/5, 25/15°C), and their interaction were the explanatory variables. Finally, replicates were treated as a covariable. According to the data collected from the literature and the germination response observed here, species were assigned a type and class of dormancy following the Baskin and Baskin classification (2014) and Silveira's diagram Silveira (2013). For example, when the FGP of seeds incubated at 3 or 5 CS were significantly higher than FGP at 0 CS, the embryo was fully developed at dispersal and the seed coat was permeable, seeds were considered PD (for further information about the classification criteria used see Table S3a,b).