Seed dormancy types and germination response of 15 plant species in temperate montane peatlands

Despite their crucial role in determining the fate of seeds, the type and breaking mode of seed dormancy in peatland plants in temperate Asia with a continental monsoon climate are rarely known. Fifteen common peatland plant species were used to test their seed germination response to various dormancy-breaking treatments, including dry storage (D), gibberellin acid soaking (GA), cold stratification (CS), warm followed cold stratification (WCS), GA soaking + cold stratification (GA+CS) and GA soaking + warm followed cold stratification (GA+WCS). Germination experiment, viability and imbibition test, and morphological observation of embryos were conducted. Of the 15 species, nine showed physiological dormancy (PD), with non-deep PD being the dominant type. Four species, Angelica pubescens, Cicuta virosa, Iris laevigata and Iris setosa exhibited morphological physiological dormancy. Two species, Lycopus uniflorus and Spiraea salicifolia, demonstrated non-dormancy of seeds. Overall, the effect hierarchy of dormancy-breaking is: CS > GA > WCS > GA+CS > D > GA+WCS. Principal component analysis demonstrated that seed traits, including embryo length: seed length ratio, seed size, and monocot/eudicot divergence, are more likely to influence seed dormancy than environmental factors. Our study suggests that nearly 90% of the tested peatland plant species in the Changbai Mountains demonstrated seed dormancy, and seed traits (e.g. embryo to seed ratio and seed size) and abiotic environmental factors (e.g. pH and temperature seasonality) are related to germination behavior, suggesting seed dormancy being a common adaptation strategy for the peatland plants in the temperate montane environment.

Methods
All statistical analyses were conducted in R v 4.2.3. Generalized Linear Models (GLMs) were used to analyze the effect of dormancy-breaking treatments (explained factor) on germination (GP on the 28th day, dependent factor). The significance of factors for each experiment was assessed by Wald Chi-Square statistics to the model. Duncan's test was used for multiple comparisons. The difference between germination percentage of control and initial viability was determined by Mann-Whitney-U-tests, to determine whether the seeds were dormant or not. Principal components analysis (PCA) was used to reduce and visualize the variability in the species’ seed germination response and to intercorrelations among seed traits or environmental factors. The PCA was carried out with the package ‘FactoMineR’ using the variance-covariance matrix. The significance level was set to α = 0.05.