**5. Conclusions**

We found that phenotypic plasticity of seed dormancy release was significantly correlated with increased gradient of aridity, suggesting that plastic responses to external stimuli provide seeds with bet-hedging capacity and the potential to cope with high levels of environmental heterogeneity. Genome-wide association analysis identified candidate genes associated with dormancy release. Gene ontology showed enrichment for genes involved in modification of the cell wall, as well as oxidative stress protection, mediating seed coat permeability and, ultimately, imbibition and germination. Knowledge of the seed dormancy regulation by environmental factors could be extended to other legume species, particularly to crop wild relatives of economically important species, such as chickpea, lentil, faba bean and soybean, as well as used in the managemen<sup>t</sup> of endangered plant species with physical seed dormancy.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2223-7747/9/4/503/s1, Figure S1: Frequency distributions of dormancy traits, Figure S2: Correlation between the phenotypic plasticity index of final dormancy (PIPY) and selected environmental variables, Figure S3: Correlation chart of the dormancy release traits and environmental variables, Figure S4: Correlation chart of the dormancy release traits, soil variables and inter-annual climatic variables, Figure S5: Manhattan plots of mapped SNP markers associated with dormancy or bioclimatic traits, Figure S6: Quantile–quantile (Q-Q) plots for all the traits obtained by standard mixed linear model (EMMA) and multi-locus linear model (FarmCPU), Figure S7: Geographic distribution of studied M. truncatula accessions, Figure S8: Final PY dormancy (FPYD) of each year under two temperature treatments (35/15 ◦C and 25/15 ◦C). Di fferent letters indicate significant di fferences among year for each temperature (Fisher's LSD test, α = 0.05), Table S1: List of tested Medicago accessions with calculated seed dormancy traits and extracted environmental variables, Table S2: Basic descriptive statistics of 23 bioclimatic variables and 10 soil variables of sites of accessions origin, Table S3: Classification of 176 M. truncatula accessions in four cluster based on environmental and climatic conditions, Table S4: Pearson coe fficients-probabilities between dormancy traits and bioclimatic variables, Table S5: Regression coe fficient (r2) between environmental variables and plasticity index by macroecological and genetic clusters [106], Table S6: Complete list of QTN identified by GWA studies for each dormancy trait, Table S7: Over-representation analysis of the 136 candidate genes potentially involved in dormancy traits.

**Author Contributions:** Conceptualization, J.P.R. and P.S.; Data curation, J.B. and P.S.; Formal analysis, J.P.R., M.D., J.B., I.H., V.P., T.V., J.M., K.H., J.V. and P.S.; Funding acquisition, M.D. and P.S.; Investigation, J.P.R., J.B. and I.H.; Methodology, M.D., J.B., V.P., K.H. and P.S.; Project administration, I.H. and P.S.; Writing – original draft, J.P.R., M.D., J.B., J.V. and P.S.; Writing – review & editing, J.P.R., M.D., J.B., V.P., J.V. and P.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Grant Agency of the Czech Republic [grant number 16-21053S].

**Acknowledgments:** We thank to Oldˇrich Trnˇený for construction of seed threshing device and Miroslav Hýbl for help with seeds cleaning.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
