Supplementary Tables for secondary dormancy study
Seeds should not germinate under conditions in which seedling development cannot be sustained. Dormancy, which allows seeds to remain inactive in an environment that would otherwise enable germination, helps optimize the timing of germination. The induction of a secondary type of dormancy after seed dispersal allows a close fine-tuning of germination to prevailing environmental conditions. In this study, we explore the relevance of secondary dormancy in response to heat for adaptation to local environments by characterizing natural variation in heat-induced secondary dormancy within a collection of 361 Arabidopsis thaliana accessions originating from across the European geographical range. We observed a latitudinal pattern in the variation of heat-induced secondary dormancy in European Arabidopsis thaliana, reflecting adaptation to variation in temperature and precipitation along this gradient. Species distribution models suggest that the phenotypic distribution of secondary dormancy has shifted in the past, in response to changes in temperature and precipitation changes, and is predicted to continue so in the coming years. We found that the acquisition of secondary dormancy varies with levels of primary dormancy and after-ripening. However, we identified several regions in the genome that specifically controlled the levels of secondary dormancy. Our findings show that secondary dormancy is a complex adaptive mechanism contributing to the dormancy trait syndromes that favors plant survival under harsh climates.
The samples used in this study originated from a collection of 361 accessions across Europe (Table S1). Genotype and accession information was obtained from the 1001 Genome database (1001genomes.org; 1001 Genomes Consortium, 2016) as well as from Wieters et al. (2021). Seed material was amplified at University of Cologne: plants were grown in growth chamber (producer, model) under long day condition 16:8 (hours) light: dark at 200C (day) : 180C night (termed “standard condition” hitherto); mature dry seeds were harvested, and packed in paper bags. The experiment was replicated on one seed batch, approximately six months (May 2022 - “Trial1”, one year (December 2022) - “Trial2”, and two years (December 2023) - “Trial3”, after seed harvest. Due to practical limitations, the first batch used a set of 295 accessions, the second batch used a full set of 361 accessions, and the third batch a set of 344 accessions.
Attached here are three Supplementary Tables that are not included in the main text.
Table S1. Information of 361 studied accessions. Table records Genotype ID (GenotypeID) according to ENA database and Genotype Name (GenotypeName) accordingly. Geographical origin (origin) and precise latitude and longitude of the genotype are recorded, together with extra information such as sequencer, collector, and CSS accession ID (AccessionID). Attached separately due to large file size.
Table S5. Regression results of heat-induced secondary dormancy with four bioclimatic variables as predictors of genetic variation in germination after three treatments, across three trials. The results of regression models that assess the influence of four bioclimatic variables on secondary dormancy and its genetic variation. The models employed a binomial likelihood with a logit link function. The bioclimatic variables included BIO3 (isothermality), BIO9 (mean temperature of the driest quarter), BIO18 (mean precipitation of the warmest quarter), and BIO19 (mean precipitation of the coldest quarter). The analysis was performed across three treatments: primary dormancy (pdorm), secondary dormancy (sdorm), and control. Secondary dormancy was induced by a 4-day 37°C treatment following a 3-day stratification at 4°C to release primary dormancy. Primary dormancy was tested without any pre-treatment, whereas control seeds were stratified at 4°C for 3 days before germination testing. The sdorm value used in this model represents secondary dormancy corrected for primary dormancy, calculated as the residual of the regression of secondary dormancy on primary dormancy. All germination tests were conducted under long-day conditions at 20°C, and germination rates were recorded after 7 days. Trial 1 is a set of 295 samples, Trial 2 has the complete set of 361 samples, and Trial 3 is a set of 344 samples. Trial 1 was performed in May 2022, Trial 2 in December 2022, and Trial 3 in December 2023. (A), (B), and (C) represent models with the control group as the baseline for Trial 1, Trial 2, and Trial 3, respectively; (D), (E), and (F) represent models in which primary dormancy is used as the baseline, for Trial 1, Trial 2, and Trial 3, respectively.
Table S6. Genome-wide association results of primary dormancy across three trials and shared genome-wide association peaks across three experimental trials. The primary dormancy treatment tested germination rates without any pre-treatment. All germination tests were conducted under long-day conditions at 20°C, and germination rates were recorded after 7 days. (A) Trial 1 is a set of 295 samples, (B) Trial 2 is the complete set of 361 samples, and (C) Trial 3 is a set of 344 samples. Trial 1 was performed in May 2022, Trial 2 in December 2022, and Trial 3 in December 2023. (D) Shared genome-wide association peaks of primary dormancy across all experimental trials, computed by Fisher’s combined probability test.
Table S7. Genome-wide association results of heat-induced secondary dormancy across three trials and shared genome-wide association peaks across three experimental trials. The secondary dormancy represents germination rates following a 3-day stratification at 4°C to release dormancy, followed by a 4-day treatment at 37°C. All germination test s were conducted under long-day conditions at 20°C, and germination rates were recorded after 7 days. (A) Trial 1 is a set of 295 samples, (B) Trial 2 is the complete set of 361 samples, and (C) Trial 3 is a set of 344 samples. Trial 1 was performed in May 2022, Trial 2 in December 2022, and Trial 3 in December 2023. (D) Shared genome-wide association peaks of heat-induced secondary dormancy across all experimental trials, computed by Fisher’s combined probability test.
