Table S8. Plasticity of early rosette diameter and flowering time in response to different environmental cues for accessions selected based on their plasticity to N.
Figure S1. Number of rosette leaves, bolting time (B) and flowering time (F) for accessions grown under limiting (lim), intermediate (int) and optimal (opt) N conditions . The seeds were stratified for 5 days at 4°C/dark, following germination for 5 days on the respective soil (short days, 20°C/16°C day/night), and vernalization for 60 days (long days, 4°C). After vernalization, the plants were acclimated for one day at long days with 21°C/17°C day/night). Next, for each accession and condition, one plant was pricked to individual pots (n = 5) containing new soil with the same N concentration. The plants were grown under long days and 21°C/17°C (day/night). The number of days was counted after pricking. Bolting time was scored when the stem reached 1 cm height; flowering time was scored when the first flower opened.
Figure S2. Flowering time (FT) performance of T-DNA lines for candidate genes related to FT plasticity in response to N. Days to the first open flower (FT) of Col-0 WT and T-DNA lines under A.limiting (Lim) N, B. intermediate (Int) N, C. optimal (Opt) N. For each line and condition, n = 10. D. FT plasticity measured as the fold change (FC) between limiting and optimal N.E. FT plasticity measured as the coefficient of variation (CV) across the three N conditions.
Figure S3 . Early rosette diameter (ERD) performance of Col-0, Gy-0, Xan-5, and proton1 in response to N. Col-0 and Gy-0 belong to the haplogroup carrying the major allele at the promoter ofPROTON1 . Xan-5 has a missense substitution that causes the loss of PROTON1 start codon, while proton1 is a strong knockdown line (Figure 4a). A. ERD in response to limiting (Lim) and optimal (Opt) N (n = 10 for each condition and line).B. ERD fold change (FC) between limiting to optimal N for each of the lines, indicating that loss of PROTON1 function reduces ERD plasticity, as observed for Xan-5 and proton1 FC.
Figure S4. Correlation of the mean levels of metabolites and the mean trait values of ERD, FRD, FT and YIE in accessions grown at limiting, intermediate and optimal N conditions. Significant Spearman correlation after Benjamini-Hochberg correction (FDR < 0.05) are represented as red for positive, or blue for negative. Abbreviations: adenosine-5-monophosphate (AMP), dehydroascorbate (DHA), fructose-6-phosphate (F6P), gamma aminobutyric acid (GABA), glucose-6-phosphate (G6P), glycerol-3-phosphate (G3P).
Figure S5. Co-expression network of At1g19880 (PROTON1).Co-expression analysis adapted from ATTED-II v.10.1 (https://atted.jp/). KEGG (Kyoto Encyclopedia of Genes and Genomes) IDs are shown. Murual ranks for the edge strength of co-expressed gene network are represented as bold (MR < 5), normal (MR < 30), or thin (MR > 30) edges.