Supplemental Material for Goad et al., 2021

Table S1. Sample information including unique genotype ID, population assignment, and collection location for samples from Goad et al. (2021).

Table S2. Number of replicates of each genotype at each sat treatment before and after data quality filters were applied.

Table S3. Sample information including IDs, run information, inclusion in final analysis, biomass measurements, sample weight, and raw ionomics data

Table S4. Associations between tissue ion concentration and terms from Equation 1 and 2. Bold text indicates significance at the Bonferroni corrected threshold of p < 0.0027.

Table S5. P values and correlation coefficients for the association between change in biomass at increased salinity and shoot ion concentrations or principle components. Bold text indicates significance at a Bonferroni corrected threshold of p < 0.0027.

Figure S1. Map of wild collection sites for each genotype in the Southeastern US from Goad et al. (2021).

Figure S2. Flood tray design. Three of the six flood trays pictured (one from each table) were used for this experiment. Each tray contained 1 cloned replicate of each named accession. All three trays were flooded from and drained into the same reservoir for the duration of the experiment.

Figure S3. (A) ADMIXTURE Plots for K values 2 to 6. (B) Cross validation error for ADMIXTURE runs with K values from 2 to 6.

Figure S4. PCA of ionome profiles for all post-filtering ionomics samples in the 2.5, 10 and 20 dS/M treatments. Samples are colored by sample weight (A) and treatment (B).

Figure S5. Comparison of raw phenotype data across treatment and population. Population 1 (red), population 2 (green), population 3 (Blue)