Figures and data
Metabolomic responses of Phragmites australis to salinity stress.
(a) Principal component analysis (PCA) score plot of the metabolome (3,413 metabolites) under control (blue) and salt (red) treatments. Ellipses represent 95% confidence intervals for each group. PERMANOVA (Euclidean distance, 9,999 permutations) with genotype as covariate revealed a significant effect of salinity (p < 0.001). (b) Volcano plot showing differential accumulation of metabolites between salt and control groups. Red points indicate metabolites that are significantly different (|log₂ fold change| > 1, false discovery rate < 0.05, variable importance in projection > 1). Selected stress-responsive metabolites (e.g., N,N-dimethylglycine, sucrose 6-phosphate, glutathione) are labeled.
Effects of salinity on biomass and sodium accumulation in Phragmites australis.
(a) Biomass of control and salt-treated plants (harvest at the final stage). Data are means ± SE. *** p < 0.001 (Welch’s t-test, n = 106 per group). (b) Sodium (Na) concentrations in green, yellow, and brown leaves under control and salt conditions. Data are means ± SE. *** p < 0.001 (Welch’s t-test for each stage, n = 105–106 per group per stage).
Geographical distribution (a) of the phylogeographical groups of Phragmites australis and leaf nutrient resorption efficiency (b–d) for nitrogen (N), phosphorus (P), and potassium (K).
Phylogeographic groups reflect deep evolutionary lineages defined by genetic markers (e.g., chloroplast haplotypes). Statistical significance was assessed using linear mixed-effects models that included genotype as a random effect to accommodate the paired experimental design. The p-values shown in panels b–d refer to the fixed effects of group, salinity and their interaction. When significant group effects were detected, post-hoc pairwise comparisons were performed separately within each salinity treatment using Tukey’s HSD test; different letters denote statistically significant differences (α = 0.05).
Geographical distribution (a) of Phragmites australis ecotypes and leaf nutrient resorption efficiency (b–d) for nitrogen (N), phosphorus (P), and potassium (K).
The colour scheme represents habitat-based adaptations and is independent of that used in Figure 3. Ecotypes categorize populations according to adaptive phenotypes to specific local habitats (freshwater wetlands, coastal saltmarsh, inland saltmarsh). Significance was tested with linear mixed-effects models incorporating genotype as a random effect to account for paired measurements. The reported p-values correspond to the fixed effects of ecotype, salinity and their interaction. Where ecotype effects were significant, post-hoc Tukey’s HSD comparisons were conducted within each salinity level; different letters indicate statistically significant differences (α = 0.05).
Assessment of three control strategies for nutrient resorption in Phragmites australis under varying salinity levels.
(a–c) Relationship between nutrient concentrations (N, P, K) in green leaves and senesced litter. (d, g) Association between phosphorus and nitrogen resorption efficiencies. (e, h) Linkage between resorbed N:P(K) ratios and green leaf N:P(K) ratios. (f, i) Correlation between log₁₀-transformed resorbed N:P(K) and log₁₀-transformed green leaf N:P(K) ratios. The gray dashed line in each panel indicates the 1:1 reference. Solid lines represent significant relationships (p < 0.05 for the SMA regression model), while dashed lines indicate non-significant relationships. Statistical annotations show results from standardized major axis (SMA) regression for each salinity level: β (with asterisks) is the SMA slope, where asterisks after β indicate significant deviation from 1 (* p < 0.05, ** p < 0.01, *** p < 0.001), and asterisks after R² indicate significance of the SMA regression model.
Influence of latitude on leaf nitrogen (N), phosphorus (P), and potassium (K) resorption efficiencies of Phragmites australis.
The analysis was performed using linear mixed-effects models, treating phylogeographical groups as a random effect.
