Figure 1.
G0 = generation 0, G1 = generation 1 etc. In blue, SSS refers to three generations of salt treatment, CSS indicates that the first generation was grown under control conditions followed by two generations in stress treatment etc., while in red HHH refers to three generations of heat treatment, etc. In G0, only one individual was grown, whereas in G1 to G3, 25 individuals per treatment were grown. Plants were propagated by single seed descent, and in case of mortality, missing plants were restocked by duplicating other lines of the same treatment.
Figure 2.
Correlation analysis to select a subset of independent phenotypic traits.
The heat map shows pairwise Pearson correlations between phenotypic traits (darker colors denote stronger correlations), with data from the heat stress experiments above the diagonal in red and data from the salt stress experiment below the diagonal in blue. Phenotypic traits written in bold letters were selected for statistical analyses.
Figure 3.
Principal component analysis of data from the heat experiment.
Colors denote different genotypes, indicating their distinct phenotypes.
Figure 4.
Effect of stress treatments on rosette leaves at FFD.
A: Col, Ler and Sha grown in G3 under salt (blue) and control (white) conditions. B: Col, Ler, Sha and Cvi grown in G3 under heat (red) and control (white) conditions. ** P-value < 0.01; * P-value < 0.05, · P value < 0.1, ns: P-value not significant. P-values were adjusted for multiple testing according to Benjamini and Hochberg [43].
Figure 5.
Transgenerational effects of two stress generations displaying three different modes of adaptation to novel environments.
A: Tracking: in genotype Sha two generations of ancestral heat treatment (orange) led to more rosette leaves after three weeks in comparison to ancestral control treatment (white). B: Transgenerational phenotypic plasticity: in genotype Sha, two generations of ancestral salt treatment (blue) led to increased number of siliques per branch when compared to ancestral control conditions (white) under G3 salt conditions, but not G3 control conditions. C: Diversified bet-hedging: in genotype Cvi ancestral heat treatment (orange) increased the variance of total siliques in comparison to ancestral control treatment (white).
Figure 6.
Transgenerational effects of heat on rosette leaves at FFD in genotype Cvi indicate diversified bet-hedging.
A: Entire data-set. No differences in number of rosette leaves at FFD were found between G3 treatments when plants experienced the same ancestral treatment. B: For each G1G2 × G3-treatment combination the five plants with the highest silique number were chosen. In plants with ancestral heat treatment, number of rosette leaves differed significantly between G3 heat (HHH) and G3 control (HHC) treatment. When plants were control treated in G1 and G2, no differences were found between G3 treatments. C: For each G1G2 × G3-treatment combination the five plants with the lowest silique number were chosen. No differences in numbers of rosette leaves were found between G3 plants when plants experienced the same ancestral treatments.