Theoretical examination of important features of biological response curves.

A) Parameters of the model include tonset: onset delay, tpeak: time of peak response, tmean: mean response time, Rpeak: the maximum measured response. These parameters are further reparametrised into Duration: metric of the length of response and Shape: our symmetry feature, which for gamma-like distributions falls between zero and one. B) In silico data highlighting the utility of model parameters, namely that discrete temporal response patterns can emerge, despite identical integrals.

Application of the model to compare volatile emission dynamics in response to a range of stimuli.

Each horizontal row of panels represents a unique experiment. A-E) responses to variable wounding intensities, F-J) responses to the same intensity of damage at different times of day (Note: tonset is standardised based on time of damage), K-O) responses compared between wounded plants and plants treated with wounding and Spodoptera exigua oral secretions (OS), P-T) responses to damage in leaves of different developmental stages (Data from Waterman et al., 2025), U-Y) wounding responses compared between genotypes with highly variable volatile emission capacity. For the first column of panels,, curves depict emission data, where the solid line represents mean of fitted emission across biological replicates and the translucent ribbon represents the baseline-subtracted raw emission data ± SE. For the remaining columns, solid points represent mean values across biological replicates (translucent points). Error bars represent SE. Within each panel,12 different letters indicate significant differences between groups as determined by multiple comparisons tests following significant (p < 0.05) one-way n = 3-5.

Impacts of herbivore-specific stimuli on volatile emission dynamics.

Emission and model parameters for wounded plants (A-E) and wounded plants treated with oral secretions (OS; F-J). For the first column of panels (A and F), curves depict emission data, where the solid line represents mean of fitted emission across biological replicates and the translucent ribbon represents the baseline-subtracted raw emission data ± SE. For the remaining columns, solid points represent mean across biological replicates (translucent points). Error bars represent SE. Within each panel, different letters indicate significant differences between groups as determined by multiple comparisons tests following significant (p < 0.05) one-way or Welch’s ANOVA. n = 4-5. Abbreviations: DMNT= 4,8-dimethylnona-1,3,7-triene, MNT = monoterpenes, SQT = sesquiterpenes, TMTT = 4,8,12-trimethyltrideca-1,3,7,11-tetraene.

Fitting responses to complex stimulus patterns.

A) Fitted emission for each curve plotted over total fitted emission and baseline-subtracted raw emission B-E) Model parameters for each peak. F) Curves depict emission data and G-I depict model parameters from Spodoptera exigua-infested plants. For A and F, the translucent ribbons represent the baseline-subtracted raw emission data ± SE from the respective, colour-coded curve. For B-E and G-I, solid points represent mean across biological replicates (translucent points). Error bars represent SE. Within each panel, different letters indicate significant differences between groups as determined by multiple comparisons tests following significant one-way ANOVA. For A-E, n = 6 and for F-I, n = 8-9. Abbreviations: DMNT= 4,8-dimethylnona-1,3,7-triene, MNT = monoterpenes, SQT = sesquiterpenes, TMTT = 4,8,12-trimethyltrideca-1,3,7,11-tetraene.

Dynamics of green leaf volatile (GLV) emissions.

For A) curves depict emission data, where the solid line represents mean of fitted emission across biological replicates and the translucent ribbon represents the baseline-subtracted raw emission data ± SE. For the remaining columns, solid points represent mean across biological replicates (translucent points). Error bars represent SE. Within each panel, different letters indicate significant differences between groups as determined by multiple comparisons tests following significant (p < 0.05) one-way ANOVA. n = 3. Abbreviations: H-al = hexenal, H-ol = hexenol, HAC = hexenyl acetate.

Volatile biosynthesis gene expression dynamics.

For A), curves depict baseline-subtracted raw expression data, where the solid line represents mean of fitted expression across biological replicates and the translucent ribbon represents the baseline-subtracted raw expression data ± SE. For the remaining columns, solid points represent mean across biological replicates (translucent points). Error bars represent SE. Within each panel, different letters indicate significant differences between groups as determined by multiple comparisons tests following significant (p < 0.05) one-way ANOVA. n = 4-5. Abbreviations: CYP92C5 = dimethylnonatriene/trimethyltetradecatetraene synthase, IGL = indole-3-glycerol phosphate lyase, TPS2 = terpene synthase 2, TPS10 = terpene synthase 10.

Fitting incomplete response curves.

Emission and fit parameters for A-E) indole, F-J) DMNT and K-O) TMTT. For the first column of figures, curves depict emission data, where the each solid line represents mean of fitted emission across biological replicates and the translucent ribbon represents the raw baseline-subtracted emission data ± SE. For the remainingg columns, solid points represent mean across biological replicates (translucent points). Error bars represent SE. Colour gradient indicates the number of hours removed from the back end of the curve. n = 5. Abbreviations: DMNT= 4,8-dimethylnona-1,3,7-triene, TMTT = 4,8,12-trimethyltrideca-1,3,7,11-tetraene.

Sampling resolution impacts fits.

Data were removed systematically in periodic intervals across the entire curve. Emission and fit parameters for A-E) indole, F-J) DMNT and K-O) TMTT. For the first row of data, curves depict emission data, where the solid line represents mean of fitted emission across biological replicates and the translucent ribbon represents represents the baseline-subtracted raw emission data ± SE. For the remaining columns, solid points represent mean across biological replicates (translucent points). Error bars represent SE. n = 5. Abbreviations: DMNT= 4,8-dimethylnona-1,3,7-triene, TMTT = 4,8,12-trimethyltrideca-1,3,7,11-tetraene.

Summary of statistical analyses presented in Figure 2 (main text).

Bold values: p < 0.05. Underlined values: p <0.1. a = analysed using one-way ANOVA, b = analysed using Kruskal-Wallis test. * = analysed on log-transformed data, † = analysed using white-adjusted ANOVA.

Summary of statistical analyses presented in Figure 3 (main text).

Bold values: p < 0.05. a = analysed using one-way ANOVA, b = analysed using Welch’s ANOVA, * = analysed on log-transformed data, † = analysed using white-adjusted ANOVA.

Summary of statistical analyses presented in Figure 4 (main text).

Bold values: p < 0.05 and underlined values: p < 0.1. a = analysed using one-way ANOVA, b = analysed using Welch’s ANOVA, * = analysed on log-transformed data, † = analysed using white-adjusted ANOVAs.

Summary of statistical analyses presented in Supplemental figures 1 and 2.

Bold values: p < 0.05. a = analysed using one-way ANOVA, b = analysed using Welch’s ANOVA.