Supporting Information Legends
Figure S1. Concentrations of caffeoylputrescine and TPI activity correlate with feeding
damage (exp. I). (A, B) Scatter plots for concentrations of caffeoylputrescine and (C, D) TPI
activity in relation to feeding damage by Spodoptera exigua larvae on ovipositionexperienced (red) and unexperienced (gray) Nicotiana attenuata plants. Lines and correlation
coefficients refer to linear regression of the metabolite concentrations and feeding damage
either 4 days (left: A, C) or 6 days (right: B, D) after larvae had started feeding.
Figure S2. Spodoptera exigua larval sub-stages within each instar and egg deposition on
the plant. During each larval instar three larval sub-stages were defined: (A) Freshly molted;
this stage is characterized by a large head capsule in relation to the body diameter; (B)
Actively feeding and consequently the body diameter increasingly exceeds that of the head
capsule; (C) Immobile larvae that stopped feeding and prepare to molt are characterized by a
very small head capsule in relation to their body. (D) Typically S. exigua moths preferred
elder leaves at the bottom of the rosette for oviposition (arrow). These leaves are very small
and hatching larvae switch quickly to younger leaves. Therefore effects were studied in leaves
systemic to the egg deposition.
Table S1. Performance parameters (exp. I – III). Mean ± SEM of mortality, mass, and
feeding damage of Spodoptera exigua larvae on oviposition-experienced and unexperienced
Nicotiana attenuata plants in three independent experimental repetitions (exp. I/II/III with
NPlant=6/12/10). Parameters were recorded at different days after larvae had started feeding.
Due to death of all larvae on a plant, number of plant replicates were reduced for the
parameter of larval mass (exp. I/II/III with NPlant=6/6, 8/6, 10/8 unexperienced/oviposited
plants). P-values refer to Fisher’s exact test in case of mortality and Welch’s t-tests in case of
larval mass and paired t-tests in case of feeding damage data. Significant differences between
oviposition-experienced and unexperienced are highlighted in bold.
Table S2. Defence parameters in Nicotiana attenuata with larval feeding (exp. I). Mean
levels ± SEM (NPlant=6) of secondary metabolites and trypsin protease inhibitor (TPI) activity
in leaves 4 and 6 days after Spodoptera exigua larvae had started feeding. We calculated the
sum of all peaks of diterpene glycosides (DTGs) to determine relative amounts in total DTGs.
P-values refer to paired t-tests between oviposition-experienced and unexperienced plants and
significant differences are highlighted in bold.
Table S3. Defence parameters of equally damaged Nicotiana attenuata (exp. IV-V). Mean
levels ± SEM of secondary metabolites and polyphenol oxidase (PPO) activity in leaves of
oviposition-experienced or unexperienced plants that received damage of equal extents. In
experiment IV, Spodoptera exigua herbivory was mimicked (NPlant=7-8) by repetitive
mechanical wounding and addition of S. exigua oral secretions (OS); leaves were harvested 2,
4, and 6 days after the first treatment. In experiment V, S. exigua feeding damage was kept
equal by adjusting number and size of larvae feeding on plants with and without oviposition
experience (NPlant=6). P-values refer to Welch’s t-tests and significant differences between
oviposition-experienced and unexperienced plants are highlighted in bold.
Table S4. GLM of defence parameters of equally damaged Nicotiana attenuata (exp. IV).
General linear mixed model of leaf metabolites and enzyme activities (NPlant=7) with the main
factors Spodoptera exigua oviposition and mimicked herbivory by repetitive wounding and
addition of S. exigua oral secretions (Wound+OS). Significant differences are indicated in
bold; a significant interaction of the main factors indicates oviposition-primed metabolites.
The induction was surveyed as repeated measurement (after 2, 4, 6 days), which was included
in the mixed model (covariance type: first order autoregressive process).
Table S5. GLM of NaMyb8 transcript and phytohormone accumulation in Nicotiana
attenuata (exp. VII). Generalised linear model (with maximum likelihood estimates) of
NaMyb8 transcripts and the phytohormones jasmonic acid (JA) and its conjugate to isoleucine
(JA-Ile) with the main factors: oviposition and larval feeding by Spodoptera exigua as well as
the interaction of both factors (N=5 untreated control plants, 9 plants with only feeding
damage, and 11 each for oviposition experienced plants with and without feeding damage).
Significant differences are indicated in bold.
Table S6. Composition of Spodoptera exigua artificial diet. Alfalfa flour pellets were
disintegrated in ~100 mL hot tap water to pulp before all ingredients except the last three
were stirred in 1.5 L boiled water. After cooling to ~60°C, oil, vitamin mix and formaldehyde
were added. The diet was stored at 4°C for up to 4 weeks. Ingredients were purchased locally
if not stated otherwise: † SIGMA-Aldrich, * Carl Roth.
Table S7. Summary of parameters in conducted experiments. An oviposition-experienced
and the corresponding unexperienced Nicotiana attenuata plant of a replicate received the
same number of larvae but the number of larvae available varied. As defence parameter, leaf
levels of nicotine, caffeoylputrescine, chlorogenic acid, rutin, diterpene glycosides, and
activities of TPI and PPO (in exp. I, IV-V) were examined. In exp. VII, NaMyb8 transcript
and JA, JA-Ile accumulations were determined.
Table S8. Deficiency of defence parameters in ir-Myb8 and ir-PI compared to WT
plants. Mean levels ± SEM of secondary metabolites in feeding-induced leaves of Nicotiana
attenuata wild type (WT) and plants silenced for NaMyb8 (ir-Myb8) or NaPI (ir-PI) gene
expression in two experiments. In Exp. VII, two independently transformed lines ir-Myb8
lines (line 1 (810): NPlant=6, line 2 (818): NPlant=10) showed no differences to WT (NPlant=16)
in other secondary metabolites but caffeoylputrescine. In Exp. VIII, deficiency of
caffeoylputrescine and TPI activity in ir-Myb8 (line 2 (818): NPlant=10) and ir-PI plants
(NPlant=14) respectively was verified relative to WT plants (NPlant=10). Leaves were harvested
4 days after Spodoptera exigua larvae started feeding. Different letters indicate significant
differences according to Welch t-test between WT and transformants (P < 0.05).
Appendix S1. Analysis of TPI activity by radial diffusion assay (exp. I).