Skip to main content
Dryad logo

Interactions between silicon and alkaloid defences in endophyte-infected grasses and the consequences for a folivore


Cibils-Stewart, Ximena et al. (2021), Interactions between silicon and alkaloid defences in endophyte-infected grasses and the consequences for a folivore, Dryad, Dataset,


1. Grasses have developed a wide range of morphological and physiological mechanisms to resist herbivory. For instance, they accumulate silicon (Si) in tissue, as physical defence, and associate symbiotically with foliar Epichloë-endophytes that provide chemical defence via antiherbivore alkaloids. Recent evidence showed that some Epichloë-endophytes increase foliar Si in forage grasses; however, it is unknown whether this impacts insect herbivores. Furthermore, while Si is primarily a physical defence, it also affects production of plant defensive secondary metabolites; Si supply might therefore affect Epichloë-alkaloids, although this remains untested.

2. We grew endophyte-free (Nil) and Epichloë-infected tall fescue and perennial ryegrass in a factorial combination with or without Si supplementation, in the absence or presence of Helicoverpa armigera. Epichloë-endophyte strains were AR584 for tall fescue, and AR37, AR1 or Wild-type (WT) for perennial ryegrass. We assessed how Si supply and Epichloë-endophytes in interaction with herbivory affected foliar Si and mutualist-derived alkaloid concentrations. Subsequently, their effects on H. armigera relative growth rates (RGR) were evaluated. 

3. Endophytes generally increased Si concentrations in Si supplied plants. In tall fescue-AR584 and perennial ryegrass-AR37, endophytes increased constitutive (herbivore-free) and induced (herbivore-inoculated) Si concentrations by at least 25%; in contrast, in perennial ryegrass the AR1 endophyte only increased constitutive levels. Si supply did not affect alkaloids produced by AR584- or AR1/WT-endophytes; however, in the presence of herbivory, Si supply decreased the induction of alkaloids produced by AR37 endophytes by 33%. For tall fescue, Si supply reduced H. armigera RGR by at least 76%, regardless of endophytic status, whereas, endophyte-alkaloids played a secondary role only reducing herbivore growth in the absence of Si supply. Conversely, in perennial ryegrass both Si and endophyte-alkaloids (regardless of Si supply) reduced herbivore RGR although not synergised. 

4. Novel interactions between constitutive and induced Si- and alkaloid-based antiherbivore defences in grasses were observed. Overall, Si had a greater effect on the folivore than endophytes in both grasses. Endophyte-defences contributed more to herbivore resistance in perennial ryegrass than tall fescue. We demonstrate that Si and endophytes were not antagonistic and highlight that the protective nature of their interaction varies with the grass-endophyte species tested.

Usage Notes

Dataset description

ALL DATA.cvs- metadata utilized for tall fescue and perennial ryegrass respectively; This metadata was subsetted using R for analysis as described in the methods of the manuscript:

"To determine individual treatment effects, both plant growth traits and primary chemistry were analyzed for each species separately, using a two-way ANOVA with Si supply and endophyte status as fixed effects." Here we utilized herbivore-free plants only. 

"Likewise, shoot Si concentration was analyzed for each grass species separately, with a three-way ANOVA using Si supply, endophyte status, and herbivory as fixed effects." Here we used the full data set for each grass species. 

"A multivariate analysis of variance (permutational MANOVA) using Si supply and herbivory treatments as fixed effect factors was utilized to determine differences in overall strain-specific alkaloid profiles (sum of all alkaloids produced by each endophyte-strain), using the ‘vegan’ package (Oksanen et al., 2019). The MANOVA function further provided results for univariate ANOVAs for treatment effects on individual alkaloids at the strain-specific level. " Here we utilized endophyte-infected plants only as endophyte-free plants do not produce alkaloids. 

"Herbivore RGR were analyzed for each species separately, utilizing herbivore-inoculated plants only with a two-way ANOVA with Si supply and endophyte status as fixed effects." Here we utilized herbivore-inoculated plants only as HGR is a herbivore parameter.

FES.ES.cvs & RYE.ES.cvs- data sets utilized to perform the effect size for tall fescue and perennial ryegrass respectively; as described in the methods of the manuscript:

" compare the magnitude of the effect of Si supply and herbivory on strain-specific alkaloid production the standardised effect sizes (SES; Hedges’ d) was calculated for each grass species separately using only endophyte-infected plants as a standardised mean difference ((SMD) = (µ1 – µ2)/s, where µ1 and µ2 are the mean trait value of µ1 = herbivore-inoculated and µ2 = herbivore-free, and s is the pooled standard deviation) using the ‘effsize’ package (Torchiano, 2020). A 95% confidence interval bar that does not overlap zero indicates a significant effect; negative effect size indicates the effect decreases mean alkaloid production, and a positive effect size indicates that the effect increases mean alkaloid production. The SES was utilized to compare each strain-specific alkaloid production regardless of the amount produced (remove units) to describe the size of the effect"


Australian Research Council, Award: DP170102278