Testing the joint effects of arbuscular mycorrhizal fungi and ants on insect herbivory on potato plants

Abstract

Ants, but not mycorrhizae, significantly affected insect leaf-chewing herbivory on potato plants. However, there was no evidence of mutualistic interactive effects on herbivory. Plants associate with both aboveground and belowground mutualists, two prominent examples being ants and arbuscular mycorrhizal fungi (AMF), respectively. While both of these mutualisms have been extensively studied, joint manipulations testing their independent and interactive (non-additive) effects on plants are rare. To address this gap, we conducted a joint test of ant and AMF effects on herbivory by leaf-chewing insects attacking potato (Solanum tuberosum) plants, and further measured plant traits likely mediating mutualist effects on herbivory. In a field experiment, we factorially manipulated the presence of AMF (two levels: control and mycorrhization) and ants (two levels: exclusion and presence) and quantified the concentration of leaf phenolic compounds acting as direct defenses, as well as plant volatile organic compound (VOC) emissions potentially mediating direct (e.g., herbivore repellents) or indirect (e.g., ant attractants) defense. Moreover, we measured ant abundance and performed a dual-choice greenhouse experiment testing for effects of VOC blends (mimicking those emitted by control vs. AMF-inoculated plants) on ant attraction as a mechanism for indirect defense. Ant presence significantly reduced herbivory whereas mycorrhization had no detectable influence on herbivory and mutualist effects operated independently. Plant trait measurements indicated that mycorrhization had no effect on leaf phenolics but significantly increased VOC emissions. However, mycorrhization did not affect ant abundance and there was no evidence of AMF effects on herbivory operating via ant-mediated defense. Consistently, the dual-choice assay showed no effect of AMF-induced volatile blends on ant attraction. Together, these results suggest that herbivory on potato plants responds mainly to top-down (ant-mediated) rather than bottom-up (AMF-mediated) control, an asymmetry in effects which could have precluded mutualist non-additive effects on herbivory. Further research on this, as well as other plant systems, is needed to examine the ecological contexts under which mutualist interactive effects are more or less likely to emerge and their impacts on plant fitness and associated communities.

Keywords: Solanum tuberosum; Ants; Arbuscular mycorrhizal fungi; Phenolic compounds; Plant defenses; Volatile organic compounds.

Fig. 1

Schematic representation of one block in the experimental design, showing the ant exclusion treatment (two levels: exclusion and presence) as the whole factor; mycorrhizal treatment (two levels: control and mycorrhization) as the split factor; and three plant genetic entries (i.e., varieties) as the split–split factor. Each experimental unit (the combination of mycorrhizal and ant exclusion treatments) consisted of nine plants in three parallel rows of three plants each (three of each potato variety)

Fig. 2

Effects of mycorrhizal treatment (two levels: control and mycorrhization) on the concentration (in mg g⁻¹ DW) of leaf flavonoids (A) and caffeic acids (B), and total amount (in nanograms h⁻¹) of VOCs (C) emitted by potato (Solanum tuberosum) plants. Bars are least square means ± SE obtained from linear models (n = 72 for phenolics and n = 24 for VOCs). Different letters above the bars indicate significant differences (P < 0.05) among mycorrhizal treatments. Statistics are shown in Table 1

Fig. 3

Effects of mycorrhizal treatment (two levels: control and mycorrhization) on A ant abundance (ants plant.⁻¹) and B the percentage of leaf area consumed by chewing insects on potato (Solanum tuberosum) plants with ant presence (white bars) or subjected to an ant exclusion treatment (grey bars). Bars are back-transformations of log-transformed least-square means ± SE obtained from linear mixed models (n = 36). Different letters above the bars indicate significant differences (P < 0.05) among mycorrhizal and ant exclusion treatments. Statistics are shown in Table 2

Fig. 4

Ant (Lasius niger) attraction (measured as odds values) to synthetic blends of VOCs which mimicked the emission rates in response to mycorrhization (blue bars) or control (grey bars) emissions based on dual-choice assay using an olfactometer (see Methods). Bars are the mean odds ratio value (± SE) for each VOC exposure treatment obtained from a generalized linear mixed model (n = 20 assay replicates, see methods of experiment 2). Odds values were calculated as the ratio between successful and unsuccessful events (i.e., likelihood of an ant being attracted vs. not attracted to a blend type) for each replicate of each treatment level