4.2 Plant diversity and soil legacy effects on herbivore-induced
responses
We also observed alterations in the herbivore-induced metabolomic
response due to plant diversity and soil legacy. Together, plant-plant
and plant-soil interactions regulated 82 metabolites in control plants
and 141 metabolites in herbivore-induced plants.
As shown above, plant-plant interactions can modulate growth-defense
trade-offs that likely vary in strength with changes in plant diversity.
In mixed communities, a combination of niche complementarity but
increased competition for light, as well as a reduction of herbivory by
specialized herbivores via dilution effects, may lead to a higher
investment of resources into growth than defense compared to
monocultures (Castagneyrol et al., 2014; Eisenhauer et al., 2019; Finch
& Collier, 2000; van Moorsel et al., 2018). In fact, earlier work
revealed that plants growing in mixed communities invested more
resources into growth than defense-related metabolites compared to
plants growing in monoculture (Broz et al., 2010), potentially reducing
herbivore resistance. While we did not find differences in the overall
metabolome composition of herbivore-induced plants in response to
increasing plant diversity, we observed induced metabolite regulation in
mixed communities. Our results suggest that the identity of the
neighboring plant species determines the extent and direction of the
plant-plant interaction. This has potential consequences for our
understanding of plant-herbivore interactions in mixed communities, but
further research is needed to confirm this hypothesis.
Plant-soil interactions, on the other hand, can prepare a plant for
future attack, also called priming (Conrath et al., 2006). Systemic
priming in plants can occur following interactions with soil microbes,
nematodes, and mycorrhizal fungi, allowing the plant to better respond
to subsequent herbivory (Kaplan et al., 2008; Martinez-Medina et al.,
2016). While we have not explicitly tested for priming, it may explain
why the absolute number of up-regulated metabolites in herbivore-induced
plants (in comparison to control plants) was highest when plants had
grown in different soil legacies. However, other possible mechanisms,
such as systemic acquired resistance to microbial pathogens, exist that
could also explain the patterns of metabolite regulation in our study
(Ryals et al., 1996).
Finally, we observed differences in the regulation of herbivore-induced
metabolites among our plant species. In R. acris plants,
plant-plant interactions resulted in a strong down-regulation of induced
metabolites, while plant-soil interactions resulted in a strong
up-regulation of induced metabolites. The response to either type of
biotic interaction was much more attenuated in G. pratense andL. vulgare , suggesting differences in the plant species-specific
adaptability which requires future research before general assumptions
can be made on the effects of plant diversity versus soil legacy
on herbivore resistance.
While the present experiment provides novel insights into how
metabolomic profiles, and thereby herbivore resistance, respond to
changes in plant and soil biodiversity, it also calls for future
studies. To allow for the comparison of plant-plant and plant-soil
interactions in our study, we inoculated sterile substrate with liquid
field soil inoculum from the Trait-Based Experiment (Ebeling et al.,
2014) in the PSI experiment. This, however, meant that the soil biota
communities were adapted and “linked” to the plot-specific plant
communities and that the sand-peat mixture that was used may have
created a different environment than the one the microbes were
accustomed to. To fully disentangle plant from soil biodiversity effects
on the plant metabolome, one would need to expose plants to artificially
constructed soil communities (see e.g. de Souza et al., 2020), also
including larger soil organisms (see e.g. Lohmann et al., 2009). While
this was not feasible in the scope of this study, it would also be
important to explore the specific effects of pre-selected functional
soil biota groups, such as nematodes (e.g. Bezemer et al., 2005).
Moreover, future studies should explore potential shifts in
growth-defense trade-offs in more detail by exploring the performance of
plants and herbivores. To our knowledge, this kind of comparable
experimental design to disentangle plant-plant and plant-soil
interaction effects has rarely been employed (but see Kos et al., 2015)
and results and conclusions can vary between studies. Hence, we advocate
for additional experiments of that kind to generate the necessary data
for more reliable conclusions.