Degree of herbivore feeding damage as an important contributor to multitrophic plant-parasitoid signaling under climate change

Abstract

Biogenic volatile organic compounds (VOCs) serve as signals mediating information between plants and their higher trophic level beneficials, such as parasitoids and predators of herbivores. We recently demonstrated with oilseed rape (Brassica napus L.) plants, herbivorous diamond-back moth (Plutella xylostella L. (Lepidoptera: Yponomeutidae)) larvae and Cotesia vestalis (Haliday) (Hymenoptera: Braconidae) parasitoids that atmospheric pollution, i.e. elevated ozone (O₃), can disturb attraction of natural enemies by plant-emitted host-induced volatile cues. Additionally, we found that the degree of herbivore feeding damage is an important contributor to this O3 interference. low feeding damage in herbivore-resistant plants was sufficient to attract C. vestalis females to host-damaged plants under ambient air, but this tritrophic signaling turned non-functional in the combination of low feeding damage and high O₃ concentration. Here we present some additional data of how climate change factors may modify herbivore feeding and their growth. We further discuss how the degree of herbivore feeding damage and the tritrophic signaling interaction relaying on the herbivore-induced VOCs from attacked plants might change through direct and indirect effects of increased levels of carbon dioxide, temperature and O₃.

This article refers to:

The global atmospheric carbon dioxide (CO₂) level is the highest met in the last 650,000 years and continues to contribute to the global average temperature rise.1 Regional ozone (O₃) concentrations have increased markedly due to environmental pollution,1 and already act phytotoxically towards sensitive plant species.2 Plants emit various types of volatile organic compounds (VOCs) with diverging ecological roles into their surrounding atmosphere.3 Higher trophic level parasitoids and predators utilize herbivore-inducible VOC cues, i.e., compounds emitted after feeding leading to quantitative or even qualitative differences in the volatile profile, to separate host-carrying from intact plants.4 Elevated CO₂ has most often caused seasonal reductions in the biogenic emissions of terpenoids,5,6 which are among main VOCs affecting plant-herbivore-parasitoid interactions.3,4 O₃ directly reacts with and degrades O₃-sensitive terpenoids, yet certain O₃-resistant VOCs, such as 1,8-cineole, will remain unaffected.7 However, disturbance of multitrophic VOC signaling by changes in CO₂ or O₃ levels have yet seldom been reported.7,8 One example exists: elevated CO₂ disturbing orientation of Cotesia vestalis to diamond-back moth (Plutella xylostella L.)-damaged cabbage (Brassica oleracea ssp. capitata) plants.9 Nevertheless, more experimentation is definitely needed. Previously, O₃ has not affected the orientation of this parasitoid to host-damaged cabbage plants,7 which demonstrated its sophisticated olfactory ability to use specific O₃-resistant cues for host finding in Brassicas. Yet another important aspect is the active role of the herbivore in plant-parasitoid interactions affected by elevated CO₂ or O₃. This was studied recently by us with the same parasitoid species and experimental system, but including herbivore-resistant and -sensitive oilseed rape (Brassica napus L.) plants.10

Plant resistance to herbivores is an important target of breeding allowing combating against rising pest problems in agriculture. The first commercial transgenic insect-resistant plants produce specific Bacillus thuringiensis (Bt) toxins contributing resistance against their target insects.11 In our recent work we used Lepidoptera-resistant Bt oilseed rape plants to compare the herbivore-induction of VOCs from these and their herbivore-sensitive parent line plants under elevated CO₂ or O₃ level.10 Interestingly, we showed that O₃ interference of volatile induction became relevant with the resistant Bt plants. Even though the resistant plants allowed feeding initiation that enabled the parasitoids to find host-damaged resistant plants under ambient air, elevated O₃ rendered these emissions undistinguishable for the parasitoid compared to intact plant emissions. As important for the sustainable use of herbivore-resistant cultivars, either of genetically modified or traditionally bred origin, their wider use in O₃-polluted areas might therefore result in lowered natural enemy population abundance (with the example of C. vestalis) by lowered VOC attraction and also by direct negative effects by reduced host density or poorer host quality.12 However, as this type of bred resistance is a means of specific herbivore control, comparative direct control by pesticides might have even more negative effects.13 Importantly, mechanical damage has induced similar VOC emissions from Bt and non-Bt line plants.14 The unchanged ability for induction by non-target herbivores might assist in attracting beneficials, which could be killed when using non-specific pesticides.

Another interesting perception raised by our previous work was that in general, the changes in the degree of herbivore feeding damage by climate change factors potentially affecting multitrophic signaling has been largely understudied. Predicting effects of changed CO₂ and temperature or O₃ on volatile-mediated multitrophic interactions is highly complex (Fig. 1), since their effects can act on more than only one level of the food chain and might further interact. For example, alterations in host plant quality can change the degree or pattern of herbivore feeding, which could lead to an altered defence response in the plant itself, such as increased release of VOCs. Then, altered host quality may affect the population development of the third trophic level parasitoid and VOC cues guiding to host may be different.

It is known that elevated CO₂ often causes compensatory feeding through reduced plant nutritional quality, e.g., by lowered nitrogen and higher carbon content,15,16 which might lead to increased VOC emissions, assuming that feeding damage is directly proportional to the amount of emissions. We detected higher induction of emissions from oilseed rape under elevated CO₂, yet the degree of feeding damage caused in 48 hours was not different between the control CO₂ and doubled CO₂-grown plants.10 However, if feeding patterns were recorded for a longer time, differences between the CO₂ treatments might have arisen. In another experiment on oilseed rape, we measured the leaf area consumed by diamond-back moth larvae and their mean relative growth rate under elevated CO₂, elevated temperature and combined elevated CO₂ and temperature. Again, there was no difference between the CO₂ treatments of this 48 h trial (Fig. 2). However, the clear increase by elevated temperature in both parameters measured suggests that direct effects by climate change-related abiotic factors may cause rapid changes in herbivory, whereas plant-mediated effects predominantly appear over longer time scales.16 These might have longer-lasting seasonal effects, yet rapid changes by direct effects can be more unpredictable and could have even dramatic regional consequences.

Effects of O₃ on VOC signaling are complicated, since O₃ has been reported to induce volatile emissions8,17,18 and thus potential interference with herbivore-induced VOCs is possible, although this was not observed in a tritrophic interaction between lima bean (Phaseolus lunatus) plants, two-spotted spider mites (Tetranychus urticae) and their specialist predators.8 O₃ usually makes leaves more palatable to herbivores by increased solubility and availability of nitrogenous compounds,19,20 though the possibility of direct negative effects by high O₃ on feeding of P. xylostella larvae was recently suggested.21 The challenge of working with highly reactive O₃ and VOCs requires sophisticated experimental set-ups,7 which to date has restricted more general findings of VOCs and multitrophic communication in field set-ups with elevated O₃.22 Recent reports that root herbivory affects shoot VOC signaling and behavior of aboveground natural enemies23 or vice versa,24 allows speculating that climate change factors might affect defence signaling between belowground and aboveground plants parts through their known effects on within-plant assimilate partitioning.25 Simultaneous analysis of belowground and aboveground VOC interactions in CO₂ and O₃ studies might help to unravel whether the growth-defence trade-off or altered within-plant resource allocation affects VOC signaling.

To conclude, we underline the importance of the degree of herbivore feeding damage responding directly to elevated CO₂ or O₃ or indirectly to plant-mediated changes, when considering functioning of multitrophic communication in future atmospheres. An increasing number of multitrophic systems, representing different genera of plants and feeding guilds of insects, as well as both direct and indirect effects of CO₂ and O₃ need to be included in order to evaluate overall effects by these abiotic factors on VOC signaling. The various ecological roles suggested for volatile signals emitted, and being detected by many types of perceivers,3 make predictions of potential changes in volatile infochemical networks by climate change an important task to be figured out for allowing understanding of multitrophic interactions in future crop and natural ecosystems.

Figures and Tables

Figure 1 Proposed effects of elevated temperature, carbon dioxide or ozone concentrations on three different trophic levels, which in turn might affect multitrophic volatile infochemical signaling. The example illustrates a Brassica plant, the specialist herbivore Plutella xylostella L. and its endoparasitoid Cotesia vestalis.

Figure 2 Mean relative growth rate ± SEM (bars) of diamondback moth (Plutella xylostella L.) larvae and the leaf area consumed ± SEM (numbers above bars, mm²) by these larvae feeding on 19-d old vegetative stage oilseed rape (Brassica napus L.) plants grown and tested under control CO₂ (360 µl l⁻¹) or elevated (720 µl l⁻¹) CO₂ and control (16/20°C) or elevated (20/24°C) temperature. Different letters indicate statistically significant differences between treatments (ANOVA followed by LSD post hoc test), n = 6. Treatments were performed as in reference 26 and mean relative growth rate and feeding area determined as in reference 21.

Acknowledgements

This work was supported by the Academy of Finland (ESGEMO Programme, decision no. 105209).

Addendum to: