Elevated CO2 affects bottom-up and top-down driver impacts in plant-insect interaction

Background/Question/Methods

Bottom-up (i.e. nutrient availability) and top-down (i.e. herbivore damage) drivers affect plant-insect interactions, however human impacts such as elevated atmospheric CO₂ concentrations and temperatures are changing the context of all ecological interactions.  Will these drivers affect systems in the same way under future CO₂ concentrations?  This work addresses how these drivers affect plant phenotype and specialist herbivore performance under current atmospheric CO₂ and whether future CO₂ concentrations will alter plant and herbivore responses.  We grew 328 mustard Brassica nigra and collard B. oleracea plants in an outdoor chamber array, applying CO₂ treatments at the chamber level and nutrients and simulated herbivory at the plant level.  We considered the effects of treatments on plant phenotypes (N%, H₂O%, and glucosinolate concentrations) under ambient and future atmospheric CO₂ levels using MANOVA.  We then conducted 235 cabbage white butterfly Pieris rapae feeding trials and evaluated the effects of plant treatments on larval behavior and performance using MANOVA.  We also evaluated the importance of plant nutritional quality, chemical defense, and mechanical defense on larval behavior and performance using exploratory structural equation modeling (SEM).  SEM allowed us to consider how plant phenotypic characteristics affect larvae independent of plant experimental treatment.

Results/Conclusions

The effects of bottom-up and top-down drivers will change under elevated CO₂ and responses are host-plant specific.  Elevated CO₂ decreased N and H₂O in both mustard and collard.  Low nutrient availability did not affect mustard N or H₂O content but decreased both in collard.  Herbivory increased N and H₂O in mustard but decreased both in collard.  Low nutrient availability decreased total glucosinolate concentrations in mustard and collard now and under elevated CO₂, but herbivory and elevated CO₂ had no effect.  Elevated CO₂ changes Pieris rapae response to both herbivory and nutrient availability.  Glucosinolate concentrations did not affect larvae, but sclerophylly decreased consumption, food utilization, and growth.  Larvae will increase consumption rates under elevated CO₂ on both mustard and collard, but only collard-hosted larvae will have lower growth.  Collard is a poorer host plant for P. rapae currently and future CO₂ concentrations will exacerbate this difference.  Elevated CO₂ will affect plant-insect interactions by increasing larval consumption rates and decreasing larval growth rates (host-specific).  Elevated CO₂ will also change mustard phenotypic responses to herbivory; induced susceptibility will become induced resistance.  Together these results suggest that elevated CO₂ will impact relative host plant suitability, bottom-up, and top-down plant-insect interactions in the next century.