Will chemical defenses become more effective against specialist herbivores under elevated CO2?

Background/Question/Methods

Elevated atmospheric CO₂ affects plant-insect herbivore interactions. Elevated CO₂ causes leaf nitrogen to decrease, the ostensible cause of herbivore compensatory feeding. CO₂ may also affect herbivore consumption by altering chemical defenses via changes in plant hormones. Elevated CO₂ can increase salicylic acid (SA) concentrations and decrease jasmonic acid (JA) and ethylene concentrations, potentially causing hormone-mediated CO₂ response (HMCR) of chemical defenses. We considered the effects of elevated CO₂ on constitutive and induced nitrogen-containing glucosinolate concentrations of mustard (Brassica nigra) and collard (B. oleracea var. acephala) at two soil nutrient levels. Constitutive concentrations of aliphatic and indole glucosinolates are promoted by SA and JA respectively; elevated CO₂ should therefore increase aliphatic and decrease indole glucosinolates. Increased concentrations of SA should also suppress upregulation of the JA pathway following herbivory, decreasing glucosinolate induction. Specialist herbivores (those that exclusively consume one plant family) are often incorrectly assumed to be unaffected by their host’s defensive chemistry. We considered the effects of glucosinolate groups on specialist herbivore Pieris rapae consumption at ambient and elevated CO₂. We used structural equation modeling (SEM) to separate the effects of leaf nitrogen and glucosinolate groups on P. rapae consumption at both CO₂concentrations.

Results/Conclusions

Elevated CO₂ affected B. oleracea but not B. nigra glucosinolates; responses to soil fertility and damage were also species-specific. B. oleracea glucosinolate response to soil fertility and damage was also CO₂-specific.  Glucosinolates did not affect P. rapae consumption at either CO₂ concentration in B. nigra, but had CO₂-specific effects on consumption in B. oleracea. At ambient CO₂, leaf nitrogen had strong effects on glucosinolate concentrations and P. rapae consumption but glucosinolate effects on consumption were weak. At elevated CO₂, direct effects of nitrogen were weaker, but glucosinolates had stronger effects on consumption.  Gluconasturtiin and aliphatic glucosinolates were feeding stimulants and indole glucosinolates were feeding deterrents. This refutes the compensatory feeding hypothesis as the sole driver of changes in P. rapae consumption under elevated CO₂. Support for HMCR was mixed: it explained few treatment effects but did explain patterns in SEM models. Further, the novel feeding deterrent effect of indole glucosinolates under elevated CO₂ in B. oleracae underscores the importance of defensive chemistry in CO₂ response. We speculate that P. rapae indole glucosinolate detoxification mechanisms may have been overwhelmed under elevated CO₂, forcing slowed consumption. Specialists may have to contend with hosts with poorer nutritional quality and more effective defenses under elevated CO₂.