Drought effects on plant direct and indirect chemical defenses and its consequences for tri-trophic interactions

Alma Carvajal Acosta, Ecology and Evolutionary Biology, University of California, Irvine, Irvine, CA, Andy Gloss, Ecology & Evolution, University of Chicago, Chicago, IL and Kailen A. Mooney, Ecology and Evolutionary Biology, University of California, Ivine, Irvine, CA

Background/Question/Methods
Drought events are predicted to increase due to climate change, yet consequences for plant–insect interactions are only partially understood. Drought stress can alter plant’s ability to produce defensive chemicals (direct defenses) as well the emission of plant volatiles associated with the attraction of herbivore’s natural enemies (indirect defenses) with potential consequences for tri-trophic interactions. We conducted a field experiment to investigate whether changes in plant chemistry mediates drought effects on tri-trophic interactions in the heartleaf bittercress (Cardamine cordifolia), the herbivorous beetle larvae (Phaedon ssp. oviformis) and its predators. To do so, we used a factorial design combining a trophic-level treatment (plants, plants+herbivores, and plants+herbivores+predator) and a water treatment (drought and ambient). Plants were in their treatments for two weeks after which we measured leaf herbivory and leaf chemicals associated with direct defenses (glucosinolates) and indirect defenses (plant volatiles) using HPLC-MS and GC-MS, respectively. We tested for bottom-up effects of drought (plant to herbivore) on leaf herbivory in plants in the two-trophic-level group. We also tested for top-down effects (predator to herbivore) on leaf herbivory in plants in the two-trophic-level and three-trophic-level groups. To investigate whether such effects are mediated by drought-induced changes in plant chemistry, we tested for the effects of trophic-levels and water treatments on plant’s glucosinolates and volatiles profiles.
Results/Conclusions
In the absence of predators, leaf herbivory (% leaf area consumed) was lower in plants assigned to the drought group than in those in the ambient group, but we detected no significant effect of water treatment on leaf herbivory (p=0.15). However, in the presence of predators, leaf herbivory was lower in those plants in the ambient (3.5%) than in the drought group (6.2%) and we detected an interactive effect of water and trophic-level treatments (p=0.06) suggesting that predators were more effective at reducing herbivory in well-watered plants. We detected 14 glucosinolates and 21 volatile compounds in Cardamine cordifolia plants, with two volatiles emitted at higher levels in plants with herbivores: Butane, 2-isothiocyanato and Isopropyl isothiocyanate. Glucosinolates content was not affected by the water treatment (PCA; p=.25) nor the trophic-level treatment (PCA; p=0.78). Differences among volatiles profiles were driven by the trophic-level treatment (db-RDA; p=0.07) and were not affected by the water treatment (p=0.99). These results suggest that plant water stress may affect predator’s ability to locate herbivorous prey, but further analysis is required to determine the underlaying causes of changes in predator behavior under drought conditions.