Plant‐mediated effects of drought stress on host preference and performance of a grass miner

Abstract

The plant stress hypothesis predicts that environmental stress increases the suitability of plants as food for herbivores, especially for senescence feeders. Yet, performance is enhanced only at moderate stress intensities in several herbivores. Even more paradoxically, a large number of insect species prefer and perform better on vigorously growing plants. In order to test plant stress theory, we conducted a laboratory experiment in which the influence of plant water stress on host preference and the performance of the grass miner Chromatomyia milii was studied. We imposed a gradient of stress intensities, i.e. 25, 50, 75, 150, and 300‐ml weekly‐administered water per grass pot, in order to study the full range of responses of C. milii to water stressed plants. Plant stress intensity was quantified by measuring individual plant mass, foliar water content and the concentration of the photosynthetic pigments chlorophyll a and b. Plant mass had decreased from the 150 and 300‐ml treatments to the lowest water treatment at the end of our experiment, which was mainly a result of a reduction in leaf area and leaf number. Foliar water content was clearly negatively affected by water shortage. Chlorophyll a and b also decreased with water shortage. Finally, the stress intensity measurements showed that plants acclimated to water stress conditions throughout the experiment. Feeding and oviposition preference of C. milii was positively related to water supply. No larvae survived on two lowest water treatments and only 38% survived on the 75‐ml treatment, while more than 80% survived on the 150 and the 300‐ml treatments. Offspring development time was longer on the 75‐ml treatment than on the 150 and the 300‐ml treatments. We also evaluated the mechanisms that could explain the response of C. milii to water stressed plants. Although no relationship between water treatment and foliar amino acid concentration was found, we observed significantly higher foliar protein concentrations in the 25 and the 50‐ml treatments. This supports the hypothesis that abiotic stress causes an increase of nitrogenous compounds in plants. Leaf senescence following self‐pruning, a process by which H. lanatus plants acclimate to drought conditions, was responsible for the dramatically high offspring mortality on the water stressed plants. The shape of the plant stress intensity–herbivore response relationship showed strong variation and depended both on the type of plant stress intensity measure and herbivore response variable involved. Yet, all relationships showed a monotonic increase of herbivore preference and performance with decreasing plant stress intensity. This indicates that C. milii prefers and performs better on vigorously growing plants. We found no support for an increased herbivore performance on moderately or severely stressed plants.