Viruses can affect plant-insect interactions by altering emission of plant volatile organic compounds (VOCs). Previous work in the lab suggested that VOCs emitted by tomato (Solanum lycopersicum) plants infected with cucumber mosaic virus (CMV) were more attractive to bumblebees (Bombus terrestris) in free choice assays. I extended this work using Arabidopsis thaliana mutants with lesions in genes encoding factors in RNA silencing. In conditioning assays, I confirmed that plant VOC emission is controlled in part by the microRNA regulatory pathway.

I used gas chromatography coupled to mass spectrometry and principal component analysis to confirm that CMV infection caused changes in VOC emission by tomato. VOCs collected from non-flowering mock-inoculated and CMV-infected plants were qualitatively distinct from each other. CMV-infected plants also released greater quantities of VOCs than mock-inoculated plants. CMV appears to be both ‘turning up the volume’ of plant volatile emission, whilst ‘tuning’ volatile blend composition so as to diminish levels of a repellant signal. These data are likely to explain how bumblebees can discriminate between VOCs emitted by mock-inoculated and CMV-infected plants.

To determine if CMV infection of tomato plants affects plant reproductive success, I carried out a series of bumblebee pollination experiments. Bumblebees pollinate tomato by ‘buzzing’ (sonicating) the flowers, which releases pollen and enhances self-fertilization and seed production as well as pollen export. First, I established that CMV-infected tomato plants produced fruits with a lower seed yield than mock-inoculated plants. When single bumblebees were allowed to buzz-pollinate flowers in a small flying arena, the fruit that developed from buzz-pollinated flowers on virus infected plants had significantly more seeds than fruit from non-visited flowers. Subsequent experiments were performed in a large flying arena. Bumblebees consistently spent longer foraging on the mock-inoculated tomato plants but seed yield was increased by bumblebee pollination in both mock-inoculated as well as virus-infected tomato plants. However, although buzz-pollination significantly enhanced seed yield from CMV-infected plants compared to fruit from non-buzz-pollinated flowers, the yield was higher in buzz-pollinated fruits from mock-inoculated plants. Similar experiments were carried out utilizing a transgenic line of tomato that constitutively expresses the green fluorescent protein in order to estimate the level of cross-pollination from either CMV-infected plants to mock-inoculated plants or vice versa. More pollen from virus-infected plants was transferred to mock-inoculated plants than the reciprocal cross. However, some caution is needed in the interpretation of the larger scale experiments because the tomato plants were affected by a fungal infection.

I investigated if the defensive plant hormone salicylic acid (SA) affects bee-perceivable VOC emission. Exogenous SA treatment renders non-flowering tomato plants more attractive to bumblebees in free choice experiments in which bees could only perceive VOCs, but bumblebees spent less time visiting SA-treated flowering tomato plants in the glasshouse (when the bumblebees were allowed unimpeded access to the flowers).

Taken together, these data provide evidence that virus infection can affect host-pollinator interactions. Speculatively, CMV infection may change the fitness of susceptible plants via changes in production of pollinator-attractive VOCs and this may affect the balance of resistant or susceptible plants within the host population.