Too hot to handle: temperature-induced plasticity influences pollinator behaviour and plant fitness.

Increased temperature can induce plastic changes in many plant traits. However, little is known about how these changes affect plant interactions with insect pollinators and herbivores, and what the consequences for plant fitness and selection are. We grew fast-cycling Brassica rapa plants at two temperatures (ambient and increased temperature) and phenotyped them (floral traits, scent, colour and glucosinolates). We then exposed plants to both pollinators (Bombus terrestris) and pollinating herbivores (Pieris rapae). We measured flower visitation, oviposition of P. rapae, herbivore development and seed output. Plants in the hot environment produced more but smaller flowers, with lower UV reflectance and emitted a different volatile blend with overall lower volatile emission. Moreover, these plants received fewer first-choice visits by bumblebees and butterflies, and fewer flower visits by butterflies. Seed production was lower in hot environment plants, both because of a reduction in flower fertility due to temperature and because of the reduced visitation of pollinators. The selection on plant traits changed in strength and direction between temperatures. Our study highlights an important mechanism by which global warming can change plant-pollinator interactions and negatively impact plant fitness, as well as potentially alter plant evolution through changes in phenotypic selection.

Herbivory and jasmonate treatment affect reproductive traits in wild Lima bean, but without transgenerational effects.

PREMISE: Plant responses to herbivores and their elicitors include changes in traits associated with phenology, defense, and reproduction. Induced responses by chewing herbivores are known to be hormonally mediated by the jasmonate pathway and can cascade and affect late-season seed predators and pollinators. Moreover, herbivore-induced plant responses can be transmitted to the next generation. Whether herbivore-induced transgenerational effects also apply to phenological traits is less well understood. METHODS: Here, we explored responses of wild lima bean plants (Phaseolus lunatus) to herbivory and jasmonate treatment and possible transgenerational effects of herbivore-induced early flowering. In a controlled field experiment, we exposed lima bean plants to herbivory by leaf beetles or methyl jasmonate sprays (MJ). We then compared plant development, phenology, reproductive fitness and seed traits among these treatments and undamaged, untreated control plants. RESULTS: We found that MJ and leaf herbivory induced similar responses, with treated plants growing less, flowering earlier, and producing fewer seeds than undamaged plants. However, seed size, phenolics and cyanogenic glycosides concentrations did not differ among treatments. Seed germination rates and flowering time of the offspring were similar among maternal treatments. CONCLUSIONS: Overall, the results confirm that responses of lima bean to herbivory by leaf beetles are mediated by jasmonate; however, effects on phenological traits are not transmitted to the next generation. We discuss why transgenerational effects of herbivory might be restricted to traits that directly target herbivores.

Host density and parasitoid presence interact and shape the outcome of a tritrophic interaction on seeds of wild lima bean.

The interaction between the seed beetle Zabrotes subfasciatus and its parasitoid Stenocorse bruchivora, was investigated on seeds of two populations of wild lima bean, Phaseolus lunatus. By manipulating the number of beetle larvae per seed and the presence of parasitoids, we determined how factors related to beetle larvae density, the seed in which they feed and the parasitoid, may interact and affect host and parasitoid survival. Results showed that an increase in larval beetle density had a negative impact on beetle performance. This effect cascaded up to parasitoids, high larval density strongly reduced parasitoid emergence. Also, parasitoid presence resulted in faster beetle development and lower female weight. An interactive effect between larval host density and parasitoid presence affected the number of insects that emerged from the seeds. Beetle performance was better in the bean population with the largest seeds, while parasitoid emergence was the lowest in these seeds. This study shows that the impact of parasitoids on seed beetles is contingent on the interaction between density-mediated (direct mortality) and trait-mediated (e.g. non-consumptive) effects. Indirect trait-mediated effects of natural enemies are likely prevalent across insect communities, understanding their role in driving host-parasitoid interactions can have important implications for biological control.