Hummingbird contribution to plant reproduction in the rupestrian grasslands is not defined by pollination syndrome.

Floral traits mediate the roles of distinct animals as effective pollinators along a generalization/specialization continuum. Many plant species are visited by different pollinator functional groups and the specific contribution of each group is expected to reflect the set of floral characteristics defined by pollination syndromes. Although considered a highly specialized nectarivorous group, hummingbirds frequently visit flowers lacking apparent specialization to bird pollination. How they contribute to the reproduction of these plants, however, has not been evaluated through field experiments considering multiple non-related plant species simultaneously. Here, we investigated hummingbirds' contributions to the pollination of ten plant species comprising a gradient of adaptation to bird pollination in the Brazilian rupestrian grasslands. We excluded hummingbirds from flowers and evaluated their relative contribution in comparison to insects (mainly bees) on conspecific/heterospecific pollen deposition and fruit set. Floral traits that are typically associated with bird pollination were associated with increased pollen deposition, but not with fruit set in the presence of hummingbirds. With hummingbirds, conspecific and heterospecific pollen deposition increased in most species, while fruit set increased in four plant species with varying degrees of fit to ornithophily. Our results show that assessing the relative contribution of specific pollinator groups may depend on when this contribution is measured, i.e. pollen deposition or fruit set. Considering fruit set, our results indicate that hummingbirds contributed to plant reproduction independently of the fit to bird pollination syndrome. This emphasizes their importance as under-appreciated generalized pollinators in some communities.

Eavesdropping on gall-plant interactions: the importance of the signaling function of induced volatiles.

The galling insect manipulates the host plant tissue to its own benefit, building the gall structure where it spends during most of its life cycle. These specialist herbivore insects can induce and manipulate plant structure and metabolism throughout gall development and may affect plant volatile emission. Consequently, volatile emission from altered metabolism contribute to eavesdropping cueing. Eavesdropping can be part of adaptive strategies used by evolution for both galling insects and the entire-associated community in order to cue some interaction response. This is in contrast to some herbivores associated with delayed induced responses, altering plant metabolites during the short time while they feed. Due to the different lifestyles of the galling organism, which are associated with different plant tissues and organs (e.g leaves, flowers or fruits), a distinct diversity of organisms may eavesdrop on induced volatiles interacting with the galls. Furthermore, the eavesdropping cues may be defined according to the phenological coupling between galling organism and host plant, which results from the development of a gall structure. For instance, when plants release volatile-induced defenses after galling insects' activity, another interactor may perceive these volatiles and change its behavior and interactions with host plants and galls. Thus, natural enemies could be attracted by different volatiles emitted by the gall tissues. Considering the duration of the life cycle of the galling organism and the gall, the temporal extent of gall-induced volatiles may include more persistent volatile cues and eavesdropping effects than the volatiles induced by non-galling herbivores. Accordingly, from chemical ecology perspective we expect that galling herbivore-induced volatiles may exhibit robust effects on neighboring-plant interactions including those ones during different plant developmental or phenological periods. Information about multitrophic interactions between insects and plants supports the additional understanding of direct and indirect effects, and allows insight into new hypotheses.