Journal of Animal Ecology (2013) doi: 10.1111/1365-2656.12158

One of the most noticeable effects of anthropogenic climate change is the shift in timing of seasonal events towards earlier occurrence. The high degree of variation in species' phenological shifts has raised concerns about the temporal decoupling of interspecific interactions, but the extent and implications of this effect are largely unknown. In the case of plant–pollinator systems, more specialized species are predicted to be particularly threatened by phenological decoupling, since they are assumed to be less flexible in the choice of interaction partners, but until now this hypothesis has not been tested.
In this paper, we studied phenology and interactions of plant and pollinator communities along an altitudinal gradient in the Alps as a model for the possible effects of climate change in time.
Our results show that even relatively specialized pollinators were much more flexible in their use of plant species as floral resources than their local flower visitation suggested. We found no relationship between local specialization of pollinators and the consistency of their visitation patterns across sites, and also no relationship between specialization and phenological synchrony of pollinators with particular plants.
Thus, in contrast to the conclusions of a recent simulation study, our results suggest that most pollinator species included in this study are not threatened by phenological decoupling from specific flowering plants. However, the flexibility of many rarely observed pollinator species remains unknown. Moreover, our results suggest that specialized flower visitors select plant species based on certain floral traits such as the length of the nectar holder tube. If that is the case, the observed flexibility of plant–pollinator interactions likely depends on a high degree of functional redundancy in the plant community, which may not exist in less diverse systems.

Specialization and phenological synchrony of plant–pollinator interactions along an altitudinal gradient