COS 116-5 - Phenological synchrony between a specialist bee and its floral host-plant is not an important driver of bee fitness in an eight-year field study

Background/Question/Methods

Changes in climate can alter the extent of seasonal overlap between interacting species. Reductions in synchrony, or mismatch, could have measurable demographic impacts, if the affected interactions are important determinants of species’ vital rates. Bees, which depend on an ephemeral (floral) food source, might be especially at risk of phenological mismatch, and specialist bees—those that depend on a narrow range of plant taxa for pollen—should be especially affected by mismatch, provided their population growth rates are sensitive to variation in the floral food supply. So far, no studies have been able to attribute any real-world population declines to changes in bee synchrony with their floral host-plants. However, it is unclear whether this absence of evidence reflects an actual absence of threat or simply lack of data. Here, I ask whether lifetime reproductive output of a specialist solitary bee, Osmia iridis, can be predicted by phenological overlap between timing of bee activity and flowering of the bee’s floral host-plants—or whether other variables are more important. To do this, I use an 8-year, multi-site dataset on the lifetime reproductive output of 298 individually marked bees, together with observations of flowering phenology and bee nesting and emergence phenology.

Results/Conclusions

The temporal gap between median emergence date of female Osmia iridis and median flowering date of the bee’s host-plants varied by approximately two weeks among years and sites, as a result of taxon-specific phenological responses to spring temperature (linear mixed model; taxon*temperature interaction, P = 0.010). However, in most sites and years, despite >6°C variation in mean spring temperature, emerging bees were well synchronized with the start of their host-plants’ flowering season. Furthermore, although bee reproductive output (number of brood produced) tended to increase with floral density of host-plants at the study sites, calculated synchrony between bee emergence and flowering was not a significant predictor of bee fitness. Instead, the site-wide rate of parasitism by a brood-parasitic wasp (Sapyga sp.) was by far the strongest predictor of bee reproductive output (R_m² = 0.52, P = 0.005)—even before accounting for direct parasite-induced mortality of offspring. These results indicate that parasite abundance—not temporal synchrony with host plants—is the primary limiting factor for O. iridis populations in our study area. More generally, these findings suggest that fears of widespread phenological mismatch between plants and pollinators are themselves mismatched with the evidence.