COS 188-2 - Bumble bee abundance is governed by climate-driven variation in floral phenology

Friday, August 11, 2017: 8:20 AM
C125-126, Oregon Convention Center
Jane E. Ogilvie1,2, Sean R. Griffin2,3, Zachariah J. Gezon2,4, Brian D. Inouye1,2, Nora C. Underwood1,2, David W. Inouye2 and Rebecca E. Irwin2,3, (1)Biological Science, Florida State University, Tallahassee, FL, (2)Rocky Mountain Biological Laboratory, Crested Butte, CO, (3)Applied Ecology, North Carolina State University, Raleigh, NC, (4)Animal Programs, Disney's Animal Kingdom, Animal Programs Administration, Lake Buena, FL
Background/Question/Methods

Given ongoing climate change and pollinator declines, it is particularly important to understand how climate influences bee populations. Climate may affect bees directly by altering survival rates and life history events like foraging activity, or indirectly by influencing the amount and phenology of floral food resources. We conducted an 8-year observational study of three bumble bee (Bombus) species, the floral assemblages they visit, and the climate of a subalpine site in southwestern Colorado where the flowering season is limited to a short snow-free period. We used structural equation modeling to examine how local climate variables influence year-to-year bumble bee abundance both directly and indirectly through effects on floral abundance and phenology. Additionally, we use 42-year local datasets to ask whether the climate and the floral abundance and phenology of the bumble bee-visited assemblages have changed directionally over time. From this we can predict how bee populations may fare given future changes in climate and the floral assemblage.

Results/Conclusions

Interannual variation in the abundance of three bumble bees (B. bifarius, B. flavifrons, and B. appositus) was driven primarily by the indirect effects of climate on the number of days above a minimum flower threshold (floral days). Bee abundance increased with more floral days, while floral days increased with more summer precipitation and later snowmelt dates. Cumulative floral abundance was also important for two species; it was included but insignificant in the model for B. appositus, and had a weak negative effect on B. bifarius abundance. Over 42-years, the snow has melted marginally earlier while summer precipitation has not changed directionally. Over the same time period, the number of floral days has increased for B. flavifrons but has not changed directionally in the floral assemblages of the two other species. Furthermore, cumulative abundance of the floral assemblages has not changed directionally through time for all three bees. Thus, at our subalpine site, continuing changes in snowmelt date and seasonal floral days may increase some bumble bee populations, while others may remain stable if their resources remain unchanged. More generally, our results suggest that bee populations will be sensitive to climate change primarily through indirect climate-driven changes in their floral resources.