COS 75-9
Impacts of a warmer drier world on bumblebee food resources: The view from the top

Wednesday, August 13, 2014: 4:20 PM
311/312, Sacramento Convention Center
Candace Galen, Biological Sciences, University of Missouri-Columbia, Columbia, MO
James D. Franklin, Division of Biological Sciences, University of Missouri, Columbia, MO
Peter G. Kevan, Department of Environmental Biology, University of Guelph, Guelph, ON, Canada
Background/Question/Methods

Experiments conducted as part of the International Tundra Experiment (ITEX; Oberbauer et al., 2013) predict that dry conditions will exacerbate impacts of global warming on the timing of tundra plant growth and development.  This prediction is especially germane to the central Rocky Mountains where climate change is already causing earlier snowmelt and more extreme summer droughts.  Phenology indicates an organism’s response to climate because it is sensitive to yearly conditions and integrates different environmental cues. We address impacts of climate change on phenology and abundance of major flower resources for Bombus (bumble bees), keystone pollinators in high elevation ecosystems. We monitored weekly changes in timing and abundance of flowering at the Pennsylvania Mountain Natural Area in the central Colorado Rocky Mountains in the mid-1970s and 2010s. Winter precipitation fell below the 22th percentile (over the past 119 yr) in 1977 and 2012, but was near normal in 1979 and 2013 (84th and 69th percentile, respectively). Average minimum monthly summer temperatures increased ~2oC over this interval.  In each year, we conducted weekly censuses of flowers for 10 plant species in 22 locations spanning krummholz, ridge, meadow, swale and summit habitats from 3600-4000 m. Plants surveyed historically accounted for ~90% of bumblebee resources at our field site.  

Results/Conclusions

Onset of flowering varied among species and with location, but overall showed mean advances under drought of 18 d in the 1970s and 11 d in the 2010s.  Flowering ceased earlier after low snowpack, shortening the average bloom time by 18% in 1977 (ns) and 45% in 2012 (P<0.05)  These results are in accord with the ITEX prediction that warm dry conditions will show the most extreme changes in timing of flowering.  For flower abundance, we found that in 1977, but not in 2012 average flower production of plant species over the seasonincreased with earlier onset of flowering. In 2012-13, flower production was low regardless of snowpack. Despite similar floral resources, bumblebee densities fluctuated dramatically between 2012-13. We hypothesize that while floral resources for bumblebees may not vary much between dry and wet years, costs of collecting those resources in a patchy landscape increase under drought.