COS 68-4 - Fire induced change in flowering phenology benefits bumble bees

Tuesday, August 8, 2017: 2:30 PM
E141, Oregon Convention Center
John M. Mola, Graduate Group in Ecology, University of California, Davis, CA and Neal M. Williams, Department of Entomology, University of California, Davis, CA
Background/Question/Methods

Fire is a dominant disturbance in plant communities worldwide, often with marked effects on floral abundance, diversity, and morphology. Post-fire changes to plant communities will likely directly impact floral visitors, particularly bees, which rely exclusively on flowers for food resources. While community-level patterns of pollinator response to fire have been explored, little inference is available on how individual species respond to fire-induced changes in the floral reward structure. To understand how fire mediates interspecific interactions, such as those of the obligate relationships between plants and pollinators, we examined how the bumble bee (Bombus vosnesenskii) population responded to changes in floral abundance, diversity, density, and phenology following fire.

The study was conducted at the University of California McLaughlin Reserve in grassland meadows. Collections of B. vosnesenskiipre-fire allowed for the selection of sites that did not differ prior to disturbance. In 2016, following fire, we captured workers at each of nine sites during two-hour long sampling rounds. Site-level inflorescence abundance and floral density were also estimated concurrent with bee surveys. Each site was visited twice, allowing us to infer changes between an early and late sampling period in the briefly flowering grasslands.

Results/Conclusions

Burned grasslands sustained an abundance of dense blooms for longer periods of time than unburned areas, resulting in prolonged habitat use by bumble bees. The effects of fire were not immediately apparent, and only through an interaction with phenology were fire-induced changes realized.

 We collected 184 and 95 bumble bees in burned and unburned sites, respectively. Bumble bee capture rates were positively predicted by log inflorescence abundance (P < 0.001), with inflorescence abundance driven by a significant interaction between burn status and sampling round (P = 0.002). Floral densities were higher in burned sites both within (P = 0.037) and between (P = 0.003) years.

Importantly, we found the longer flowering season was not due to increased plant diversity in burned relative to unburned sites, but rather through the lengthening of flowering for individual species. Proportional change in inflorescence abundance declined across all sites, but the decline was smaller for plants in burned (-22.8%) compared to unburned (-49.4%) areas.

Our results suggest fire can prolong flowering, increasingly the temporal availability of floral resources for interaction partners. Further studies should examine the extent to which fire-induced changes in phenology can modify the timing of species interactions.