Tuesday, August 4, 2009 - 3:40 PM

OOS 19-7: Will climate change disrupt synchrony between subalpine plants and pollinators?

Jessica Forrest and James D. Thomson. University of Toronto

Background/Question/Methods

There are concerns that species-specific shifts in phenology in response to climate warming may lead to temporal mismatches between interacting species. As yet, there have been few empirical data to evaluate whether (or which) plant-pollinator interactions might be vulnerable to phenological disjunctions brought about by climate change. The risk of mismatch is greatest when interacting organisms use different cues to regulate phenology, so identifying these cues and understanding how they covary is critically important.  We studied trap-nesting Hymenoptera across an elevational gradient in the Colorado Rocky Mountains and asked 1) what cues control insect emergence, and 2) to what extent do these same cues predict flowering phenology across sites? To do this,  we established standard, artificial trap-nests at 10 sites that spanned an 8-week range in the date of spring snowmelt. A reciprocal transplant experiment was conducted by switching a subset of the occupied nests at the highest- and lowest-elevation sites prior to over-wintering. Air-temperature loggers recorded hourly temperatures at each site. At 2-3 day intervals the following spring and summer, insect emergence and flowering phenology were recorded at each site.

Results/Conclusions

Trap-nests were colonized by a variety of solitary, flower-visiting Hymenoptera—mainly megachilid bees, eumenine wasps, and their predators. For most insects, we found no evidence of local adaptation in the duration of diapause: emergence time was accurately predicted by temperatures at the overwintering site, regardless of site of origin. In general, degree-days above 10 degrees Celsius was the best predictor of insect emergence; for plants, degree-days above 0 Celsius was usually the best predictor. These results suggest that emergence time is highly plastic in these insects and depends on similar—but not identical—cues to those regulating flowering time in the local plant community. For the bee species best represented in our dataset, emergence time was better predicted by temperature variables than by local flowering phenology, indicating that some decoupling between plants and pollinators is possible. However, the use of similar temperature cues and generalist habits should make plants and pollinators in these habitats resilient to phenological changes driven by climate warming.