Tuesday, August 4, 2009 - 10:10 AM

COS 28-7: Pollinator diversity and abundance in a fragmented oak-savannah ecosystem

Lisa A. Neame and Elizabeth Elle. Simon Fraser University

Background/Question/Methods Habitat fragmentation is widely thought to be a threat to biodiversity. However, the responses of pollinators to habitat fragmentation are still poorly understood, as pollinator communities are variable in time and space. In a highly fragmented oak-savanna ecosystem in the southeast corner of British Columbia, Canada, there is a diverse flowering plant community. As many of the flowering plants in this ecosystem require insect pollinators, we investigated the pollinator community in some of the remaining ecosystem fragments. I sampled 19 oak-savanna fragments in an area that is highly affected by urbanization. Habitat fragments ranged in size from 0.3 ha to 31 ha and are surrounded by a variety of land-use types, including forest, low density suburban and urban neighbourhoods. I sampled using two methods, pan-trapping and netting, as different methods are known to sample different subsets of the pollinator community. 

Results/Conclusions Surveys captured approximately 5400 bees, flies and wasps in 166 species and 51 genera. Contrary to predictions that species richness should increase with increasing area, species richness of fragments was not predicted by the size of the fragment (R2 = 0.093, p = 0.202). Overall species richness was also not significantly different between surrounding land-use categories. However, examination of rank-abundance plots suggests that larger sites do tend to have a greater number of rare species compared to smaller sites. Preliminary analysis suggests that the proportion of brood cleptoparasitic bees has a complex relationship with degree of urbanization. As degree of urbanization increases, the proportion of cleptoparasitic bees has a hump-shaped distribution, where the highest proportions are found at sites with intermediate urbanization. Since brood cleptoparasites are the highest trophic level in this pollinator community, I expect that they should be the most sensitive to fragmentation. Further analysis will include a more sensitive measure of land use surrounding each fragment and will also incorporate a measure of connectivity between fragments. Clarification of the relative importance of site and landscape factors can inform conservation measures for pollinators and pollination services.