丁香园AV

Areas of Research

Determinants and Impacts of Pollinator Diversity

Classic island biogeography theory predicts that biodiversity and abundance should increase with island size;  we make similar predictions for fragmented ecosystems.  We鈥檝e shown that pollinator biodiversity tends to be higher in larger fragments of the Garry Oak Ecosystem, but this is largely due to the response of some (but not all) nesting guilds of bees.  Pollen limitation of at least some wildflowers is lower in fragments that support greater diversity.  Currently, we are investigating how pollen limitation of multiple wildflower species varies among years and fragments that differ in bee diversity, and we are building plant-pollinator interaction webs to investigate how connectance varies with plant and bee diversity.  We are also beginning to investigate how pollinators use areas surrounding habitat fragments ("the matrix") to better understand responses to fragmentation.

Human impacts go beyond fragmentation of habitat to other modifications.  We are interested in how anthropogenic impacts from farming to logging to cattle grazing affect bee communities and pollination services.  For example, we are currently investigating how rangeland practices affect pollinator diversity and abundance, expecting both positive (nest site creation) and negative (removal of forbs) impacts.

Pollinator Diversity and Crop Pollination

We have recently begun investigation of the contribution of wild bees to crop production in British Columbia, and how aspects of farm management and the surrounding landscape affect both the diversity of wild bees in crops and the pollination services they provide.  In some cases, wild bees do not appear to contribute much to crop pollination, with honeybees doing most of the work.  In other cases (such as squash in the Okanagan valley) native bees have relatively high visit rates, and their visit rates are higher when farms are in proximity to natural areas.  Current and future work in blueberry aims to both to improve our understanding of how farm and landscape factors affect the diversity and abundance of wild bees in fields, and to experimentally investigate whether we can enhance habitat to support native bee diversity (by providing nest sites and forage outside the crop bloom). The long-term goal is to help growers use both native and managed bees to provide reliable and economical crop pollination, but this research also highlights our basic philosophy that conservation and agriculture can work hand-in-hand.

The Evolution of Selfing

Current models predict that selfing should be advantageous for plants experiencing unreliable pollination (selfing increases reproductive assurance and decreases pollen and seed discounting) and low inbreeding depression. To test these models, we are comparing functional aspects of the mating system among populations of Collinsia parviflora (Blue-eyed Mary) from Vancouver Island that differ in flower size.  We鈥檝e found that autonomous selfing ability decreases as flower size increases; that pollinator visit rates increase with flower size;  that selfing provides greater reproductive assurance in small-flowered populations and less in large-flowered populations; and that although inbreeding depression is low overall it tends to be higher in large-flowered populations.  

Why is Collinsia flower size so variable in BC?  All BC populations of Collinsia are tetraploid, as are some populations of different flower sizes in the US;  diploid populations in the US are either large- or small-flowered (never intermediate) and ploidy provides barriers to interbreeding.  We hypothesize that among-population variation in flower size is due to divergent selection, both by pollinators (as noted above) and by abiotic factors.  Rainfall is significantly lower in locations where Collinsia populations are small-flowered, and these populations reach reproductive maturity earlier, perhaps because time to complete the life cycle is limited in ephemeral environments.  Ongoing experiments aim to evaluate local adaptation due to flower size variation, and contemporary selection due to pollinators (which should select for larger flowers when present) and rainfall (which should select for faster reproduction when in short supply). 

Multi-trophic Interactions

Interactions between members of a community can be important determinants of fitness, and when one of the players is a plant we need to transcend the 鈥渢rophic鈥� part of those interactions and include mutualisms in our conceptual framework.  JC Cahill of the U. of Alberta and I showed that disrupting plant-mychorrhizal interactions affected the composition of floral visitors, due to shifts in the floral display of the interactors.  Former postdoc Alida Janmaat (currently at UFV) and I are investigating whether spittlebugs, which reduce flower size in large- but not small-flowered Collinsia, affect pollinator visit rates.  And at some point the lab needs to follow up on the observation that small-flowered populations of Collinsia produce more extra-floral nectar than mid- or large-flowered populations (is it due to differences in ant or herbivore communities?  Simple re-allocation of resources?), and ant protection, and plant fitness, is correlated with extra-floral nectar production.