Research and Teaching Interests
General Research Interests
Thetremendous variation in structure, number and arrangement of flowers pollinated by animals suggests adaptation to strong and diverse selection. Indeed, Darwin used flowers to refine his ideas about natural selection and evolution. My research program investigates floral diversity using perspectives from pollination biology and life-history evolution.
In recent years I have become increasingly interested in the community context of reproductive success and the balance between mutualistic and antagonistic relationships between interacting species.
Opportunities for Study
If you are interested in graduate or undergraduate research, please E-mail or write to me indicating your academic background and interests, and why you are interested in research.
Main Research Interests (Details below)
- Evolution of Floral Design
- Community context of reproductive success in rewardless flowers
- Pollinator networks in the tall grass prairie
Detailed Description of Main Research
1) Evolution of Floral Design
The flowers of many plant species are visited by a wide variety of animals. These visitors may differ greatly in how much pollen they transfer, and some may consume pollen without contacting female sex organs. Historically, research in pollination biology has emphasized selection by one or a few effective pollinators, i.e. visitors that contribute to seed set. Recent interest has focused on the full range of visitors in natural populations.
Our model system for these studies is Polemonium brandegeei, a perennial herb native to the Rocky Mountains of Colorado and Wyoming. Visitors to Polemonium brandegeei flowers are diverse and include hummingbirds, hawkmoths, small bees and pollen-consuming syrphid flies.
Ph.D. student, Mason Kulbaba (2012), used experimental populations to quantify natural selection on floral design by hummingbirds and hawkmoths in Polemonium brandegeei. Mason’s work suggested that some floral traits might “specialize” toward the optimum for one of the pollinators (e.g., width of the floral tube) whereas others may evolve toward a generalist “compromise” phenotype (e.g., female and male sex organs close together).
B.Sc. Honour’s student, Jian-fei Shao (2014), showed that prior deposition of sterile self-pollen may reduce outcrossed seed production. His work confirmed a potential cost of having sex organs close together.
M.Sc. student, Dawn Wood (2017), worked in natural populations to evaluate how pollen movement and seed production caused by large, nectar-foraging visitors (hawkmoths & hummingbirds) compared to the influence of smaller, pollen (and nectar) foraging bees and flies.
2) Reproduction & Mating Patterns in Rewardless Flowers
“Generalized food mimics” produce no nectar or pollen reward for visitors but have colours and scents similar to those in rewarding flowers. Because pollinators learn to avoid food mimics, visit rates and pollinator fidelity are often low, resulting in pollen limitation of fruit set and increased movement between species. The co-flowering community is particularly important for food mimics because rewarding flowers sustain pollinator populations, but may act as competitors on a local scale. Fruit set in food mimics may be enhanced by colour similarities to rewarding species, or earlier flowering than in rewarding species. The relative importance of these strategies is unknown.
In collaboration with Dr. Bruce Ford, we are studying how flower shape combines with the diversity of plants and pollinators to determine reproductive success in lady’s slipper orchids (Cypripedium). These rewardless orchids trap insect visitors in their slipper-shaped petal, and escape routes lead the insects past the sex organs. Our work to date has been on the threatened tall-grass prairie specialist, C. candidum, and widespread C. parviflorum.
M.Sc. student, Melissa Pearn (2012) (now Grantham) showed that the size of flowers & pollination routes, visit rates, and fruit set were all lowest in early-flowering C. candidum, intermediate in hybrids and highest in later-flowering C. parviflorum. Thus early-flowering does not confer a fruiting advantage to C. candidum, at least in Manitoba.
B.Sc. Honours student, Jessica Guezen (2013), confirmed that fruit set by C. candidum is pollen limited. In contrast C. parviflorum in Manitoba experienced a balance between pollen and resource limitation of reproduction. This may reflect the larger flowers of C. parviflorum, although C. candidum and C. parviflorum also differ in flower colour (white versus yellow) and scent.
B.Sc. Honours student, Kaman Choi (2014), applied UV-absorbing sunscreen to the yellow flowers of C. parvilforum and a heavily-visited rewarding species (Dasiflora fruiticosa) to change the visual stimulus to pollinators. The treatments reduced insect visitation to Dasiflora fruiticosa but did not change fruit set in C. parviflorum, perhaps because heavy rains washed off the sunscreen.
B.Sc. Honours student, Steven Anderson (2015), showed that insects visit C. parviflorum in proportion to their abundance as estimated by pan traps, indicating neither attraction nor avoidance. Comparison of visits to white, yellow and blue pan traps indicated that flies, but not bees prefer yellow over white.
M.Sc. student, Steven Anderson (2017), surveyed populations of C. candidum across a latitudinal gradient from Iowa to Manitoba. He will compare fruit set across populations in relation to the diversity and abundance of potential insect pollinators and co-flowering rewarding species. This work will begin to explore the potential for competition versus facilitation in this system.
Hybridization occurs between threatened tall grass specialist, C. candidum and widespread C. parviflorum where they co-occur and may be due to low pollinator fidelity in these food mimics. Genetic fingerprints in 16 populations confirmed first and later generation hybrids. Interestingly, these data indicated greater gene flow from the endangered to the widespread species. This work was with former summer student, Lauren Sawich, and research associate, Habib Ghazvini.
3) Pollinator networks and pollination services in the tall grass prairie
Pollination networks describe community wide interactions between plants and animal pollinators. Early work on networks (including theoretical treatments of community stability) involved qualitative descriptions of interaction patterns with the simplifying assumption that each visit from an animal to a plant has the same pollination benefit. Recent studies have begun to examine the validity of this assumption.
Former MSc student, Sarah Semmler (2015) described the short-term effects of fire pollination network structure by examining changes in plant and insect communities. Sarah also investigated how pollination services are affected by the identity and activity of pollinators. Sarah’s work was in Manitoba’s tall grass prairie, a critically endangered ecosystem.
This work was done in conjunction with CANPOLIN (Canadian Pollination Initiative), a strategic NSERC network.