Are different bee species attracted to different plant and floral traits in alfalfa (Medicago sativa)?

Monday, June 1, 2015
Big Basin (Manhattan Conference Center)
Austin Bauer , Entomology, University of Wisconsin Madison, Madison, WI
Johanne Brunet , Entomology, USDA - ARS, Madison, WI
Alfalfa (Medicago sativa) represents a great system for studying plant-pollinator interactions and insect-mediated gene flow. Alfalfa is an outcross, perennial plant visited by different bee species and exhibits variation in floral traits such as flower color and floral display size. This research focuses on the effects floral traits have on bee foraging behavior, comparing three bee species which include two social bees, honey bees (Apis mellifera) and bumble bees (Bombus impatiens) and one solitary bee, the alfalfa leaf-cutting bee (ALCB) (Megachile rotundata).  The main goals of this project were first to identify the floral traits of alfalfa that are most attractive to each bee species; and second to determine how such floral traits affect bee movements.  In summer 2014, I measured floral traits of alfalfa in four patches (81 plants per patch) set up at the West Madison Agricultural Research Station. I measured the number of racemes per plant, number of flowers per raceme and flower color (reflectance, chroma, and hue) for each plant within a patch. I also measured bee visitation within patches by following bees as they moved between racemes and plants. Multiple regression analyses were used to identify floral traits that best explain variation in plant attractiveness for each bee species. For honey bees and bumble bees, number of racemes per plant significantly explained variation in plant attractiveness. Plants with a greater number of racemes scored higher on the plant attractiveness index than plants with fewer racemes. Flower color is also a significant factor for bumble bees. Though there is no clear pattern for which aspect of color bumble bees are attracted too (reflectivity, chroma, hue) as the three components are highly correlated.