Floral complexity and bumble bee foraging: A test of the efficacy backup hypothesis

Flowers in plants pollinated by animals have complex signals with visual and olfactory components. One hypothesis for the function of such multimodal signals is the efficacy backup hypothesis, which states that a component in one modality provides backup when environmental noise reduces the efficacy of a component in another modality. This was tested experimentally using bumble bees (Bombus impatiens) foraging for nectar in laboratory arrays. Bumble bees were trained at moderate light intensity to discriminate between rewarding and non-rewarding targets of different colors (visual only), different odors (olfactory only), or pairs of different colors and odors (bimodal). Choice tests were subsequently performed at high, moderate, and low light intensity, using targets of the same type to which bees had been trained. We first verified that light intensity affected the efficacy of color in the visual only treatment. We then tested the essential prediction of the backup hypothesis, namely that adding an olfactory component in the bimodal treatment compensated for loss of efficacy of the visual component. Discussion of results is placed in a general context of complex signal displays in animal behavior.