Effect of flower visual angle on flower constancy: a test of the search image hypothesis

Pollinators often sequentially visit 1 flower type while bypassing other equally rewarding flower types, behavior known as flower constancy. One explanation for flower constancy is that pollinators use the search image of a specific flower type to efficiently find flowers because they are often cryptic. The so-called search image hypothesis predicts that temporal specialization to 1 flower type by an individual pollinator declines as flower conspicuousness increases because if flowers are conspicuous, pollinators can divide their attention among different flower types. To test this prediction, we investigated visitation sequences of individual bumble bees (Bombus ignitus) to a patch of blue and yellow artificial flowers. We used different flower sizes and interflower distances as independent determinants of flower visual size and noisiness of the background and, thus, of flower crypsis. Flower color selectivity and flower search time decreased and subsequently increased with visual size of the nearest flowers, with a minimum of approximately 15° of the visual angle. Within-bout constancy, measured as the tendency of flying to the same flower types in succession compared with the expectation of a given flower selectivity within a bout, increased with visual size of the nearest flowers. Flower selectivity and within-bout constancy did not differ among arrays that shared the same floral visual angle, indicating that visual size, rather than flight distance or flower size itself, had a substantial effect on flower choices. Our results suggest that flower visual size and background noisiness affect flower crypsis and constancy, providing support for the search image hypothesis.