The number of flowers open at one time on a plant can alter pollinator behavior, which may in turn alter plant reproductive success and mating system. To investigate this, we experimentally manipulated the daily floral display of monkeyflower plants in four clonally replicated arrays. Plants received one of an exponential series of floral display treatments: 2, 4, 8 or 16 flowers. These treatments were performed in a Latin Square Design, so that each genet experienced all 4 floral display sizes on each day, and each ramet experienced all display sizes over the four days. We then quantified patterns of pollinator visitation and the opportunity for self-pollination in these experimental populations. Bombus spp constituted over 95% of all visitors, and these pollinators visited three times more flowers on 16 flowered displays than on 2 flowered displays. There were clear differences between and within pollinator species in several aspects of behavior that could influence plant mating patterns, including not only number of flowers probed, but interplant movement distances and grooming behavior. Flowers in all floral display treatments were equally likely to be probed, with a mean per flower visitation rate near 0.6 probes/flower/hour. Because flowers last only a single day and stigmas close by mid-day, there is a notably short window of receptivity for pollination in this species, which has important consequences for its mating biology. Overall, plants with large displays should experience much more self- pollination through geitonogamy as a result of these patterns of pollinator behavior.

Key words: breeding system, floral display, geitonogamy, inflorescence size, Pollination, pollinator behavior