Randall J. Mitchell, University of Akron, Jeffrey D. Karron, University of Wisconsin-Milwaukee, and Karsten G. Holmquist, University of Wisconsin-Madison.
A mechanistic understanding of the factors causing differences in selfing rate is needed to evaluate the evolution of mating systems in plants. Recent work indicates that selfing rates vary markedly among individual flowers on the same plant. This might reflect pollinator behavior, pollen carryover, and patterns of pollen movement. When species have multi-flower displays, pollinators often visit several flowers consecutively on each plant. Although the first flower probed is likely to receive substantial outcross pollen, if pollen carryover is limited receipt of self pollen would increase rapidly for later flowers. Here we report the first direct test of this hypothesis. In experimental arrays of Mimulus ringens, selfing rates increase significantly with successive probes on a multi-flowered plant. The first flower probed typically had a selfing rate of 0-20%. By contrast, the fourth flower probed typically had a selfing rate of 80-100%. This pattern resulted from an increase in the proportion of geitonogamous pollen transfer. These findings are consistent with known rates of pollen carryover, pollinator behavior, and variation in mating system.