The number of flowers open on a plant can influence the level of among flower selfing or geitonogamy. We compared selfing rate, estimated with allozyme data from progeny array, between two populations of Aquilegia coerulea in southwest Utah. In the Archery population 55 % of the plants had a single flower, in contrast to 21 % of the plants in the Cedar Breaks population. As no geitonogamous selfing can occur in single-flowered plants, we hypothesized that selfing rate would be lower at the Archery population. This assumes that geitonogamous selfing contributes to selfing rate in these populations. To control for other variables that could potentially influence selfing rate between these two populations, we measured the level of protandry, recorded pollinator abundance, flower size and plant density. We also bagged plants in each of these two populations to determine the potential for autogamous selfing in these populations. There was no difference in the multilocus estimate of selfing rate, using PGM and MPI, for Cedar Breaks (selfing rate = was 0.22 +/- 0.08 , 41 families and 549 progenies) and Archery (selfing rate = 0.20 +/- 0.8, 38 families and 284 progenies). At both sites, bees did the majority of pollination. Our results suggest that plant architecture and geitonogamy do not strongly influence selfing rate in these two populations of A. coerulea.

Key words: Aquilegia coerulea, autogamy, geitonogamy, plant architecture, pollination, selfing rate