The number of open flowers displayed simultaneously by individual plants (floral display size) influences pollinator attraction and mating outcomes. Plants that display more flowers generally attract more pollinators, which can enhance overall pollination. However, on large displays pollinators tend to visit more flowers, possibly increasing among-flower self-pollination and associated inbreeding depression. The reproductive consequences of display size are further complicated because it varies daily during a plant’s annual flowering period. Consequently, a plant’s seasonal dynamics of display size could influence its interactions with pollinators and resulting mating outcomes. To assess this possibility, we censused the daily flowering phenologies of 90 Delphinium glaucum plants. This bumble-bee pollinated species produces mostly bisexual flowers that have distinct male and then female phases (protandry). We also recorded fruit production and assayed the proportion of outcrossed seeds with molecular markers. Analysis of among-plant variation in the relation of fruit production and female outcrossing rate to display-size dynamics involved functional regression. This approach uses a spline function characterizing each individual’s display dynamics as an independent predictor of a relevant dependent variable.
Results/Conclusions
During the 30-day flowering period of the D. glaucum population the overall floral sex ratio shifted from male-phase biased to female-phase biased, owing to protandry. Consequently, early flowers had access to more pollen during their female phases than late flowers. In this context, individuals flowered for 8 to 25 days, and average display size increased from 2 flowers on a plant’s first day to a peak of 13 flowers before declining to 3 flowers on the last day. Overall, display-size dynamics during the initial halves of plant’s flowering periods had little effect on among-plant variation in fruit production or outcrossing rate. This result reflects the fact that plants initially displayed mostly male-phase flowers, which were not receiving pollen. In contrast, display dynamics during the final halves of individual flowering periods affected female reproduction. Plants with large final displays generally produced more fruit than average if they had brief flowering periods, but they produced fewer fruit than average if they had long flowering periods. In contrast, large final displays generally promoted female outcrossing rate, particularly for plants with brief flowering periods. All these effects intensified as flowering periods progressed. These results are generally consistent with the declining opportunities for pollinator-mediated cross-pollination during the overall flowering period caused by the shift in floral sex ratio and would not be evident had display dynamics been ignored.