Abstract

Floral traits and the number of visitors are expected to change with different mating systems. We tested this hypothesis by comparing flowers of Prunella vulgaris (Lamiaceae) with inserted and exserted styles across a strongly exserted style biased, an exserted style biased, and a strongly inserted style biased subalpine population.

We examined flowering phenology, floral morphology, flower visitation rate, capacity for autonomous self-pollination and visitor contribution to seed production for each style type and in each population. We also examined inbreeding depression (ID) by comparing the relative performance of progeny from self- and cross-pollination. Exserted style plants had larger and more open flowers, increased pollen production, higher amounts of nectar rewards and higher visitation and outcrossing rates than inserted style plants.

Similarly on the population level, the visitation rates were higher in the exserted style-biased populations than in the inserted style-biased population. Inserted style plants provided a stronger reproductive assurance (RA) through autonomous selfing than exserted style plants. RA and outcrossing rates did not differ among populations, showing low visitation rates may be sufficient for adequate seed production in P. vulgaris.

Although inserted style plants had a lower ID level than exserted style plants, the ID of both was less than 0.5, suggesting that an ID should not counteract the evolution of selfing in this species. Inserted style plants provide RA through autonomous selfing, and exserted style plants ensure outcrossing through pollinator services, supporting a stable mixed mating system in this subalpine plant.