Habitat effects on reproductive phenotype, pollinator behavior, fecundity, and mating outcomes of a bumble bee–pollinated herb

PREMISE: Fecundity and mating outcomes commonly differ among plant populations occupying contrasting environments. If self‐pollination occurs primarily among flowers within plants, contrasting reproductive outcomes among populations must reflect environmental effects on plant‐pollinator interactions. Specifically, local conditions could affect features of plant phenotypes that influence pollinator behavior, in turn modifying plant reproductive outcomes. METHODS: We compared phenotypes, pollinator abundance and behavior, and female fecundity and mating in two meadow populations and two forest populations of Aconitum kusnezoffii within 3 km of each other. Mating outcomes were assessed using microsatellites. RESULTS: Meadow plants generally produced more, shorter ramets with more, larger flowers, but less nectar per flower than forest plants. These differences likely largely represent phenotypic plasticity. Individual bumble bees visited more flowers on forest plants, likely because the more abundant bees in the meadows depleted nectar availability, as indicated by briefer visits to individual flowers. Despite similar fruit set in both habitats, forest plants set more seeds per fruit. Nevertheless, meadow plants produced more seeds overall, owing to sevenfold greater flower production. Consistent with individual bees visiting fewer flowers on meadow plants, more of their seeds were outcrossed. However, the outcrossed seeds of forest plants included more male mates. CONCLUSIONS: Reproductive outcomes can vary among populations of animal‐pollinated plants as a result of differences in the availability of effective pollinators and environmental effects on plant phenotypes, and their functional consequences for pollinator behavior that governs pollen dispersal.