Many environmental factors impact plant and pollinator communities. However, variation in soil moisture and how it mediates the plant–pollinator interactions has yet to be elucidated. We hypothesized that long-term variation in soil moisture can exert a strong selective pressure on the floral and vegetative traits of plants, leading to changes in pollinator visitation. We demonstrated that there are three phenotypic populations of Gentiana aristata in our study alpine region in the Qinghai-Tibetan Plateau that vary in floral colour and other traits. Pink (dry habitat) and blue (intermediate habitat) flower populations are visited primarily by bumblebees, and white (wet habitat) flower populations are visited by flies. These patterns of visitation are driven by vegetative and floral traits and are constant when non-endemic plants are placed in the intermediate habitats. Additionally, the floral communities in different habitats vary, with more insect-pollinated forbs in the dry and intermediate habitats versus the wet habitats. Through a common garden and reciprocal transplant experiment, we demonstrated that plant growth traits, pollinator attractiveness and seed production are highest when the plant population is raised in its endemic habitat. This suggests that these plant populations have evolved to pollinator communities associated with habitat differences. This article is part of the theme issue ‘Natural processes influencing pollinator health: from chemistry to landscapes’.

Long‐term variation in the population density of introduced honey bees (Apis mellifera) has been shown to be associated with variations in floral traits in alpine lotus (Saussurea nigrescens). However, it remains to be determined whether a high density of honey bees affects the abundance of nectariferous plants and the species richness and abundance of native bumble bees. We predicted that a high density of introduced honey bees lasting three decades would decrease the species richness and abundance of native bumble bees but increase the abundance of honey bee host plant species. Here, field experiments were conducted to examine the diversity of nectariferous plants and native bumble bees along the typical gradients of honey bee density (high density of honey bees close to the apiary and low density of honey bees distant from the apiary). We investigated nectariferous plant abundance, floral and seed traits, and bumble bee species richness and abundance at sites with either high or low honey bee densities in an alpine meadow. Our results demonstrate that an increased population of introduced honey bees is associated with increased host plant abundance and flower/capitulum number per plant but decreased nectar volume per flower, seed mass, and species richness and abundance of native bumble bees. The bumble bee visitation rate was positively correlated with nectar volume per flower at sites close to and far from apiaries. The honey bee visitation rate was positively correlated with flower/capitulum number per plant at sites close to apiaries and nectar volume per flower at sites far from apiaries. Seed mass was negatively correlated with nectariferous plant abundance. Our findings show that introduced honey bees decrease the species richness and abundance of native bumble bees, attributed to an evolutionarily decrease in nectar resources among honey bee host plant species, but increase the abundance of nectariferous plants, attributed to the production of many small seeds by plants. This suggests that long‐term high‐density beekeeping affects the biodiversity of honey bee host plants and native bumble bees. Our results provide new insights into the mechanisms of maintaining the biodiversity of nectariferous plants and native bumble bees.

The global expansion for massively introduced managed species has profoundly impacted biodiversity and ecosystem functions. However, it is unclear how introduced pollinators and mass-flowering crops simultaneously affect the structure and function of pollination networks. Here, we conducted field experiments to investigate the ecological impacts of introduced honey bee (Apis mellifera) during the flowering period of a mass-flowering crop (Brassica rapa var. oleifera) on plant-pollinator interaction networks and reproductive performance of wild plants in 48 alpine grasslands. We showed that the spillover of introduced honey bees had weak effects on the pollinator communities. In addition, honey bee spillover strengthened the pollination network structure’s stability and improved plant communities’ pollination function. These effects can be observed when the introduced honey bees and mass-flowering plants coexist in the alpine grasslands, highlighting the positive effects of honey bee spillover on the pollination-limitation alpine ecosystem.

No abstract