Studies at local spatial scales have shown that the generalist honeybee Apis mellifera L. can strongly affect the structural organization and properties of pollination networks. However, there is still little knowledge on how the connectivity of the honeybee within networks (i.e., interactive role) could affect pollination networks at a global scale. To fill this gap, we evaluated how the interactive role of A. mellifera could affect niche overlap, specialization, and robustness of pollination networks. We used 109 weighted pollination networks spread across about 94 degrees of latitude and 227 degrees of longitude. We developed a new standardized framework based on species removal to assess the impact of the honeybee on plant-pollinator networks. We found that when the honeybee was removed from the networks, plant species had less niche overlap (i.e., shared fewer interactions via their pollinators) and the networks became more specialized and more robust to species extinctions. Our findings indicate that A. mellifera’s effects on pollination networks vary geographically and could influence several ecological and evolutionary factors acting at local scales, including pollination services. We hope this contribution will stimulate new macroecological studies involving abundant and generalist species and their functional roles within ecological communities.

Bombus pauloensis Friese is a promising managed bumblebee that can pollinate crops in South America. Highbush blueberry (Vaccinium corymbosum L.) is a recently introduced and economically promising crop cultivated in open fields or greenhouses in Colombia. Although this crop is known to be pollinator-dependent, the efficiency of local pollinators in this geographic area has yet to be established. This study aimed to establish the pollination effectiveness and foraging behavior of B. pauloensis in two of the most common cultivars of highbush blueberries planted in a high Andean region of Colombia. We hand-reared and located B. pauloensis colonies in two different plantations of Sharpblue (open field) and Biloxi (greenhouse) cultivars. The time spent per flower and the number of flowers visited per minute of 300 foragers were registered to characterize the foraging behavior in both cultivars. Pollen analysis was performed once from corbicular loads, larvae, larval cells, and colony feces to identify the pollen sources collected by colonies located in the Sharpblue crop. Analysis of fruit quality was performed to establish the effect of B. pauloensis pollination. Foragers visited blueberry flowers with a corolla aperture of more than 3mm in both cultivars but spent more time visiting Biloxi than Sharpblue flowers. B. pauloensis pollination reduced the ripening time and increased the fresh and dry weight and the number of true seeds in both cultivars. Our results provide evidence that B. pauloensis pollination can improve the fruit quality in the blueberry cultivars evaluated here in Colombia, making it an efficient native pollinator for a promising commercial species.

Plant–pollinator interactions are essential to ecosystem functioning, yet the mechanisms that determine why some plant species become highly connected within interaction networks remain insufficiently understood, particularly in tropical coastal systems. Here, we examine how multiple plant traits predict the interactive role of species within a bee–plant network in a coastal ecosystem in the Gulf of Mexico. Using an existing dataset comprising 35 plant species and 47 bee species, we quantified each plant’s interactive role through species degree, betweenness, and closeness centrality. We then evaluated how six traits (i.e., flower number, flower size, flower color, number of stamens, plant height, and life form) influence these network positions. Our results show that four traits significantly predicted plant interactive roles. Plants surrounded by more open flowers and those with larger flowers interacted with a greater diversity of bee species, indicating that resource detectability and accessibility strongly shape visitation patterns. Herbaceous species also exhibited higher interactive roles than woody plants, likely due to their rapid growth, abundant and synchronous flowering, and predictable resource availability in dynamic coastal environments. Additionally, yellow-flowered species received disproportionately more visits and achieved higher interactive roles, consistent with known sensory biases of bees toward yellow wavelengths. In contrast, plant height and stamen number showed no detectable influence on network position. Overall, our findings demonstrate that a combination of vegetative and floral traits (particularly those signaling abundant, accessible, and visually detectable resources) drives the emergence of key plant species within bee–plant networks. Integrating plant traits with network metrics provides a powerful framework for identifying species that sustain pollinator diversity and for predicting community responses to environmental change.