Fine-tuning the buzz: comparing visitation frequency and pollination effectiveness in plant-pollinator networks.

Ecological network approaches have advanced our understanding of how species interactions influence community and evolutionary dynamics. However, a key limitation is that most network analyses rely solely on visitation data, often overlooking functional aspects of interactions. Here, we combined quantitative (visitation frequency) and qualitative (pollen removal and deposition) components to assess bee interactions with buzz-pollinated flowers in the field. We recorded bee visitation to Chamaecrista (Fabaceae) flowers to represent the quantitative component and conducted single-visit experiments to evaluate qualitative components related to male and female reproductive performances. Data were integrated into ecological networks to explore the structure of plant-pollinator interactions. Across 1838 interactions involving 10 plant species, flower-buzzing bees were the most effective pollen depositors, while robbers removed large amounts of pollen but frequently damaged floral structures. Network analyses revealed that male performance components generated more specialised and modular networks than those based on visitation or female performance, highlighting functional differences among bee groups. While visitation networks offered partial insights, inclusion of pollination effectiveness metrics revealed the importance of specialised vibration behaviours in plant reproductive performance. We emphasise integration of both quantitative and qualitative data to better predict ecological and evolutionary dynamics in specialised pollination systems.