Optimising daily sampling of plant-bee interaction networks

Area of Science:

• Ecology
• Network analysis
• Pollination biology

Background:

• Plant-pollinator interactions exhibit temporal dynamics beyond seasonal changes, including hourly variations.
• Capturing daily temporal variability in pollination networks is challenging, leading to incomplete understanding and potential inaccuracies in sampling protocols.
• Limited sampling coverage can result in missing interaction links, hindering the identification of ecological processes and patterns.

Purpose of the Study:

• To investigate the fine-scale temporal organization of diurnal plant-bee interactions.
• To examine how interaction composition and network structure fluctuate throughout the day.
• To evaluate the impact of temporal sampling design on network representativeness.

Main Methods:

• Extensive sampling of plant-bee interactions in a Brazilian Cerrado over two flowering seasons.
• Utilizing six 2-hour sampling periods throughout the day to capture diurnal variability.
• Comparing interaction beta-diversity and network structural properties across sampling periods using a factorial approach.

Main Results:

• Plant-bee networks showed reduced modularity and specialization, but increased nestedness from dawn to dusk.
• Interaction beta-diversity increased with wider temporal intervals, driven by interaction rewiring among shared species.
• Combining four 2-hour sampling periods recovered ~80% of interactions and generated networks closely resembling full-day structures, with central hours being most informative.

Conclusions:

• Plant-bee interaction networks are highly time-sensitive at the intraday scale.
• Optimizing sampling effort by distributing it across informative periods or focusing on central daily hours improves network representativeness.
• A framework is provided to evaluate temporal sampling designs for time-dependent interaction networks.