Ecology Letters (2026)
Quantifying biotic interactions at scale is critical to understanding ecosystem functions, such as plant-pollinator interactions. Here, we demonstrate how woody plant functional traits and vegetation structure gathered with remote sensing can aid in predicting plant-pollinator network indices from fine to broad spatial extents. We analysed 209 tropical plant-pollinator networks, where vegetation characteristics were generated using spectral and LiDAR data. We found that pollination network structures were correlated to plant traits reflective of drought-adaptive strategies and nutrient availability. Networks were predicted to be more nested and less specialised with increasing leaf nitrogen content. Further, relationships between leaf thickness, photosynthetic capacity, and water content indicate that networks may be more modular and less connected under arid, high-light conditions where plants invest in water-conserving tissues yet maintain high photosynthetic rates. Our findings reveal plant functional strategies as underlying environmental mechanisms that structure plant-pollinator networks, paving the way to predict interactions using remote sensing.
