Novel approaches for quantifying pollen chemistry and insect-pollen networks over broad spatial and temporal gradients

Research on the diets of insect pollinators and how diet chemistry affects the ecology and evolution of both plants and pollinators could benefit from an integrated mix of new or repurposed methodological approaches, including recent advances in organic chemistry and modern applications of network science to ecology. Analyses of pollinator diet at local and landscape scales can be dramatically improved by enhancing these two approaches. First, all chemical ecology still suffers from poor chemical analyses, or poorly articulated metabolomics >data. This can be corrected via improved spectroscopy, structure determination, and careful statistical and machine learning models for analyzing non-targeted metabolomics data and identifying chemical features important for insect nutrition. Second, pollination networks need help – many are simply rough visitation networks that suffer from limited natural history data, sparse experimental validation, and a lack of standardized spatial scales. Existing observational studies can be complemented with pollen-insect associations uncovered from museum specimens, selected experiments, and focused network analysis, yielding much more informative pollination and visitation networks. Refining these enhancements to complement existing chemistry and network approaches allows for clear hypothesis tests about how networks change in space and time, and the role that pollen and nectar chemistry play in pollinator diet and network dynamics across different scales.