The oldest direct evidence of insect pollination is a 99-million-year-old beetle preserved in Burmese amber with pollen grains attached to its body. The pollen is from a cycad or gymnosperm, and the beetle belongs to a lineage associated with flower-visiting. But 99 million years ago, flowering plants were diversifying, not dominant. The pollinator is older than the ecosystem it now characterizes.
The amber preserves not just the insect but the interaction. The pollen grains are not scattered randomly — they cluster on specific body parts, in patterns consistent with deliberate feeding behavior rather than incidental contact. The beetle was visiting the reproductive structures of a plant and carrying pollen between them. The mutualism was already structured: the insect gained food, the plant gained transport.
What the amber captures is the modularity of pollination. The beetle didn't evolve to pollinate flowers — flowers didn't dominate yet. It evolved to exploit plant reproductive resources, and the pollen-carrying was a byproduct of feeding. When angiosperms later diversified, they recruited existing insect visitors who already had the behavioral repertoire. The pollination syndrome didn't start from scratch with flowering plants; it was assembled from pre-existing parts.
The oldest pollinator pollinated something that wasn't a flower. The relationship between insect and reproductive structure predates the structure we associate with it. Flowers co-opted a workforce that was already employed elsewhere. The amber doesn't record the origin of pollination. It records pollination before its most famous context existed.