Interactions between climate and forest cover may dictate the development of forest understory species distributions, and thus the delivery of important ecosystem services such as pollination. Despite this importance, current knowledge is insufficient to accurately project the combined interactive effects of climate and forest cover on pollinator networks, and their combined implications for species dispersal and climate refugia under future climate change scenarios. In this project we will address these knowledge gaps by applying DNA meta-barcoding techniques to identify flower-visiting insects in areas of a forest stand diverging in tree species composition and density. This will be done by analyzing environmental DNA (eDNA) that insects leave on flowers when foraging for understory plant pollen and nectar. This technique allows us to study pollination non-invasively in a context where such can be considered as a rare event and thus is hard to study by direct observations only. Flower samples were collectes within a double gradient of forest density and overstory species composition (from pure birch to pure spruce plots). Our goal is to assess how pollinator diversity changes depending on the forest context in which a flower is found.