Semi-Quantitative Monitoring of Plant-Arthropod Interactions by eDNA Metabarcoding of Individual Flowers and Leaves.

Environmental DNA (eDNA) analysis has revolutionized our ability to study plant-arthropod interaction diversity. However, the method has one main limitation: while it accurately recovers the presence of individual species, it has limited quantitative recording capabilities. Recent work suggests that this problem can be overcome by processing replicate eDNA samples at individual sites. The naive occupancy of a species within a replicate sample approximates its relative abundance or its interaction strength with the sampled matrix. However, this approach will lead to a significant increase in the necessary workload and costs, especially as the usually applied filtering of eDNA samples is laborious and expensive. Here, we explore the option to directly extract eDNA from individual leaves and flowers omitting a filtration step. We test different isolation protocols and compare the results with a conventional filtration-based protocol. We show that simple precipitation suffices to recover community-level eDNA of arthropods from individual flowers, significantly reducing costs and efforts. We then test our optimized protocol in two apple orchards in Germany, where we sample individual leaf and flower samples. Our protocol reliably recovers typical orchard-associated arthropod species across broad ecological guilds. The naive occupancy-based abundances of these species reflect the expected community composition in orchards very well. For comparative purposes, we also performed a conventional, quantitative bee monitoring in the orchards using window traps. Based on this monitoring, we show that replicate sampling of flowers is even a promising option to semi-quantitatively recover bee pollinators and their ecological interaction diversity from eDNA. Simplified replicate eDNA sampling will considerably improve our ability to characterize arthropod communities.