Environmental DNA metabarcoding reveals complex interactions between natural capital in orchards: insights into arthropod community dynamics and foraging diversity

Natural vegetation fragments in agroecosystems often support relatively diverse arthropod communities and the ecosystem services they deliver. Unfortunately, these natural fragments are not typically accounted for in farm management, partly due to the challenges in applying conventional approaches (e.g., sweep netting) to monitor arthropod-plant interactions at the vast scale of agroecosystems. Increasingly, environmental DNA (eDNA) metabarcoding—a molecular method capable of amplifying trace amounts of DNA—has been applied in agriculture to monitor arthropod-plant interactions. Here, we used pan traps and eDNA metabarcoding of pan-trap water to measure the response of two environmental indicators—orchard arthropod community diversity and arthropod-plant foraging diversity—to the presence of adjacent natural vegetation and crop flowering intensity. We sampled six orchards, adjacent to either pasture habitat or natural vegetation, across four time points during crop flowering. In total, we collected more than 2000 arthropod specimens and identified over 250 plant taxa from eDNA, including crops, weeds and native Australian flora. Adjacent habitat type did not significantly influence arthropod species richness or abundance. However, orchards adjacent to natural vegetation showed increased capture rates of beneficial native bee (Lasioglossum sp.) and hoverfly (Melangyna viridiceps) species. Additionally, pan traps in orchards adjacent to natural vegetation contained more than double the number of foraging plant genera at low crop flowering compared to pan traps at orchards adjacent to pasture. Our findings highlight that eDNA-based monitoring can enhance our understanding of the biological processes that influence crop productivity while also revealing the complex interactions between different components of natural capital in agroecosystems.