Environmental DNA (eDNA) is a tool that is increasingly used for biomonitoring. The rapid uptake of eDNA for monitoring in aquatic ecosystems has focused on invasive species detection and qualitative biodiversity assessments, despite its potential to be used more substantively. Beyond basic monitoring applications, eDNA can be combined with a "metaweb" approach, allowing for the reconstruction of theoretical food webs in study systems. This approach has the potential to better map biodiversity to ecosystem processes using the structure and properties of food webs, helping to address the growing recognition of multitrophic biodiversity as an important determinant of ecosystem functioning. However, despite this potential, there remain relatively few studies that have attempted to construct stream food webs from eDNA data and relate their properties to environmental drivers. To investigate the feasibility of using eDNA and metawebs for food web reconstruction, we collected eDNA from native forest streams and streams in exotic pine (Pinus radiata) plantations across New Zealand. We expected that stream food webs in pine plantations would be less complex with altered connectance and shorter food chain lengths, reflecting increased disturbance and lower terrestrial plant diversity. Initial analyses using whole community indicators (the Taxon-Independent Community Index) suggest that stream food webs in pine plantations are more impacted than their native forest counterparts. Further analysis will explicitly consider key food web properties such as complexity, and food chain length. These properties will be compared with food webs constructed from benthic data collected using traditional methods. Assessing these different approaches will help inform our understanding of how eDNA data can be used and what limitations need to be considered. This improved knowledge will help to upscale the method using the extensive national eDNA data now available to test hypotheses about biodiversity and ecosystem functioning in stream-riparian networks at broad spatial scales.