1. The early detection of invasive non-native species (INNS) is important for informing management actions. Established monitoring methods require the collection or observation of specimens, which is unlikely at the beginning of an invasion when densities are likely to be low. Environmental DNA (eDNA) analysis is a highly promising technique for the detection of INNS – particularly during the early stages of an invasion. 
2. Here, we compared the use of traditional kick-net sampling with two eDNA approaches (targeted detection using both conventional and quantitative PCR, and passive detection via metabarcoding with conserved primers) for detection of quagga mussel, Dreissena rostriformis bugensis; a high priority INNS, along a density gradient on the River Wraysbury, UK. 
3. All three molecular tools outperformed traditional sampling in terms of detection. Conventional PCR and qPCR both had 100% detection rate in all samples, and outperformed metabarcoding when the target species was at low densities. Additionally, quagga mussel DNA copy number (qPCR) and relative read count (metabarcoding) were significantly influenced by both mussel density and distance from source population, with distance being the most significant predictor. 
4. Synthesis and application. All three molecular approaches were more sensitive than traditional kick-net sampling for the detection of the quagga mussel in flowing water, and both qPCR and metabarcoding enabled estimates of relative abundance. Targeted approaches were more sensitive than metabarcoding, but metabarcoding has the advantage of providing information on the wider community, and consequently impacts of INNS. 

The following files have been produced as part of the manuscript: Targeted and passive environmental DNA approaches outperform established methods for detection of quagga mussels, Dreissena rostriformis bugensis in flowing water

1. qPCR data from mesocosm experiments for the detection of quagga mussel (D. r. bugensis)
2. qPCR data from field experiments for the detection of quagga mussel (D. r. bugensis)
3. Metabarcoding data generated using the COI primer pairs (Leray+Geller) (Leray et al 2013) for the 54 field samples and 3 positive samples, including OTU ID and taxonomy generated from the MetaBEAT pipeline.
4. Sub-sample of those sites metabarcoding and qPCR for the detection of quagga mussel (D. r. bugensis)
5. Data Descriptions of the columns found in each dataset and the contents.