Sediment collection
Sediment samples were collected in 2016 at six marine finfish farms located in BC, Canada: Baxter Islet (BI, May 10), Bedwell Sound (BS, April 13), Midsummer Island (MI, June 8), Mussel Rock (MR, April 12), Plover Point (PP, May 11), and Venture Point (VP, June 14). Samples were collected at seven stations along a transect in a dominant current direction initiated from the fish cage edge. The seven stations were intended to be 0 m, 15 m, 30 m, 60 m, 90 m, 125 m, and 500 m from cage edge. Samples were collected from those distances from cage edge where possible; the exception was the reference station at MI, which was 2260 m from cage edge as it was the closest reference station found with soft (and thus ‘grab-able’) sediments. Detailed site information was provided in Table S1.
A total of three sediment grab samples were collected at each station using a Van Veen Grab Sampler (0.1 m2, 24L) (Blomqvist, 1991; Gage & Bett, 2005). Pore-water sulphide concentration and redox potential for each grab were measured prior to the collection of other environmental variables to avoid oxygen injection into the sediment. Sediments were classified into four different sulphide concentration categories according to the Canadian Aquaculture Activities Regulations (AAR), as follows: anoxic (> 4,500 µM), hypoxic (1,300 – 4,500 µM), oxic B (700 – 1,300 µM), and oxic A (< 700 µM) (AAR, 2016). How each station was sampled is presented schematically in Figure S1 and summarized here. One grab sample was devoted entirely to macrofauna analysis (details below); each of the other two grabs were used to collect samples for geochemical analyses (porosity and organic content) and three eDNA samples, which were collected from the top 2 cm of sediment. The samples for geochemical analyses were collected using spatulas and eDNA samples were collected using modified 60 cc syringe-cores according to the sediment and meiofaunal methods detailed in Sutherland et al . (2007). Pore-water sulphide concentration, redox potential, porosity and organic content were measured following protocols used in Sutherland et al . (2007). To avoid cross-contamination during sample collection, the grab sampler was rinsed in salt water and visually inspected for cleanliness between grabs, samples were collected from the center of the grab only, and syringes were cleaned using diluted bleach solution (final concentration of 3% sodium hypochlorite) between farms. In summary, total numbers of the various sample types collected from a total of 42 stations were: 42 benthic macrofauna samples; 84 samples measured for porosity and organic content; 126 samples analyzed for sediment pore-water sulphide concentration and redox potential; and 252 samples analyzed for eDNA metabarcoding.