Abstract
Environmental impact assessment and regular environmental monitoring are prerequisites for the construction, operation, and decommissioning of offshore installations such as the Hywind Scotland wind park. Molecular approaches are increasingly being considered as a possible complement or alternative to currently used marine baseline and monitoring methods, both for pelagic and benthic organism studies. The following report is a proof-of-concept study where two molecular methods, metabarcoding and quantitative assays, have been used to characterize the pelagic environment at the Hywind Scotland wind park based on filtered water samples from the installation and a reference area. The purpose of the report is to showcase the use of molecular methodology in future studies of the pelagic ecosystem. Metabarcoding was employed for a community view of a) fish species specifically, using the MiFish primer set, and b) a universal eukaryote dataset based on 18S V1-V2 primers. Quantitative assays were employed for two commercially important pelagic fish species: mackerel and herring.
MiFish results comprised the detection of 39 fish species. Atlantic mackerel, sprat, Atlantic herring, haddock, pouting, and lemon sole were the most abundant in terms of sequence reads. Mackerel abundance was higher at 10 m depth compared to 50 m, equally distributed in installation and reference areas, for sprat and herring, abundance was high at both 10 m and 50 m, with higher abundance in the installation. The 18S data were dominated by alveolates, then metazoans, where copepods represented most reads. Beta diversity analysis of both MiFish and 18S data showed a clear and significant separation in data according to depth. For the fish specific MyFish marker the signals for typical pelagic species were consistently stronger at 10 m while demersal species had a stronger signal at the 50 m depth. A small but less clear difference in diversity data were also found between the installation and reference areas, but in the case of e.g., pelagic fish composition and their relative abundance, this difference could also be dependent on random placement of schools at the time of sampling. Sampling over a longer time frame than one day would strengthen any conclusions regarding these differences. The results show that metabarcoding has high potential to be used as an environmental monitoring method for the pelagic ecosystem and validate the ability of metabarcoding data to reflect differences in underlying organism community composition.
The test of the quantitative assays for mackerel and herring showed clearly that they worked with no indication of unspecific amplification (false positives). The results were further corroborated by the number of reads in the metabarcoding dataset. In the park area, there were significant differences in the signal from the two depths for mackerel with a higher biomass of mackerel at the 10 m depth compared to the 50 m depth. There was also an indication of higher biomass of mackerel at 10 m depth in the reference area. There was no significant difference in biomass of mackerel between the installation and the reference area when considering both sampling depths combined. The data for herring showed a slightly different pattern with a significantly higher biomass in the installation compared to the reference area, but also for this species, there were indications of higher biomass in the 10 m samples than in the 50 m samples, especially in the reference area.
We conclude that ddPCR using species specific assays applied on water samples is a powerful tool to assess biomass of pelagic species using filtered samples of water. To account for temporal and spatial variation in the behavior of these species, a full-scale project would benefit from samples taken at night and samples taken during other seasons. The statistical power would also benefit for samples taken over more days than what was possible here (one day only). In that way any coincidence in the distributions of shoals that may have contributed to the indicated increased biomass of herring in the installation would be ruled out. In this pilot project we included a reference area at a distant of 10 km away from the installation in a direction perpendicular to the current. To better understand the degradation of DNA over time samples also in the direction of the current could be considered. One major benefit of eDNA sampling is the restricted use of pelagic trawl inside an offshore wind farm. To better understand the correlation between eDNA results and actual fish biomass, eDNA samples should be taken and trawling conducted simultaneously in the same area. One could consider trawling to be conducted in the reference area that would allow for a ground proofing of the data also in a near-by installation.