Practical implementations of the eDNA concept for marine resources monitoring

Monitoring marine resources to ensure healthy and productive oceans requires cost- and time efficient methods. The concept of environmental DNA was presented over 20 years ago for the monitoring of rare and invasive species, and since the method has encapsulated the essence of what is required. The term “eDNA” describes all genetic material present in the environment that can be sampled and analyzed using molecular methods to primarily reveal species identity but can also be used for other applications. After several years of development and application, the range of potential uses of eDNA is wide, however, practical implementation of the approach for management activities still are limited. Among the challenges to wider application is the lack of understanding how eDNA abundance measured in eDNA samples translates to organism quantity (e.g. species abundance or biomass), primarily caused by limited knowledge of eDNA ecology and biased in molecular methods applied.

The overarching aim of this thesis is to investigate the prospects for the implementation of eDNA-based methods in practical ecosystem and fisheries monitoring. To achieve this objective, this thesis introduces a novel approach for precise eDNA-based monitoring of catch composition, and thus in the bigger picture, assesses the feasibility of improved monitoring of bycatch in fisheries in terms of cost and time efficiency. This thesis combines experimental approaches for testing and estimating the relationships between eDNA and biomass in mixtures of commercial fish species, with practical implementations of the concept to samples derived from large scale practical fisheries. The experimentally derived statistical models for converting eDNA proportions to biomass are subsequently used to assess species proportions from “production water” eDNA samples collected onboard fishing vessels and at catch processing factories. In this thesis Chapter II lays the groundwork by studying suspected primary factors, which potentially could influence eDNA availability and proportions, such as surface area of fish, water type (factory or vessel), and water temperature. Chapter III and IV elaborate on the concept and the pipeline for practical eDNA-based bycatch monitoring. These chapters study bycatch estimation of herring in the sprat fishery from the Baltic Sea (Chapter III) and bycatch of mackerel in herring fisheries from the North Sea (Chapter IV) respectively. The empirically developed models for converting eDNA proportions to biomass proportions show high precision in bycatch fraction estimates in both fisheries. These eDNA derived estimates are compared to visual assessments, i.e. manual sorting and weighing of species from catch subsamples, made by fishers, fisheries control, and 3^rd party independent control. This comparison shows that eDNA-based methods consistently provide more precise and lower estimates compared to the visual methods. The study from the sprat fishery additionally illustrated the importance of molecular approach choice (qPCR versus DNA metabarcoding) on the performance of the method. It also highlights the value of small mixture experiments conducted alongside, for calibrating the actual catch to different previously derived eDNA to biomass conversion models – when available. Finally, the study on herring fishery with bycatch of mackerel highlights the need for understanding how the mixing of water onboard fishing vessels and at the factories influences the results for eDNA-based bycatch assessment.

Expanding beyond the estimation of bycatch quantities, I also used eDNA in an attempt to improve region wide monitoring of non-indigenous species (NIS) in the Arctic and Sub-Arctic (Chapter V), the second objective of this thesis. A large-scale sampling effort consisting of over 900 eDNA samples from many different geographical regions applying both qPCR and DNA-metabarcoding provided evidence of NIS occurrence and assessed their seasonal and geographical spread. In general, the detection rate was low, with two NIS found using qPCR and eleven using DNA-metabarcoding. The NIS cladoceran Evadne nordmanni, was detected by both methods with qPCR being more sensitive in detections. NIS spread was associated with populated areas, and in general more NIS were found in the Barents Sea than in the Greenlandic waters. Moreover, NIS presence was highest in the Arctic summer period, mainly August. Following, the study provides recommendations for future eDNA-based NIS monitoring in the Arctic.

Overall, this thesis expands the applicability of eDNA for both bycatch estimation and large-scale NIS monitoring. Particularly, the application to fisheries catches and the observed quantitative precision is unparalleled in the scientific literature. The positive results and the high sensitivity of the studies presented here give reason to hope the eDNA samples collected from fishing vessels can also be used to identify even more rare species in catches such as Protected Endangered and Threatened (PET species) for example sharks and rays, and to take an additional step in terms of genetic resolution by identifying the population affiliation of the fish in the catch.