Authors: Shale Rosen and Nils Olav Handegard
There is great uncertainty surrounding the global abundance of mesopelagic species (generally distributed from 200 – 1000 m deep). Estimates vary between one  and ten  billion metric tonnes, and methods to address this uncertainty are needed.
Worldwide, fisheries research institutions use trawl sampling and acoustics to estimate fish distribution, abundance and population parameters for a range of different pelagic species. Monitoring programmes invest a substantial amount of effort worldwide, but mostly address larger, commercially important, fish species like herring, pollock and anchovies, among others.
Different trawls have different catchability and are designed to capture a specific length distribution of nekton. Fine-meshed trawls may effectively catch the smaller sizes, but are generally too small and towed too slowly to catch the larger fish species. Larger-meshed trawls may effectively sample the larger fish at the expense of not retaining smaller species or damaging fragile and gelatinous ones in the cod end (last part of the net).
In cases where the trawl herds the nekton towards the cod end but they are lost due to mesh selectivity or too badly damaged to identify or measure, a camera system is one solution. However, the image quality and resolution must be sufficient to count and identify the organisms.
The Deep Vision system was designed by Scantrol to determine species and sizes of fish in trawl catches , and is being further developed through the Research Council of Norway (RCN) funded CRISP centre for research innovation . Its most important components are a colour stereo camera system which takes 5 image pairs per second and an imaging chamber which guides all objects into the field of view of the camera and lights them indirectly with strong strobed light. Analysis software uses the time- and depth-referenced stereo images to identify and measure the organisms with high accuracy.
During a survey run by the Institute of Marine Research, through the RCN funded REDUS project, we used Deep Vision to capture time- and depth-referenced images of mesopelagic fish during trawling. The results confirmed that important mesopelagic species could be easily identified and measured with the system. In addition, the sampling provided fine-scale distribution data, including overlap with larger pelagic fish species such as mackerel, herring and blue whiting.
- Gjøsæter J, Kawaguchi K. A review of the world’s resources of mesopelagic fish. 1980 p. 151. Report No.: FAO Fisheries Technical Paper 193.
- Irigoien X, Klevjer TA, Røstad A, Martinez U, Boyra G, Acuña JL, et al. Large mesopelagic fishes biomass and trophic efficiency in the open ocean. Nat Commun. 2014;5. doi:10.1038/ncomms4271
- Rosen S, Jörgensen T, Hammersland-White D, Holst JC. DeepVision: a stereo camera system provides highly accurate counts and lengths of fish passing inside a trawl. Can J Fish Aquat Sci. 2013;70: 1456–1467. doi:10.1139/cjfas-2013-0124