Danish brown shrimp fishery

Swept area information for Danish beam trawlers targeting brown shrimp has been updated for 2021-2024 with an improved method. The updated data can be downloaded from here while we are updating our online platform on spatial information of the fishing pressure from the different mobile bottom-contacting gears: https://ono.dtuaqua.dk/DDFAM/

The data can be downloaded as a raster stack in 0.001° x 0.001° resolution (about 60x100 m) and can be viewed here: https://ono.dtuaqua.dk/hesterejer.html

27 vessels have targeted brown shrimp between 2021 and 2024 on a total of 8,325 trips. It was possible to link 99.3 % of the trips (8,268 trips) to VMS or AIS positions with estimated fishing. When zooming in to individual hauls it was possible to link 36,082 out of 36,840 (97.9%) to positional data. The small proportion (2.1%) of un-positioned hauls may partly be explained by lack of fishing estimate during the logged time for the recorded haul, and partly by incomplete position registrations on some trips.

On average the brown shrimp vessels swept an area west of Jutland of 3091 km² per year during the period. But since they often impact the same area (up to 78 times per year in one cell in 2022, see also https://ono.dtuaqua.dk/DDFAM/ -> Information -> Fishing Pressure Indices), the total extent of impacted seabed area (the footprint) was on average 1208 km² yearly (see table 1).

Table 1. Swept area and impacted area between 2021 and 2024 from Danish beam trawler targeting brown shrimp.

Method for linking landings to positions and estimating swept area.

Data sources:

• Sales notes data contain information on landed weight and value per vessel, landing date, species, and fishing area, covering all vessel lengths.
• Logbook data are recorded by the fisher and include information on landings per vessel, landing date, catch date and/or haul start and end, species, fishing area, ICES rectangle, and gear type. Vessels over 10 meters in length in the North Sea and over 8 meters in length in the Baltic Sea are required to keep a logbook. Smaller vessels can instead complete a fishing area declaration. For these smaller vessels, DTU Aqua estimates the gear type based on the composition of the landings combined with available knowledge. Landings are estimated by the fisher but are not always as accurate as the sales notes.
• Positional data are collected from three sources:

o VMS data: For Danish vessels, one position per hour is sent to this satellite-based monitoring system. From 2005 to 2011, all vessels over 15 meters were required to have VMS; after 2012, this applies to all vessels over 12 meters.
o AIS data: A safety system mandatory for all fishing vessels over 15 meters. Many vessels under 15 meters have it voluntarily installed, and the proportion of vessels with AIS has steadily increased during the period 2007–2023. Positions have varying frequency depending on speed and system, typically less than 1 minute, but depend on signal reception by a land-based mast and can be manually turned off, so there may occur gaps in the data. All brown shrimp fishing trips are usually within the range of land based masts.
o Blackbox data: Positional data typically every 10 seconds, installed on mussel vessels and experimental fisheries, including camera vessels.

Spatial distribution of landings and gear deployment:

Estimating where fish are caught requires several steps, where the different data sources are linked.

1. Logbook and sales notes are linked so that each trip and haul is assigned landing values and weights.

2. All three types of positional data are merged into one dataset, and vessel tracks are estimated using cubic hermite spline interpolation (based on the vessel course in the positional data) if there is less than 2 hours between positions. and/or reduced to one position per minute.

3. The combined logbook and sales notes data is linked to the merged and processed positional data to assign a gear type to each position. Based on knowledge of typical vessel speeds when fishing with the different gear types, the positions where vessel speed falls within a defined interval of gear-speeds are defined as fishing positions.

4. Fishing positions are combined into trawl tracks by assigning straight lines between neighboring positional points.

5. Gear width for bottom contacting gears are estimated based on vessels size and gear type, using the method described in Eigaard et al 2016. Subsequently width is added to the estimated trawl tracks/lines using a flat buffer to create a polygon for each haul. For gears with no bottom contact, no width is estimated, and only the line is used for spatially distributing catches.

6. Landing weight and value are then distributed across the hauls based on information from the logbooks, either by trip, date or haul-period, depending on how the logbook is filled out.

7. Hauls (polygons and lines) are divided into a 0.01° x 0.01° grid, and landings are assigned based on the polygon area (bottom contacting gears) or the line length (non-bottom contacting gears) in each grid cell.

Known caveats:

- Not all trips have positional data, which means that the linked dataset does not reflect the full fishing effort. Especially the fleet of smaller vessels below 12 meters fishing inshore has low coverage.

- It is not possible to know exactly when a vessel conducts fishing. Estimating the activity based only on vessel speed leads to some positions being estimated as fishing, where no fishing is taking place (false positives). On the contrary, some positions where fishing is taking place will be estimated as not fishing (false negatives). But by and large the assumption that a certain gear type is fishing at certain speeds is mostly accurate.

- Sometimes errors in logbooks and sales notes lead to false outputs. But since the data is quality controlled by the Fishery control and enforcement, the Danish Agricultural and Fisheries Agency and screened by DTU aqua, these errors can be considered minor.

- Positional data errors also occur, especially with AIS data. Linking MMSI number to vessel ID and problems with position at the wrong place are the most common error. Most issues are identified and solved in the data handling and refinement process.

- Discards and illegal fishing are not considered in this method.

References:

Ole R. Eigaard, Francois Bastardie, Mike Breen, Grete E. Dinesen, Niels T. Hintzen, Pascal Laffargue, Lars O. Mortensen, J. Rasmus Nielsen, Hans C. Nilsson, Finbarr G. O’Neill, Hans Polet, David G. Reid, Antonello Sala, Mattias Sköld, Chris Smith, Thomas K. Sørensen, Oliver Tully, Mustafa Zengin, Adriaan D. Rijnsdorp, Estimating seabed pressure from demersal trawls, seines, and dredges based on gear design and dimensions, ICES Journal of Marine Science, Volume 73, Issue suppl_1, January 2016, Pages i27–i43, https://doi.org/10.1093/icesjms/fsv099

van der Reijden, Karin J.; Eigaard, O.R. (Ole) (2023). The seafloor footprint of Danish fishing. Technical University of Denmark. Online resource. https://doi.org/10.11583/DTU.23617944.v1