Unravelling the scientific potential of high resolution fishery data

Abstract

Fisheries science and fisheries management advice rely on both scientific and commercial data to estimate the distribution and abundance of marine species. These two data types differ, with scientific data having a broader geographical coverage but less intensity and time coverage compared to commercial data. Here we present a new type of commercial data with high resolution and coverage. To our knowledge, the dataset presented in this study has never been used for scientific purposes. While commercial datasets usually include the total weight by species on per haul basis, the new data also include the commercial size class for the species landed, recorded directly on a haul-by-haul basis. Thus, this dataset has the potential to provide knowledge on landed fish with as high spatio-temporal resolution as when coupling logbooks and sales slips but with the addition of detailed knowledge on the size distribution. Such information may otherwise be obtained through on-board observer programmes but unlike the observers’ data, the dataset presented here is routinely collected on most of the trips of the vessels involved, which means that the coverage of the data for the individual vessel is larger than observers’ data. Furthermore, the risk of changes in fishing behaviour due to the presence of an observer on-board is avoided. This paper describes the coverage and completeness of the dataset, and explores the reliability of the data available. We conclude that the main limitation is the small number of fishing vessels covered by the program, but that the data from those vessels are accurate, detailed, and relatively reliable.

More data for the money: Improvements in design and cost efficiency of electronic monitoring in the Danish cod catch quota management trial

“…Depending on the speed of video transmission and audit, such maps could potentially help fishers improve their spatial selectivity by consulting with continuously updated maps. Additionally, linkage of EM with video data to eLog data allows, at least in Denmark, for a linkage between EM with video data and grading machine data collected from sea-packing vessels (Plet-Hansen et al, 2018). This will allow for investigations of species' landings and discard size composition at the haul level, potentially shedding further light on the drivers behind the occurrence of unwanted catches.…”

The value of commercial fish size distribution recorded at haul by haul compared to trip by trip

“…Commercial size classification follows the requirements of the EU (EU, 1996). Plet-Hansen et al (2018) previously described the SIF dataset in details and investigated its usefulness for scientific purposes through comparison to logbook and sales slips data. The dataset was considered suitable for further scientific analyses, notwithstanding some variability in data quality across years, vessels, species and size classes (Plet-Hansen et al 2018).…”

“…The final dataset after the validation according to Plet-Hansen et al, 2018 is the baseline for comparison as records of species and sizes are directly available at the individual haul level. The dataset is henceforth referred to as "SIF".…”

“…This mean that any size information is collected at the level of the full fishing trip (EU, 2011(EU, , 2016. This trip-level information can be redistributed back to the haul level by using the size composition of each species for the fishing trip, under the assumption that size composition at the trip-level reflects the size composition at the haul level (Plet-Hansen et al, 2018). Possible discrepancies when deducing spatial distribution from analysing commercial fishing data might not be an issue when using data from small-scale vessels which perform only few hauls within short distance, but could be pronounced for large-scale vessels conducting several hauls per day and weeklong fishing trips, making a mismatch between trip and haul fish size composition highly likely.…”

Fishing for euros: how mapping applications can assist in maintaining revenues under the Landing Obligation

“…A further barrier to operational use of this type of analysis lies in the very limited amount of observer 582 data collected, at less than 1% of fishing operations, and the risk of an observer effect influencing this 583 data (Schaeffer and Hoffman, 2004;Plet-Hansen et al, 2018). Certainly the variation in observer coverage on different types of vessels, in different areas and across different seasons and years may 585 influence the resultant hotspot maps.…”