GINR reports commercial data from the Greenland halibut fishery in Disko Bay; Catches peaked 2004-2006 above 12,000 t

Commercial data for the Greenland halibut fishery in the Disko Bay.

Northwest Atlantic Fisheries Organization

NAFO SCR Doc. 26/017

SCIENTIFIC COUNCIL MEETING –JUNE 2026 Commercial data from the Greenland halibut fishery in Disko Bay Rasmus Nygaard1, Henrik Christiansen1 1Greenland Institute of Natural Resources, P.O. Box 570, 3900 Nuuk, Greenland

NYGAARD, R., & CHRISTIANSEN, H. 2026. Commercial data from the Greenland halibut fishery in Disko Bay. NAFO Scientific Council Research Document, SCR Doc. 26/017: 1-19.

Abstract

The commercial fishery targeting Greenland halibut in Disko Bay started around 1910 with the introduction of longlines in Greenland. The fishery is traditionally performed with longlines from small open boats, small vessels or through a hole in the sea ice and transported by dog sledge. This document presents catch statistics and data from the commercial catch from various resources. The document includes statistics of commercial sampling effort done by the Greenland Institute of Natural Resources (GINR) and calculations of mean size in the landings and a preliminary catch-at-age estimation. Also provided are three CPUE indices from Disko Bay based on logbooks (one for longline fishery and one for gillnets) and factory landings data (longline).

Introduction

The commercial fishery in Disko Bay started around 1910 with the introduction of longlines in Greenland. The fishery is traditionally performed with longlines from small open boats, small vessels or through holes in the sea ice and transported by dog sledge. In the 1980s, small vessels entered the fishery and the use of gillnets increased in the following years. In the late 1990s, the first regulations limiting areas open to gillnet fishery were introduced, limiting gillnet fishery to the winter season. The main fishing areas are the deep Ilulissat Icefjord and the shallower bank next to Ilulissat where icebergs strand and Torsukattaq, an icefjord located in the northeastern part of Disko Bay.

Materials and methods

Recent catch statistics (factory landing and logbooks) are available from a centralized database managed by the Greenland Fisheries and Hunting Control Authority (GFJK). Both logbook (haul by haul) and factory landings (daily individual fishing event landings) are reported as individual fishing events containing metadata, such as date, field code or position, effort, gear type, sorting category, bait, and more. Data quality has been high since 2012 with very little missing data.

Commercial sampling

Commercial samples are collected by GINR. During surveys or in sampling campaigns factories are visited and the size of the landed fish by species and gear is registered. However, due to the logistic challenges in Greenland (size of Greenland and mainly transport by air or sea), sampling the catch is challenging. To ensure sufficient

Northwest Atlantic Fisheries Organization

www.nafo.int 2 length information from commercial catches is available, GINR previously sampled length measurements in factories during the winter months and during the survey period in July-August. However, since 2022 the length distributions in the landings have been based on data from automated sorting machines, i.e., “grader data”, provided by Halibut Greenland.

Age-length key (ALK)

Age information is occasionally obtained from commercial landings, but most otoliths in the area are collected through biological surveys with GINR research vessel RV Sanna and RV Tarajoq.

Logbook CPUE calculation

A general linear model (GLM) with year, month and boat as factors is applied to the longline and gillnet fishery logbook data since 2007. Only longline sets with more than 200 hooks and gillnets with catches between 50 and 200 kg/gillnet are included to omit obvious outlier values and limit the influence of potential data errors on the analysis. CPUE observations are log-transformed prior to the GLM analysis. Least-mean square estimates were used as standardized CPUE series. For more information about the standardized logbook CPUE see Riget and Nygaard (2018). A new CPUE based on factory landings was introduced in 2022 (Bjare and Nygaard, 2022). The new CPUE covers almost all longline fishery, including logbook-providing larger vessels, small boat fishery, as well as ice fishery.

Results

Catches

Catch statistics date back to 1904, but total landings remained at a low level until the beginning of the 1980s (Table 1 and Fig. 1). In Disko Bay, catches increased during the 1980s and peaked in 2004 to 2006 with catches of more than 12,000 t per year (Fig. 1). Thereafter, catches decreased and the TAC was not reached anymore. Since 2009 catches have gradually increased and in 2016 catches were 10,760 t. This was followed by a poor season in 2017 where only 6,409 t were caught in Disko Bay. After 2017 catches have gradually increased and in 2023 reached 11,435 t, but decreasing to 9,644 t in 2024 and further to 8,053 t in 2025.

Distribution of catch

The fishery in Disko Bay is concentrated near the mouth of the Ilulissat Icefjord (Kangia) near Ilulissat city and typically more than one third of Disko Bay catches are from this small area (Fig. 2). Ilulissat Icefjord is normally filled with icebergs during the summer. The Ilulissat Icefjord is deeper (>900m) than the central part of Disko Bay and larger fish are normally concentrated there. The fishery in the northern part of Disko Bay is concentrated around the settlements Saqqaq and Qeqertaq and the icefjord Torssukattak east of the settlements. In the most recent years the fishery has increased in the western part of Disko Bay between Aasiaat and Qeqertarsuaq, where deep trenches are located.

Breakdown of catch

A breakdown of catch by gear and area is provided in Table 2. The catch by area (Disko Bay or Ilulissat Icefjord), gear (longline or gillnet), and month is combined with the length frequencies from the commercial landings (Table 3) to calculate mean size in the landings and estimate catch-at-age (CAA). Due to logistical challenges in Greenland not all months have commercial length information (Table 3). Nevertheless, in most years the sampling covers the three different categories (Disko Bay longline, Ilulissat Icefjord longline, and gillnet fishery). However, since 2022 data from automated sorting machines (“graders”) was used to assess the length distribution of the landings. Graders are sorting machines that sort every individual fish into size groups by weighing each individual. The sorting machine weighs every individual fish typically with an accuracy of 5 grams. The individual weights from millions (2.6 mio in 2025) were recalculated to total length using the length-weight relationship from the surveys. Halibut Greenland notes the geographical origin, and a separate length frequency is available from both Kangia and central Disko Bay. Grader data furthermore contain a

Northwest Atlantic Fisheries Organization

www.nafo.int 3 vendor number and daytime stamp and in principle a length frequency can be calculated on the individual landing. However, to simplify, length frequencies are aggregated on a monthly basis for each subarea within the area. Only gear is not registered in the grader data, but given the large amount of data, gear is indirectly weighted in the length frequencies. In all circumstances, length frequencies based on grader data since 2022 should be far more accurate than when based on eventual factory visits. Mean length in the landings and CAA was estimated by applying a monthly length distribution to the age-length key (ALK). Grader data is available also prior to 2022 and awaiting analysis.

Size of the landed fish

In Disko Bay, mean individual size in the landings gradually decreased for more than a decade in both the winter longline fishery (a mixture of Greenland halibut from Disko Bay and large fish from Kangia) and the summer longline fishery (Fig. 3). The overall mean length when accounting for differences in distribution of the catch and gear have decreased gradually since 2010 reaching a record low in 2025 (Fig. 3 and 4).

Age-length key (ALK)

Age reading of Greenland halibut was suspended from 2011 to 2017 at GINR due to low quality of age readings and lack of an internationally agreed method. However, age readings have since been reinitiated and an ALK is currently being constructed back in time.

Catch-at-age (CAA)

CAA tables were constructed using age readings from whole frozen otoliths from all three inshore areas collected from 2008, 2009 and 2010 until 2019 (Table 4). The 2020 to 2025 CAA tables were constructed with individual year’s ALK from Disko Bay based on the new method. Despite the ALK still being preliminary and unverified, CAA tables reveal the dominance of the 2015 year class (YC) and younger in the CAA bubble (Fig. 5). The fishery in 2025 was mainly based on ages 5-8.

Factory landings CPUE (longline)

A general linear model (GLM) with year, month, vessel type, and catch area as factors was applied to the longline landings in the factory provided landing slips from 2012 to 2024 (see Bjare and Nygaard, 2022). The new CPUE based on factory landings data consists of more than 10,000 observations per year in all years (Table 5). The CPUE decreased until 2017 but recovered until 2023. After 2023 the CPUE decreased reaching the second lowest level in the time series in 2025 (Fig. 6).

Logbook CPUE (longline)

A general linear model (GLM) with year, month, and boat as factors was applied to the longline fishery logbook data since 2007. The CPUE index gradually decreased from 2007 to 2017 (Fig. 7). From 2020 to 2023 the CPUE gradually increased. However, both in 2024 and 2025 the index decreased further to the lowest levels in the time series only surpassed by 2017.

Logbook CPUE (Gillnet)

The gillnet logbook CPUE covers about 50% of the gillnet fishery (Table 7). The gillnet CPUE based on logbooks shows a gradual decreasing trend, despite legal mesh size reductions from 110 mm half mesh to 95 mm half mesh in Greenland halibut gillnets, as well as reductions for cod gillnets of 80 mm half mesh, which are occasionally used to target Greenland halibut (Fig. 8). Despite these issues, which make the gillnets more effective, the CPUE gradually decreased over the period, except for a short rebound in 2022.

Northwest Atlantic Fisheries Organization

www.nafo.int 4 Discussion

Although the catches have been fairly stable over the past three decades the number of Greenland halibut taken annually continues to increase. The catches in numbers of Greenland halibut were record high during the most recent three years, and the fishery is currently relying on younger year classes than in the past. The use of grader data has proven to be a valuable source of information, and it does not appear to have changed the results compared to the previous procedure, where length information was collected opportunistically. All CPUEs decreased in 2025. The factory based longline CPUE and the logbook CPUE are in good agreement even though they rely on different data sources. Although CPUE analyses should generally be treated with caution, it should also be noted that the CPUE index from the factory landings presented here is based on all individual landings and typically constitutes more than 10,000 individual observations per year.

References

Riget, F. and Boje, J. (1989). Fishery and some biological aspects of Greenland halibut (Reinhardtius hippoglossoides) in West Greenland waters. NAFO Sci. Council Studies (13): 41-52.

Riget, F. and Nygaard, R. (2018). An analyses of logbooks of Greenland Halibut Stock Component in NAFO Division 1A Inshore. NAFO SCR 18/023.