General information
Icelandic slope beaked redfish (Sebastes mentella) is a redfish species which is similar in appearance to golden redfish (Sebastes norvegicus). There are some characteristic features that distinguish those two species apart, and the depth is one of them, with Icelandic slope beaked redfish inhabiting deeper waters (>400 m). Around Iceland the species is mainly found in the warmer waters in the western, southern, and south-eastern parts of continental slope. Beaked redfish is a slow growing, long-lived, and late-maturing species.
Icelandic slope beaked redfish on the continental shelf and slope of Iceland (the Icelandic waters ecoregion, which is defined to be within the Icelandic 200 NM EEZ and includes ICES Division 5.a and part of ICES Subarea 14), is treated as distinct biological stock and management unit. Mainly fish larger than 30 cm are found in Icelandic waters. The East Greenland shelf is most likely the main nursery area for the Icelandic slope stock.
Scientific data
The Icelandic autumn survey (IS-SMH) on the continental shelf and slope in Icelandic waters covers depths down to 1500 m. Data on Icelandic slope beaked redfish are available from 2000–2023. The survey was not conducted in 2011.
Survey indices
The total biomass and abundance indices were highest in 2000 and 2001, declined in 2002 and have, since then, fluctuated at that level without clear trend (Figure 1). The biomass index of fish 45 cm and larger has, however, increased from the lowest value in 2007 to the highest one in 2021 but has since then decreased (Figure 1). The abundance index of fish 30 cm and smaller (recruits) has been at very low level since 2010. No fish smaller than 30 cm was observed in the 2021 and 2022 surveys and very little in 2023 (Figure 1).
Distribution
Icelandic slope beaked redfish from Icelandic autumn survey is caught along the south-east to the west slope of the Icelandic continental shelf (Figure 2) but is most abundant south-west along the Reykjanes ridge and west of Iceland (Figure 3). Icelandic slope beaked redfish is mainly caught at depths between 400–800 m (Figure 4).
Length and age
The length of the Icelandic slope beaked redfish in the autumn survey is between 25 cm and more than 50 cm (Figure 5). Since 2000, the mode of the length distribution has shifted to the right or from 36–39 cm in 2000 to about 42–45 cm in 2012–2023. During this period the mean length of the fish caught has increased from 37.4 cm in 2000 to 43.2 cm in 2023. This is a large increase in mean length for a species which annual growth is around 1–2 cm and where very few individuals larger than 50 cm are observed. This confirms the recruitment failure.
Otoliths from the autumn survey have been sampled since 2000 and otoliths from the 2000, 2006, 2009, 2010, 2014, 2017–2019 and 2021 surveys have been age read (Figure 6). The age reading shows that the stock consists of many cohorts and the age ranges from 5 to over 50 years. The 1985 and 1990 cohorts were large and were still relatively strong in the 2021 survey. No fish 10 years old and younger were observed in the 2021 survey, yet another indication of a recruitment failure.
Information from the fishing industry
Landings
Total annual landings of Icelandic slope beaked redfish from the Icelandic Waters Ecoregion 1950–2023 are presented in Figure 7.
During the 1950–1977 period, before the extension of the Icelandic EEZ to 200 NM, Icelandic slope beaked redfish was mainly fished by West-Germany. The catches peaked in 1953 to about 87 000 t but gradually decreased to about 23 000 t in 1977. After the extension of the Icelandic EEZ in 1978 the fishery has almost exclusively been conducted by Icelandic vessels. Annual landings gradually decreased from 57 000 t in 1994 to 17 000 t in 2001 and were at that level until 2010. Annual landings in the years 2011–2023 were between 8 300 and 12 000 t. The total catch in 2023 were 6 676 t, a decrease of 2 781 t from previous year.
Fisheries and fleets
The fishery for Icelandic slope beaked redfish in Icelandic waters is a directed bottom trawl fishery along the shelf and slope southwest and west of Iceland at depths between 500 and 800 m (Figure 8). The total number of boats that account for 95% of fishery have been declining steadily (Figure 9). In 1995 about 50 vessels fished for the stock but were around 20 in 2023.
Sampling from the commercial fishery
Table 1 shows biological sampling from the catch of Icelandic slope beaked redfish in the Icelandic Waters ecoregion 2000–2023. Number of samples and the number of length measurements have decreased since 2012.
Figure 10 shows the number of samples taken by month 2012–2023 and clearly shows the decreased sampling over time. The reason is reduced sampling effort of onboard observers from the Directorate of Fisheries.
Location of sampling in relation to the location of the fishery 2020-2023 is shown in Figure 11 and indicates that sampling effort is where the most of the fishery takes place.
Year | Samples | Measurements |
|---|---|---|
2000 | 202 | 42253 |
2001 | 103 | 19737 |
2002 | 179 | 32864 |
2003 | 168 | 29318 |
2004 | 140 | 22309 |
2005 | 207 | 34233 |
2006 | 256 | 40261 |
2007 | 142 | 22689 |
2008 | 200 | 33880 |
2009 | 184 | 30606 |
2010 | 168 | 28463 |
2011 | 138 | 21239 |
2012 | 68 | 11118 |
2013 | 64 | 9468 |
2014 | 93 | 15380 |
2015 | 58 | 9089 |
2016 | 92 | 13715 |
2017 | 57 | 10453 |
2018 | 26 | 4787 |
2019 | 41 | 7676 |
2020 | 27 | 5408 |
2021 | 23 | 4005 |
2022 | 7 | 241 |
2023 | 19 | 1880 |
Length distribution from the commercial catch
Length distributions of Icelandic slope beaked redfish from the bottom trawl fishery show an increase in the number of small fish in the catch in 1994 compared to previous years (Figure 12). The peak of about 32 cm in 1994 can be followed by approximately 1 cm annual increase in 1996–2002. The length distribution in 2004–2023 peaked around 39–42 cm. The length distribution of Icelandic slope beaked redfish from the pelagic fishery, where available, showed that in most years the fish was on average bigger than taken in the bottom trawl fishery (Figure 12).
Catch per unit effort
Trends in non-standardized CPUE (kg/hour) and effort (thousand hours fished) are shown in Figure 13. CPUE of tows where more than 50% and 80% of the catch was Icelandic slope beaked redfish gradually decreased from 1978 to a record low in 1994. Since then, CPUE has been steadily increasing and was in 2020 and 2021 at the highest level observed in the time series. From 1991 to 1994, when CPUE decreased, the fishing effort increased significantly. Since then, effort has decreased and is now at a similar level as in 1980. CPUE and effort data were not available for 2022.
Discard
Although no direct measurements are available on discards, it is believed that there are no significant discards of Icelandic slope S. mentella in the Icelandic redfish fishery.
Analytical assessment
The stock was benchmarked in 2023 (ICES, 2023) and is now assessed as a category 1 stock using an age- and length-based assessment model (Gadget). During the meeting the reference points were defined.
An error in the procedure for calculating biomass from the model output was found in 2024, which affected the estimates of total biomass and spawning stock biomass, and thus the reference points Blim and Bpa. Correcting the biomass levels therefore necessitates re-calculation of the stock’s reference points. The error did not affect estimated recruitment levels nor exploitation rates. The results are described in more details in Comparison with previous assessment and forecast
Maturity
Maturity-at-age are shown Figure 14. Males mature at earlier age than females. Most of the fish is mature at age around 20 years old.
Natural mortality
Natural mortality 𝑀 for long-lived species is considered low. In the assessment model presented here, 𝑀 was set as 0.05.
Assessment
Data
The model uses multiple disparate datasets. The input data includes:
Length disaggregated survey indices from the Autumn Survey IS-SMH (2000–2023, excluding 2011).
Length distributions from the Icelandic commercial bottom trawl fleet (1975–2023).
Landings per 6-month period from Iceland (1975–2023).
Age-length distributions from the Autumn Survey.
Maturation data from the Autumn survey.
An overview of the input data and their annual availability is shown in Figure 14.
Model settings
The model runs from 1975 to 2024. Each year is divided into two 6-month time-steps.
Two sub-stocks are modelled:
An immature stock that has an age range of 3–20 years.
A mature stock that has an age range of 5–50 years.
The oldest age is treated as a plus group (50 years and older).
Movement from the immature stock to mature stock occurs via:
Maturation (using a length-based ogive)
Ageing (20-year-old fish automatically move to the mature stock at the end of the year).
Modelled length ranged from 5–60 cm (with no mature individual 50 cm). Each length group was 1 cm.
Recruitment to the immature stock occurs at age 3.
The length increments in the survey were 10–30 cm, 30–35 cm, 34–40 cm, 41–45 cm, and 46–55 cm (in total five length bins).
One commercial fleet (bottom trawl).
Model processes
Natural mortality:
- Natural mortality, Ma, fixed at 0.05 for all ages. The value chosen was based on settings in other redfish stocks.
Growth:
Length-based Von Bertalanffy growth function, k, L∞, informed by age–length frequencies.
Parameter β of the beta-binomial distribution controlling the spread of the length distribution.
Maximum length group growth was set to 5 cm per timestep.
Length-weight relationship αs, βs, were fixed based on the means of log-linear regression of Autumn survey data.
Maturity:
- The logistic length-based maturity ogive αm, l50 was estimated from Autumn survey data.
Recruitment:
Annual recruitment occurs in the first timestep, one parameter per year, Ry, and y ∈ (1970, 2020).
Recruitment scalar, Rc, is multiplied against all Ry to help optimization.
Mean length at recruitment, l0, is estimated.
Length at recruitment has a CV of 0.1, based on Autumn survey.
Initial population:
Total initial abundance of both stocks, N0, is estimated.
Initial numbers-at-age calculated via 𝑁0,𝑎 = 𝑁0 × ^𝑒−𝑎(𝑀𝑎+𝐹0 )^
The additional mortality parameter 𝐹0 determines the steepness of the initial numbers-at-age reflecting previous effects of fishing (estimated).
Initial numbers-at-age is subsequently split between stocks using an age-based ogive. The age at which 50% of the stock was mature, 𝑎50, was estimated from the Autumn survey data and was fixed in the model, the alpha parameter of the ogive 𝛼𝑎 was estimated.
Initial mean length at age were based on the Von Bertalanffy growth function (see above).
Variance in initial length at age was fixed and based on length distributions obtained in the autumn survey for each stock.
Fleet operation:
Two fleets: commercial bottom trawl and Autumn survey fleet.
Logistic fleet selection, 𝛼f, l50; one set for each of the fleets (Autumn survey or Commercial).
Length-weight relationship
The conversion from length to weight uses the following formula:
\[W_{l} = \ \alpha*l^{\beta}\]
In the model, the alpha and beta parameters are fixed and estimated from biological information collected during the Icelandic autumn survey. The observed values and estimated relationship are shown in Figure 15.
Diagnostics and model fit
Survey indices can be variable for the Icelandic slope beaked redfish due to its tendency to be influenced by a few very large hauls. The index data used as input here are the total raw numbers of fish caught (within length slices) in the entire autumn survey. Although they are expected to represent the entire stock, they are also expected to be highly variable because no treatment or data pre-processing has been performed to reduce this variability. This variability is reflected in the model’s fit to the survey index data (Figure 17). In general, the model appears to follow the stock trends historically, although abundance is underestimated from 2000 to 2003 for the 10–30 cm, 30–35 cm, and 35–40 cm length groups. The terminal estimates do not deviate from the observed value for the for any of the length groups (Figure 16).
Length and age distribution
The model estimated catch composition is illustrated in Figure 17 to Figure 21, with corresponding residual plots for each catch composition component shown in Figure 22. The model fits both length distributions good (Figure 18, Figure 20 and Figure 21), although in some years, it is noticeable that the model is not capturing the peaks (ca. 40–45 cm fish) in the Autumn survey data (see 2012 to 2015 in Figure 18). The fits to the age distribution data from the autumn survey show that the fit is not particularly good for the oldest ages (30+) where the model underestimates these ages (Figure 17). Furthermore, the model overestimates certain age classes which can be followed through years, first in 2009 as 12-19 years old fish and then again in 2017 and 2018 as 20–28 years old fish. The fit to the commercial age-length distributions is worse; however, this is likely because there are few age readings in each time step (Figure 19). There are no discernible patterns in the residuals for any of the catch composition components (Figure 22).
Growth
For the Autumn survey, the growth patterns predicted by the model closely follow the observed growth from approximately age 10 onwards; however, prior to age 10, growth is underestimated (Figure 23). This noticeable shift is consistent between years suggesting that allowing for age specific variation in growth will improve the model. The model also fits the growth data from the bottom trawl consistently, although a similar trend of underestimating the growth rate in the younger ages is also apparent in 2001 and 2002 (Figure 24). This suggests that the model is overestimating the recruitment length, although it should be noted that (1) the age-length data is sparser for the younger ages, and (2) that because the stock does not enter the fishery until later ages, the beta-binomial length update will have created plausible standard deviations in the length at age by that time.
Maturation
The model’s fit to the maturation data is shown in Figure 25.
Fleet Selectivity
Estimated length-based selection by fleet is shown in Figure 26.
Model results
Annual output from the final model is shown in Figure 27. A steep decline in the spawning stock is seen from the late 1980s to the early 2000s. This is followed by a period of stability in the 2000s and a gradual decline in the 2010s. The SSB is currently at its lowest point in the time-series. Since a recruitment spike in 2003, annual recruitment has also steadily declined, and furthermore, since 2010 recruitment has remained at exceptionally low values resulting in a declining total stock size and a stock composition that is increasingly dominated by older, mature fish. Fishing mortality has declined since the 1990s and was fairly stable around 0.9 from 2013–2019 and 1.1 from 2020–2022.
Retrospective analysis
The analytical retrospective analysis is shown in Figure 28. The model is consistent for the first three peels. An upward revision in biomass (and thus downward revision in 𝐹) occurs on the 4th peel 5th peel. This suggests uncertainty in the model output; however, it should be noted that the larger revisions also coincide with the removal of age data. Notably, the last three years of age data from the Autumn survey are removed in the 4th and 5th peels ( Figure 28).
The Mohn’s rho values for Fbar and the spawning stock biomass are relatively low or -7.8% and 8.4% respectively. Mohn’s rho for recruitment is on the other hand very higher and is a result of high uncertainty due to low selectivity at the smallest age (3) detectable by the surveys. Mohn’s 𝜌 values are within the range recommended (< 0.2).
Conclusion
Overall, the gadget model presented here captures the overall trends in the data and offers a significant improvement over the current category 3 ‘survey trend’ empirical rule used in assessments. The main issues identified with the model, for instance, the consistent trend in the analytical retrospective analysis, and the fits to the age-length distributions (particularly to younger ages) will likely improve as more age data becomes available in the coming years.
Reference points
This year all reference point were estimated and revised. This re-estimate was needed after an error was discovered during the update assessment of Greenland halibut and beaked redfish. The routine that collated SSB from the model output incorrectly calculated mean weight at length resulting in a upward bias in total and spawning stock biomass estimates (see Figure 29). As the estimate of B\(_{lim}\) for Icelandic slope beaked redfish was based on B\(_{loss}\) it was affected by this error and thus the biomass and fishing pressure reference points needed to be revised. Figure 29 illustrates the difference between last years estimates and the corrected last years assessment.
The revision of reference points followed the same procedure as used at WKBNORTH(ICES 2023) and resulted in higher estimated of B\(_{lim}\) and B\(_{pa}\) and lowered estimates of F\(_{lim}\) and F\(_{pa}\) (see Table 2). F\(_{MSY}\) was also lowered from 0.061 to 0.041.
Approach | Reference point | Value | Old values | Basis |
|---|---|---|---|---|
MSY approach | MSY Btrigger | 217 563 | 192 119 | Bpa |
FMSY | 0.041 | 0.061 | Fishing mortality that leads to MSY; estimated using stochastic simulations | |
Precautionary approach | Blim | 156 568 | 138 257 | Bloss. Median SSB (2000–2005) |
Bpa | 217 563 | 192 119 | Blim × e1.645σ, σ = 0.2. | |
Flim | 0.079 | 0.079 | Fishing mortality that in stochastic equilibrium will result in median SSB at Blim | |
Fpa | 0.041 | 0.061 | Maximum F at which the probability of SSB falling below Blim is < 5% |
State of the stock
The state of the stock is at a low level. Since 2007, survey estimates of have consistently shown very low abundance of pre-fishery juveniles (< 30 cm). This raises concerns about the productivity of the stock. Without substantial recruitment biomass levels will likely continue to decline.
Short term forecast
Maturity, growth, and the length-weight relationship in the forecast are based on the processes estimated within the model. Similarly, the commercial fleet selectivity is the same as estimated by the model. Intermediate catch is equal to the leftover of the current fishing year (January-August) and for September-December F=FMSY. Recruitment in the forecast is the average of the last five years (2020–2023). Values for the interim year is shown in Table 3. The results of the prognosis are shown in Table 4.
Variable | Value | Notes |
|---|---|---|
F (2024) | 0.014 | F that corresponds to assumed catch in 2024 |
SSB (2025) | 91 426 | Projected from the assessment; in tonnes |
Recruitment age 2 (2025) | 3.7 | From the assessment; in thousands |
Recruitment age 2 (2026) | 3.5 | From the assessment; in thousands |
Catch (2024) | 1 500 | Based on catches in the first eight months of the 2023/2024 fishing year; in tonnes |
Basis | Catch (2025) | F (2025) | SSB (2026) | % SSB change1) | Advice change2) |
|---|---|---|---|---|---|
MSY approach | 0 | 0 | 90 571 | -0.9 | |
1) SSB in 2026 relative to SSB in 2025 | |||||
2) Advice value for 2025 relative to advice value for 2024 (0 t) | |||||
Uncertainties in the assessment and forecast
Only the fishable biomass of the stock is found in Icelandic waters and recruitment comes most likely from East Greenland. Connection of the Icelandic slope stocks with the stocks found in East Greenland and the deep pelagic stock is not known. Currently, little age data is available, therefore, when years are removed in the retrospective analysis the model results substantially. We anticipate reduced uncertainty when more age data are added.
Comparison with previous assessment and forecast
Figure 29 shows biomass trajectories before and after the correction. There is an upward revision in the estimated SSB and TB. The difference is largest (~12%) when the SSB stabilized from 2000 to 2008. However, throughout the time series the uncorrected estimates of SSB and TB both fall within the uncertainties (yellow shaded area) of the respective revised estimates. Furthermore, the corrected terminal estimates of both SSB and TB are very similar to the uncorrected equivalents (3% increase) so the status of the fishery and the advice for the 2023–2024 fishing year remain unchanged.
Basis for advice
ICES MSY approach agreed during the WKBNORTH meeting (ICES 2023).
Management considerations
Beaked redfish is a slow growing, late maturing deep-sea species and is therefore considered vulnerable to overexploitation and advice must be conservative.
Regulations and their effects
There are no explicit management for Icelandic slope beaked redfish. The species is managed under the ITQ system.
Management
Ministry of Food, Agriculture and Fisheries (MFAF) in Iceland is responsible for management of the Icelandic fisheries, including the Icelandic slope beaked redfish fishery, and for the implementation of the legislation in the Icelandic Exclusive Economic Zone (EEZ). There is, however, no explicit management plan for the Icelandic slope beaked redfish.
The Ministry issues regulations for commercial fishing for each fishing year (1 September–31 August), including allocation of the TAC for each of the stocks subject to such limitations. Redfish (golden redfish (Chapter 19) and Icelandic slope beaked redfish) has been within the ITQ system from the beginning. Icelandic authorities gave, however, until the 2010/2011 fishing year a joint quota for these two species, and Icelandic fishermen were not required to divide the redfish catch into species. MFRI has since 1994 provided a separate advice for the species. The separation of quotas was implemented in the fishing year that started 1 September 2010.
Figure 30 shows the net transfer of quota to or from Icelandic slope beaked redfish by fishing year. For the first five fishing years (2010/2011–2014/2015), very little was transferred from (negative values) or to (positive values) beaked redfish and the set quota was taken. During the 2015/2016–2020/2021 fishing year there were substantial quota of slope beaked redfish was transferred to other species or about 10–20% per fishing year and the set quota not fished. But in the last two fishing years (2021/2022 and 2022/2023) this pattern reversed, that is, quota was transferred from other species to Icelandic slope beaked redfish. The catch, therefore, exceeded the set quota of more then 20%.
References
ICES. 2023a. Benchmark workshop on Greenland halibut and redfish stocks (WKBNORTH). ICES Scientific Reports. 5:33. https://doi.org/10.17895/ices.pub.22304638