Garrido, González-Troncoso and González-Costas assess cod stock in NAFO Division 3M; SSB above Blim, below Btrigger in 2025
Assessment of the Cod Stock in NAFO Division 3M
Northwest Atlantic Fisheries Organization
Serial No. N7743 NAFO SCR Doc. 26/023
SCIENTIFIC COUNCIL MEETING –JUNE 2026
Assessment of the Cod Stock in NAFO Division 3M
by
Irene Garrido, Diana González-Troncoso and Fernando González-Costas
Instituto Español de Oceanografía, Vigo, Spain
GARRIDO, I., GONZÁLEZ-TRONCOSO, D., & GONZÁLEZ-COSTAS, F. 2026. Assessment of the Cod Stock in NAFO Division 3M. NAFO Scientific Council Research Document, SCR Doc. 26/023: 1-55.
Abstract
An assessment of the cod stock in NAFO Division 3M was conducted using a Bayesian SCAA (statistical catch- at-age) model. The STACFIS catch estimates and the Flemish Cap survey indices were used to fit the model. In 2025, SC approved the Reference Points to be applied under the approved NAFO Precautionary Approach Framework. Bmsy is defined for this stock as the equilibrium SSB that corresponds to the F35%SPR estimated with the mean values of the biological parameters and recruitment of the 2007-2023 period. Blim is 30% Bmsy (Blim = 15 724 t). Btrigger was approved to be 75% Bmsy (Btrigger = 39 310 t). Flim corresponds to F35%SPR estimated with the mean values of the biological parameters of the 2007-2023 period (Flim = 0.171) and Ftarget is the 85% of Flim (Ftarget = 0.145). Results indicate that SSB declined rapidly since 2017 but has remained stable during the last five years and is estimated to be above Blim and below Btrigger (Cautious Zone) in 2025. Since 2013, recruitment has varied at intermediate levels but much lower than those observed in 2011-2012. In 2021 and 2022, good recruitments were observed, while since 2023 is at a very low level. F has remained below Flim since the fishery reopened in 2010, and has been below Ftarget during that period except in 2011. In 2021, the minimum level of F since the re-opening was reached, increasing since then.
Introduction
The 3M cod stock was under fishing moratorium from 1999 to 2009 following a decline to well below Blim (Vázquez and Cerviño, 2005). The stock collapse has been attributed to three simultaneous circumstances: 1) overfishing, 2) increased catchability at low abundance levels and 3) a series of very poor recruitments starting in 1993. The relatively good recruitments (age 1) observed after 2005 allowed the reopening of the fishery in 2009.
Since 1974, when a TAC was established for the first time, estimated catches ranged from 48 000 tons in 1989 to 5 tons in 2004. Annual catches were about 30 000 tons in the late 1980’s (notwithstanding the fact that the fishery was under moratorium in 1988-1990) and diminished since then as a consequence of the stock decline. Between 1998 and 2009, almost coinciding with the last fishing moratorium, yearly catches were below 1 161 tons. The results of the 2009 assessment led to a reopening of the fishery with a TAC of 5 500 tons in 2010. With the results of the following years assessments established TACs for 2010-2025 ranged between 17 500 tons in 2019 and 1 500 tons in 2021, being 12 613 tons for 2025 and 15 360 for 2026. The STACFIS estimated catches for 2010-2025 were between 17 520 tons in 2019 and 2 055 tons in 2021, being 12 485 tons in 2025 (Table 1A and Figure 1).
2 Northwest Atlantic Fisheries Organization
www.nafo.int A VPA based assessment of the cod stock in Flemish Cap was approved by NAFO Scientific Council (SC) in 1999 for the first time and was annually updated until 2002. However, catches between 2002 and 2005 were very small undermining the VPA based assessment, as its results are quite sensitive to assumed natural mortality when catches are at low levels. Cerviño and Vázquez (2003) developed a method which combines survey abundance indices at age with catchability at age, the latter estimated from the last reliable accepted XSA. The method estimates abundances at age with their associated uncertainty and allows calculating the SSB distribution and, hence, the probability that SSB is above or below any reference value. The method was used to assess the stock in the period 2003-2007. In 2007 results from an alternative Bayesian XSA model were also presented (Fernández et al., 2007) and in 2008 this XSA model was further developed and approved by the NAFO SC (Fernández et al., 2008), being used between 2008 and 2017 in the assessment of this stock.
In April 2018 a benchmark on the 3M cod was carried out by the NAFO Scientific Council (NAFO, 2018a). During that meeting it was decided to replace the Bayesian XSA with a Bayesian SCAA (statistical catch-at-age), that has been used since then. Another important change introduced at the benchmark is the prior median value of the natural mortality, which the benchmark agreed to base on biological and multi-species considerations; this has resulted in considerably higher values of M than estimated in previous assessments. The results of the Bayesian SCAA model are presented here, including the updated input data until 2025.
In 2020 the Commission adopted technical measures, in force since January 2021 (NAFO, 2021), to try to protect the productivity of Division 3M cod stock. These measures included the closure of the directed fishery of the 3M cod during the first quarter of the year, as well as the mandatory use of sorting grids when using a net having a mesh size of 130 mm in this fishery.
Besides these technical measures, no vessel shall fish cod in Division 3M with a net having a mesh size smaller than 155 mm when not using sorting grids, and the minimum landing size for this stock is 41 cm (fork length) (NAFO, 2025).
Material and Methods
Data used Commercial data
Total Catch
In 2010 the fishery on this stock was reopened after the moratorium period between 1999 and 2009. Since then, STACFIS/ CESAG estimated catches were used for the stock assessment (González-Costas et al., 2018; NAFO, 2018a). Between 2010 and 2012, only trawler vessels were present in the fishery; since 2013, longliners from Faroes and Norway were also periodically active. Since 2017, the Faroese fishery has been exclusively conducted by longliners. Since 2016, Norwegian vessels alternate the use of trawl and longline, depending on the year. This causes the proportion of trawlers and longliners to be variable among the years, ranging between 16% and 53% (Table 1B).
In 2025 the CESAG estimated catch was 12 485 tons (Table 1A, Figure 1). Information on cod catches from the following countries were available for the estimation in 2025: EU-Portugal, EU-Spain, Faroe Islands (Denmark), Norway, Russia and United Kingdom. The proportion of longline catches in 2025 was 22% (Faroe Islands catches).
Length distributions
In 2025 there were commercial length distributions from EU-Portugal (Alpoim et al., 2026), EU-Spain (González-Costas et al., 2026), Russia (Fomin and Pochtar, 2026), United Kingdom (NAFO Secretariat communication), Faroes (Ridao-Cruz, 2026) and Norway (Nedreaas, personal communication). Length sampling data by EU-Portugal were available only after the assessment was performed; therefore, this information was not considered. The available length distributions for trawlers weighted to the total trawl catch, on one hand, and the length distribution for the longliners weighted to the total longliner catch, on the
3 Northwest Atlantic Fisheries Organization
www.nafo.int other hand, were added to get the total commercial length distribution. The length frequency distributions in 2025 from the commercial catch by country and total and from the EU survey (González-Troncoso et al., 2026) are shown in Figure 2A.
Table 1C shows the number of individuals measured as well as the length range, the mean and the mode for each of the countries with samples and for the total commercial length distribution, and Table 1D the mean and mode of the final total commercial length distributions and the survey length distribution for 2010-2025.
Figure 2B shows the total commercial length distribution since 2010. In 2021 there is a main mode in the length distribution between 63-70 cm and a secondary mode, much weaker, between 45-50 cm. In 2022, there are two modes at the same level, one around 51-55 cm and another around 66 cm. Since 2023, the length distribution shows a unimodal shape, with the mode being around 54 cm. The mean length (Figure 2C) in the 2021-2025 period was fairly constant around 60 cm.
Indices by age
As no age-length keys (ALK) were available for commercial catch from 1988 to 2008, each year the corresponding ALKs from the EU survey (read by the IIM in Vigo) were applied in order to calculate annual catch-at-age. An ALK was available for 2009-2011 only from the Portuguese fishery and was applied to the total commercial catch length distribution to derive the total age distribution of the commercial catches.
Since 2012 the ALK from the EU survey has been used for both commercial and survey indices, although some years ALK from the Portuguese and/or the Spanish catches were available. The reason not to use the commercial ALKs to the commercial distribution is that these commercial ALKs have not been validated and more research is needed to completely identify the source of discrepancies observed.
Due to administrative problems, the ALK from the EU survey since 2023 has not been available for the assessment. The ALK resulting from the average of the last 3 available years (2020-2022) was used to derive both commercial and survey indices in 2023, 2024 and 2025.
Catch-at-age
Catch-at-age in numbers for 1988 to 2025 is presented in Table 2. These numbers were obtained by applying the ALK to the total commercial catch length distribution each year.
The catch-at-age ranges from 1 to 8+. No catch-at-age was available for 2002-2005 due to the lack of length distribution information because of low catches. Catch proportions at age over time (Figure 3A) indicate that the bulk of the catch was comprised of 3-5 years age cod until 2015, although between 2006 and 2014 the catches contained mostly ages 3 and 4 individuals; in the period 2015-2023, ages 5 to 8+ were the most dominant in the catches. Since 2024, the most fished ages were 4 and 5.
Figure 3B shows standardised catch proportions at age (each age standardised independently to have zero mean and standard deviation 1 over the range of years considered). Assuming that the selection pattern at age is not too variable over time, it should be possible to follow cohorts from such figure. Some strong and weak cohorts can be followed, although the pattern is not too evident. The 2009 and 2010 cohorts can be easily followed, reaching age 8+ in 2019 and 2020. Cohort from 2011 started with a good recruitment in 2012 but then disappeared until age 5, in 2016. The cohorts since 2012 were poor. As a consequence, since 2015 all the values of the ages less than 4 have been negative. The 2011 cohort show a recruitment over the average in 2012, but remains below the average for all the other ages. Since then, low recruitments have been observed as well as low abundances until age 4. It is remarkable the big catches starting at age 6 for these cohorts, except for the one from 2015, recruited in 2016, which has all the values negative until 8+.
4 Northwest Atlantic Fisheries Organization
www.nafo.int Mean weight-at-age
For 2025, there are two commercial length-weight relationships available: EU-Spain and Faroes. Both are presented in Figure 4 alongside the 2025 EU bottom survey one. This year, the length-weight relationship from the EU bottom trawl survey has proved to be the one that better characterizes the length distribution of the commercial fleet and, for this reason, was the one used to obtain the mean weight-at-a