Canadian Science Advisory Secretariat (CSAS) Research Document 2025/020 Maritimes Region May 2025 Framework Review for 4X5Y Haddock: Part 1 - Review of the Data Inputs Barrett, T.J., and Barrett, M.A. Fisheries and Oceans Canada Saint Andrews Biological Station 125 Marine Science Drive St. Andrews, New Brunswick, E5B 0E4 Foreword This series documents the scientific basis for the evaluation of aquatic resources and ecosystems in Canada. As such, it addresses the issues of the day in the time frames required and the documents it contains are not intended as definitive statements on the subjects addressed but rather as progress reports on ongoing investigations. Published by: Fisheries and Oceans Canada Canadian Science Advisory Secretariat 200 Kent Street Ottawa ON K1A 0E6 http://www.dfo-mpo.gc.ca/csas-sccs/ DFO.CSAS-SCAS.MPO@dfo-mpo.gc.ca © His Majesty the King in Right of Canada, as represented by the Minister of the Department of Fisheries and Oceans, 2025 This report is published under the Open Government Licence - Canada ISSN 1919-5044 ISBN 978-0-660-76389-7 Cat. No. Fs70-5/2025-020E-PDF Correct citation for this publication: Barrett, T.J., and Barrett, M.A. 2025. Framework Review for 4X5Y Haddock: Part 1 - Review of the Data Inputs. DFO Can. Sci. Advis. Sec. Res. Doc. 2025/020. iv + 85 p. Aussi disponible en français : Barrett, T.J. et Barrett, M.A. 2025. Examen du cadre pour l’aiglefin des divisions 4X5Y : Partie1 – Examen des données d’entrée. Secr. can. des avis sci. du MPO. Doc. de rech. 2025/020. iv + 92 p. http://www.dfo-mpo.gc.ca/csas-sccs/ mailto:DFO.CSAS-SCAS.MPO@dfo-mpo.gc.ca https://open.canada.ca/en/open-government-licence-canada#:%7E:text=Open%20Government%20Licence%20-%20Canada%201%20Using%20Information,Governing%20Law%20...%208%20Definitions%20...%20More%20items iii TABLE OF CONTENTS ABSTRACT .................................................................................................................................. iv INTRODUCTION .......................................................................................................................... 1 FRAMEWORK REVIEW AND OBJECTIVES .......................................................................... 1 BACKGROUND ............................................................................................................................ 1 HISTORY OF THE 4X5Y FISHERY ......................................................................................... 1 HISTORY OF THE 4X5Y ASSESSMENT ................................................................................ 2 REVIEW OF STOCK STRUCTURE ......................................................................................... 2 REVIEW OF DIFFERENCES IN GROWTH BY DFO STATISTICAL UNIT AREA .................. 3 FISHERY ...................................................................................................................................... 5 CATCH ..................................................................................................................................... 5 SPATIAL AND TEMPORAL TRENDS OF THE CATCH .......................................................... 6 CATCH-AT-AGE (CAA) ............................................................................................................ 6 Catch at Length ..................................................................................................................... 7 Age Length Key ..................................................................................................................... 8 Catch at Age ......................................................................................................................... 8 Catch Uncertainty in 4Xp ...................................................................................................... 8 Catch-at-Age for the Alternative Catch Scenarios .............................................................. 11 FISHERY LENGTH-AT-AGE (LAA) ....................................................................................... 11 FISHERY WEIGHT-AT-AGE (WAA) ...................................................................................... 12 SURVEYS ................................................................................................................................... 12 INDIVIDUAL TRANSFERABLE QUOTA SURVEY ................................................................ 12 DFO SURVEY INDEX ............................................................................................................ 13 DFO SURVEY NUMBERS-AT-AGE AND LENGTH .............................................................. 13 DFO SURVEY LENGTH-AT-AGE AND WEIGHT-AT-AGE ................................................... 14 DFO SURVEY MATURITY ..................................................................................................... 15 DFO SURVEY DISTRIBUTION, CONDITION, SURVEY Z, AND RELATIVE F .................... 15 ECOSYSTEM CONSIDERATIONS ............................................................................................ 16 CONCLUSIONS .......................................................................................................................... 17 ACKNOWLEDGEMENTS ........................................................................................................... 18 REFERENCES CITED ................................................................................................................ 18 TABLES ...................................................................................................................................... 22 FIGURES .................................................................................................................................... 34 iv ABSTRACT Haddock (Melanogrammus aeglefinus) are caught as part of a multi-species groundfish fishery concentrated on the western Scotian Shelf (SS) and in the Bay of Fundy (BoF) in the Northwest Atlantic Fisheries Organization (NAFO) Divisions 4X5Y. This document is a review of the data inputs for the modeling framework for the 4X5Y Haddock stock that is expected to be completed in 2024. Stock structure was reviewed and an evaluation of spatial differences in growth was conducted to identify appropriate boundaries for the separation of the data inputs for faster growing Haddock in the BoF and slower growing Haddock on the SS. Fleet structure and the spatial and temporal distribution of catches were reviewed. The method to estimate catch-at- age was revised and the catch history was estimated for two alternative catch scenarios that assume catches in the south of Fisheries and Oceans (DFO) statistical unit area (DFO unit area) 4Xp are from the 5Z Haddock stock. These catch scenarios can be used to capture uncertainty in stock mixing and to explore the potential causes of retrospective patterns in model fits. The DFO summer ecosystem survey biomass index for 4X5Y Haddock was estimated based on a weighted-mean biomass and a weighted-mean biomass assuming a delta-lognormal distribution. These two indices will be evaluated in sensitivity analyses when models are fit. The fishery and survey length-at-age, weight-at-age, age-length keys, and the survey maturity-at-age were estimated by region (BoF and SS) using methods to fill missing data. Stomach contents data and a number of ecosystem indicators from the DFO Atlantic Zonal Monitoring Program were identified to be considered for exploring ecosystem considerations for the stock. 1 INTRODUCTION Haddock (Melanogrammus aeglefinus) occur in the northwestern Atlantic from southwest Greenland to Cape Hatteras. The species is a bottom dwelling member of the gadid family that occurs most commonly at depths of 30 to 350 m and at bottom temperatures above 2°C (Scott and Scott 1988). Their diet consists mainly of small invertebrates and fish. A major stock exists on the western Scotian Shelf (SS) and in the Bay of Fundy (BoF) in the North Atlantic Fisheries Organization (NAFO) divisions 4X5Y (Figure 1). Major spawning aggregations are found on Browns Bank (Figure 2) and peak spawning occurs from April to May, although spawning may occur as early as February if conditions are favorable (Head et al. 2005). The most recent analytical assessment of 4X5Y Haddock was based on a virtual population analysis (VPA) model (Wang et al. 2017). In 2018, projections from the VPA model showed large retrospective patterns and there was a mismatch between the model predicted biomass and the survey biomass (Finley et al. 2018). The VPA model has therefore not been used to provide catch advice or estimate stock status since 2018, and stock status updates have been provided qualitatively by examining temporal trends in the biomass index estimated from the Fisheries and Oceans Canada (DFO) Maritimes Summer Ecosystem Survey, hereafter the “DFO ecosystem survey”. FRAMEWORK REVIEW AND OBJECTIVES This document is Part 1 of the Framework Review for 4X5Y Haddock and represents the data inputs and considerations for the modelling framework that is expected to be completed in 2024. The specific objectives of this document are to: • Review current stock structure and evaluate whether there is a scientific basis for any changes in stock structure or the management area for 4X5Y Haddock. • Review the basis for separating the stock into regions (BoF and SS) based on growth rates and review how fishery fleets are defined. • Review the commercial fishery data inputs: spatial and temporal distribution of the catch, fishery catch-at-age (CAA), age-length keys (ALK), and fishery weight-at-age (WAA). • Review DFO ecosystem survey data inputs: biomass index, ALK, CAA, WAA, maturity, fish condition, relative annual fishing mortality (relF), and relative annual total mortality (relZ). • Identify potential datasets that can be used to explore ecosystem considerations for the stock. BACKGROUND HISTORY OF THE 4X5Y FISHERY A total allowable catch (TAC) for 4X5Y Haddock was first introduced in 1970 and a seasonal spawning closure was implemented on Browns Bank (February 1–June 15, Stone and Hansen 2015). The minimum mesh size used in fishing nets has varied throughout the fishery, but 130 mm square mesh was made mandatory in 1992. Limited entry licensing, first introduced for the large trawler fleet, was extended to all groundfish vessels in 1976. In 1977, Canada extended its jurisdiction from 12 nautical miles to 200 nautical miles from the coast, and foreign vessels could now only fish under a Canadian licence (DFO 2018). In the early 1990s, management measures were implemented for dockside monitoring, small fish protocols, and 2 conservation harvesting plans (DFO 2018). For a more detailed review of the history of the 4X5Y Haddock fishery, see Stone and Hansen (2015). Starting in the 2015–16 fishing season, a minimum size of 38 cm was established for a small fish protocol. Areas are closed when the number of undersized Haddock (<38 cm) exceed a percentage of the catch (25–40% depending on the year). At-sea observer coverage has been low in 4X5Y with a target of 5–20% for observed trips; however, the realized number of observed trips has been lower (<4.3%) in the last 5 years. HISTORY OF THE 4X5Y ASSESSMENT Over the past decade, two models were used for the 4X5Y Haddock assessment. The first was a Sequential Population Analysis (SPA) tuned to the DFO summer ecosystem survey and a joint industry and DFO led Individual Transferable Quota (ITQ) survey (1995–2012, Hurley et al. 2009). The second was a VPA model with varying natural mortality (M) at ages >10 for different time blocks (Stone and Hansen 2015, Wang et al. 2017). In both cases, a strong retrospective pattern in the model results (i.e., a tendency to systematically overestimate spawning biomass when additional years of data were added) and poor model fit to survey indices occurred within 5 years. Consequently, both the SPA and VPA model results were not considered reliable to produce meaningful projections and catch advice. In 2010, the fishery was managed using a removal reference fishing mortality rate (Fref=0.25). The limit reference point (LRP) was defined as 0.4 SSBMSY (spawning stock biomass at maximum sustainable yield) and the upper stock reference point (USR) as 0.8 SSBMSY based on biomass estimates from a Sissenwine-Sheppard stock production model (Mohn et al. 2010). During the 2016 framework, reference points were re-evaluated, and a fishing mortality limit reference (Flim) of 0.25 was defined to be applied when the stock is in the healthy zone (i.e., SSB > USR), and a Fref of 0.15 was defined to be applied when the stock is in the cautious zone (i.e., LRP < SSB < USR). The LRP was revised and defined based on Brecover (lowest biomass that produced recruitment that led to stock recovery) and the USR was changed to approximately twice the LRP (Wang et al. 2017). Since 2018, the VPA model has not been used to provide catch advice, and stock status updates have been provided qualitatively by comparing the annual survey biomass index to the long-term median biomass index (e.g., DFO 2020, DFO 2021a). REVIEW OF STOCK STRUCTURE The 4X Haddock assessments from 1974–1997 considered catches in 4Xs and the Canadian portion of 5Yb and survey strata 492–494 (Figure 3) as part of the Gulf of Maine (GoM) stock in 5Y (Hurley et al. 1998). The 4Xs and 5Yb areas were first combined with the 4X Haddock assessment in 1998 after a re-evaluation of stock definition (Hurley et al. 1998). In the northwest Atlantic Ocean, there are likely to be partially discrete groups of Haddock on Georges Bank, northern GoM and BoF, western SS and Browns Bank, and the eastern SS (Grosslein 1962, Page and Frank 1989, Begg 1998) based on physical and oceanographic factors (e.g., Fundian Channel, Browns Bank gyre) that serve as semi-permeable barriers. Eggs and larvae of Haddock can episodically cross these barriers with changing environmental conditions (Campana et al.1989) and movement by juveniles and adults typically occurs seasonally (Schroeder 1942, Frank 1992, Begg and Weidman 2001, Brickman 2003, Fowler 2011). The main Haddock spawning areas in the region are on Georges Bank and Browns Bank (Figure 2, Wise and Jensen 1960). The timing of spawning depends on temperature, with spawning occurring earlier in New England and on Georges Bank (February–March) compared 3 to the SS (April–June, Lapolla and Buckley 2005, Begg 1998). Larvae that hatch earlier in the season are predicted to have higher survival due to lower predation (Lapolla and Buckley 2005). The spawning areas have strong gyres that retain fishes and their prey on the Banks (Campana et al. 1989). The gyre on Browns Bank releases larvae onto the SS where currents transport them inshore and towards the BoF (Campana et al. 1989, Hurley and Campana 1989). Biophysical modeling has suggested that ocean currents on Browns Bank may episodically export a significant amount of larvae to Georges Bank (Brickman 2003) and vice versa (Campana et al. 1989). When abundance is high, Haddock may move from high density areas to less suitable habitats with lower intraspecific competition (Brodziak et al. 2008, Stone and Hansen 2015). Otolith stable isotope analyses have provided evidence that Haddock shift their distribution and home range throughout their life history (Begg and Weidman 2001). It is hypothesized that for large year classes, juvenile Haddock may move from the eastern SS (Western Bank) to the central and western SS, ultimately leading to mixed stocks (Frank 1992, Brickman 2003). Adults typically return to their natal origin to spawn, and larger, older Haddock migrate more than smaller Haddock (Needler 1930). Early tagging studies of Haddock provided evidence of seasonal mixing of adults between the BoF, GoM, Great South Channel, and Georges Bank (Figure 2, Needler 1930, Schroeder 1942, McCracken 1960). Haddock tagged off Digby (BoF) were recaptured on Georges Bank, and in some years a small proportion of tagged individuals moved from Georges Bank to GoM (Brodziak and Col 2006), while in other years the tagging data suggested movement was minimal (Brodziak et al. 2008). Fowler (2011) proposed two remaining migratory populations of Haddock on the Scotian Shelf: 1. western SS which overwinter on Browns Bank and move inshore (4Xr) during the summer and; 2. eastern SS (4TVW) which overwinter in 4W offshore and move to the southern Gulf of St. Lawrence in the summer. Genetic studies focused on the population structure of Haddock in the northwest Atlantic are limited. A study examining genetic variation of Georges Bank Haddock found significant differences between samples from 1975 and 1985, suggesting genetic heterogeneity and variation in the annual contributions of these genetic components (Purcell et al. 1996). Lage et al. (2001) found no significant genetic differences in four microsatellite loci among Haddock caught on Georges Bank, Browns Bank, and the SS; however, a study using single nucleotide polymorphisms (SNPs) found that samples from the western SS were significantly different from Georges Bank but not the GoM (Berg et al. 2021). Further research is needed to understand the potential genetic differentiation of Haddock populations. REVIEW OF DIFFERENCES IN GROWTH BY DFO STATISTICAL UNIT AREA Differences in growth rates between the BoF (DFO ecosystem survey strata 482–495) and western SS (DFO ecosystem survey strata 470–481) regions were reported by Hurley et al. (1998). These differences in growth rates formed the basis of the separation of the survey biomass index by region (BoF vs. SS) and estimation of the CAA using separate ALKs by region in the most recent assessment framework (Stone and Hansen 2015). However, spatial areas used to define the regions for the catch and survey did not align (see Figure 1 and Figure 32 in Stone and Hansen 2015), such that catches in 4Xp were grouped as SS and the portions of survey strata 482 to 485 in 4Xp (Figure 3) were grouped as BoF. An evaluation of growth rates was conducted by DFO statistical unit area (hereafter DFO unit area) to assess whether there is 4 still support for the status quo preparation of data inputs by region and to determine the most appropriate spatial boundaries for the definition of regions (BoF vs. SS). Von Bertalanffy (vonB) growth models were fit to the DFO ecosystem survey and National Marine Fisheries Service (NMFS) bottom trawl survey for GoM age and length data by cohort for DFO unit areas in 4X, GoM grouped as 5Y (NMFS survey strata 26–28 and 36–40; see Figure 4), the Canadian portion of 5Z (i.e., 5Zjm), and the western four DFO unit areas of 4W (i.e., 4Whjkl). There were insufficient NMFS survey data to fit vonB growth models at a finer geographic scale than 5Y. All available survey data were used (generally summer data for 4X, summer and winter for 5Z, and spring and fall for 5Y) and growth relationships were fit when there were at least 15 observations per area and cohort. There were insufficient data in the Canadian portion of 5Yb and in 4Xm to estimate growth relationships for these areas. Ages were adjusted to a fraction of the year to account for the month the fish was sampled (e.g., 𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 = 𝑎𝑎 + 1 12 for February and 𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 = 𝑎𝑎 + 11 12 for December where 𝑎𝑎 is age as an integer in years). VonB growth models (modelling length as a function of age) were fit as: 𝐿𝐿 = 𝐿𝐿𝑖𝑖𝑖𝑖𝑖𝑖�1 − 𝑒𝑒−𝑘𝑘(𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎−𝑎𝑎0)� Eqn 1 using maximum likelihood estimation to minimize residuals where parameter 𝐿𝐿𝑖𝑖𝑖𝑖𝑖𝑖 represents the asymptotic length, parameter 𝑘𝑘 represents the growth rate (a measure of how fast 𝐿𝐿𝑖𝑖𝑖𝑖𝑖𝑖 is reached), and parameter 𝑎𝑎0 is the theoretical length-at-age (LAA) zero. The models were estimated with and without the 𝑎𝑎0 parameter and the final selected models excluded 𝑎𝑎0 due to the limited data available to reliably estimate 𝑎𝑎0, and using the age adjustment, a length of zero at age zero was deemed a reasonable assumption. The relationship between 𝐿𝐿𝑖𝑖𝑖𝑖𝑖𝑖 and 𝑘𝑘 (Figure 5) and the changes in 𝐿𝐿𝑖𝑖𝑖𝑖𝑖𝑖 and 𝑘𝑘 by cohort (Figure 6) showed differences among areas (e.g., slower growth in 4W and faster growth in 5Z). A loess smoother (span = 0.5) was used to smooth the interannual variability in the 𝐿𝐿𝑖𝑖𝑖𝑖𝑖𝑖 and 𝑘𝑘 estimates and help identify differences among areas. Differences in 𝐿𝐿𝑖𝑖𝑖𝑖𝑖𝑖 among DFO unit areas can be described qualitatively as: 4Whjkl < 4Xnop < 4Xqrs ~ 5Y ~ 5Zjm Although 𝐿𝐿𝑖𝑖𝑖𝑖𝑖𝑖 and 𝑘𝑘 are correlated (Figure 5), the relationship is not 1:1 such that differences in 𝑘𝑘 among DFO unit areas can be described qualitatively as: 4Whjkl > 5Zjm > 4Xqrs > 4Xnop and 5Y ~ 4Xqrs5Zjm Ninety five percent confidence intervals were added to the loess smoothers for 4Xqrs and 4Xnop (Figures 7 and 8) to evaluate the status quo assumption that there are growth differences between BoF (4Xqrs) and western SS (4Xnop) regions. In general, there is support for this assumed difference in growth (e.g., non-overlapping confidence intervals in Figure 7), although 𝐿𝐿𝑖𝑖𝑖𝑖𝑖𝑖 in 4Xp is higher than 4Xno beginning in the mid-1980s and the relationship between 𝐿𝐿𝑖𝑖𝑖𝑖𝑖𝑖 and 𝑘𝑘 for 4Xq deviates from 4Xrs. To determine the appropriate spatial bounds for the separation of the BoF (faster growing) and SS (slower growing) regions, the vonB growth models were fit separately by survey strata in 4Xp (Figures 9 and 10). The growth parameters for survey strata 480 and 481 were more similar to 4Xno so these survey strata were included in the SS region and growth parameters for survey strata 482 and 483 were more similar to 4Xqr so these survey strata were included in the BoF region (Figure 1, Figure 9, Figure 10). This spatial definition also formed the basis for 5 defining regions for the catch history and is supported by length frequency distributions and growth of Haddock from the fishery catch (data obtained from port samples). FISHERY Haddock in 4X5Y are caught as part of a multi-species groundfish fishery. The science advice and the management of the fishery is specific to each major harvested species (i.e., Haddock, Halibut, Cod, Pollock, Redfish, and Silver Hake). The Haddock fishery is limited by the incidental catch of Cod which has a TAC that is usually reached first among the TACs for the various groundfish species. Haddock are primarily caught using bottom trawls; however, fixed gears are also used (e.g., longline and handline). The directed Haddock fishery bottom trawls have used a 130 mm square mesh cod end net since 1992. Haddock are also landed in 4X5Y from the directed Redfish fishery (100–115 mm diamond mesh, DFO 2021b), Silver Hake fishery (55–60 mm, Stone et al. 2013), as well as the Sculpin (90–100 mm mesh) and Winter Flounder (155–165 mm mesh) fisheries (Andrushchenko et al. In press). The 4X5Y Haddock fishing season is regulated by an annual TAC and runs from April 1 to March 31. Catch monitoring requirements for the fishery include logbooks,100% dockside monitoring, vessel monitoring systems (VMS), hail in and hail out requirements, and targeted observer coverage. A regulated spawning closure occurs on Browns Bank annually from February 1st to June 15th. CATCH The fishery catch was estimated by extracting landings data from the COMLAND database (1970–2001) and the MARFIS database (2002–2022). Catches were summarized by NAFO division, quarter (Q), region (BoF and SS), DFO unit area, and fleet (Tables 1–2, Figures 11– 16). The COMLAND and MARFIS databases only include catch data for Canadian fleets. Foreign catches were reported from 1967 to 2002 (Table 1, Figure 11) and were included in the total fishery landings and CAA by using catch multipliers (ratio of the combined Canadian and foreign catch to the Canadian catch) that were applied to the individual Canadian catches. This assumption results in the foreign catches being assigned to fleets proportionally to the estimated Canadian catches by fleet. Catches without coordinates but with an identified DFO unit area were assigned the average latitude and longitude for catches in that DFO unit area for the same year, month, and gear type. Catches assigned to 5Yu (u = unknown) or any other DFO unit area in 5Y were assigned to 5Yb. 4Xu catches (Figure 12) were assigned to a region (BoF or SS) based on past fishing behaviour, using the combinations of factors listed below. Regions were assigned based on combinations 1–3 if catches were from a single region; if not then the region for 4Xu catches were assigned proportional to the catch by region based on the factors in combination 4. 1. Year, Vessel, and Gear 2. Vessel and Gear 3. Year, Port, and Gear 4. Year, Month, and Gear Fleets were initially defined based on two regions (BoF and SS) and four gear categories: i) fixed gear, ii) groundfish (GF) trawl: 120–150 mm mesh size, iii) Redfish (RF) trawl 101– 120 mm mesh size, and iv) other (other mobile gears) (Figure 13). When mesh size was not reported for the trawl gear in COMLAND, the fleet was assigned as follows: 6 1. Based on the data column “MAIN_SPECIES_SOUGHT” for which Cod, Haddock, Pollock, and “unspecified groundfish” were assigned to GF and Redfish was assigned to RF 2. Based on the percentage of the catch as the specified species: >50% catch as Cod, Haddock, Pollock, and “unspecified groundfish” was assigned to GF and >50% catch as Redfish was assigned to RF Although the catch history was initially generated for the 8 fleets (four gear categories defined above and two regions, Figure 13), the catches for the RF fleets and “other” fleets were relatively small and lacked sufficient port sampling data needed to estimate the CAA. The number of fleets was therefore reduced to two gear categories (Fixed and Mobile) and the two regions (BoF and SS) (Figure 14), consistent with the last framework for this stock (Stone and Hansen 2015). SPATIAL AND TEMPORAL TRENDS OF THE CATCH Landings of Haddock were highest in the late 1960s before the implementation of a TAC in 1970 (Table 1, Figure 11). Since the late 1980s, landings have generally been below 10,000 mt and have been consistent over the most recent time period. The landings by area varied over the historical catch time series, with a shift in contribution from 4Xo to 4Xp since the mid-1990s (Figure 12), in particular in survey strata 482 and 483 in the south of 4Xp in some years (Figure 16). This shift results in an increased proportion of the catch coming from the faster growing BoF region in more recent years (Figure 14, Figure 16). In the last decade, the majority of landings were from mobile gear with a shift away from fixed gears (Figure 13, Figure 14, Table 2), and the proportion of annual landings in Q1 has increased (Figure 15). The Q1 landings were primarily from 4Xp and 4Xn (Figure 17 a–f) and the greatest landings were generally observed in February and March. This is consistent with the temporal shift in landings from 4Xo to 4Xp (Figure 12). The spatial distribution of landings has been variable throughout the time series. In the early 2000s, higher landings were observed in the BoF (4Xs and 4Xr), spread throughout 4Xp, the southwestern portion of 4Xn, and in concentrated areas of 4Xq (Figure 17 a–f). Following a decline in 2010, catches increased from 2017–2020 in BoF and 4Xq; however, most catches in 2021–2022 were observed in 4Xp and 4Xn (Figure 12, Figure 17 a–f). CATCH-AT-AGE (CAA) The age composition of Haddock catches are estimated using otoliths collected by port sampling, where a random sub-sample of Haddock are selected and measured to estimate the length frequency distribution of the catch, and otoliths are taken for two fish per 2 cm length bin. The 4X5Y and 5Z Haddock ages were estimated by a new ager beginning in 2021, who replaced the ager that estimated ages from 2016 to 2020. During a quality control exchange with the US, as part of the assessment process for the transboundary 5Z Haddock stock in 2021, it was identified that there was low (59.8%) agreement between ages estimated in quarters 3 and 4 of 2020 by the DFO ager compared to the US ager. Upon inspection of the otoliths with discrepancies between agers, it was determined that the interpretation of otoliths differed between DFO and the US. According to the standard rules for age interpretation of DFO groundfish, a wide or narrow hyaline edge should not be counted as a year of growth in the months from August–January (Table 3). The hyaline rings were incorrectly counted on the edge, leading to the interpretation that two rings were aged as a 2 year old fish (as opposed to the correct age 1) for 5Z Haddock in quarters 3 and 4 in 2020. This finding triggered a re-aging in 2023 of all 4X5Y Haddock collected in August–January from 2016 to 2020. 7 A percent agreement between readers was estimated and the Evans and Hoenig (1998) test for symmetry was conducted between the old and revised ages following the re-aging. The annual percent agreement in 2018 was 86.8% with no significant bias (p = 0.11); however, the percent agreement ranged between 67.2 to 73.9% with significant (p <0.01) bias for 2016–2017 and 2019–2020. This suggested a significant difference between readers and the revised ages for 2016–2020 were used in this document. A comparison between the present ager and the ager before 2016 was conducted to see whether there was a significant bias in quarters 3 and 4 using otoliths from 2014. The percent agreement between the initial ager and the new ager was 89.4% with no significant bias (p = 0.08), so no additional years were considered for re-aging. The fishery CAA was estimated by first estimating the catch-at-length (CAL) using data on the length composition of catches from port samples and then estimating CAA by applying a forward ALK (i.e., distribution of age in each length bin). Length samples from the observer program were not used to estimate LF distributions. The fishery CAA has traditionally been estimated using DFO’s CAA application (e.g., Stone and Hansen 2015). The CAA was previously calculated annually with some undocumented decisions to fill gaps in sampling (e.g., missing LF samples or missing ages in ALKs) making reproducibility of the CAA difficult. Here we apply a structured algorithm for estimating the CAA following a similar approach to that used for 3Pn4RS Atlantic Cod (Ouellette-Plante et al. 2022) with the objective of documenting the assumptions made for filling gaps in sampling and allowing for reproducibility of the calculations used to estimate the CAA. The algorithm is outlined below. Catch at Length The CAL was estimated by assigning a representative LF distribution to each individual reported catch. The CAL was generated using 2 cm length bins to be consistent with the length bins used for age sampling. At least five unique LF samples were used to represent an individual reported catch, with equal weight put on each LF sample. The estimated weights-at-length for each LF distribution were estimated using the DFO ecosystem survey weight-length relationships (Table 4, Figure 18). The representative LF samples were identified for each individual catch record by sequentially going through the following list of factors until at least five unique LF samples were identified: 1. Year, Quarter, Fleet, DFO unit area 2. Year, Quarter, Fleet, Region 3. Year, ± 1 Quarter, Fleet, DFO unit area 4. Year, ± 1 Quarter, Fleet, Region 5. ± 1 Year, Quarter, Fleet, DFO unit area 6. ± 1 Year, Quarter, Fleet, Region 7. ± 1 Year, ± 1 Quarter, Fleet, Region 8. Year, Quarter, Fleet 9. Year, ± 1 Quarter, Fleet 10. ± 1 Year, ± 1 Quarter, Fleet 11. Year, Fleet, Region 12. Year, Fleet 13. ± 1 Year, Fleet, Region 8 14. ± 1 Year, Fleet This approach led to identifying at least five unique LF samples for each catch record, with the exception of some catches in 1970, for which only one unique LF sample was identified. The single LF sample was used for these catches in 1970. The 4Xp DFO unit area was divided into two areas (BoF and SS; see Figure 1). Coordinates were not available for some port samples in 4Xp (e.g., all years prior to 1991). All LF samples and landings in 4Xp in these years were assumed to be from the SS region for the estimation of the CAL. This is consistent with the assumption used in the last framework (Stone and Hansen 2015). The number of LF samples by DFO unit area and quarter is plotted in Figure 19. The CAL was over a broader size range in the 1970s and 1980s, with a declining and narrowing of length of the catch since the mid 1990s (Figure 20). Age Length Key Forward ALKs were generated using 2 cm groupings by year, quarter, and region (SS and BoF) that estimate the proportion of fish at age for a given length using ages collected from the port and observer sampling programs. Missing ages for length bins that were observed in the fishery (from port samples) were filled as follows: 1. ± 1 Length, Quarter, Year, Region 2. Length, ± 1 Quarter within a year, Year, Region 3. Length, Quarter, ± 1 Year, Region 4. ± 1 Length, Quarter, Year, Region [repeated after steps 1–3] 5. Length, Quarter, ± 2 Years, Region When there were no ages for a year, quarter, and region to generate an ALK, an ALK from an adjacent quarter was used as a fill (consistent with number 2 above) and then an adjacent year was used as a fill (consistent with number 3 above) when required. The final steps in filling gaps in the ALK were assigning an age of 12+ to any length bin greater than or equal to 77 cm and manually filling 15 gaps for older fish (Age 12+) and younger fish (Ages 0–3) at the beginning and end of the length distributions. Catch at Age The CAA was estimated from the combination of the CAL and ALK (defined separately by region) (Figure 21, Figure 22, Table 5, Table 6). In general, the CAA estimated here was similar to that estimated for 1970–2013 by Stone and Hansen (2015) (Figure 23). Catch Uncertainty in 4Xp The CAA was estimated for three catch scenarios to be compared in sensitivity analyses when models are fit: 1. The status quo catch area: all catches in 4X5Y 2. Catches (as well as length frequency and age samples) in survey strata 483 and 5Z9 are excluded 3. Catches (as well as length frequency and age samples) in survey strata 482, 483, and 5Z9 are excluded 9 The spatial bounds for these catch scenarios were defined based on the similarity in growth (relative LAA) of Haddock and LF distributions of Haddock catches from port samples in strata 482, 483, and 5Z9 to Eastern Georges Bank (EGB). Catches in the south of 4Xp were previously hypothesized to include Haddock from EGB (Stone and Hanson 2015). In the last assessment framework, an alternative catch scenario was examined in a sensitivity analysis that excluded catches within five nautical miles of the 4X5Z boundary line based on a hypothesis that when there were strong EGB Haddock year-classes (e.g., 2000 and 2003), EGB Haddock extend into the Fundian Channel (Stone and Hanson 2015). Data from the commercial fishery and the surveys were examined to evaluate this hypothesis by exploring four different data sources: 1. Spatial distribution of fishery catch vs. survey biomass 2. Fishery CAA vs. survey CAA (cohort strengths) 3. LF distributions 4. Growth (Length-at-age) Spatial Distribution of Fishery Catch vs. Survey Biomass There has been an increase in the proportion of total stock landings from 4Xp beginning around the year 2000, and this increase coincides with a decrease in the proportion of landings from 4Xo (Figure 12). The proportion of stock landings in survey strata 482 and 483 (including 5Z9) exceed 25% in the late 2000s (Figure 16) where significant landings were observed just north of the 4X and 5Z border (Figure 17 a–f). While relative catches in strata 482 and 483 were on average 25% in the 2000s, the mean proportion of survey biomass in these strata was only 7% in the 2000s (Figure 24). Fishery CAA vs. Survey CAA The largest cohorts contributing to the 4X5Y fishery CAA in the last two decades are the 1998, 2003, 2010, and 2013 cohorts (Figure 21). While the 2013 cohort is the largest in the survey CAA and the 1998 cohort is the second largest cohort in the last two decades, the 2003 and 2010 cohorts are approximately average in size (Figure 25). The 2003 and 2010 cohorts are ;however, large cohorts on EGB (Figure 26), suggesting that these cohorts may be contributing to the 4X5Y catch. Classification based on size (LF) distribution of catch Differences in growth have been identified between BoF, SS, and EGB (see Review of Differences in Growth by DFO Statistical Unit Area section). The southern portion of 4Xp is the spatial area where these three regions converge and was divided into smaller areas based on survey strata (strata 480–483, 5Z9). Cumulative LF distribution functions (CDFs) from port samples were used to estimate the probability of being drawn from each (statistical) population or group (BoF, SS, EGB), for each survey stratum. Given a single observed CDF from a survey stratum and a single CDF from each group, the predicted CDF for the survey stratum was estimated as the proportions (�̂�𝑝𝐵𝐵𝐵𝐵𝐵𝐵, �̂�𝑝𝑆𝑆𝑆𝑆, �̂�𝑝𝐸𝐸𝐸𝐸𝐵𝐵) of each of the groups that minimize the differences in the squared cumulative proportions at length (nearest cm) between the observed and predicted CDFs. Predicted probabilities of belonging to each group were estimated by year, quarter, and fleet by averaging probabilities across 1,000 simulations, where a single simulation involved randomly selecting a single CDF for the survey stratum and each group (within the year, quarter, and fleet). 10 Probabilities were estimated for survey strata when at least one CDF was available. When a CDF was not available for each group for a given year, quarter, fleet, a set of CDFs was substituted as follows until at least one CDF was identified: 1. ± 1 Quarter 2. ± 1 Year (same Quarter) 3. ± 1 Quarter and ± 1 Year 4. Any Quarter within Year 5. Any Quarter ± 1 Year 6. Any Quarter ± 2 Year Probabilities of belonging to each group were plotted by quarter and stratum, and a loess smoother (span = 0.75) was used to visualise the general patterns in the probability over time (Figure 27). Looking at trends from the loess smoothers, survey strata 480 and 481 generally had the highest probability of belonging to the SS group, with the exception of stratum 481 in quarter 3, where the predicted probabilities were similar among the three groups (Figure 27). Stratum 482 generally had low predicted probabilities of belonging to SS, and generally a higher probability for EGB at the beginning of the time series and then higher probability for BoF at the end of the time series (Figure 27). Survey strata 483 and 5Z9 generally had the highest probability of belonging to the EGB group (Figure 27). Classification based on growth of catch using empirical LAA The LAA from port and observer samples for each survey stratum (strata 480–483, 5Z9) were used to estimate the probability that the sample belongs to each (statistical) population or group (i.e., BoF, SS, or EGB) where the LAA “populations” for each group were defined based on LAA data from the DFO summer ecosystem survey (BoF and SS) and the DFO summer and winter ecosystem surveys (EGB). The BoF group was defined as survey strata 484 to 495, SS was strata 470 to 477, and EGB was strata 5Z1 and 5Z2. Only LAA data for ages 4 and older were used to estimate the probabilities to 1) reduce the bias in LAA due to fishery selectivity for younger fish and, 2) reduce the influence of growth within a year on the LAA for younger fish. Given a single LAA port or observer sample of fish (mean 𝑛𝑛 = 20 fish per sample) from a trip (port) or set (observer), the predicted probability of each individual fish (𝑖𝑖) belonging to each group (𝑔𝑔) was estimated by calculating the likelihood that the sample was drawn from each population (group) distribution of LAA. The likelihood (𝐿𝐿) that an individual fish with length (𝑙𝑙𝑖𝑖) from sample 𝑗𝑗 with age (𝑎𝑎) in year (𝑦𝑦) was drawn from group (𝑔𝑔) was defined as: 𝐿𝐿𝑎𝑎,𝑦𝑦,𝑔𝑔,𝑎𝑎,𝑖𝑖 �𝑙𝑙 = 𝑙𝑙𝑖𝑖|𝑁𝑁�𝜇𝜇𝑎𝑎,𝑦𝑦,𝑔𝑔,𝜎𝜎𝑎𝑎,𝑦𝑦,𝑔𝑔 2 �� Eqn 2 where 𝑁𝑁(𝜇𝜇,𝜎𝜎2) is a normal distribution with mean and variance defined as the mean and variance of the DFO ecosystem survey lengths-at-age 𝑎𝑎 in year 𝑦𝑦 for group 𝑔𝑔. The probability of each individual fish (𝑖𝑖) belonging to each of the 3 groups (�̂�𝑝𝑖𝑖,𝑎𝑎∊𝑔𝑔) was defined as: �̂�𝑝𝑖𝑖,𝑎𝑎∊𝑔𝑔 = 𝐿𝐿𝑎𝑎,𝑦𝑦,𝑔𝑔,𝑎𝑎,𝑖𝑖�𝑙𝑙=𝑙𝑙𝑖𝑖|𝑁𝑁�𝜇𝜇𝑎𝑎,𝑦𝑦,𝑔𝑔,𝜎𝜎𝑎𝑎,𝑦𝑦,𝑔𝑔 2 �� ∑ �𝐿𝐿𝑎𝑎,𝑦𝑦,𝑔𝑔,𝑎𝑎,𝑖𝑖�𝑙𝑙=𝑙𝑙𝑖𝑖|𝑁𝑁�𝜇𝜇𝑎𝑎,𝑦𝑦,𝑔𝑔,𝜎𝜎𝑎𝑎,𝑦𝑦,𝑔𝑔 2 ���𝑔𝑔 Eqn 3 The overall probability of a sample (j) belonging to a group (𝑔𝑔) was defined as the mean probability of each individual fish in the sample belonging to that group: �̂�𝑝𝑎𝑎∊𝑔𝑔 = ∑ �𝑝𝑝�𝑖𝑖,𝑎𝑎∊𝑔𝑔�𝑛𝑛 𝑖𝑖=1 𝑖𝑖 Eqn 4 where 𝑛𝑛 is the number of individual LAA observations (𝑖𝑖) in sample 𝑗𝑗. 11 The results were displayed as the mean probability of belonging to each population by quarter (weighted by sample size 𝑛𝑛). A loess smoother (span = 0.75) was used to show the temporal trends in probability across groups for each stratum (Figure 28). The variability in the predicted probabilities based on LAA (Figure 28) were much lower than for the LF distributions (Figure 27). Looking at trends from the loess smoothers, survey strata 480 and 481 generally had a high overlap of predicted probabilities across the time series (Figure 28). Stratum 482 generally had similar probabilities for BoF and EGB (higher than SS), except for quarter 4 in the 2000s where EGB had a higher probability that exceeded 50% in some years (Figure 28). Survey strata 483 and 5Z9 generally had similar trends over time with a higher probability of belonging to EGB at the beginning of the time period, similar probability for EGB and BoF after 2010, and lower probability for SS (but increasing over time) (Figure 28). The apparent decrease in the probability of belonging to the EGB group for 5Z9 (a stratum in EGB) appears to be caused by the similarity in growth between EGB and BoF in the later years (e.g., Figure 29) and not that Haddock in 5Z9 are believed to be from the BoF region. Classification based on growth of catch using von Bertalanffy estimated LAA by cohort The overall probability of a sample (j) of LAA belonging to a group (𝑔𝑔) as described in 4a (above) was also estimated using a mean and variance for the populations (groups BoF, SS, and EBG) estimated from two parameter vonB growth models fit to length-at-adjusted age (see Review of Differences in Growth by DFO Statistical Unit Area section) by group and cohort. The mean and variance in Equation 2 were defined in this case to be the vonB model predicted mean length at the adjusted age of the individual fish and the variance of the residuals from the vonB model, respectively. The predicted probabilities based on predicted LAA using the vonB models (Figure 30) were very similar to those for LAA using the empirical mean LAA (Figure 28). Catch-at-Age for the Alternative Catch Scenarios Excluding catches and age composition data in survey strata 483 and 5Z9, in general had little influence on the CAA estimation, with the exception of a reduction of the strength of the 2000 and 2003 cohorts with the numbers-at-age 4 in 2007 being reduced the most (Figure 31). The strength of the 2000 and 2003 cohorts are further reduced in the CAA when the catches and data in stratum 482 are also excluded (Figure 32) where the size of the 2003 cohort became closer to an average sized cohort in the time series. FISHERY LENGTH-AT-AGE (LAA) A fishery LAA matrix was estimated by year and region using LAA data collected from port and observer samples. Lengths were adjusted to reflect a mid-year length to account for growth within the year. This was done by estimating the growth from the month the fish was caught to a month value of 6 (i.e., July) from a three-parameter vonB growth model fit by region and cohort (for cohorts 1966–2016) using an adjusted age that incorporated month (see Review of Differences in Growth by DFO Statistical Unit Area section). The vonB model from the nearest cohort was used to estimate the incremental growth for cohorts outside the range of 1966–2016. LAA was then estimated as the mean adjusted (July) LAA. Missing LAA values were filled as: 1. LAA-1 was filled when LAA-2 for that cohort was available using the mean rate of growth from age-1 to age-2 (i.e., LAA-2/LAA-1) from the cohorts above and below that cohort. 2. LAA-1 was filled when LAA-2 for that cohort was available using the mean rate of growth from age-1 to age-2 from the three cohorts above and below that cohort. 12 3. LAA was filled using the mean LAA from years above and below. 4. LAA was filled as a linear interpolation of log-transformed length over one age along a cohort. 5. LAA was filled as a linear interpolation of log-transformed length over two ages along a cohort (i.e., LAA[i,j] and LAA[i+1,j+1] are filled using LAA[i-1,j-1] and LAA[i+2,j+2] where i is year and j is age). 6. LAA-11 and LAA-12+ were filled using the maximum LAA in that cohort. The average fishery LAA has declined for older ages of Haddock for both regions throughout the time series (Figure 33). A final LAA matrix for the stock (combined BoF and SS) was estimated as a mean LAA, weighted by the CAA for BoF and SS (Table 7). FISHERY WEIGHT-AT-AGE (WAA) A fishery WAA matrix was estimated by year and region by converting the mid-year LAA matrix to WAA using the weight-length relationship from the survey (Table 4). The average fishery WAA has declined for older ages of Haddock for both regions throughout the time series (Figure 33). A final WAA matrix for the stock (combined BoF and SS) was estimated as a mean WAA, weighted by the CAA for BoF and SS (Table 8). SURVEYS A mobile gear fixed station survey in NAFO division 4X was conducted by the ITQ mobile gear <65 ft fleet from 1996 to 2012. The survey covered a broader area (including nearshore areas) than the DFO summer ecosystem survey (see Stone and Hansen 2015) and was conducted in July using a standardized Balloon 300 trawl equipped with a cod end liner of the same mesh size as the DFO survey. The ITQ survey was discontinued in 2013 and the index was not estimated for 2011 and 2012 (Stone and Hansen 2015). DFO has conducted a stratified random bottom trawl survey of the BoF and SS every summer since 1970 using seven research vessels: the A.T. Cameron from 1970–1981, the Lady Hammond in 1982, the CCGS Alfred Needler from 1983–2003, 2005–2006, 2009–2015, 2017, and 2019, the CCGS Teleost in 2004, 2007, 2016, 2018, 2020, and 2022, the CCGS Templeman in 2008, the CCGS Cartier in 2021, and the CCGS Cabot in 2022. Based on an analysis of comparative fishing experiments by Fanning (1985), a conversion factor of 1.2 for Haddock has been applied to the total abundance, total biomass and age-specific abundance series prior to 1982 (i.e., for 1970–1981) to account for the effect of vessel and gear changes (Yankee 36 to Western IIA bottom trawl) between the A.T. Cameron and the Hammond/Needler (Note: this is not a length-based conversion). A more recent analysis of comparative fishing experiments between the Alfred Needler and the Teleost showed that no conversion factor was required for 4X5Y Haddock (Fowler and Showell 2009). There are currently no conversion factors established for either the Cartier or Cabot between the Needler/Teleost so the data from these vessels are currently excluded from this document but will be integrated into the modelling framework when conversion factors become available in 2024. The average number of tows per year per strata for the DFO summer ecosystem survey over the last two decades has been 3.7 for the BoF strata and 4.0 for the SS strata (Figure 34). INDIVIDUAL TRANSFERABLE QUOTA SURVEY The ITQ survey biomass index and the estimated numbers-at-age for the survey from 1996– 2010 is provided in Table 9 and Figure 35, and is unchanged from Stone and Hansen (2015). A 13 comparison between the biomass index and the relative numbers-at-age estimated from the ITQ index and the DFO summer ecosystem survey are provided in Figures 36 and 37. The ITQ survey shows a larger decline in biomass after the year 2001 compared to the DFO survey and higher proportions of fish at age 1 and 2, suggesting the ITQ survey has higher selectivity of smaller fish. DFO SURVEY INDEX An index of stock biomass was estimated as the mean biomass per standardized tow, defined as a 1.75 nautical mile (nm) tow. Using a stratified random design, the annual mean biomass per tow was estimated as a weighted mean with weights (𝑤𝑤) proportional to the strata area divided by the number of tows in that strata (𝑛𝑛) and the weighted standard error of the mean was estimated as (Kish 1992): 𝑆𝑆𝑆𝑆 = �𝑠𝑠2 𝑖𝑖 (∑𝑤𝑤2)/𝑖𝑖 (∑𝑤𝑤/𝑖𝑖)2 Eqn 5 where 𝑠𝑠2 is the unweighted sample variance. The mean biomass per standardized tow differed by region, with generally higher density of Haddock for SS compared to the BoF (Figure 38). The distribution of biomass per standardized tow was explored by plotting the residuals of a linear model with a response variable of biomass/tow and categorical factors year and strata (Figure 39a). The distribution appeared skewed to the right so the residuals from a model with ln-transformation of biomass (removing zeros) were plotted and appeared bell-shaped (Figure 39b). The index was therefore also estimated assuming a delta-lognormal distribution where the mean and SE were estimated following Pennington (1996). Although this method can provide less biased estimates of the mean when there are extreme observations (e.g., a large biomass estimate from a single tow), it is not robust to small departures from the assumed lognormal distribution of positive tows (Syrjala 2000). Small positive values (tows with biomass per standardized tow of less than 0.5 kg) were therefore replaced with zero following the suggestion of Pennington (1991). The two survey biomass indices were similar (Figure 40) with the main differences being the lower biomass in 1977 for the delta distribution (smaller influence of a single large tow in 1977), and more stable coefficient of variation over time (Figure 41). The Gini index (Gini 1921) was calculated in each year as an indicator of the relative distribution of survey biomass among survey stations (Figure 42, index based on arithmetic mean only). The Gini index is commonly used as a summary of income inequality and is used here as a statistic to summarize the dispersion of biomass across tows. A value of zero reflects equal biomass at each survey station and a value of one reflects a single station with all the biomass. Over the last decade there has been a decline in the index from a time series maximum of 0.86 in 2009 to about 0.6 from 2016–2020. This could be related to the strength of the 2013 cohort which also has resulted in a low percentage (approximately 10%) of tows with zero biomass between 2014 and 2020 (Figure 43). DFO SURVEY NUMBERS-AT-AGE AND LENGTH The survey numbers-at-length (NAL) were estimated using the LF distribution for each tow using 2 cm length bins. When body weights were not available for a length bin, they were estimated using the survey weight-length relationships by region (BoF and SS) (Table 4). Forward ALKs were generated by year and region (BoF: survey strata 482 to 495 and SS: survey strata 470 to 471). Missing ages for lengths that were observed in the survey LF samples were filled as follows: 1. ± 1 Length bin, Year, Region 14 2. Fish <12 cm are age 0 3. ± 1 Year, Region The final steps in filling gaps in the ALK were assigning an age of 12+ to older fish in length bins >77 cm and manually filling two gaps. The overall survey NAL were higher for SS compared to BoF and a truncation of the length distribution was observed over time for both regions (Figure 44). The survey numbers-at-age (NAA) were estimated by applying the ALKs to the NAL to obtain the NAA separate by region which was summed to obtain the stock NAA (Figure 25). Based on the survey NAA, only the 2013 cohort made a substantial contribution to the survey catch (Figure 25) with the largest estimated recruitment based on NAA-1 in 2014 (Figure 45). DFO SURVEY LENGTH-AT-AGE AND WEIGHT-AT-AGE The survey LAA was estimated by region as the mean LAA from all fish caught in the survey sampling. Missing LAA values were filled as follows: 1. LAA-0 and LAA-1 were filled by taking the mean LAA from years above and below. 2. LAA-0 at the beginning of the time series was filled using the mean LAA-0 from the first 5 years with data. 3. LAA was filled using model estimates from a three-parameter von Bertalanffy growth model of mean LAA by cohort for cohorts 1966 to 2016. 4. LAA was filled as a linear interpolation of log-transformed length over one age along a cohort. 5. LAA was filled using the rate of growth from the previous cohort (LAA[i-1,j]/LAA[i-2,j-1]) and multiplying by the LAA for the previous age in that cohort (i.e., LAA[i-1,j-1]) where i is year and j is age. 6. LAA-12 in 1970 for BoF was filled as the mean LAA-12 from the next 5 years. The survey WAA was estimated by region as the mean WAA from all fish caught in the survey sampling. Missing WAA values were filled as follows: 1. WAA-0 and WAA-1 was filled by taking the mean WAA from years above and below. 2. WAA-0 at the beginning of the time series was filled using the mean WAA-0 from the first 5 years with data. 3. WAA was filled as a linear interpolation of log-transformed length over one age along a cohort. 4. WAA was filled as a linear interpolation of log-transformed length over two ages along a cohort (i.e., WAA[i,j] and WAA[i+1,j+1] are filled using WAA[i-1,j-1] and WAA[i+2,j+2] where i is year and j is age). 5. WAA was filled by taking the mean WAA from years above and below. 6. WAA was filled using the rate of growth from the previous cohort (WAA[i-1,j]/WAA[i-2,j-1]) and multiplying by the WAA for the previous age in that cohort (i.e., WAA[i-1,j-1]) where i is year and j is age. The final survey LAA and WAA matrices for the stock (combined BoF and SS) were estimated as a mean, weighted by the survey NAA for BoF and SS (Table 10, Table 11). Both survey mean LAA and WAA show an overall decline in older ages (4+) throughout the time series, with minor improvements over the last 3 years in the Bay of Fundy (Figure 46). The LAA by cohort shows this decline with the length of Haddock much smaller in the 2000s compared to the 15 1960s–1990s (Figure 47). WAA matrices were estimated to represent January 1st stock WAA and SSB (April 1st) WAA by adjusting the survey WAA from month 7 to month 1 and 4, respectively using the Rivard (1982) method which uses a log-linear interpolation between a WAA 𝑎𝑎 in year 𝑦𝑦 and the WAA 𝑎𝑎 − 1 in year 𝑦𝑦 − 1. DFO SURVEY MATURITY Maturity data were only collected on the DFO summer ecosystem survey from 1970–1985 and then sporadically from the summer and winter surveys afterwards. Sufficient maturity data (n >20 observations by year and region) were available from the survey from 1970–1985 and 1988, 1993, 1994, 2016, 2019, and 2020 (Figure 48). Data from the NMFS surveys (Spring and Fall for strata 29, 30, 34, 35, 36) were explored as an additional data source for Haddock maturity; however, they were not incorporated based on similar gaps in the time series for the NMFS spring surveys and high variability in the estimation of maturity from the NMFS fall surveys likely due to survey timing. Maturity data were available from EGB (strata 5Z1, 5Z2, 5Z9) from 1987–2021 and were used to predict the length-at-maturity and age-at-maturity in years with no data in the stock area. The length and age at 50% and 90% maturity (𝐿𝐿50, 𝐿𝐿90, 𝐴𝐴50, 𝐴𝐴90; hereafter, maturity statistics) were estimated by year and region using binomial logistic regression models (Figure 48a, Figure 48b). The values of these maturity statistics for BoF and SS for years with missing data were estimated from the predicted values from a linear model with the predictors year (categorical) and region (Figure 48c, Figure 48d). This effectively estimated the mean difference in each maturity statistic between regions in years when data were available, and this difference was used to predict the maturity statistics for BoF and SS from the EGB values. The maturity-at-age data will be used as a model input to estimate spawning stock biomass from total stock biomass. The focus for the data inputs is therefore on estimating a maturity-at-age matrix. Sharp changes with magnitude of approximately one year in the age-at-maturity in EGB from 2004–2005 and from 2009–2010 could be related to strong 2000 and 2003 cohorts observed on EGB (Figure 26), which were not observed to be as strong in 4X5Y. An alternative method to estimate age-at-maturity was also explored that was not dependent on EGB. The 𝐴𝐴50 and 𝐴𝐴90 from 1986–present were filled as the mean values from 1986–present (Figure 48e) and these were used to generate the maturity-at-age matrices from the logistic regression equation: 𝑃𝑃(𝑥𝑥) = 1 1+𝑒𝑒−(𝑏𝑏0+𝑏𝑏1𝑥𝑥) Eqn 6 where the regression coefficients were defined from the predicted maturity statistics as: 𝑏𝑏1 = ln(9) /(𝐴𝐴90 − 𝐴𝐴50) and 𝑏𝑏0 = ln(9) − 𝑏𝑏1 𝐴𝐴90. A final maturity-at-age matrix for the stock (combined BoF and SS) was estimated as a mean, weighted by the survey NAA for BoF and SS (Table 12). DFO SURVEY DISTRIBUTION, CONDITION, SURVEY Z, AND RELATIVE F The spatial distribution of survey catches has been consistent throughout the time series with large tows of Haddock more commonly observed on Browns Bank, Roseway Bank, Baccaro Bank, and areas of the BoF (Figure 49 a–d). Survey catches remain much lower in 4X5Y in the summer compared to winter catches on EGB. Fulton’s condition factor (K) was estimated for each region using a ratio of length (L) and weight (W) as K =100*W/L3 from the DFO summer ecosystem survey data. In the BoF, mean annual condition of Haddock declined for both sexes until 2004, then fluctuated below the time series mean until a decline for females in 2010, and was followed by an increase in condition in recent 16 years for both sexes (Figure 50). On the SS, mean annual condition of Haddock has fluctuated around the time series mean until 2010, and has remained below the mean in the most recent time period (Figure 50). Relative total mortality (Z) was estimated using the DFO summer ecosystem survey NAA of fully recruited age groups (ages 3–8), and relative fishing mortality (relF) was estimated as the ratio of fishery catch to survey biomass to explore potential changes in natural mortality. For both BoF and SS, relative total mortality remained consistent over time, while relative fishing mortality declined (Figure 51, Figure 52). This decline was much more pronounced for the SS, suggesting a potential increase in natural mortality since 2000 (Figure 52). ECOSYSTEM CONSIDERATIONS Haddock adjust their depth distribution based on changing water temperatures throughout the year. They typically inhabit inshore waters but may overwinter in deeper waters and then move into shallower areas as temperatures rise in the summer months (Scott and Scott 1988, Rogers et al. 2016, Perry and Smith 1994). This behaviour may occur more readily in cooler waters associated with the SS and off Newfoundland compared to more temperate waters (Murawski and Finn 1988). Optimal water temperatures for adult Haddock range from 4–7 ˚C, with all life history stages of Haddock typically avoiding waters with temperatures above 10 ˚C (Bigelow and Schroeder 1953, Cargnelli et al. 1999). Both the magnitude and the timing of algal blooms may impact the recruitment of Haddock (Friedland 2021, Platt et al. 2003). On Georges Bank, the fall phytoplankton bloom is hypothesized to provide energy to pre-spawning adult Haddock and years with higher algal blooms have been associated with recruitment of exceptional year classes (Friedland 2021). On the SS, the survival of Haddock larvae is dependent on the timing of the spring phytoplankton bloom, and when spawning time corresponds with algal blooms, higher survival may occur allowing for a more abundant food source (Platt et al. 2003). Thus a reduction in algal blooms may ultimately impact both reproductive success and recruitment. DFO’s Atlantic Zone Monitoring Program (AZMP) program was implemented in 1998 to collect and analyze biological, chemical and physical oceanographic field data. Data from this program are summarized and made available for use in tables from the azmpdata package in R (https://github.com/casaultb/azmpdata 2022). Sampling stations in the stock area include stations along a transect through Browns Bank and fixed stations in the BoF and near the 4X/4W border. Additional data are available in the azmpdata package and include the North Atlantic Oscillation (NAO) index, temperature, chlorophyll-a concentrations, and zooplankton abundance. As larvae, Haddock consume plankton, and transition to a diet mostly of benthic invertebrates and fish as juveniles and adults (Kane 1984, Mahon and Neilson 1987, Brodziak 2005). Stomachs of commercially important fish species are collected annually on the DFO summer ecosystem survey. Haddock stomachs have been routinely sampled since 2007; however, stomach content analyses are currently available only up until 2017. From 1997–2017, 3,789 Haddock stomachs were analyzed, and the majority of the stomach contents contained crustaceans (e.g., shrimp, amphipods), echinoderms (e.g., brittle stars), marine worms (e.g., bristle worms) and bivalves (e.g., cockles, clams). The only abundant fish found in the stomachs of Haddock were sand lance (Ammodytes dubius). In general, adult Haddock are consuming more fish, crustaceans, and echinoderms than smaller sized (<38 cm) Haddock. Stomachs that contained Haddock as a prey item occurred in the stomachs (n=30) of 11 species of fish from 1995–2020. Greater occurrences of Haddock in predator stomachs 17 occurred earlier in the time series; however, this observation may be based on delayed processing of samples as opposed to a reduction in predation. Based on the limited data available, Pollock (Pollachius virens), followed by Atlantic Cod (Gadus morhua) had the highest occurrences of Haddock in their stomachs. CONCLUSIONS There is no evidence from the recent literature that would support a change in stock structure for 4X5Y Haddock. The basis for the separation of the fishery and survey data inputs by region (BoF: faster growing vs. SS: slower growing) was reviewed in this document. The vonB growth parameters estimated using data from the DFO summer ecosystem survey for strata 480 and 481 were more similar to 4Xno so these survey strata were included in the SS region and the vonB growth parameters for survey strata 482 and 483 were more similar to 4Xqr so these survey strata were included in the BoF region. This spatial definition also formed the basis for defining regions for the catch history (revised from the last framework which considered all catches in 4Xp as SS). Landings have been consistent over the last 30 years; however, over the last two decades the majority of catches has been with mobile gear, temporally the fishery has shifted to relatively more landings in the winter months, and the spatial distribution of the catch has shifted from 4Xo to southern 4Xp which borders the EGB stock. The methods to estimate fishery CAA for 4X5Y Haddock have been updated from the last assessment framework using a structured algorithm to allow for reproducibility of results. This resulted in only minor changes to the historical CAA. The stock origin of catches in the south of 4Xp remains a large uncertainty. Analyses were conducted in the document to evaluate the hypothesis that catches in the south of 4Xp are of EGB origin. Multiple lines of evidence are consistent with this hypothesis: • Catches in survey strata 482 and 483 are large in some years (approximately 25% of the stock catch) but survey biomass estimates in these strata are low (approximately 7% of stock biomass). • The 2003 and 2010 cohorts are strong in the 4X5Y fishery CAA and EGB survey NAA but not the survey NAA for 4X5Y. • The LF distribution of the catches in survey strata 482 and 483 are generally more similar to EGB than BoF and SS, except for some recent years where there is overlap between BoF and EGB. • The LAA of the catches in survey strata 482 and 483 are similar to EGB although not different from BoF in recent years. The CAA was estimated by excluding landings and length/age composition data from a) survey strata 483 and 5Z9 and b) survey strata 482, 483, and 5Z9. The influence of removing these catches was a reduction of the strength of the 2000 and 2003 cohorts such that the 2003 cohort became closer to an average strength cohort for catch scenario that excluded survey strata 482, 483, and 5Z9. The three different catch scenarios will be considered in the modelling framework. Fishery and survey WAA and LAA show an overall decline in older ages (4+) throughout the time series, with minor improvements observed from data over the last 3 years in BoF. Two methods for calculating the DFO summer ecosystem survey index were explored; 1) the status quo (annual mean biomass per tow as a weighted mean proportional to size of stratum and number of tows) and, 2) a weighted mean assuming a delta lognormal distribution. Although the 18 indices estimated using the two different methods were similar, the delta lognormal index reduced the influence of extreme tows and had a more stable coefficient of variation over time. Both indices will be considered in the modelling framework. Estimates of relative fishing mortality indicate a decline on the SS since 2000 indicating a potential change in natural mortality. Some environmental data sets have been identified from DFO’s AZMP program and will be used in the modelling framework to explore relationships with model parameters such as recruitment and growth. ACKNOWLEDGEMENTS D. Frotten and D. D’Entremont of DFO and at-sea observers from Javitech Ltd. and Atlantic Catch Data provided the port and observer samples from the Canadian fishery. K. Kraska provided age estimates for the DFO ecosystem surveys and commercial fishery and S. Sutherland provided age estimates for the NMFS surveys. I. Andrushchenko, J. McIntyre, and E. Brunsdon provided background information, edits to the document, and code review. The section on the review of stock structure was supported by an unpublished literature review by G. Puncher. M. Cassista-Da Ros provided data from the DFO survey stomach collection database. Members of the Scotia-Fundy Groundfish Advisory Committee provided background on the fishery and contributed to the conceptualization of the 4Xp catch uncertainty analysis. Comments during the peer-review meeting by B. Hubley, C. Perretti, and members of the Scotia-Fundy Groundfish Advisory Committee improved the working paper. REFERENCES CITED Andrushchenko, I., Martin, R., Doherty, P., Debertin, A., McCurdy, Q., MacEachern, E., Clark, D., and Clark, C. In press. Western Component Pollock - Data Inputs. DFO Can. Sci. Advis. Sec. Res. Doc. Begg, G.A. 1998. A review of stock identification of Haddock, Melanogrammus aeglefinus, in the northwest Atlantic Ocean. Mar. Fish. Rev. 60(4): 1–15. Begg, G.A., and Weidman, C.R. 2001. Stable d13C and d18O isotopes in otoliths of Haddock Melanogrammus aeglefinus from the northwest Atlantic Ocean. Mar. Ecol. Prog. Ser. 216: 223–233. Berg, P.R., Jorde, P.E., Glover, K.A., Dahle, G., Taggart, J.B., Korsbrekke, K., Dingsør, G.E., Skjæraasen, J.E., Wright, P.J., Cadrin, S.X., Knutsen, H., and Westgaard, J. 2021. Genetic structuring in Atlantic Haddock contrasts with current management regimes, ICES J. Mar. Sci. 78(1): 1–13. Bigelow, H.B., and Schroeder, W.C. 1953. Fishes of the Gulf of Maine. U.S. Fish Wildl. Serv., Fish. Bull. 53: 577 p. Brickman, D. 2003. Controls on the distribution of Browns Bank juvenile Haddock. Mar. Ecol. Prog. Ser. 263: 235–246. Brodziak, J.K.T. 2005. Haddock, Melanogrammus aeglefinus, life history and habitat characteristics, Second Edition. NOAA Tech. Memo. NMFS-NE-196, 64 p. Brodziak, J.K.T., and Col, L. 2006. Northeast Consortium Cooperative Haddock tagging project: summary of reported Haddock tag recaptures through November, 2006. 19 Brodziak, J.K.T., Col, L., Palmer, M., and Brooks, L. 2008. Northeast Consortium Cooperative Haddock tagging project: summary of reported Haddock tag recaptures through November, 2008. Campana, S. E., Smith, S.J., and Hurley, P.C.F. 1989. A drift-retention dichotomy for larval Haddock (Melanogrammus aeglefinus) spawned on Browns Bank. Can. J. Fish. Aquat. Sci. 46: 93–102. Cargnelli, L.M., Griesbach, S.J., Berrien, P.L., Morse, W.W., and Johnson, D.L. 1999. Haddock, (Melanogrammus aeglefinus) life history and habitat characteristics. NOAA Tech. Memo. NMFS-NE-128, 31 p. DFO. 2018. 4VWX5 groundfish integrated fisheries management plan, Maritimes Region. Fisheries and Oceans Canada. DFO. 2020. Stock Status Update of Haddock (Melanogrammus aeglefinus) in NAFO Divisions 4X5Y. DFO Can. Sci. Advis. Sec. Sci. Resp. 2020/021. DFO. 2021a. Stock Status Update of Haddock (Melanogrammus aeglefinus) in NAFO Divisions 4X5Y for 2020. DFO Can. Sci. Advis. Sec. Sci. Resp. 2021/021. DFO. 2021b. Stock Status Update of Unit 3 Redfish for 2020. DFO Can. Sci. Advis. Sec. Sci. Resp. 2021/026. Evans, G.T., and Hoenig, J.M. 1998. Testing and viewing symmetry in contingency tables, with application to readers of fish ages. Biometrics. 54: 620–629. Fanning, L.P. 1985. Intercalibration of Research Survey Results Obtained by Different Vessels. CAFSAC Res. Doc. 85/3. 43 p. Finley, M., Wang, Y., and Stone, H.H. 2018. Assessment of 4X5Y Haddock (Melanogrammus aeglefinus) in 2016. DFO Can. Sci. Advis. Sec. Res. Doc. 2018/041. iv + 54 p. Fowler, G.M. 2011. Old and older perceptions of the migrations and distribution of Haddock, Melanogrammus aeglefinus, in northwest Atlantic waters from tagging conducted in the Bay of Fundy, Georges Bank, Scotian Shelf, and the Southern Gulf of St Lawrence. J. Northw. Atl. Fish. Sci. 43: 137–157. Fowler, G.M., and Showell, M.A. 2009. Calibration of bottom trawl survey vessels: Comparative fishing between the Alfred Needler and Teleost on the Scotian Shelf during the summer of 2005. Can. Tech. Rep. Fish. Aquat. Sci. 2824: iv +25 p. Frank, K.T. 1992. Demographic consequences of age-specific dispersal in marine fish populations. Can. J. Fish. Aquat. Sci. 49: 2222–2231. Friedland, K.D. 2021. A test of the provisioning hypothesis of recruitment control in Georges Bank Haddock. Can. J. Fish. Aquat. Sci. 78: 655–658. Gini, C. 1921. Measurement of inequality of incomes. The Economic Journal. 31: 124–126. Grosslein, M.D. 1962. Haddock stocks in the ICNAF convention area. ICNAF Redbook III: 124– 131. Head, E.J.H., Brickman, D., and Harris, L.R. 2005. An exceptional Haddock year class and unusual environmental conditions on the Scotian Shelf in 1999. J. Plank. Res. 27(6): 597– 602. Hurley, P. C. F., and Campana, S.E. 1989. Distribution and abundance of Haddock (Melanogrammus aeglefinus) and Atlantic cod (Gadus morhua) eggs and larvae in the waters off southwest Nova Scotia. Can. J. Fish. Aquat. Sci. 46: 103–112. https://www.dfo-mpo.gc.ca/fisheries-peches/ifmp-gmp/groundfish-poisson-fond/groundfish-poisson-fond-4vwx5-eng.html https://www.dfo-mpo.gc.ca/csas-sccs/Publications/ScR-RS/2020/2020_021-eng.html https://www.dfo-mpo.gc.ca/csas-sccs/Publications/ScR-RS/2020/2020_021-eng.html https://www.dfo-mpo.gc.ca/csas-sccs/Publications/ScR-RS/2021/2021_021-eng.html https://www.dfo-mpo.gc.ca/csas-sccs/Publications/ScR-RS/2021/2021_021-eng.html https://www.dfo-mpo.gc.ca/csas-sccs/Publications/ScR-RS/2021/2021_026-eng.html https://www.dfo-mpo.gc.ca/csas-sccs/publications/resdocs-docrech/1985/1985_003-eng.html https://www.dfo-mpo.gc.ca/csas-sccs/Publications/ResDocs-DocRech/2018/2018_041-eng.html https://www.dfo-mpo.gc.ca/csas-sccs/Publications/ResDocs-DocRech/2018/2018_041-eng.html 20 Hurley, P.C.F., Comeau, P., and Black, G.A.P. 1994. Assessment of 4X Haddock in 1993. DFO Atl. Fish. Res. Doc. 94/39. 42 p. Hurley, P.C.F., Black, G.A.P., Mohn, R.K., and Comeau, P. 1997. Assessment of 4X Haddock in 1996 and the first half of 1997. DFO Can. Stock Assess. Sec. Res.Doc. 97/108. 101 p. Hurley, P.C.F., Black, G.A.P., Comeau, P.A., Mohn, R.K., and Zwanenburg, K. 1998. Assessment of 4X Haddock in 1997 and the first half of 1998. DFO Can. Stock Assess. Sec. Res. Doc. 98/136. 96 p. Hurley, P.C.F., Black, G.A.P., Simon, J.E., Mohn, R.K., and Comeau, P.A. 2002. Assessment of the Status of Div. 4X/5Y Haddock in 2002. DFO Can. Sci. Advis. Sec. Res. Doc. 2002/098. 77 p. Hurley, P.C.F., Black, G.A.P., Young, G.A., Mohn, R.K., and Comeau, P.A. 2009. Assessment of the Status of Divisions 4X5Y Haddock in 2005. DFO Can. Sci. Advis. Sec. Res. Doc. 2009/024. vi + 86 p. Kane, J. 1984. The feeding habits of co-occurring Cod and Haddock larvae from Georges Bank. Mar. Ecol. Prog. Ser. 16(1): 9–20. Kish, L. 1992. Weighting for unequal Pi. Journal of Official Statistics. 8: 183–200. Lage, C., Purcell, M., Fogarty, M. and Kornfield, I. 2001. Microsatellite evaluation of Haddock (Melanogrammus aeglefinus) stocks in the northwest Atlantic Ocean. Can. J. Fish. Aquat. Sci. 58(5): 982–990. Lapolla, A., and Buckley, L.J. 2005. Hatch date distributions of young-of-year Haddock Melanogrammus aeglefinus in the Gulf of Maine/Georges Bank region: implications for recruitment. Mar. Ecol. Prog. Ser. 290: 239–249. Mahon, R. and Neilson, J.D. 1987. Diet changes in Scotian Shelf Haddock during the pelagic and demersal phases of the first year of life. Mar. Ecol. Prog. Ser. 37: 123–130. McCracken, F.D. 1960. Studies of Haddock in the Passamaquoddy Bay Region. J. Fish. Res. Bd. Can. 17(2): 175–180. Mohn, R.K., Trzcinski, M.K., Black, G.A.P., Armsworthy, S., Young, G.A., Comeau, P.A., and den Heyer, C.E. 2010. Assessment of the Status of Division 4X5Y Haddock in 2009. DFO Can. Sci. Advis. Sec. Res. Doc. 2010/085. vi + 61 p. Murawski, S. A., and Finn, J. T. 1988. Biological bases for mixed-species fisheries: species co- distribution in relation to environmental and biotic variables. Can. J. Fish. Aquat. Sci. 45: 1720–1735. Northeast Fisheries Science Center. 2014. 59th Northeast Regional Stock Assessment Workshop (59th SAW) Assessment Report. US Dept. Commer. Northeast Fish. Sci. Cent. Ref. Doc. 14–09: 782 p. Needler, A.W.H. 1930. The migrations of Haddock and the interrelationships of Haddock populations in North American waters. Contributions to Canadian Biology and Fisheries. 6(1): 243–313. O'Boyle, R. 1981. An assessment of the 4X Haddock stock for the 1962-80 period. CAFSAC Res. Doc. 81/24. 54 p. O'Boyle, R.N., Frank, K., and Simon, J. 1989. An evaluation of the population dynamics of 4X haddock during 1962-88 with yield projected to 1990. CAFSAC Res. Doc. 89/58. 59 p. https://www.dfo-mpo.gc.ca/csas-sccs/publications/resdocs-docrech/1994/1994_039-eng.html https://www.dfo-mpo.gc.ca/csas-sccs/publications/resdocs-docrech/1997/1997_108-eng.htm https://www.dfo-mpo.gc.ca/csas-sccs/publications/resdocs-docrech/1997/1997_108-eng.htm https://www.dfo-mpo.gc.ca/csas-sccs/publications/resdocs-docrech/1998/1998_136-eng.htm https://www.dfo-mpo.gc.ca/csas-sccs/publications/resdocs-docrech/2002/2002_098-eng.htm https://www.dfo-mpo.gc.ca/csas-sccs/publications/resdocs-docrech/2002/2002_098-eng.htm https://www.dfo-mpo.gc.ca/csas-sccs/publications/resdocs-docrech/2009/2009_024-eng.htm https://www.dfo-mpo.gc.ca/csas-sccs/publications/resdocs-docrech/2009/2009_024-eng.htm https://www.dfo-mpo.gc.ca/csas-sccs/publications/resdocs-docrech/2010/2010_085-eng.htm https://www.dfo-mpo.gc.ca/csas-sccs/publications/resdocs-docrech/1981/1981_024-eng.html https://www.dfo-mpo.gc.ca/csas-sccs/publications/resdocs-docrech/1989/1989_058-eng.html https://www.dfo-mpo.gc.ca/csas-sccs/publications/resdocs-docrech/1989/1989_058-eng.html 21 Ouellette-Plante, J., Van Beveren, E., Benoît, H.P. and Brassard, C. 2022. Details of catchR, an R package to estimate the age and length composition of fishery catches, with an application to 3Pn4RS Atlantic cod. DFO Can. Sci. Advis. Sec. Res. Doc. 2022/015. iv + 69 p. Page, F.H., and Frank, K.T. 1989. Spawning time and egg stage duration in northwest Atlantic Haddock (Melanogrammus aeglefinus) stocks with emphasis on Georges and Browns Bank. Can. J. Fish. Aquat. Sci. 46: 68–81. Pennington, M. 1991. On testing the robustness of lognormal based estimators. Biometrics. 47: 1623–1624. Pennington, M. 1996. Estimating the mean and variance from highly skewed marine data. Fish. Bull. 94: 498–505. Perry, R. I., and Smith, S. J. 1994. Identifying habitat associations of marine fishes using survey data: an application to the northwest Atlantic. Can. J. Fish. Aquat. Sci. 51: 589–602. Platt, T., Fuentes-Yaco, C., Frank, K.T. 2003. Marine ecology: Spring algal bloom and larval fish survival. Nature. 423: 398–399. Purcell, M.K., Kornfield, I., Fogarty, M., and Parker, A. 1996. Interdecadal heterogeneity in mitochondrial DNA of Atlantic Haddock (Melanogrammus aeglefinus) from Georges Bank. Molecular Marine Biology and Biotechnology. 5(3): 185–192. Rivard, D. 1982. APL programs for stock assessment (revised). Can. Tech. Rep. Fish. Aquat. Sci. No. 1091. Rogers, R., Rowe, S., and Morgan, J. 2016. Depth and temperature associations of Haddock Melanogrammus aeglefinus off southern Newfoundland. J. Fish Bio. 89(5): 2306–2325. Schroeder, W.C. 1942. Results of Haddock tagging in the Gulf of Maine from 1923 to 1932. J. Mar. Res. 5(1): 1–19. Scott, W.B., and Scott, M.G. 1988. Atlantic fishes of Canada. University of Toronto Press. Can. Bull. Fish. Aquat. Sci. 219: 731 p. Syrjala, S.E. 2000. Critique on the use of the delta distribution for the analysis of trawl survey data. ICES J. Mar. Sci. 57(4): 831–842. Stone, H.H., Themelis, D., Cook, A.M., Clark, D.S., Showell, M.A., Young, G., Gross, W.E., Comeau, P.A., and Allade, L.A. 2013. Silver Hake 2012 Framework Assessment: Data Inputs and Exploratory Modelling. DFO Can. Sci. Advis. Sec. Res. Doc. 2013/008. v + 133 p. Stone, H.H., and Hansen, S.C. 2015. 4X5Y Haddock 2014 Framework Assessment: Data Inputs and Exploratory Modelling. DFO Can. Sci. Advis. Sec. Res. Doc. 2015/022. iv + 90 p. Wang, Y., Stone, H.H., and Finley, M. 2017. 4X5Y Haddock 2016 Framework Assessment: Modelling and Reference Points. DFO Can. Sci. Advis. Sec. Res. Doc. 2017/026. v + 69 p. Wise, J.P. and Jensen, A.C. 1960. Stocks of the important commercial species of fish in the ICNAF Convention area. Int. Comm. NW Atl. Fish. Ann. Meet. Doc. 25, ser. No 743: 1–14. https://www.dfo-mpo.gc.ca/csas-sccs/Publications/ResDocs-DocRech/2022/2022_015-eng.html https://www.dfo-mpo.gc.ca/csas-sccs/Publications/ResDocs-DocRech/2022/2022_015-eng.html https://www.dfo-mpo.gc.ca/csas-sccs/Publications/ResDocs-DocRech/2022/2022_015-eng.html https://www.dfo-mpo.gc.ca/csas-sccs/Publications/ResDocs-DocRech/2013/2013_008-eng.html https://www.dfo-mpo.gc.ca/csas-sccs/Publications/ResDocs-DocRech/2013/2013_008-eng.html https://www.dfo-mpo.gc.ca/csas-sccs/publications/resdocs-docrech/2015/2015_022-eng.html https://www.dfo-mpo.gc.ca/csas-sccs/publications/resdocs-docrech/2015/2015_022-eng.html https://www.dfo-mpo.gc.ca/csas-sccs/Publications/ResDocs-DocRech/2017/2017_026-eng.html https://www.dfo-mpo.gc.ca/csas-sccs/Publications/ResDocs-DocRech/2017/2017_026-eng.html 22 TABLES Table 1. Estimated total landings and total allowable catch (TAC) in metric tonnes by calendar year and region (SS = Scotian Shelf; BoF = Bay of Fundy) where the spatial definition of regions is defined in Figure 1. A dash (—) indicates no data or not applicable. Year Foreign Landings a BoF (other) BoF (strata 482) BoF (strata 483 b) Total BoF Total SS Total 4X5Y Stock Area Total Allowable Catch 1967 3,214 c 20,119 — — 20,119 19,783 39,902 — 1968 3,345 c 11,787 — — 11,787 20,521 32,308 — 1969 2,204 c 6,676 — — 6,676 23,653 30,329 — 1970 2,022 d 4,329 — — 4,329 13,743 18,072 18,000 1971 1,099 d 3,703 — — 3,703 13,888 17,592 18,000 1972 890 d 3,411 — — 3,411 10,072 13,483 9,000 1973 419 d 2,470 — — 2,470 10,636 13,106 9,000 1974 792 d 5,183 — — 5,183 8,195 13,378 0 1975 2,159 d 5,570 — — 5,570 12,727 18,298 15,000 1976 1,072 d 4,000 — — 4,000 13,497 17,498 15,000 1977 1,662 d 3,524 — — 3,524 17,757 21,281 15,000 1978 1,164 d 5,562 — — 5,562 21,647 27,209 21,500 1979 88 d 6,061 — — 6,061 18,863 24,925 26,000 1980 332 d 8,052 — — 8,052 21,087 29,139 28,000 1981 481 d 7,605 — — 7,605 23,753 31,358 27,850 1982 858 d 8,749 — — 8,749 16,952 25,701 32,000 1983 518 d 9,338 — — 9,338 18,023 27,361 32,000 1984 206 d 7,120 — — 7,120 14,013 21,133 32,000 1985 26 d 5,909 — — 5,909 10,222 16,131 15,000 1986 50 d 5,316 — — 5,316 10,257 15,573 15,000 1987 17 d 2,609 — — 2,609 11,172 13,781 15,000 1988 55 d 2,057 — — 2,057 9,231 11,288 12,400 1989 34 e 1,273 — — 1,273 5,559 6,833 4,600 1990 52 e 1,565 20.7 1.38 1,587 5,966 7,553 4,600 1991 41 e 2,319 101 31.2 2,451 7,377 9,828 0 1992 17 e 2,218 89.9 14.1 2,322 8,203 10,525 0 1993 21 e 1,849 40.3 8.76 1,898 5,070 6,968 6,000 1994 1 f 1,598 14.3 6.37 1,619 2,787 4,406 4,500 1995 9 f 1,938 357 189 2,484 3,180 5,664 6,000 1996 8 f 2,556 318 170 3,044 3,200 6,244 6,500 1997 8 f 2,817 410 281 3,508 3,031 6,539 6,700 1998 1 g 2,620 659 296 3,576 4,303 7,878 8,100 1999 0 g 2,443 719 751 3,914 2,702 6,616 8,100 2000 0 g 2,052 631 421 3,105 3,852 6,956 8,100 h 2001 0 g 2,736 505 991 4,231 4,251 8,483 8,100 h 2002 0 g 3,235 741 698 4,674 3,329 8,003 8,100 h 2003 — 4,078 747 1,141 5,966 2,727 8,693 8,100 h 2004 — 2,529 432 1,039 3,999 2,511 6,510 10,000 h 2005 — 1,627 444 1,276 3,348 2,315 5,663 8,000 h 2006 — 1,343 405 585 2,333 2,399 4,732 7,000 h 2007 — 1,235 672 2,482 4,388 2,483 6,871 7,000 h 2008 — 1,000 1,147 837 2,984 2,377 5,361 7,000 h 2009 — 767 436 986 2,189 3,289 5,478 7,000 h 2010 — 613 419 957 1,989 3,662 5,651 6,000 h 2011 — 449 385 601 1,435 2,295 3,730 6,000 h 2012 — 761 296 188 1,244 2,883 4,127 5,100 h 2013 — 811 741 206 1,758 1,775 3,533 5,100 h 2014 — 895 158 395 1,448 1,276 2,724 5,100 h 2015 — 1,112 79.6 279 1,471 1,296 2,767 5,100 h 2016 — 1,752 206 346 2,304 1,105 3,409 5,100 h 2017 — 3,428 232 186 3,846 1,163 5,009 7,650 h 2018 — 3,358 365 145 3,868 945 4,813 7,650 h 2019 — 3,046 171 138 3,355 1,496 4,851 9,000 h 2020 — 2,007 281 309 2,597 3,294 5,891 6,877 h 2021 — 985 904 114 2,003 2,264 4,267 6,877 h 2022 — 767 700 97.5 1,564 2,552 4,116 6,198 h Notes: a Foreign landings were proportionally assigned to SS and BoF in this table and therefore are included in the total 4X5Y stock area column. b The portion of survey strata 5Z9 in DFO unit area 4Xp is included with strata 483. c O'Boyle 1981, d O'Boyle et al. 1989, e Hurley and Comeau 1994, f Hurley et al. 1997, g Hurley et al. 2002, h TAC for fishing season (April 1–March 31 of following year). 23 Table 2. Estimated total landings in metric tonnes by fleet (SS = Scotian Shelf; BoF = Bay of Fundy; F = fixed gear; M = mobile gear) and calendar year. Year BoF (F) BoF (M) SS (F) SS (M) Total 4X5Y Stock Area 1967 1,825 18,294 0 19,783 39,902 1968 733 11,054 1,647 18,874 32,308 1968 802 5,874 1,908 21,745 30,329 1970 684 3,644 2,897 10,846 18,072 1971 530 3,173 3,087 10,802 17,592 1972 562 2,849 4,123 5,949 13,483 1973 452 2,017 5,920 4,716 13,106 1974 565 4,618 6,369 1,826 13,378 1975 600 4,971 5,199 7,528 18,298 1976 284 3,716 5,120 8,378 17,498 1977 211 3,313 4,405 13,352 21,281 1978 369 5,193 6,445 15,202 27,209 1979 250 5,811 4,402 14,461 24,925 1980 392 7,660 6,024 15,063 29,139 1981 265 7,340 7,422 16,332 31,358 1982 315 8,434 7,425 9,527 25,701 1983 348 8,990 8,233 9,791 27,361 1984 183 6,937 6,456 7,557 21,133 1985 137 5,772 4,077 6,145 16,131 1986 119 5,197 5,339 4,917 15,573 1987 166 2,444 4,917 6,255 13,781 1988 134 1,923 3,452 5,779 11,288 1989 121 1,152 2,746 2,814 6,833 1990 169 1,418 3,924 2,043 7,553 1991 278 2,173 5,129 2,248 9,828 1992 633 1,689 6,157 2,046 10,525 1993 464 1,434 3,741 1,329 6,968 1994 154 1,465 2,073 714 4,406 1995 415 2,069 2,073 1,107 5,664 1996 373 2,671 2,030 1,169 6,244 1997 390 3,118 1,841 1,190 6,539 1998 761 2,815 1,882 2,421 7,878 1999 696 3,218 1,325 1,377 6,616 2000 518 2,587 2,204 1,647 6,956 2001 367 3,864 1,891 2,361 8,483 2002 649 4,025 1,682 1,647 8,003 2003 666 5,300 1,374 1,353 8,693 2004 377 3,622 785 1,726 6,510 2005 401 2,947 560 1,755 5,663 2006 444 1,889 857 1,542 4,732 2007 547 3,841 1,031 1,452 6,871 2008 297 2,687 872 1,505 5,361 2009 383 1,806 532 2,758 5,478 2010 542 1,447 714 2,949 5,651 2011 338 1,097 565 1,730 3,730 2012 195 1,049 596 2,287 4,127 2013 38.4 1,720 378 1,397 3,533 2014 11.9 1,436 250 1,026 2,724 2015 10.6 1,460 101 1,196 2,767 2016 3.37 2,301 82.4 1,022 3,409 2017 2.85 3,843 49.0 1,114 5,009 2018 2.04 3,866 21.6 923 4,813 2019 1.49 3,354 16.7 1,479 4,851 2020 2.23 2,595 40.6 3,253 5,891 2021 2.45 2,001 44.6 2,219 4,267 2022 2.15 1,562 29.9 2,522 4,116 24 Table 3. Aging protocol by month and quarter (Q) for 4X5Y Haddock otoliths. Opaque sections indicate summer growth and hyaline (annuli) rings indicate slower winter growth. Based on the edge of the otolith and the month sampled, 1 year is added to the age (+1), 1 year is subtracted from the age (-1) or no changes are made (=). Edge Q1 Jan Q1 Feb Q1 Mar Q2 Apr Q2 May Q2 June Q3 July Q3 Aug Q3 Sept Q4 Oct Q4 Nov Q4 Dec wide opaque = +1 +1 +1 = = = = = = = = narrow opaque = = = = = = = = = = = = wide hyaline -1 = = = = = = -1 -1 -1 -1 -1 narrow hyaline -1 = = = = = = -1 -1 -1 -1 -1 25 Table 4. Estimated regression coefficients (a = intercept and b = slope, n = sample size) from the regression of log10(weight in kg) on log10(length in cm) by year and region (BoF = Bay of Fundy; SS = Scotian Shelf) from Haddock at least 20 cm in length collected from the DFO summer ecosystem survey. Year BoF n BoF a BoF b SS n SS a SS b 1970 192 -2.10539 3.084532 1,088 -2.44547 3.286905 1971 170 -2.07463 3.058974 867 -2.14671 3.093931 1972 162 -1.81137 2.923960 743 -2.23240 3.161174 1973 283 -2.05359 3.045901 565 -2.07990 3.066916 1974 324 -2.21714 3.145169 1,195 -2.12957 3.093631 1975 176 -2.00546 3.027663 550 -2.22950 3.153312 1976 419 -1.88172 2.938045 776 -2.13160 3.083517 1977 396 -2.06530 3.061609 759 -2.14581 3.103643 1978 365 -2.07951 3.081040 1,374 -2.27198 3.187906 1979 797 -2.21995 3.153333 1,242 -2.07712 3.063980 1980 880 -1.97929 3.005732 1,128 -2.01470 3.021643 1981 814 -1.91159 2.972035 1,219 -2.11162 3.082051 1982 456 -1.97447 3.010088 443 -2.12447 3.083612 1983 498 -1.88914 2.955086 893 -2.17436 3.100123 1984 412 -1.96667 2.998904 539 -2.17466 3.103408 1985 247 -1.88118 2.940965 572 -2.05412 3.030162 1986 212 -1.91085 2.967923 594 -2.10578 3.065523 1987 176 -1.97578 3.002859 503 -1.99127 3.005943 1988 200 -2.07309 3.062436 401 -2.28609 3.185882 1989 143 -1.97923 3.011611 441 -2.13304 3.085471 1990 210 -1.82158 2.922080 601 -1.97482 3.003685 1991 233 -1.93553 2.978873 559 -2.13663 3.082461 1992 144 -2.01079 3.021011 475 -1.94041 2.972012 1993 106 -1.74555 2.856694 412 -2.16575 3.096996 1994 164 -2.01753 3.021993 674 -2.12665 3.078891 1995 390 -2.08198 3.048451 638 -2.09047 3.047831 1996 357 -2.01853 3.015170 780 -2.06580 3.033495 1997 288 -1.91492 2.955286 764 -2.01672 3.011504 1998 260 -2.00805 3.012953 630 -2.11865 3.068482 1999 268 -2.00643 3.010496 787 -2.02224 3.017227 2000 443 -2.09292 3.056297 706 -2.20620 3.120640 2001 275 -1.91631 2.955476 907 -2.11672 3.069341 2002 445 -1.99233 2.990165 956 -2.04167 3.012269 2003 281 -1.85884 2.910944 730 -1.86694 2.911359 2004 239 -2.12751 3.061750 575 -2.29205 3.171636 2005 185 -1.99395 2.990214 907 -2.22288 3.135336 2006 265 -2.05702 3.036577 743 -2.11452 3.060106 2007 205 -1.91170 2.954522 717 -2.12756 3.078897 2008 158 -2.11512 3.066586 684 -2.21251 3.128347 2009 159 -2.03245 3.024403 559 -2.21013 3.139367 2010 189 -2.15173 3.081395 530 -2.28920 3.164332 2011 253 -2.06607 3.026162 633 -2.05209 3.020152 2012 215 -1.92561 2.948711 688 -1.97636 2.961146 2013 260 -1.98631 2.991648 648 -1.93290 2.946570 2014 385 -2.13575 3.081349 494 -1.99529 2.984684 2015 563 -2.02866 3.008123 828 -2.00063 2.985196 2016 762 -2.02121 2.993971 783 -1.99615 2.968785 2017 611 -1.99831 2.986591 660 -2.14259 3.077341 2018 373 -1.97576 2.981431 549 -2.19886 3.123689 2019 413 -2.09695 3.055958 681 -2.15529 3.098387 2020 327 -2.06140 3.033629 495 -2.27289 3.171666 2021 239 -2.00022 3.011811 325 -2.15442 3.097145 2022 320 -2.08845 3.059790 511 -2.15200 3.088947 26 Table 5. Estimated fishery numbers at age (000s) of Haddock from the Scotian Shelf. Year 1 2 3 4 5 6 7 8 9 10 11 12+ 1970 0 298 712 1,438 277 296 3,759 1,287 227 78.0 84.9 27.1 1971 0 136 2,080 935 1,160 460 42.6 2,922 1,000 172 108 172 1972 0.267 54.6 1,892 1,533 477 559 88.8 22.7 936 387 16.3 242 1973 0.640 170 462 2,650 892 397 588 349 279 385 68.8 24.0 1974 0 103 1,250 242 1,412 396 132 201 72.7 205 304 36.8 1975 0 149 1,646 3,561 595 1,095 279 173 54.8 43.0 102 171 1976 0 138 788 2,699 3,066 395 905 191 79.6 91.6 23.8 145 1977 0 765 2,166 1,642 3,254 2,559 324 378 43.4 72.4 30.9 63.3 1978 0 78.1 2,784 5,295 1,616 2,119 740 140 112 15.3 12.6 45.4 1979 0 61.0 828 5,351 3,127 817 952 228 40.7 34.6 10.9 18.9 1980 0 55.3 1,574 2,601 4,244 2,576 520 431 162 36.1 25.7 18.0 1981 0 55.9 597 4,244 3,316 3,024 1,234 381 342 110 21.9 44.5 1982 0 19.3 911 1,550 3,737 1,526 1,345 263 118 87.3 29.2 30.0 1983 0 9.61 851 3,686 3,351 1,919 797 357 160 98.1 50.4 34.0 1984 0 7.60 210 2,952 2,595 2,025 963 420 143 89.7 29.7 39.0 1985 0 23.0 520 985 2,965 1,293 515 464 406 252 99.0 78.6 1986 0 181 341 1,831 1,495 2,323 593 315 223 88.4 70.3 74.4 1987 0.501 20.6 252 674 2,513 1,030 2,170 566 215 205 53.6 80.4 1988 2.15 11.0 835 865 938 1,520 648 553 196 130 107 84.8 1989 0 64.5 317 467 775 318 607 355 389 118 45.3 84.6 1990 0 135 635 211 284 436 406 624 237 180 94.6 44.9 1991 0 2.98 368 1,634 463 261 316 188 326 272 124 315 1992 4.52 112 139 2,044 1,608 230 158 322 263 278 76.3 206 1993 0 12.0 391 317 1,532 771 138 68.5 73.4 29.4 65.3 77.9 1994 0 75.3 135 343 188 814 157 35.8 25.7 3.87 28.8 18.7 1995 0 31.1 351 388 402 115 326 373 111 20.2 10.3 37.9 1996 0 1.23 242 475 339 204 216 384 324 78.3 6.92 3.33 1997 0 0 242 1,057 390 247 110 57.0 72.2 76.8 29.2 2.19 1998 0 5.60 68.3 540 1,185 787 334 161 105 56.8 68.0 20.3 1999 0 21.4 106 218 558 488 301 89.2 35.8 21.2 16.8 15.7 2000 0 82.9 505 511 432 774 688 322 67.7 24.9 15.3 6.01 2001 0 30.5 573 533 444 384 912 565 211 41.9 25.0 38.5 2002 0.130 10.6 186 835 306 286 263 590 275 93.3 57.4 58.0 2003 0.029 0.873 45.5 730 942 328 193 61.0 96.3 128 34.6 4.94 2004 0 10.6 92.1 415 643 904 193 87.3 88.8 61.8 18.1 14.2 2005 0.154 8.20 36.3 318 943 483 303 386 26.2 44.2 17.7 4.34 2006 0 9.17 244 135 431 548 579 536 79.5 8.53 29.8 2.20 2007 0 12.6 114 971 157 389 381 330 341 64.3 35.3 13.5 2008 0 18.2 209 277 1,070 266 153 280 169 163 27.4 18.5 2009 0 10.5 299 353 360 1,053 214 140 217 173 67.4 34.9 2010 0 4.66 85.3 472 402 494 1,015 226 140 238 120 86.6 2011 3.16 34.7 58.9 140 1,071 376 209 483 36.6 25.7 3.77 59.8 2012 0 21.8 211 314 333 1,284 304 284 273 111 20.4 57.1 2013 0.922 49.1 653 303 240 105 289 339 77.9 69.7 31.9 9.63 2014 0.149 148 449 1,014 120 46.2 70.5 57.4 43.3 15.1 9.78 1.84 2015 0.547 139 667 604 556 37.2 10.6 13.4 57.1 2.55 0.589 5.75 2016 0 126 390 462 441 269 21.2 2.88 56.7 0.617 0.269 0.060 2017 0 4.09 176 713 402 132 62.9 5.66 8.65 0 0.310 0.066 2018 0 2.42 14.9 661 506 16.5 8.06 37.1 0.345 0.348 0 0 2019 0 6.96 139 86.2 533 1,305 30.0 12.6 2.81 0 0 0.003 2020 0.193 26.1 670 372 266 724 2,487 132 55.5 0.894 0 0.013 2021 6.47 103 695 858 825 180 177 959 81.1 9.04 0.397 0.010 2022 2.50 151 111 1,650 766 212 291 207 512 18.4 0 2.54 27 Table 6. Estimated fishery numbers at age (000s) of Haddock from the Bay of Fundy. Year 1 2 3 4 5 6 7 8 9 10 11 12+ 1970 6.07 459 147 119 77.1 142 898 238 33.5 13.8 39.7 25.6 1971 0.250 445 403 148 151 8.15 7.88 593 140 15.2 4.65 200 1972 8.62 305 1,432 206 72.9 64.0 13.2 37.5 236 121 0.788 52.9 1973 61.9 2,162 68.4 414 79.5 29.0 75.6 10.9 42.3 109 9.07 2.29 1974 38.6 771 3,476 147 388 92.7 24.5 46.2 27.6 36.7 107 1.50 1975 0.683 1,689 2,212 1,441 50.7 102 34.9 5.80 1.84 1.70 8.27 34.0 1976 1.11 1,204 1,699 871 355 11.5 73.6 0 0 0 0 16.0 1977 12.8 1,091 768 305 279 211 35.8 23.1 21.6 0.153 0 5.74 1978 0.003 11.0 1,116 1,287 321 518 139 11.5 19.0 6.38 0 31.7 1979 6.67 25.3 325 599 695 786 182 121 69.7 35.3 2.16 37.9 1980 1.19 294 1,070 984 1,411 828 120 225 60.1 12.0 3.37 0.419 1981 0.408 647 1,106 1,632 715 553 154 77.5 54.5 18.7 1.75 1.63 1982 0 940 2,135 1,093 1,013 385 345 50.8 68.7 21.3 6.23 1.29 1983 0 116 3,115 2,111 764 672 157 61.9 71.2 41.5 35.6 25.7 1984 2.72 1,083 1,323 2,240 780 397 148 41.0 22.7 15.5 13.8 1.38 1985 7.31 740 2,481 462 710 289 207 83.6 48.3 64.9 17.9 14.7 1986 0 349 703 2,251 311 467 77.8 47.9 25.4 37.8 4.55 16.5 1987 0 120 495 361 623 188 94.1 45.0 17.8 17.7 11.4 12.8 1988 2.24 67.0 179 150 167 287 109 90.3 45.5 34.6 18.8 20.3 1989 0.068 111 265 107 115 18.6 45.1 25.5 19.4 14.6 28.0 8.03 1990 0 159 437 97.0 59.2 53.0 41.0 63.9 38.2 23.4 8.89 2.69 1991 2.31 20.5 596 542 131 37.0 38.6 27.0 33.0 33.5 14.5 24.2 1992 0.598 83.8 66.9 497 415 27.5 29.6 47.6 19.6 25.7 2.05 27.9 1993 0.577 98.6 264 70.4 306 258 43.8 12.8 14.9 12.1 11.1 9.64 1994 2.46 56.1 207 193 41.8 249 90.0 5.79 1.43 11.1 2.48 9.05 1995 0.282 45.6 381 426 222 90.2 128 112 60.6 4.57 5.95 12.8 1996 0 17.8 658 656 254 145 63.9 109 125 37.7 9.95 7.91 1997 0 2.47 290 1,024 567 281 90.3 40.7 34.4 17.7 14.5 2.35 1998 0 43.2 82.3 592 754 431 180 116 29.0 25.7 38.7 16.3 1999 0 10.2 222 295 556 472 304 158 11.5 16.0 53.3 14.0 2000 0 72.2 176 293 239 454 274 211 110 54.0 8.05 20.7 2001 0 43.6 828 721 391 144 269 312 167 72.3 29.5 8.42 2002 0.486 32.9 271 1,461 612 286 216 205 170 96.9 60.6 6.28 2003 0 19.0 860 822 1,824 383 111 162 45.0 39.3 24.7 22.8 2004 0 1.27 112 734 414 871 483 188 45.1 62.6 54.6 26.4 2005 0 8.01 16.3 142 1,086 477 399 126 40.6 21.4 14.4 7.48 2006 0 21.9 473 147 287 510 299 132 33.5 3.51 1.25 15.2 2007 0.180 50.5 184 3,015 124 115 338 183 72.4 32.2 13.9 4.96 2008 0 39.4 166 316 1,654 106 95.4 158 104 44.1 3.80 3.24 2009 0.578 17.3 74.8 202 316 789 320 55.4 31.0 15.1 4.05 6.12 2010 0 6.14 4.65 74.0 101 287 878 190 18.8 28.1 32.5 8.30 2011 0.166 20.8 39.3 34.2 170 154 170 443 138 19.6 11.5 4.83 2012 2.62 122 84.1 68.5 58.4 156 134 116 263 87.3 4.54 24.2 2013 17.2 92.6 969 215 82.5 29.2 70.3 60.9 56.8 185 86.9 7.10 2014 5.78 143 310 765 148 52.1 26.0 44.1 27.5 4.11 22.2 14.9 2015 0 156 329 396 650 39.0 12.5 20.2 10.5 7.35 2.94 10.1 2016 1.63 281 1,448 381 544 446 36.1 8.66 8.09 4.63 0.426 6.49 2017 1.83 18.8 540 4,283 300 260 181 5.96 0 3.55 0.535 0 2018 1.48 108 164 729 3,897 193 29.9 81.3 0.921 0 0 0 2019 1.63 93.7 188 249 501 3,167 83.5 16.4 3.07 3.28 0 0 2020 6.52 60.5 195 456 265 305 1,670 49.0 21.7 8.48 1.21 0 2021 24.3 104 461 216 355 109 107 473 193 5.75 1.12 0 2022 76.9 436 304 200 344 217 83.8 185 177 5.16 0 0 28 Table 7. Fishery length-at-age (LAA) in cm, estimated as the weighted mean LAA from Bay of Fundy (BoF) and Scotian Shelf (SS), weighted by the catch-at-age for BoF and SS, and adjusted for growth to the month of August. Year 0 1 2 3 4 5 6 7 8 9 10 11 12+ 1970 12.7 25.4 36.0 42.7 47.2 49.5 56.5 57.1 61.9 65.6 67.5 67.0 74.5 1971 12.4 24.8 35.0 43.0 49.5 52.0 54.0 54.6 60.1 63.9 66.8 67.6 69.1 1972 12.9 25.8 34.8 43.6 51.7 54.5 58.9 59.8 63.3 62.8 64.8 68.6 68.1 1973 12.7 25.4 35.2 40.2 48.7 56.3 60.4 61.4 64.3 64.2 70.1 69.0 75.4 1974 13.5 27.0 34.1 43.2 49.0 56.5 61.6 64.9 64.5 66.1 67.3 69.6 74.2 1975 12.5 25.0 36.6 42.6 50.2 57.7 61.1 65.3 67.1 68.2 67.9 68.8 71.8 1976 14.0 28.0 35.7 42.4 49.0 56.1 61.7 64.6 66.5 66.7 69.1 69.6 71.8 1977 14.5 29.0 36.1 44.3 49.5 54.4 60.4 65.7 67.3 68.5 70.1 71.0 72.5 1978 14.5 29.0 35.4 43.6 51.1 57.3 61.4 65.9 69.3 71.5 73.6 72.4 72.8 1979 11.4 22.7 35.1 42.4 49.8 56.9 61.3 65.5 68.9 72.6 72.5 74.6 73.1 1980 12.0 24.0 35.3 42.1 49.1 55.2 62.1 65.2 68.8 70.7 74.0 74.2 78.6 1981 15.0 30.0 37.1 43.2 49.3 55.4 60.7 64.0 67.0 70.1 71.2 74.3 76.1 1982 13.9 27.7 35.2 42.9 50.1 54.6 60.1 64.4 67.8 70.4 72.8 73.7 76.9 1983 13.9 27.8 32.5 40.9 48.5 56.2 61.4 66.1 67.9 69.1 71.0 72.1 73.5 1984 13.5 27.0 35.2 40.9 45.9 52.3 58.7 62.4 65.5 68.2 70.6 71.9 74.9 1985 14.8 29.7 36.7 43.1 46.3 50.0 55.1 59.7 61.6 62.7 64.6 66.8 69.4 1986 15.2 30.4 35.2 42.7 46.8 49.7 53.4 58.2 61.5 64.1 65.7 67.3 70.1 1987 15.5 25.6 37.2 41.3 46.4 49.1 53.5 56.0 59.5 61.3 63.9 67.9 69.1 1988 15.6 31.1 38.8 43.8 46.5 52.7 53.7 56.7 58.7 62.6 65.2 65.1 68.1 1989 14.5 29.1 40.2 47.1 50.4 53.1 56.6 59.6 60.5 60.4 60.4 65.5 67.9 1990 16.0 31.9 42.3 47.1 50.6 57.4 59.7 61.3 62.5 63.3 63.5 67.3 70.8 1991 16.4 32.7 41.4 45.0 52.1 57.8 60.8 63.9 64.9 65.3 65.1 65.2 66.6 1992 17.0 29.1 38.7 44.8 49.6 56.9 60.4 62.1 62.6 63.1 65.8 65.6 63.8 1993 17.4 34.7 39.6 44.1 46.8 52.6 58.6 61.7 64.9 62.3 69.4 64.5 64.1 1994 16.9 33.8 42.0 46.7 50.3 54.2 57.7 62.4 64.1 60.2 67.1 63.9 68.4 1995 11.5 23.0 39.6 46.6 51.8 55.4 58.7 60.7 62.7 67.6 66.1 67.3 68.5 1996 12.6 25.2 40.0 44.7 49.1 53.6 57.8 56.4 59.9 61.7 62.8 67.1 68.2 1997 10.9 21.8 43.7 44.4 49.0 54.1 59.8 63.0 62.6 65.5 66.8 69.0 68.7 1998 10.8 21.6 37.9 45.0 45.0 51.5 55.8 60.4 63.6 63.9 66.4 65.7 68.8 1999 11.5 23.0 40.8 46.6 50.6 51.6 56.7 61.6 65.6 64.6 64.7 68.8 68.2 2000 12.3 24.7 37.8 42.1 48.8 49.0 52.4 56.7 59.5 63.9 65.7 64.4 69.