Management Track Assessments Spring 2023

Some of the important sources of uncertainty relate to assessment data inputs and the availability of information that would help better understand the dynamics of bluefish. Research recommendations from the recent research track assessment fully detail these uncertainties and data needs. A list of some of the research ideas designed to improve the bluefish stock assessment and reduce some of the uncertainties include:

1. 1.
   Expanding the collection of recreational release length frequency data. The bluefish assessment stratifies recreational release lengths by region, and data in the southern region is lacking. These southern fish tend to be smaller and improved information pertaining to the size distribution of the southern fish would help refine the estimate of recreational discard weight.
3. 2.
   Addressing the uncertainty around temporal availability of bluefish to the fisheries and surveys. The research track assessment made significant advancements in developing an index of bluefish availability based on forage fish in the diets of bluefish like predators. This forage fish index was incorporated into a companion assessment model as a covariate on  MRIP   CPUE  catchability. Further developing this index will help improve the assessment model fit to the  MRIP   CPUE  information, which is an important index that helps scale biomass estimates from the model.
4. 3.
   Develop fishery dependent or independent sampling programs to provide information on larger, older bluefish. The dynamics of this size class are not well sampled or understood.
5. 4.
   Develop an updated recreational release mortality study to derive a more informed estimate of recreational discard mortality. Recreational discards are a significant proportion of the total catch so reducing the uncertainty around the release mortality is important.

 The 7-year Mohn’s ρ, relative to SSB, was 0.14 in the 2022 assessment and was 0.22 in this assessment. The 7-year Mohn’s ρ, relative to F, was 0.10 in the 2022 assessment and was 0.14 in this assessment. This is considered a minor retrospective pattern for both SSB and F because the ρ-adjusted estimates of 2022 SSB (SSB_ρ= 43,235) and 2022 F (F_ρ= 0.177) were within the approximate 90% confidence regions around SSB (36,194–76,871) and F (0.105–0.219).

Population projections for Atlantic Bluefish are reasonably well determined. Shifting to WHAM for model projections has allowed for the incorporation of model uncertainty, auto-regressive processes, and the uncertainty in recruitment and numbers-at-age. The retrospective pattern in F and SSB is considered minor (within the 90% CI of both F and SSB), however, the ρ values of F and SSB have increased when compared to the previous research track assessment.

The Atlantic Bluefish stock is in a rebuilding plan with a rebuild date of 2028. F_Rebuild was re-calculated using a projection that assumes a constant F strategy, such that biomass in 2028 has a 50% chance of exceeding the SSB_{MSY proxy}; F_Rebuild was calculated to be 0.183. The MAFMC risk policy was applied using this F_Rebuild strategy in short term projections to generate ABC values that are consistent with the rebuilding schedule for the next two years.

 A change to the way the age-length keys (ALK) were developed from the research track, which used full multinomial age-length keys, was implemented for this Atlantic Bluefish assessment update. Instead of using full multinomial age-length keys, a hybrid approach was used, and the holes in the ALK were filled with the multinomial model fits. This approach to filling ALK holes isnow consistent with the methodology used for other NEFSC stock assessments and with the NEFSC StockEff program. This new method resulted in minor changes to the results of SSB and F compared to the 2022 research track assessment results.

Stock status of Atlantic Bluefish has not changed from the status determined in the research track assessment.

The Atlantic Bluefish stock has experienced a slight increase in SSB over the past 5 years, coinciding with a decrease in F. Recruitment has increased each year since 2019, and the terminal year recruitment (137 million fish) is the highest value since 2005. Both commercial and recreational fisheries have had low catches since 2018, all well below the time series average of 26,386 mt. With the low catches since 2018, fishing mortality has decreased and remained well below F_{MSY proxy} (0.239). The low catches in recent years are partially a result of bag limit implementation as part of the rebuilding plan. However, these lower catches could also be due to decreased bluefish availability. Anecdotal evidence suggests larger bluefish stayed offshore and inaccessible to most of the recreational fishery in recent years.

The recent bluefish research track identified several new research recommendations that would improve our understanding of bluefish dynamics and help better assess the population through the current or future models. These recommendations include: expand collection of recreational release length frequency data, continue development and refinement of the forage fish / availability index as well as incorporation of this index into a base model for bluefish management advice, initiate additional fisheries-independent surveys or fishery-dependent sampling programs to provide information on larger, older bluefish, continue coastwide collection of length and age samples from fishery-independent and -dependent sources, refinement and development of indices of abundance, and develop a recreational demand model.

WHAM allows for incorporation of environmental covariates on the catchability of survey indices, and a companion model was developed for the research track that leveraged this capability. The companion model investigated a forage fish index as a covariate on catchability of the MRIP CPUE and showed promise for continued development. The covariate led to an overall decreasing trend in catchability over time. This model will be further developed leading up to the 2025 management track assessment, at which time it could be considered for the primary model.

 Estimates of LPUE are made in two ways: per trap hauled, and per day fished. The two estimates generally track each other well. In the case of per day fished, steam time is not considered which increases uncertainty. In the case of traps hauled, the number of traps per string noted in the VTR trip entry is not always the number of traps emptied per haul due to a number of causes normal for fishing, which increases uncertainty.

Especially in a potentially food-limited environment, there also may be uncertainty in calculating LPUE in a baited trap fishery. If the pot is left to soak long enough, it can attract enough crabs to fill the pot and appear as if the population has not changed in density over time. The crabs must travel longer distances to reach the bait.

There is very little uncertainly in landings due to the specialized fleet and processing needs.

  N/A

  N/A

  N/A

  N/A

  N/A

 After implementation of the FMP in 2002, the red crab fishery was restricted to five full-access permits. The vessels fish cooperatively and over a wide area. The footprint of the fishery stretches from the Hague line to Virginia. The fishery is, by regulation, male-only. Landings include some females landed under an EFP during the years 2010 and 2011 but this did not continue. Red crabs are known to segregate loosely by sex and perhaps even size, with the smallest individuals being found further down the continental slope in deeper water. This is one aspect of the fishery that makes discarding highly variable, as the fishing vessels find areas where there is a high concentration of large males but there are always a certain number of females and undersized males which are discarded depending on where the pots are set.

Red crab is considered a data-poor species. They live outside the range of NEFSC surveys (they are sometimes caught in the northern shrimp survey and non-random deep stations in the bottom trawl survey, but in very small numbers) so there is no fishery-independent index of abundance.

Almost nothing is known about deep-sea red crab growth and longevity, and methods of crustacean ageing are still in development and highly uncertain. There have been some studies on mating behavior of captive red crabs, but they do not live long in captivity so controlled growth studies have not been possible. A tagging study for growth was attempted in 2010 with poor results. Very few tagged crabs were ever recovered, for reasons unknown.

There have been two surveys of the deep-sea red crab resource in the past; one in 1974 (Wigley, Theroux and Murray, 1975) and another during the summers of 2003–2005 (Wahle et al. 2008). The first survey was undertaken soon after the fishery began and the second shortly after the deep-sea red crab FMP was implemented in 2002. The methods used by the two surveys were the same, and the 2003–2005 survey attempted to replicate the survey gear used in 1974 as closely as possible. At some stations a benthic sled equipped with a camera and strobe light was towed along the bottom for 30–75 minutes, taking a picture every 10 seconds. The area of bottom illuminated in the images was estimated, and counting the red crabs visible in each image, then extrapolating thatnumber out to the area of the survey resulted in an estimate of swept-area abundance. At other stations, a trawl net was deployed to determine red crab sex ratios, weights and length frequencies by sex. Mean weights were used to estimate swept-area biomass.

After the second survey, it was possible to see the effect the fishery had on red crab population structure. As would be expected, the number of large male crabs was reduced. However, the number of smaller males and females appeared to have increased. After the results of the second survey were presented at the Northeast Data Poor Stocks Working Group in 2008 (citation below), the potential for sperm limitation in the red crab population caused concern. It was noted in the mating studies and seen in the benthic sled images that male red crabs needed to be about 50% larger than the females to be able to mate, since the male carries the female underneath his body during the process. If the larger males were being removed from the population, there was concern that the largest females with the highest repoductive potential would not be able to find mates and recruitment would diminish.

There have been efforts to estimate MSY for red crab, most recently also at the Northeast Data Poor Working Group meeting. The group tried several models and methods at the time, and reviewed estimates that had taken place in the past. Specifically the models were the depletion-corrected average catch (DCAC) model and the two-point boundary model. The DCAC model works from the idea that for a new fishery, in calculating average catch for an estimate of sustainable yield, the windfall at the start of a fishery will represent a certain number of units of sustainable yield and increase beyond the actual number of years of fishing, with that increase being based on an estimate of natural mortality. The two-point boundary model estimates the average recruitment needed to sustain the catch time series based on mature biomass at two points in time defined as the initial and final survey values, then estimates equilibrium catch based on recruitment, final biomass and survival. At the end of the day there was not one widely accepted value due to many factors, and the group determined there was substantial uncertainty in any potential BRP estimates that came out of the workshop based on current information about the stock. Survey-based BRPs were also relying on assumptions that the population was stationary and the fishing area did not extend beyond the surveyed area.

In terms of other effects of the fishery on the population, the percent mortality for discarded crabs is unknown. Crabs apparently survive being brought to the surface and returned to the bottom still contained in a special crab hotel at a high percentage, but method of discard (dropping vs. sliding) can increase mortality (Tallack 2007). Little is known about the survival of crabs discarded from a vessel that must descend down through hundreds of meters of water without protection after handling.

Since the FMP was put in place, the information originally available to assess the sustainability of the fishery has been total landings and port sampled carapace widths from which estimates of LPUE could be made and the size of landed crabs could be monitored. Estimates of LPUE used as a proxy for abundance can be uncertain in a pot fishery (see above). There are also economic and/or market factors that may affect culling patterns, fishing effort, location and catch for red crab. The fishery is market-driven and annual variations in landings reflect this.

In recent years (2016+) observer coverage of the red crab fishery has increased substantially and the data observed trips provide have allowed us to look at other aspects of red crab biology and thefishery. Since we never had information about undersized males that were discarded at sea, their numbers and length frequencies recorded by observers give us some information about recruitment. Since we never had information about females as they were all discarded at sea, their length frequencies and reproductive status (egg) data recorded by observers give us some information about the potential for sperm limitation. Most of the egg observations have been presence/absence but there have been some observers who stage the eggs, which determines if they are viable; important to note on the larger females especially where sperm limitation is a concern. The measurements of females, which have not had the same level of fishery removals as the males, have provided evidence that the crabs caught at the southern end of the fishery footprint reach greater maximum sizes than those caught in the more northern regions. Anecdotally, the red crabs found in the Gulf of Maine are the smallest, so there is likely a latitudinal size gradient.

Based on this information and previous recommendations, questions for research are listed below. They include the recommendations from the 2008 Northeast Data-poor Stocks Working Group report, past assessments, Tallack (2007) and Wahle et al. (2008).

• –
  What is the lifespan of a red crab?
• –
  How often do red crabs molt, and what is the percent increase in body size after a molt? Do red crabs, especially females, have a terminal molt? Molt frequency of females would be especially important too, as they mate during the molt period.
• –
  Do red crabs store sperm and if so for how long? Although some crabs are known to store sperm from a single mating to produce several clutches of eggs, it is unclear whether red crabs have this ability.
• –
  What percent of females with eggs are carrying viable eggs? Unmated females may produce clutches of unfertilized eggs.
• –
  How long do red crabs incubate their eggs? This would be useful to know in conjunction with molt frequency and sperm storage capability as it could relate to how many clutches of eggs a female could have in an intermolt period.
• –
  How many eggs are carried in a clutch and how does this vary by the size of the crab? Are there patterns that might indicate sperm storage?
• –
  Can we gather additional information on the relative sizes of mating pairs and any possible effects on reproductive potential? Could simulation modeling be used to explore the response of population sex ratios and size ratios to different fishing patterns? Is there a way to evaluate the importance of large male red crabs in reproduction considering the size distribution of females?
• –
  How can we design a successful tagging study to explore red crab growth rates, fishing mortality rates and molt frequencies in situ?
• –
  Are there any new, innovative ageing methods that might work for red crab?
• –
  What is the main food source for red crabs? Are they cannibalistic?
• –
  How much and when do red crabs move about their range? Is there a seasonal migration of one or both sexes?
• –
  Can we improve estimates of total discard mortality by considering seasonality, predation and displacement?
• –
  Could gear design studies help reduce discard and increase efficiency in the red crab fishery?
• –
  Could soak time studies help reduce uncertainty in  LPUE  as an estimate of abundance?
• –
  Traditional reference points for the deep-sea red crab stock are difficult to estimate due to lack of basic information. Are there non-traditional reference points that could be used to determine stock status? Could  BRP s based on size and sex ratios be useful due to the importance of preventing sperm limitation? It would require regular surveying.
• –
  How do economic factors alter the distribution and interpretation of fishing effort for the red crab fishery?
• –
  Can we collect biological data from a large number of red crabs over a wide area, crabs that have not been selected by a commercial trap (say, with a trawl) and are therefore more representative of the population at large? What questions could that help us answer: percentage of females with eggs, sperm storage, percentage in different shell conditions and states of molting, degree of segregation of the population by sex and size?
• –
  Would regular surveys of the red crab population provide useful information on the continuing effects of the fishery and the current population structure? What methods could be used? Can we get a better understanding of the habitat–abundance relationship for allocation of effort for any potential survey?

 None.

 For the estimation of biomass using NEFSC bottom trawl survey data, the most important source of uncertainty is the apparent productivity differences between the two intra-annual cohorts. The biomass of the cohort caught during the fall surveys is about five-fold higher than the biomass of the cohort caught in the spring surveys (NEFSC, 2011a; 2011b). However, the mean exploitation rate for the Jan.–June fishery was more than three times higher on the apparently less productive spring survey cohort. Using annualized biomass and exploitation rates to determine stock status does not account for these differences, and therefore, may impact resource sustainability. The review panel for the 2020 Level 3 assessment of this stock agreed and concluded that Annual averaging of the spring and fall survey biomasses assumes that a single population is being exploited and does not account for the large difference in apparent productivity of the two intra-annual cohorts. Cohort-specific stock size estimates and Reference Points are required to determine the stock status of cephalopod species with subannual lifespans because maturation and growth rates of intra-annual cohorts exhibit a high degree of seasonal and interannual variability (Arkhipkin et al. 2020). Because the generation time for longfin squid is only 6–8 months, overfishing of a single cohort potentially could jeopardize stock sustainability due to recruitment overfishing.

During the 2020 assessment, cohort-specific biomass Reference Points were derived separately for squid caught in the NEFSC spring versus fall bottom trawl surveys because annualized biomass estimates and Reference Points (i.e., averages of the spring and fall survey biomasses) do not account for the apparent productivity differences that exist between the two intra-annual cohorts caught during these surveys.

 These questions are not applicable to the subject assessment because an analytical model was not utilized.

Projections were not possible, because there is no anaytical model from which to do so. The stock is not currently subject to a rebuilding plan but has been in the past.

 The 2009–2020 NEFSC spring and fall survey biomass indices now include actual rather than nominal tow distances, at the request of the NTAP. This resulted in minor changes to the indices that did not affect stock statuses for either cohort. For example, the difference between the two-year moving average biomass estimates for 2019 during this assessment versus the 2020 assessment represented a 3.5% increase of 2,290 mt which is inconsequential in relation to the magnitude of the 2019 biomass estimate in relation to the threshold B_{MSY proxy}.

The other assessment methodology change involved the discard estimates for 2020–2022. Since the 2007 implementation of trimester-based quotas, discards have represented a minor portion of the catch; an average of 1.6% during 2007–2019. Nevertheless, the 2020–2022 discards were estimated using a new method with data from a new database, both of which differ from those used to estimate the 1989–2019 discards for the 2020 assessment. The new database, the CAMS database, was created and is maintained by GARFO. Some of the major differences between the two discard estimation methods and comparisons between the discard estimates for each method, during 2019–2021, are summarized in Tables 3 and 4 of the 2023 longfin squid tables file at SASINF (see2023_DORY_UNIT_TAB.pdf ). The 2019 landings were compared between the two discard estimation methods, as were the 2019–2021 discards, but only the CAMS discard estimates were available for 2022. The 2019 CAMS landings (12,489 mt) were only 31 mt (0.25%) higher than the 2019 landings from the Area Allocation database formerly used in the discard estimation method as the denominator (kept weight of all species). The higher 2019 CAMS discard estimate (357 mt, CV= 0.18) was more precise than the discard estimate based on the former discard estimation method (315 mt, CV= 0.32. However, only the CAMS discard estimates for 2020–2022 were used in this assessment, because 2019 was the terminal year of the previous assessment and the 2019 discard estimate was accepted for use in the stock status determination. When the 2019–2021 CAMS discard estimates were compared with those from the former discard estimationmethod, most of the CAMS estimates were slightly higher, but most fell within the 90% confidence interval of the discards estimated using the former method.

 Overfishing status is unknown for this stock because an analytical model could not be developed by the most recent benchmark assesment Working Group given the available data. There has been no change in the overfished status of longfin inshore squid since the 2020 operational assessment, for either of the two cohorts or for both cohorts combined (i.e., based on annualized biomass and biomass Reference Points). Under the current assessment process, however, stock status for this subannual species is always reported for the prior year (i.e., for multiple generations past) and this stock status is assumed to remain the same for the next three years when the next Management Track Assessment occurs. It is imperative that both squid stocks be assessed annually so that the timing of their stock status updates are appropriate for their short lifespans.

The use of conventional stock assessment models to assess squid stocks such as longfin inshore squid (Doryteuthis (Amerigo) pealeii), is inappropriate because they do not account for the species’ subannual lifespan and semelparous reproduction. This neritic species has a lifespan of 6–8 months, and like many other loliginids, there are two dominant intra-annual cohorts; winter-hatched and summer-hatched cohorts that have different growth rates and median sizes-at-maturity (Brodziak and Macy 1996; Macy and Brodziak 2001). Length-based models are not generally used to assess squid stocks because time-consuming, expensive aging (counting of daily increments in statoliths) must be used to identify the intra-annual cohorts due to the high plasticity in individual growth rates (Arkhipkin et al., 2015). Like most squid stocks, stock size estimates of longfin inshore squid exhibit high interannual variability because each year, the population relies on new recruits to each intra-annual cohort, and recruitment levels depend on the favorability of environmental conditions (Brodziak and Hendrickson 1999; Hatfield et al. 2001).

 Based on the NRCC assessment schedule, this stock is currently assessed every three years. Instead, it is recommended that this subannual species be asssessed on an annual basis, as was done historically. A Research Track Assessment is scheduled for review in 2026, following two years of research dedicated to developing a stock assessment model and Reference Points that account for the life history of this semelparous species and its two intra-annual cohorts. The research topics identified here and in previous assessments should be reviewed and prioritized prioritized by the Working Group prior to conducting any research on the stock.

During the upcoming longfin squid Research Track Assessment, the top priority should be the development of a cohort-specific assessment model that incorporates, for example, their different growth rates and median ages-at-maturity, along with spawning and non-spawning spawning natural mortality rates.

Simulation testing should be conducted to evaluate the model’s ability to accurately and precisely estimate stock conditions under a wide range of scenarios, including process and measurement errors.

A method that accounts for the species’ semelparous reproduction should be used to estimate separate MSP based Biological Reference Point proxies for each of the two intra-annual cohorts to ensure adequate spawner escapement.

Annual pre-season biomass estimates for each cohort will be necessary to set cohort-specific Annual Biological Catches and quotas. This requires a streamlined regulatory process that allows for rapid implementation of the cohort-specific quotas in order to avoid foregone yield during years of high stock size and to avoid recruitment overfishing of either cohort during years when stock size is low.

Extend the work that has been conducted on estimation of the NEFSC bottom trawl catchability of this species in both the spring versus fall bottom trawl surveys (NEFSC 2011a; Jacobson et al. 2015). Additional empirical data for estimating these catchabilities are needed to investigate the apparent productivity differences between the two cohorts caught in the NEFSC spring versus fall bottom trawl surveys. Biomass estimates for the fall NEAMAP surveys were only a small percentage of the total fall biomass, averaging 5.5% during 2009–2022, but a comparison of longfin squid catchability differences between the NEAMAP and NEFSC fall surveys might also be useful.

Life history, life history, life history!

Declining trends in growth rates and changes in the sex-ratio at age may change the productivity of the stock and in turn affect estimates of the biological reference points. Changes in growth, maturity, and recruitment may be environmentally mediated but mechanisms are unknown.

 The 7-year Mohn’s ρ, relative to SSB, was 0.03 in the 2021 assessment and was 0.06 in this assessment. The 7-year Mohn’s ρ, relative to F, was 0.01 in the 2021 assessment and was 0.03 in this assessment. No retrospective adjustment of SSB or F in 2022 was required.

Population projections for Summer flounder are reasonably well determined.

 No major changes, other than the addition of three years of data, were made to the Summer flounder assessment for this update. Minor changes to the survey input CVs and fishery and survey input Effective Sample Sizes improved model diagnostics but had limited affects on the model results.

Overfishing status has changed since the last assessment for Summer flounder. The stock status remains as not overfished but overfishing is occurring.

The current fishing mortality rate is near the threshold, and so recent near-average recruitment has resulted in relatively stable SSB. SSB is projected to remain relatively stable in the short term at current fishing rates.

The Summer flounder assessment could be improved with more intensive and comprehensive sampling of the fishery catch by sex.

Sufficient length and age sampling of the fishery catch needs to be maintained.

Declining trends in growth rates and maturity at age may change the productivity of the stock and in turn affect estimates of the biological reference points. Changes in growth, maturity, and recruitment may be environmentally mediated but mechanisms are unknown.

 The 7-year Mohn’s ρ, relative to SSB, was −0.14 in the 2021 assessment and was −0.21 in this assessment. The 7-year Mohn’s ρ, relative to F, was 0.20 in the 2021 assessment and was 0.42 in this assessment. There was a major retrospective pattern for this assessment because the ρ-adjusted estimates of 2022 SSB (SSB_ρ= 193,087) and 2022 F (F_ρ= 0.098) were outside the approximate 90% confidence regions around SSB (131,720–192,050) and F (0.14–0.208). A retrospective adjustment was made for both the determination of stock status and for projections of catch and biomass in 2024 and 2025. The retrospective adjustment changed the 2022 SSB from 159,050 to 193,087 and the 2022 F_Full from 0.171 to 0.098.

Population projections for Scup are reasonably well determined.

 No major changes, other than the addition of three years of data, were made to the Scup assessment for this update. Minor changes to the survey input CVs and fishery and survey input Effective Sample Sizes improved model diagnostics but had limited affects on the model results.

As in recent assessments for Scup the stock status remains as not overfished and overfishing not occurring.

The current fishing mortality rate is relatively low, but recent below average recruitment has resulted in a decrease in SSB. SSB is projected to continue to decrease in the short term.

The Scup assessment could likely be improved with more intensive sampling of the fishery catch.

Sufficient length and age sampling of the fishery catch needs to be maintained.