Reference Document Home | Publications Home
The Fishery
Research Surveys
Yield per Recruit

Replacement Yield Model

Biological Reference Points
Stock Status

Rebuilding Projections

Sources of Uncertainty

Northeast Fisheries Science Center Reference Document 09-08

The 2008 Assessment of Atlantic Halibut in the Gulf of Maine-Georges Bank Region

Laurel A. Col and Christopher M. Legault
National Marine Fisheries Service, Woods Hole Lab., 166 Water St., Woods Hole MA 02543-1026

Web version posted May 22, 2009

Citation: Col LA, Legault CM. 2009. The 2008 Assessment of Atlantic Halibut in the Gulf of Maine-Georges Bank Region. US Dept Commer, Northeast Fish Sci Cent Ref Doc. 09-08; 39 p. Available from: National Marine Fisheries Service, 166 Water Street, Woods Hole, MA 02543-1026.

Information Quality Act Compliance: In accordance with section 515 of Public Law 106-554, the Northeast Fisheries Science Center completed both technical and policy reviews for this report. These predissemination reviews are on file at the NEFSC Editorial Office.

Acrobat Download complete PDF/print version


Atlantic halibut were fished heavily during the mid-1800s to early 1900s, and due to current low population sizes there are minimal data on halibut in the Gulf of Maine-Georges Bank region.  Previous assessments have relied on Northeast Fisheries Science Center (NEFSC) autumn bottom trawl indices to provide relative measures of population abundance (NEFSC 2001; Brodziak 2002; Brodziak and Col 2005).  However, during the 2008 Groundfish Assessment Review Meeting (GARM) (NEFSC 2008), a simple production model, termed Replacement Yield Model here, was accepted as an appropriate method to assess Atlantic halibut.  The fishing mortality reference point proxy (0.073) from the re-estimated yield per recruit analysis was used to inform the intrinsic rate of growth for the Replacement Yield Model, and the model was tuned to the NEFSC autumn survey swept-area biomass index.  The resulting fishing mortality (0.065) was slightly below the FMSY proxy, indicating that overfishing is not occurring for Atlantic halibut.  On the other hand, the resulting biomass (1300 mt) from the Replacement Yield Model was well below both the BMSY proxy (49000 mt) and ½ BMSY proxy (24000 mt), indicating that Atlantic halibut continues to be in an overfished condition. 

Although there is a great deal of uncertainty associated with using a simple production model like the Replacement Yield Model, all of the available data for Atlantic halibut were incorporated, and the model was accepted as the best available science at the time of the 2008 GARM meeting (NEFSC 2008).  The Replacement Yield Model utilized the extensive time series of commercial halibut catch without requiring the use of age data or survey indices that matched the full time series of the commercial fishery data set.  This was particularly useful for Atlantic halibut where the early years of catch data provide information on the potential maximum biomass levels whereas the NEFSC survey time series started roughly 80 years after the collapse of the commercial halibut fishery.  The Replacement Yield Model provides annual estimates of relative F, biomass, and replacement yield in addition to revised biomass reference points, enabling overfishing status to be determined for the first time. 

Two major considerations for Atlantic halibut are minimum size regulations and transboundary movements.  A growth and maturity study by Sigourney et al. (2006) estimated the L50 for females to be 103 cm which is well above the current minimum size limit (91 cm).  An increase in the minimum size limit may increase the survival of halibut to spawning size, however greater compliance with minimum size regulations would need to occur for the benefits to be realized.  Additional information from recent Atlantic halibut longline tagging studies (Kanwit 2007) has indicated transboundary movement rates of 33% from US to Canadian waters.  This indicates that in the future, Atlantic halibut should be assessed as a transboundary US/Canadian stock, and that further work should be conducted to clarify stock boundaries.


Atlantic halibut (Hippoglossus hippoglossus) is the largest species of flatfish in the northwest Atlantic Ocean.  It is a long-lived, late-maturing species distributed from Labrador to southern New England (Bigelow and Schroeder 1953).  Atlantic halibut within the Gulf of Maine-Georges Bank region (NAFO Divisions 5Y and 5Z, Figure 1) have been exploited since the early 1800s, with major abundance declines noted as early as the 1870s (Goode 1884, Hennemuth and Rockwell 1987, Grasso 2008).  Grasso (2008) provides a comprehensive summary of the historical fishery, and explains that prior to 1836, halibut were incredibly abundant and were largely discarded due to the meat being poor for salting.  During this time halibut were considered to be an unpalatable species, a substantial inconvenience and even a danger to cod fishermen using small dories, and a voracious predator of the cod that they targeted.  A small market occurred for halibut fins prior to the 1830s, which gradually became a delicacy as cod declined.  During the late 1830s-1840s improvements in icing techniques and refrigeration technology led to many fleets converting from salt to ice to preserve catch, and halibut was found to be ideal for preserving on ice.  This, along with a cultural shift of preferring fresh to salted fish led to the height of the halibut fishery during the 1840s to 1870s.  The height was quickly followed by a series of local depletions.  This was due to general overfishing as well as recruitment overfishing since small halibut were heavily targeted, and halibut reach marketable sizes well before they reach maturity.  During the 1880s halibut was considered to be commercially extinct as populations had declined to low levels from southern New England north to Greenland.  Halibut suffered further declines during the 1930s due to their susceptibility to trawl gear at young ages (Figure 2).  It was not until 1999 that halibut were specifically managed under the Northeast Multispecies Fisheries Management Plan (Amendment 9) which limited the landing of Atlantic halibut in Federal waters to one halibut per trip, with a minimum size limit of 91 cm.

In previous index-based assessments (NEFSC 2001; Brodziak 2002, Brodziak and Col 2005), Northeast Fisheries Science Center (NEFSC) autumn weight per tow survey indices were expanded to swept-area biomass estimates, and the 5-year average biomass index was compared to BMSY proxy reference points for status determination (Figure 3).  Reference points for Atlantic halibut were originally determined by the New England Fisheries Management Council (Applegate et al. 1998) using Canadian Atlantic halibut length-weight equations (McCracken 1958) and von Bertalanffy growth curves (Nielson and Bowering 1989) to perform yield per recruit (YPR) and biomass per recruit analyses.  Natural mortality was assumed to be 0.1, and a Maximum Sustainable Yield (MSY) proxy was chosen to be 300 mt, yielding a BMSY proxy = 5400 mt, a ½ BMSY proxy = 2700 mt, and an FMSY proxy (threshold) = F0.1 = 0.06.  Based on the Groundfish Assessment Review Meeting (GARM) 2005 assessment of Gulf of Maine-Georges Bank Atlantic halibut, the stock was overfished (B2004 was 5% of BMSY proxy) and it was unknown whether overfishing was occurring (Brodziak and Col 2005). 

In the Atlantic halibut assessment presented here, NEFSC survey and commercial fishery data were updated through 2007 and estimates of discards from the United States (US) commercial fishery were included in total catch estimates to reflect the GARM Data Meeting recommendations.  Reference points were re-evaluated by updating YPR analyses using recent estimates of growth (Sigourney 2002) and maturity parameters (Sigourney et al. 2006).  The resulting FMSY proxy was used to define the intrinsic rate of growth in a Replacement Yield Model as recommended by the GARM Biological Reference Points meeting panel (O’Boyle 2008b).  The Replacement Yield Model incorporates the entire time series of catch data, tunes to the autumn survey swept-area biomass index, and results in BMSY and MSY proxy reference points, as well as annual estimates of biomass and relative fishing mortality.


Consistent records of Atlantic halibut landings from the Gulf of Maine-Georges Bank region (Statistical Areas 511-515, 521-522, 525-526, 561-562) began in 1893 (ICNAF 1952, Table 1, Figure 2).  Current US landings were extracted from the NEFSC commercial fisheries database (CFDBS) AA tables, and current Canadian landings (Division 5Zc) were extracted from the NAFO 21A database[1].  Historical distant water fleet landings are also included from 1962-1974 (Table 1). Landings have continued to decrease since the 1890s as components of the resource have been sequentially depleted.  Annual landings averaged 663 mt between 1893 and 1940, declined to an average of 144 mt during 1941-1976, and declined further to an average of 91 mt during 1977-2000 (Table 1, Figure 2).  Total reported commercial landings of halibut increased somewhat from record lows of 17-20 mt during 1998-2000 to 52 mt in 2007.  Of the 2007 landings, 22 mt (42%) were landed by US fishermen and 30 mt (58%) were landed by Canadian fishermen (Table 1, Figure 4). 

US discards from the Northeast Fisheries Observer Program database were estimated for the period 1989 to 2007 based on the Standardized Bycatch Reporting Methodology combined ratio estimation (Wigley et al. 2007).  The 1999 implementation of a one halibut per trip limit as well as a 91 cm minimum retention size increased the discard to kept ratio from 17% during 1989-1998 to 147% during 1999-2007 (Table 2, Figure 4).  Due to the low occurrence of Atlantic halibut in the observer database, the 1989-1998 average discard ratio was applied to the landings from 1893 to 1998 and the 1999-2007 average discard ratio was applied to landings in those years (Table 1, Figure 2).  Discards were not estimated by gear due to low encounter rates and high variability, however it should be noted that primary gear types for halibut landings have changed over time.  Including US discards, total catch increased from 18 mt in 1998 to 84 mt in 2007 (Table 1, Figure 4).  Canadian discard estimates were not available.


The NEFSC spring and autumn bottom trawl surveys provide measures of relative abundance of Atlantic halibut within the Gulf of Maine-Georges Bank region (offshore survey strata 13-30 and 36-40, Table 3, Figure 5).  Both indices have high interannual variability since the surveys capture low numbers of halibut, and in some years there are no halibut caught (Figure 6), indicating that halibut abundance is close to being below the detectability levels of the surveys.  The autumn survey biomass and abundance indices show little contrast or trend (Figures 7a and b), whereas the spring survey biomass and abundance indices (Figures 7a and b) suggest a relative increase during the late 1970s to early 1980s, a decline during the 1990s, and an increase since the late 1990s.  It is unknown whether these trends in survey indices for the Gulf of Maine-Georges Bank region have been influenced by changes in the seasonal distribution and availability of Atlantic halibut.  There is some evidence of environmental forcing in the spring survey indicated by a negative correlation with spring bottom water temperature anomalies (Figure 8).  However, the differences in trends between the two surveys also likely reflect the high variability due to low encounter rates of halibut in the NEFSC surveys. 

Due to the lack of alternative population estimates, the expansion of autumn survey indices to swept-area biomass has been used to estimate Atlantic halibut biomass and compared to previous MSY reference points for status determination in previous assessments (Figure 3).  Relying entirely on surveys for stock status determination is not ideal for most species.  However, this method is particularly problematic for Atlantic halibut since the NEFSC autumn bottom trawl survey started roughly 80 years after the fishery collapsed, and encounter rates of halibut in consistently sampled survey strata are very low (Figure 5 and Figure 6).  In previous assessments a survey catchability coefficient of one was assumed for swept-area biomass estimates, which is likely high.  There have also been changes in doors, nets and vessels throughout the time series which may affect catchability of Atlantic halibut over the time series.  Since the surveys encounter so few halibut, conversion factors have not been estimable.  The inability to calculate conversion factors for halibut will become a much greater problem in 2009 when the NEFSC survey will change to the R/V Henry Bigelow, which is likely to have vastly different catchabilities than the R/V Albatross IV or Delaware II for most species.


Currently the NEFSC does not age Atlantic halibut samples from either the commercial fishery or NEFSC bottom trawl surveys.  Fortunately age samples from the experimental halibut lonline fishery (Kanwit 2007) and NEFSC spring and autumn bottom trawl surveys through 2000 have been aged for growth analyses (Sigourney 2002), and maturity analyses (Sigourney et al. 2006).  This information was used along with NEFSC length and weight data to update yield per recruit (YPR) analyses for Atlantic halibut.

Combined years (1992-2007) of NEFSC spring and autumn length and weight data over all strata were used to estimate length-weight parameters:

W = αLβ

 α was estimated to be 0.00415 (using cm and grams) and
 β was estimated to be 3.23040.

This length-weight relationship was very similar to the McCracken (1958) equation used for previous Gulf of Maine-Georges Bank Atlantic halibut reference point determinations (Applegate et al. 1998, Figure 9).  A von Bertalanffy growth equation was used to estimate length at age by sex for aged Atlantic halibut from NEFSC surveys and the halibut experimental longline fishery (Sigourney 2002).  The growth equation resulted in somewhat larger halibut at age than the previously used Nielson and Bowering (1989) equation, and confidence intervals reflect high uncertainty in the estimates (Figure 10).  However, this was considered to be the best available information for the Gulf of Maine-Georges Bank region by the August, 2008 GARM review panel (NEFSC 2008), and the length-weight equation was applied to the female lengths at age to determine weight-at-age inputs for YPR analyses (Table 4).

Maturity percentiles at age from Sigourney et al. (2006) were used to calculate a maturity ogive for female halibut (Figure 11):

S(a) = (1+e(-α -βa))-1

 a is age,
 β is a parameter assumed to be equal to (2ln3)/(L75-L25), estimated to be 0.518, and
 α is a parameter assumed to be equal to –βL50, estimated to be -3.778. 

The resulting maturity at age and weight at age were used in YPR analyses with a plus group for ages 41 to 50 (Table 4, Figure 12).  Sigourney et al. (2006) recorded halibut from the recent NEFSC survey time series up to age 40, and it is likely that larger halibut landed in the earlier part of the fishery time series were at least 50 years of age.  No estimates of natural mortality rates for Atlantic halibut or Greenland halibut are included in previous assessments (Brodziak and Col 2005, DFO 2006, DFO 2007, DFO 2008).  Pacific halibut have similar growth patterns and maximum age, and in recent reports, M was estimated to be 0.15 for Pacific halibut based on catch curve analysis and energetic models of growth and reproduction (Clark and Hare 2006).  Therefore M was assumed to be 0.15 for the Gulf of Maine-Georges Bank Atlantic halibut, however it should be cautioned that this estimate is somewhat higher than using maximum age as a proxy to estimate M (for –ln(0.05)/(max age of 50), M ~ 0.06). 

As in the previous reference point determination (Applegate et al. 1998) a knife-edge selectivity at age 4 (~60cm and 2.4kg) was used for YPR analyses.  Since Amendment 9 was implemented in 1999, regulations have prohibited landing halibut less than 91cm.  However there is evidence from Northeast Fisheries Observer Program data that smaller halibut are continuing to be landed (Table 5).  Kept halibut from observer data indicate that even after implementation, mean lengths of kept halibut generally ranged from 80-90cm (~ages 5.5-6.5), with minimum sizes of kept halibut generally ranging from 40-50cm (~ages 2.5-3.5, Table 5).  Discarded halibut mean lengths have ranged from 27-70cm (~ages 2-5), with minimum discard lengths generally ranging from 20-40cm (~ages 1-3, Table 5).  Survival of Atlantic halibut discarded from longline gear is estimated to be 77% whereas survival of discards from otter trawl gear was estimated to be substantially lower at 35% (Neilson et al. 1989).  Thus, selectivity of Atlantic halibut likely starts around age 2 (30cm) for bottom trawl gear, which corresponds to the size selectivity of other flatfish (NEFSC 2008), whereas selectivity from longline gear likely occurs at older ages around 6-7 years.  This disparity in gear selectivity should be researched further, however with limited data to compare the NEFSC survey Yankee 36 otter trawl gear to commercial fishing gear, age 4 was chosen as a reasonable midpoint for knife-edged selectivity (Table 4, Figure 12).  NFT YPR version 2.7.2[2] was used to perform the YPR analysis, which resulted in an F0.1 of 0.073.  This is slightly higher than the previous F0.1 of 0.06, using M = 0.1 (Applegate et al. 1998). 


Available models are limited for data poor species such as Atlantic halibut.  An age-structured production model as described in Brandao and Butterworth (2008b) was not considered to be a reasonable approach given the lack of available data (O’Boyle 2008a).  A simplistic LOSS model without constraining the intrinsic rate of growth to YPR output or tuning to survey catchability yielded a wide range of results with little information on which to inform model selection.  By using F0.1 to inform the intrinsic rate of growth in a Replacement Yield Model, and penalizing results that differed greatly from an assumed NEFSC autumn survey catchability, model results were considered to be more reliably estimated.  This approach also incorporated the most available data for Atlantic halibut and was recommended by the GARM Biological Reference Points review panel (O’Boyle 2008b). Therefore, the resulting FMSY proxy (F0.1) from the YPR analysis was used to inform the intrinsic rate of growth (defined as 2*F0.1 or 0.146) for the Replacement Yield Model.  Since Atlantic halibut catch predates reliable landings statistics beginning in 1893 (ICNAF 1953, Grasso 2008), a linear increase in catch was assumed from 1800-1893 following the advice of the August, 2008 GARM review panel (NEFSC 2008, Table 6).  Although this estimate is crude, it was considered preferable to assuming that 1893 biomass was representative of an unfished population and thus equal to carrying capacity.

A Replacement Yield Model similar to that described in Brandao and Butterworth (2008a) was used to provide annual estimates of biomass, replacement yield and fishing mortality.  In this model, estimated biomass is defined as:

By = B y-1 + R y-1 – Cy-1

 By is the biomass at the start of year y,
 By-1 is the biomass at the start of the previous year,
 R y-1 is the replacement yield in the previous year, and
 Cy-1 is the total catch in the previous year.

Replacement yield is defined as:

Ry = rBy (1- By /K)

 r is the intrinsic rate of growth, and
 K is the carrying capacity (assumed to be equal to the model estimated biomass in 1800).

The model was fitted to the NEFSC autumn survey swept-area biomass index, and the following negative log-likelihood (-lnL) was used to determine the model with the best estimates of carrying capacity and predicted survey catchability coefficient parameters:

-lnL = log (δ) + 0.5∑(ln(Iy)-ln(Byq)) 22 + p1 + p2

 δ is a constant,
 Iy is the swept-area biomass index in year y,
 q is the catchability of the NEFSC fall survey defined as the exponent of the average of
 p1 is the sum of the penalties for biomass going to the defined minimum boundary in a given year, and
 p2 is a penalty for the difference between the model-estimated q and the assumption that
  the NEFSC autumn survey q is 0.5

Replacement Yield Model Results

The estimated biomass for the Replacement Yield Model indicated a sharp decline from around 4,000-5,000 mt during the early 1900s to around 1,000 mt during the mid-1900s.  Atlantic halibut hit record low biomass levels of less than 500 mt during the 1990s and has since increased to 1,300 mt in 2007 (Table 6, Figure 13).  Relative F (catch/biomass) has been highly variable with spikes close to 0.7 in the late 1800s, and around 0.4 in 1940 and 1967.  However, relative F has been comparatively low since the mid-1990s, with a slight increase to 0.065 in 2007 (Table 6, Figure 14).  Replacement yield decreased sharply in the 1870s to a low of 500 mt in 1900, increased slightly to 700 mt around 1920, gradually decreased to 60 mt in the early 1990s, and is currently close to 190 mt (Table 6, Figure 15).

For the Replacement Yield Model, only the most recent 45 years can be included for residual pattern analyses, where survey swept-area biomass estimates are available.  The predicted survey index from the Replacement Yield Model is fairly flat compared to the noisy NEFSC survey index (Figure 16a).  The residuals for the Replacement Yield Model (Table 7, Figure 16b) indicate that there was minor patterning in the residuals, with the Replacement Yield Model slightly overestimating biomass during the mid-1960s and greatly underestimating biomass in some other years due to the high variability in the autumn survey index.  However there are no periods of consistently strong residual patterns.

Sensitivity Analyses for Replacement Yield Model

Two sensitivity analyses were run for the Replacement Yield Model based on panel recommendations from the GARM Biological Reference Points meeting (O’Boyle 2008b).  The first was to test using a parabolic increase of catch instead of a linear increase to represent 1800-1892 catch in the Replacement Yield Model.  The resulting biomass estimates were essentially identical using either method, indicating that the Replacement Yield Model was not highly sensitive to the method of estimating historic catch (Figure 17a).

The second sensitivity recommended by the review panel (O’Boyle 2008b) was to test various natural mortality rates for Atlantic halibut in the Replacement Yield Model based on published values from halibut assessments in other regions.  No alternative natural mortality rates were available from published assessments.  However, three natural mortality rates were tested in the YPR analyses to generate three F0.1 estimates used to determine the intrinsic growth rates in Replacement Yield Models.  The natural mortality estimate of 0.15 was the preferred M since this was based on current Pacific halibut estimates (Clark and Hare 2006), resulting in F0.1 = 0.073.  A natural mortality estimate of 0.10 was tested since this was used in the previous YPR analysis for Atlantic halibut (Applegate et al. 1998), resulting in F0.1 = 0.053.  However, it should be noted that the natural mortality used for the 1998 YPR analysis was simply based on Pacific halibut assessments at that time (Applegate et al. 1998).  Finally, a natural mortality estimate of 0.08 was tested based on a maximum age of 40 years, resulting in F0.1 = 0.046.  FMSY proxies based on YPR analyses with alternative estimates of M are presented below with the resulting biomass reference points, Maximum Sustainable Yield, current relative fishing mortality, and current biomass estimates from the Replacement Yield Model.

Text Table 1

As natural mortality rates decreased, resulting biomass from the Replacement Yield Model increased (Figure 17b).  The reference point tables above indicate that biomass reference points from Replacement Yield Models also increased with decreasing natural mortality rates.  Although initially counter-intuitive, this was due to defining the intrinsic rate of growth in the Replacement Yield Model as being 2*FMSY proxy from the YPR analysis.  As the intrinsic growth rate decreased with M, carrying capacity (estimated biomass in 1800) and thus biomass reference points had to be increased in the Replacement Yield Model in order to keep biomass from decreasing to zero over the time series of the catch.  Since biomass reference points increased proportionally with biomass, all sensitivity runs for natural mortality rates resulted in current biomass levels of 5-6% of ½ BMSY proxies.  Therefore, the review panel (NEFSC 2008) concluded that the natural mortality estimate essentially was a scaling mechanism, and in relation to biomass reference points, was not highly influential on the assessment results.


The review panel (NEFSC 2008) concluded that the fishing mortality reference point based on M = 0.15 for YPR analyses was most appropriate since the Pacific halibut assessment is the only halibut assessment that assumes a natural mortality rate based on empirical research (Clarke and Hare 2006).  Biomass reference points were based on Replacement Yield Model estimated carrying capacity (97,000 mt), which was informed by the FMSY proxy (F0.1) from the YPR analysis.  Target biomass (BMSY proxy) was defined as half of K and threshold biomass was equal to ½ of the BMSY proxy (note that only two significant digits were included for the ½ BMSY proxy to reflect input data).  A maximum sustainable yield was calculated as the FMSY proxy multiplied by the BMSY proxy from the Replacement Yield Model.  Therefore, the Atlantic halibut reference points as accepted by the 2008 GARM review panel were as follows:

FMSY                0.073
BMSY                49,000 mt
½ BMSY            24,000 mt
MSY                3,500 mt

In comparison to previous index-based assessments, BMSY, MSY and current biomass from all of the Replacement Yield Model scenarios are substantially higher since they include the implied higher biomass levels that enabled large amounts of catch in the late 1800s.  However, current biomass as a percent of the threshold is similar for the two methods.  Below are the biological reference points from the previous index-based assessment (Brodziak and Col 2005) and NEFSC survey swept-area biomass updated through 2007:

Text Table 2

Depletion-Adjusted Average Catch

Calculating a Depletion-Corrected Average Catch (DCAC) is an alternative method to estimate MSY for data poor stocks where a time series of catch is the most dependable data source (MacCall 2007).  Using NFT DCAC version[3], sensitivities were run to calculate average DCAC as a proxy for MSY using various time series of catch and assumptions for the relationship of FMSY to M (Figure 18).  The time series used for the DCAC model runs included the total time series of the available catch data (1893-2007), a period of steady decline in biomass (1893-1943), the period of sharpest decline in biomass (1895-1899), the entire time series used for the Replacement Yield Model (1800-2007), and the period from the start of directed fishing effort on Atlantic halibut to the first sharp decline in biomass (1830-1899).  As seen in both Figure 18a (using FMSY and M values consistent with the Replacement Yield Model) and Figure 18b (assuming that FMSY is roughly equal to M), average DCAC estimates for Atlantic halibut are substantially lower than the 3,500 mt MSY value estimated by the Replacement Yield Model.

            MacCall (2007) cautions that the yields calculated through the DCAC model are only sustainable if the biomass is at or above BMSY and that if the resource is below BMSY, the currently sustainable yield (Ycurrent) can be approximated using the following equation:

Ycurrent = Ysust(Bcurrent/BMSY)

where Ysust = (∑C)/(n + (DELTA/0.2*M))

where C = catch
            n = number of years of catch
            DELTA = the biomass at the beginning of the time series minus the biomass at the end of the time series

Using the entire time series of data in the Replacement Yield Model, and output of current biomass and BMSY estimates from the model: ∑C = 317,390 mt, n = 208, DELTA = 0.987, and Ycurrent = 35 mt.  Using the entire time series of recorded catch: ∑C = 45,665 mt, n = 115, DELTA = 0.098, and Ycurrent = 10 mt.  Since the Replacement Yield Model estimates MSY for a population at BMSY, this indicates that given the current low biomass of Atlantic halibut, the current yield should be much lower than the estimated MSY.


The review panel for the GARM 2008 assessment recommended going forward with using M = 0.15 for the YPR analysis, resulting in a FMSY proxy of 0.073.  Current relative F from the Replacement Yield Model was 0.065 in 2007, indicating that overfishing is not occurring for Atlantic halibut.  However, it should be cautioned that relative F is 89% of the proxy F threshold.  The 2007 estimated biomass from the Replacement Yield Model was 1,300 mt, or 5% of the biomass threshold, indicating that Atlantic halibut continues to be in an overfished condition (Figure 19).



In 2001 the Gulf of Maine-Georges bank portion of the Atlantic halibut stock was determined to be overfished (Brodziak 2002) and in 2004 Amendment 13 to the Magnuson-Stevens Fisheries Conservation and Management Act required a rebuilding plan for Atlantic halibut.  However, a trajectory for rebuilding could not be calculated until the acceptance of the Replacement Yield Model (NEFSC 2008) which resulted in the first fishing mortality estimates for Atlantic halibut.  Therefore, the rebuilding time period for Atlantic halibut was determined to be from 2004 to the estimated year in which halibut would rebuild to BMSY at F = 0, plus one mean generation time from the updated YPR analyses.  Based on panel recommendations (NEFSC 2008), projections were run for Atlantic halibut using the Replacement Yield Model, assuming M = 0.15 and a linear increase in catch from 1800-1893.  The resulting rebuilding time frame for Atlantic halibut was 2056, and Frebuild = 0.044.

There are a number of reasons to suggest that both the rebuilding time frame and the Frebuild are highly optimistic, the first being that the Replacement Yield Model assumes maximum growth rate of the population even at low abundance.  There are currently no indications that Atlantic halibut are either reproducing or growing at their maximum potential in the currently depleted state.  The second is that the Replacement Yield Model does not incorporate age structure.  This is of particular concern for Atlantic halibut since the mean age of maturity for females is 7.3 years (Sigourney 2006), creating both a lag time of initial response to management measures and a slower rebuilding trajectory which are not realized in the current projections.  The final source of concern for calculating rebuilding trajectories is that the currently assessed Gulf of Maine-Georges Bank component is likely a small portion of a larger US-Canadian Atlantic halibut stock (Kanwit 2007, see sources of uncertainty below).  This substantially increases uncertainty in the current projections since the Replacement Yield Model does not incorporate the entire dynamics of the stock.

The Frebuild for the current Replacement Yield Model is only slightly lower than the average model-estimated relative fishing mortality for the 1995-2007 period (0.052).  Under this Frebuild the projected biomass is estimated to roughly double over the next seven years and to continue with roughly exponential growth throughout the rebuilding time period (Figure 20).  This rate of increase has not been shown in the 200+ years of model estimated biomass and is thus unlikely to be biologically feasible.  Further, there are no indications in the NEFSC survey indices that significant recent increases in population abundance or biomass are occurring.  Therefore, both the rebuilding time frame and the Frebuild from the Replacement Yield Model are highly optimistic.

2009 Catch Estimates

Three scenarios of relative F in 2009 were calculated for Fstatus quo, FMSY and Frebuild.  In each case the catch in 2008 was set equal to the catch in 2007.  The results for 2009 catch estimates based on the three scenarios were as follows: Fstatus quo= 100 mt, FMSY = 112 mt, and Frebuild = 68 mt (Table 8).


Model Uncertainty

Limited biological data lead to uncertainty in growth and maturity at age estimates for the YPR analysis, although recent research and the experimental halibut fishery have allowed for updated estimates to be based on Atlantic halibut in the Gulf of Maine-Georges Bank region.  A lack of reported landings prior to 1893 lead to rough estimates of catch during 1800-1892; however, the Replacement Yield Model does not appear to be highly sensitive to these estimates.  A lack of available natural mortality estimates for Atlantic halibut necessitates the use of Pacific halibut estimates, and this leads to uncertainty in MSY and biomass reference point estimation in the Replacement Yield Model.  Previous estimates of MSY as well as MSY approximations from Depletion-Corrected Average Catch analyses indicate that MSY may be overestimated, and that given the current low biomass of Atlantic halibut, current yield should be substantially lower than MSY.  However, the resulting status of Atlantic halibut being near the overfishing level and far below ½ BMSY remains regardless of the assessment model or assumptions of M.  Another problem for any Gulf of Maine-Georges Bank Atlantic halibut model is providing informative tuning indices.  Although the NEFSC autumn survey swept-area biomass index has been considered to be the best available estimate of biomass in previous assessments, there is a great deal of uncertainty as to whether this index is reliable for detecting population biomass trends due to the low encounter rates of Atlantic halibut.  Finally, the greatest uncertainty with the Replacement Yield Model is the rebuilding projections.  These projections are likely to be highly optimistic, especially considering the model inclusion of maximum intrinsic growth despite recent low biomass.

Minimum Size Regulations

Atlantic halibut maturity at length estimations from Sigourney et al. (2006) indicate that the female L50 (103 cm) is well above the current minimum size limit (91 cm), and that recruitment overfishing is likely occurring.  Increasing the minimum size limit may increase survival of halibut to spawning size, especially in the longline fishery where discarded survival rates are higher than in the otter trawl fishery (77% and 35% respectively, Neilson et al. 1989).  However, regulation of the minimum size is a concern since it appears that even with the current regulations, there are undersized halibut being landed (Table 5).  In order to ensure that the fishery minimizes the take of immature Atlantic halibut and thus promotes rebuilding, size limits would not only need to be increased to 103 cm, but compliance with the minimum size limit would need to be increased.

Transboundary Movement

Another source of uncertainty is the stock boundary determination for Atlantic halibut.  For management purposes, the Gulf of Maine-Georges Bank region is considered to be a separate stock from Canadian Scotian Shelf-Southern Grand Banks and Gulf of St. Lawrence stocks.  However in 2007, Kanwit showed that substantial transboundary movements occurred in Atlantic halibut tagged off of the coast of Maine.  As of January, 2009, 1,547 Atlantic halibut had been tagged from the experimental halibut fishery (2000-2004), DFO longline surveys (2007-2008), and an ongoing voluntary Maine state waters tagging project (starting in 2002).  Of these tagged halibut, 227 tags had been returned, resulting in a 15% return rate, and 33% of the 195 tagged halibut with reliable release and recapture locations had crossed into Canadian waters (Figure 21).  By using the haversine great-circle method (Sinnot, 1984) to estimate the minimum distance traveled, we found that several individuals traveled over 1,000 km north-east to Newfoundland, the Gulf of St. Lawrence, and the Grand Banks.  The distribution of distances that tagged halibut traveled was skewed (Figure 22) and recaptures may be biased due to landings restrictions on halibut in US waters as well as to high rates of recaptures from the above mentioned programs (Kanwit, 2007).  However, there is strong evidence that Atlantic halibut are capable of both long distance movements and crossing US-Canada boundaries in substantial numbers.  The days at large for the tagged halibut varied (Figure 23), and no significant relationship was found between days at large and distance traveled (Figure 24), indicating that there may be both residential and migratory populations.  Further examination of season-specific distances traveled and directions of movement as well as length-specific distances traveled could be informative.  Although previous tagging results from Canada (McCracken 1958, Stobo et al. 1988) do not show substantial movements into US waters, it would be informative to update and expand on this work as well. Overall, the recent tagging results in Maine inshore waters clearly indicate trans-boundary movement of Atlantic halibut, and future assessments should consider combining the Gulf of Maine-Georges Bank region with Canadian stocks.


We would like to thank Dr. Jon Brodziak for his previous work on Atlantic halibut assessments and his insightful feedback on the 2008 assessment.  We would also like to thank Kohl Kanwit and the rest of the staff from the Maine Department of Marine Resources as well as the participating fishermen for their initiation and involvement in the experimental halibut fishery, the DFO longline surveys, and the voluntary Maine state waters tagging project.  The data provided by these sources have not only added to important growth and maturity estimates, but have provided a much needed insight into the movement patterns of Atlantic halibut.  We are also indebted to the participants and reviewers throughout each stage of the GARM process, who have provided constructive and valuable feedback on the Atlantic halibut assessment.  Finally, this assessment could not have been completed without the dedicated Ecosystems Survey Branch staff and all of the scientists and volunteers who have collected data on NEFSC surveys, the Northeast Fisheries Observer Program staff, the Data Management Systems staff, and all who participated in collecting commercial fisheries data from New England ports. 


Applegate A, Cadrin S, Hoenig J, Moore C, Murawski S, Pikitch E. 1998. Evaluation of existing overfishing definitions and recommendations for new overfishing definitions to comply with the Sustainable Fisheries Act. Final Report, Overfishing Definition Review Panel. New England Fishery Management Council, Newburyport, MA; 179 p.

Brandao A, Butterworth DS. 2008 a. A “replacement yield” model fit to catch and survey data for the south coast kingklip resource of South Africa . WG-Dem:K:05.

Brandao A, Butterworth DS. 2008 b. An updated assessment of the South African kingklip resource including some sensitivity tests.  WG-Dem:K:02

Brodziak J.  2002. Atlantic halibut. In NEFSC, Assessment of 20 Northeast groundfish stocks through 2001, pp.306-313. NEFSC Ref. Doc. 02-16, 509 p. Available at:

Brodziak J, Col L.  2005. Atlantic halibut. In: Assessment of 19 Northeast groundfish stocks through 2004. 2005 Groundfish Assessment Review Meeting (2005 GARM), Northeast Fisheries Science Center, Woods Hole, MA, 15-19 Aug. 2005. NEFSC Ref. Doc. 05-14; p 424-431. Available at:

Clark WG, Hare SR. 2006. Assessment and management of Pacific halibut: data, methods, and policy.  IPHC Sci Report No. 83.  Available at:

Collette BB, Klein-MacPhee G, Editors. 2002. Bigelow and Schroeder's Fishes of the Gulf of Maine (3rd Edition). Washington (DC): Smithsonian Institution; 748 p.

Department of Fisheries and Oceans (DFO). 2006. Atlantic halibut on the Scotian Shelf and Southern Grand Banks (Div. 3NOPs4VWX). DFO Can Sci Advis Sec Sci Advis Rep. 2006/038.

DFO. 2007. Stock assessment of Atlantic halibut of the Gulf of St. Lawrence (NAFO Divisions 4RST) in 2006. DFO Can Sci Advis Sec Sci Advis Rep. 2007/007.

DFO. 2008. Assessment of the Greenland halibut stock in the Gulf of St. Lawrence (4RST) in 2006. DFO Can. Sci. Advis. Sec. Sci. Advis. Rep. 2007/044.

Goode, G.B.1884. The fisheries and fishing industries of the United States . U.S. Commission of Fish and Fisheries. Section V, Vol. 1, Part 1: 1-119.

Grasso GM. 2008. What appeared limitless plenty: the rise and fall of the nineteenth-century Atlantic halibut fishery. Environ Hist. 13: 66-91.

Hennemeth RC, Rockwell S. 1987. History of fisheries conservation and management. In: Georges Bank. Backus R, Price R, Bourne D, eds. MIT Press, Cambridge MA; p 431-446.

International Commission for the Northwest Atlantic Fisheries (ICNAF). 1952. United States landings of groundfish from the convention area, 1893-1950. In: ICNAF annual meeting 1952. St. Andrews, New Brunswick, June 30 – July 10, 1952. Vol. 4, Doc. 7.

Kanwit JK. 2007. Tagging results from the 2000-2004 Federal Experimental Fishery for Atlantic halibut (Hippoglossus hippoglossus) in the Eastern Gulf of Maine. J Northw Atl Fish Sci. 38:37-42.

MacCall A. 2007. Appendix D – Depletion-Adjusted Average Catch. Report of the Lenfest Working Group on Annual Catch Limits, p 27-32.

McCracken FD. 1958. On the biology and fishery of the Canadian Atlantic halibut, (Hippoglossus hippoglossus L.). J Fish Res Bd Canada . 15(6):1269-1311.

Neilson JD, Bowering WR. 1989. Minimum size regulations and the implications for yield value in the Canadian Atlantic halibut (Hippoglossus hippoglossus) fishery. Can J Fish Aquat Sci. 46: 1899-1903.

Neilson JD, Waiwood KG, Smith SJ. 1989. Survival of Atlantic halibut (Hippoglossus hippoglossus) caught by longline and otter trawl gear. Can J Fish Aquat Sci. 46: 887-897.

Northeast Fisheries Science Center (NEFSC). 2001. Assessment of 19 Northeast groundfish stocks through 2000. NEFSC Ref Doc. 01-20; 217 p.

NEFSC. 2008. Assessment of 19 Northeast groundfish stocks through 2007: report of the 3rd Groundfish Assessment Review Meeting (GARM III), Northeast Fisheries Science Center, Woods Hole, MA, August 4-8, 2008.  US Dep Commer, NOAA Fisheries, Northeast Fish Sci Cent Ref Doc. 08-15; 884 p + xvii.  Available at:

O’Boyle R, Chair. 2008a. Part 2. Assessment Methodology (Models). In: Appendix to the Report of the 3rd Groundfish Assessment Review Meeting (GARM III): assessment of 19 Northeast groundfish stocks through 2007. Northeast Fisheries Science Center, Woods Hole MA, August 4-8, 2008.  US Dep Commer, NOAA Fisheries, Northeast Fish Sci Cent Ref Doc. 08-16; 1056 p.

O’Boyle R, Chair. 2008b. Part 3. Biological Reference Points.  In: Appendix to the Report of the 3rd Groundfish Assessment Review Meeting (GARM III): assessment of 19 Northeast groundfish stocks through 2007. Northeast Fisheries Science Center, Woods Hole, MA, August 4-8, 2008.  US Dep Commer, NOAA Fisheries, Northeast Fish Sci Cent Ref Doc. 08-16; 1056 p.

Sigourney DB. 2002. Growth, sexual maturity and distribution of the Atlantic halibut (Hippoglossus hippoglossus L.) in the Gulf of Maine-Georges Bank region. M.Sc. Thesis, Univ. Mass. Amherst, Amherst, MA.

Sigourney DB, Ross MR, Brodziak J, Burnett J. 2006. Length at Age, Sexual Maturity and Distribution of Atlantic Halibut, Hippoglossus hippoglossus L., off the Northeast USA.  J Northw Atl Fish Sci. 36:81-90.

Sinnott KW. 1984. Virtues of the haversine. Sky and Telescope. 68(2):159.

Stobo WT, Neilson JD, Simpson PG. 1988. Movement of Atlantic halibut (Hippoglossus hippoglossus) in the Canadian North Atlantic. Can J Fish Aquat Sci. 45: 484-491.

Wigley SE, Rago PJ, Sosebee KA, Palka DL.  2007. The analytic component to the Standardized Bycatch Reporting Methodology omnibus amendment: sampling design and estimation of precision and accuracy. NEFSC Ref Doc. 07-09; 156 p.


[2] NOAA Fisheries Toolbox Version 3.0, 2008.  Age Based Yield per Recruit Version 2.7.2

[3] NOAA Fisheries Toolbox Version 3.1, 2009.  Depletion Corrected Average Catch Version
NMFS Search
Link Disclaimer
Privacy Policy
(File Modified Jul. 01 2016)

This page has had 2 visits today, 22 visits this week, 74 visits this month, 825 visits this year