Skip Maine state header navigation

State Search:

Agencies | Online Services | Help

Page Tools Page Tools Email page Watch page Add link to MyMaine Map addresses En español En français English

Project Rational 

Atlantic herring (Clupea harengus) is one of the most biologically and economically important fish species in the western Atlantic.  With an estimated complex-wide biomass of 1.8 million metric tons (Overholtz et. al., 2004), herring provide a significant forage base for other fish species, marine mammals, and birds (NEFMC, 2003b), as well as supporting the second largest commercial fishery on the east coast (NMFS 2002).  Atlantic herring landings in 2003 were reported at 100,677 mt or approximately 222 million pounds, with an estimated value in excess of $15.5 million (NEFMC 2004).  In addition to the direct economic contribution of herring landings, this fishery supports a domestic value added industry (canned sardines and frozen whole fish) worth approximately $50 million and the North Atlantic lobster fishery estimated at $260 million. 

The New England Fisheries Management Council (NEFMC) and the Atlantic States Marine Fisheries Commission (ASMFC) manage Atlantic herring cooperatively on a meta-stock basis.  This management strategy was originally adopted due to a lack of scientific data supporting the assumed theory of distinct spawning stocks.  Anecdotal information and exploratory scientific studies (i.e. morphometrics and spatial analysis of recolonization) have lead most biologists to conclude that sub-stocks exist (Armstrong and Cadrin, 2001; Overholtz, 2002).  However, the current lack of understanding regarding the intermixing of components in the fishery precludes separate assessment and management of the stocks.  Therefore, the entire complex is assessed as a whole with subsequent consideration of the individual components (Overholtz et. al., 2004).

Assumptions of stock intermixing play an important role in the allocation of the resource under the two existing Fishery Management Plans (FMPs).  NEFMC and ASMFC both employ stock mixing ratios based partially on 20-30 year old tagging data to determine annual Total Allowable Catch (TAC) values in each of the four management areas.  Although past and present stock assessment models estimate biomass for the entire herring complex, the inshore stock component is considered the limiting factor (NEFMC 2003a).  In recognition of this and the general uncertainty regarding the status of the inshore stock, a risk assessment approach to determining the TACs is being proposed through Amendment 1 to the NEFMC Atlantic sea herring FMP.  The tagging project is capable of contributing current data to the risk assessment process by updating stock intermixing rates and migration patterns.  This information will contribute to scientifically based determinations of the area TACs and result in better management of the herring fishery. 

Background
ME DMR initiated an Atlantic herring coded-wire tagging study in 2001.  Although field-tagging methods were very successful leading to high rates of tagging, the project eventually had to be abandoned.  Recoveries could only be made at facilities with automated detectors, and the detectors were extremely expensive to purchase and not well suited to a processing plant environment.  DMR, with the cooperation of the herring industry evaluated and redesigned the tagging project to employ anchor tags and started fieldwork again in the spring of 2003.  Industry members almost exclusively supported the project in its first year through in-kind contributions and cash donations, collectively representing about $75,000.  Due to the success of the anchor tagging project in 2003, efforts were continued in 2004 again through the support of the industry. 

Study Design 
Tagging events and tag returns are categorized in the context of spatial and temporal strata in this study.  These spatial strata were initially identified with the cooperation of the herring industry and from a review of previous tagging studies.  There are five major spatial strata: Nova Scotia, New Brunswick, the Gulf of Maine, Georges Bank, and Southern New England (Figure 1).  Nova Scotia and New Brunswick are included because previous tagging studies indicated intermixing between herring on both sides of the international boundary.  Additionally, several returns from fish tagged in the Gulf of Maine and Southern New England in 2003 and 2004 were caught in Canadian waters illustrating the importance of DMR and DFO pursuing this work collaboratively.  The Canadian Department of Fisheries and Oceans (DFO) is currently tagging herring from weirs in New Brunswick and investigating the possibility of tagging on purse seiners in Nova Scotia.   

Figure 1: Spatial strata designated for the Atlantic herring tagging project

The four temporal strata identified for this study are largely based on previous tagging work (Creaser et. al., 1984 and Overholtz et. al., 2004), the existing NEFMC Atlantic Sea Herring Fishery Management Plan, and information provided by industry members.  These strata are: spring migration (May-June), summer feeding/spawning (July-October), fall migration (November-December), and winter feeding (January-April).    

The major focus of this study is tagging herring in the Gulf of Maine during the summer feeding/spawning temporal stratum and tagging in southern New England during the winter feeding stratum.   

Field Methods
Tagging events in 2004 occurred on both commercial purse seine and contracted midwater trawl vessels.  Tagging that took place on commercial purse seine vessels used a modified xactix box as a holding tank and a submersible pump designed to supply a constant flow of fresh sea water (Figures 2A and 2B). The herring were captured using a dip net while they were freely circling the perimeter of the seine, prior to the pocket being formed. 

Figure 2A: Hauling back the seine     Figure 2B: Xactix holding tank

                                    

Herring were also captured with an aquarium codend that was laced into the commercial net on a midwater trawler (Figures 3A and 3B).  The specific depths and durations of the tows depended on the fishing conditions, but tow times were kept as minimal as possible (10-15 mins.).  Regulated seawater flow was introduced into the holding tank of the aquarium codend after each set was hauled back.  Tagging occurred directly from the holding tank to minimize handling of the herring.    

Figure 3A: Aquarium codend (AQCE)      Figure 3B: Deploying AQCE

                       

Tagging personnel used bare hands to handle the fish during the tag application procedure to minimize scale loss (Figures 4A and 4B).  All herring were selected for fitness and only tagged if they had an excess of 80% scale coverage.  Fish judged to be missing more than 20% of their scales were discarded without a tag.   

Figure 4A: Tag placement                         Figure 4B: Tagging from AQCE

            

Predation was minimized by fishing at night when possible and releasing tagged herring under a protective covering.  In the spring of 2004 a method was developed for releasing the tagged herring under a tarp off the stern of the participating fishing vessel (Figure 5A).  This simple solution proved extremely successful in eliminating avian predation (Figure 5B), by providing cover while the fish escaped to depths in excess of 20+ feet.  Additionally, the pink color of the anchor tags (reds diffuse at very shallow depths) was specifically selected to reduce underwater predation by marine mammals and other fish species while affording maximum visibility after harvesting. 

Figure 5A: Protective tarp                       Figure 5B: Northern Gannet

           

Returns
Reward posters, tag return forms, and brief project descriptions, designed to inform the fishermen about the project were hand delivered to primary bait dealers throughout the coast of New England. Reward posters were also distributed by DMR personnel to fishery related businesses and other waterfront locations.  An extensive outreach effort targeting primary lobster dealers began in the fall of 2004 after several reports of lobstermen finding but not reporting tagged fish in their bait.  Information packages including posters and mailers were delivered to 85 lobster dealers from Kittery to Stonington, Maine.  This effort will continue in the spring of 2005 and cover all primary lobster dealers between Stonington and the Canadian border along with major dealers in other New England states.

An annual lottery was established with three rewards, one $1,000 and two $500 prizes In order to encourage tag returns. The winners will be presented with their prizes at the Maine Fisherman’s Forum in March 2005. 

Results
The total number of herring tagged since the start of the project in 2003 stands at 39,211.  Almost 20,000 herring were tagged in each year of the project (Table 1).  In 2004, fish were tagged in the target spatial strata (Gulf of Maine and Southern New England) and temporal strata (Summer Feeding Spawning and Winter Feeding/Spring Migration).

 

Tagging did not occur in every spatial or temporal stratum for many reasons.  For example, logistic and financial restrictions did not allow for tagging on Georges Bank or in Nova Scotia in 2003 or 2004.  The timing of tagging events is also linked to the behavior of the herring, making tagging in every spatial stratum during every temporal stratum impractical or impossible. 

 

Table 1: Number of herring tagged by spatial and temporal strata*

* - A key to the strata abbreviations can be found in Appendix A

 

The first field-tagging event of 2004 occurred in February near the “Mud Hole” off the coast of New York. Additional tagging in southern New England occurred in April off Block Island, RI and eastern Long Island, NY.  Tagging trips then commenced again in June and continued through October in the Gulf Of Maine.  The following table (Table 2) shows the number and type of all tagging trips taken in 2004.

 

Table 2:  Number and type of tagging trips in 2004

A total of 13,525 herring were tagged and released aboard purse seine vessels in 2004, with an average of 1,352 herring tagged per trip. Aboard midwater trawl vessels 5,875 herring were tagged and released for an average of 1,958 per trip.  The following map (Figure 6) shows the location of each set (some tagging trips had as many as six sets) with the markers representing the number of herring tagged per set.

 

Representative samples (N=50) were taken on each tagging trip in order to categorize the size and maturity of the captured herring (refer to Appendix B for length frequency plots by trip). 

 

Figure 6:  Location of tagging sets in 2004

Tag Returns

Fifty eight tags were returned in 2004 for a total of 105 returns since the start of the project. The longest distance traveled by a tagged herring was 414 nautical miles (measurement represents a straight line between the point of release and the point of recovery) from Block Island, RI to Scots Bay, NS.  The time at large for the recaptured fish ranged from 2 to 551 days with an average of 92 days. 

 

The following table and figures show movement between the tagging strata and the recovery strata.  Nine tags were returned without adequate recovery information and could not be traced to a reliable catch location (indicated as “UNK” in Table 3).

Table 3:  Movement from tagging strata to recovery strata

Figure 8: Tag returns (blue dots) from herring released in the GOM and their average movement pattern (blue arrows) to each recovery stratum

 

Figure 9: Tag returns (blue dots) from herring released in the SNE and their average movement pattern (blue arrows) to each recovery stratum

 

It is important to note that the interpretations of the observed movement patterns in the following discussions are uncertain and should not be construed as fact.  The range of variables (temperature, weather, food supply, predation, fishing pressure, etc) potentially affecting herring movement is vast and largely unknowable.  Tag/recovery studies using non-archival tags only provide information on the release location and a single recovery location for a given fish.  No information on any movement or behavior between the time of tagging and recovery is ever available from this type of study.  Results from this project have shown that a single herring can migrate from SNE to NS in 26 days, emphasizing the highly mobile behavior of Atlantic herring.  Therefore, all interpretations of these results should be viewed only as possible explanations based on the collective knowledge of herring biology, general behavior patterns, commercial fishing patterns and abundance surveys.

Tag returns from fish released in the GOM were recovered from each of the five defined spatial strata.  The majority of recoveries (70%) were made in the same stratum of release with 83% recovered within 2 months.  These returns indicate short-term residency of pre-spawning herring in the GOM.  The remaining 17% of returns imply either long-term residency and/or migration back to the GOM on an annual basis.  Returns from Canadian waters ranged from 17 to 367 days at large, suggesting some of the fish tagged in the GOM were migrants on their way to NS spawning grounds (several of the fish were recovered in spawning condition in Scots Bay and on German Bank).  Significantly, there is currently no assumption in the herring stock assessment that the GOM fishery exploits the NS spawning component. 

 

Only two fish from this project have been recovered on Georges Bank and both were tagged in the GOM in July on the same trip near Seguin Island.  These recoveries appear to support the mixing assumption presented in NEFMC’s FMP (Table 4: red highlights).  Finally, fish tagged in the GOM were also recovered in the SNE stratum supporting the assumption that GOM fish migrate to winter feeding grounds of RI and the mid-Atlantic coast (Table 4: gray highlights).

 

Fish tagged in the SNE stratum were recovered in four of the five defined strata (no tags were recovered on GB).  Only five recoveries were made in the SNE stratum, three were within one week at large while the other two were caught ¾ of a year later.  These two long-term recoveries suggest that herring return to the winter feeding grounds after migrations to summer spawning areas.  There are no known resident spawning components in the SNE stratum so annual emigration is very likely (Overholtz et al., 2004).  Canadian (NB and NS) recoveries collectively accounted for 35% of the SNE returns.  There recoveries are significant because there is also no assumption in the current FMPs that the winter fishery exploits fish from the NS spawning component (Note, there is no inclusion of the NS spawning component at all in Table 4).

 

Table 4:  Seasonal distribution of spawning components

                * - GB spawning component includes Nantucket Shoals

The tag return information can also be interpreted using time at large/distance plots.  These charts incorporate a temporal element into the data interpretation.  The red lines on each chart indicate 1 year at large.  Fish released in the GOM during the summer feeding/spawning stratum show a clear pattern of short term residency (<100 days), followed by longer distance migrations (Figure 10).  After 1 year at large, recoveries were made very close to the original tagging location suggesting possible site fidelity. 

 

Fish released in SNE during the winter feeding stratum left the area quickly (300+ miles in <100 days).  Recaptures made 200+ days after the initial tagging suggest migrations back to areas near the release site (Figure 11).  This pattern is likely due to the fact that most herring in 2003 and 2004 were tagged at the very end of the winter feeding temporal stratum.  Therefore, they departed the area quickly without showing any initial residence period and returned to the same area in less than 1 year.  Had fish been tagged in SNE earlier in the winter feeding period, tag returns might have shown longer residence times.  Additional recoveries made after 400 days at large indicate a repetition of the long distance migration pattern observed in the previous year.

 

Figure 10: Days at large/distance traveled plot for fish released in the GOM

 Figure 11: Days at large/distance traveled plot for fish released in SNE

Return Rate

The unadjusted return rate for the two-year herring tagging project is approximately 0.3%.  This return rate is not unexpected for a herring tagging project targeting an offshore fishery pursued by mobile gear.  Significantly, 86% of the recaptures were made after at least 2 weeks at large.  The convention with herring tag returns is to assume fish recovered after 2 weeks have had an opportunity to mix and redistribute according to normal movement patterns.  Therefore, although the unadjusted tag return rate is low, the percent of significant returns is extremely high.  These returns contribute to our understanding of migration patterns, site fidelity and stock intermixing. 

 

There appears to be no significant difference between return rates from fish tagged in the GOM during the SFS stratum (0.26%) or the SNE during the WF stratum (0.27%).  A simple chi-square test showed no significant difference in return rates from fish originally captured for tagging by purse seiners or midwater trawlers employing the aquarium codend.

 

Methods for adjusting returns based on catch and effort are being explored.  This adjustment factor to the return rate is important because herring recaptures are made when and where the commercial herring fleet is fishing.  Therefore, where there is little or no fishing effort, few or no returns at all can be expected.  Conversely, where there is significant fishing effort artificially high return rates are possible in proportion to the number of fish released.

Reporting Rate

Seeding studies were conducted in the summer of 2004 to develop an estimate of the total tag reporting rate.  The resulting estimate incorporates a measure of the reporting rate (meaning tags that are seen, recovered and reported) and the “sighting” rate.  The sighting rate is an artifact of a high volume, bulk processed fishery like herring where there are limited opportunities to see and recover tags.  The seeding study produced a single measure of these combined variables.   

 

Tags were surreptitiously seeded into landed herring catches on 7 occasions (3 trials with fish going to processors and 4 trials with fish going to bait dealers) between June and July.  On each occasion, 10 tags were seeded into trucks while the fish were pumped off the catcher vessel or into hoppers entering the processing facilities.  Thirteen of the 70 tags seeded into the herring catch during this period were recovered (19%).  Returns from the processing facilities were significantly better (33%) then returns from the bait sector (8%). 

 

The extremely low estimate of the tag reporting rate resulting from this seeding study was surprising.   The reporting rate from the bait sector was of special concern and resulted in a significant outreach program.  DMR is currently trying to raise awareness, increase incentives and make it easier, for lobstermen especially, to return tags.  The seeding study will be repeated in the summer of 2005 to see if any of these efforts have increased the estimated reporting rate.

 

Conclusions 

The initial results of this project prove that mark/recovery experiments are successful with Atlantic herring even in the context of an offshore fishery dominated by mobile gear.  Despite the low unadjusted return rate, the majority of recoveries are long term and indicate important migrations patterns and possible annual site fidelity.  The recoveries from spawning grounds in NS are of special significance given that no assessment currently accounts for mixing between GB, GOM and NS stock components. 

 

Long Term Project Goals

The Atlantic herring anchor tagging project is designed to be a long-term, comprehensive program.  Data collected over the next five years should contribute to our understanding of migration and movement patterns, spawning site fidelity and stock intermixing.   Project plans for the next year include repeating tagging efforts in Southern New England during the winter feeding stratum and tagging in the Gulf of Maine and on Georges Bank during the summer feeding/spawning stratum.  These efforts will be accomplished through the support of the Northeast Consortium which committed significant funding to the project for the 2005 season.  Cooperative efforts between the DMR and DFO, including tag retention studies and data pooling will continue in 2005.  We anticipate conducting a collaborative effort in the summer of 2005 to tag herring in Scots Bay and on German Bank, NS.

 

Acknowledgements
The success experienced by the Atlantic herring project in 2004 was due in great part to the continued commitment and support of the herring industry.  We are grateful for the willing cooperation of the captains and crews of the F/V Thunder Bay, F/V Western Sea, F/V Ocean Venture and F/V Western Wave.  Their contributions ranged from allowing us to accompany them on commercial fishing trips, conducting contract work and providing insight on herring movement and behavior.  We would also like to thank the East Coast Pelagics Association, The Downeast Fixed Gear Fisherman’s Association, Connors Brothers, Purse Line Bait, Cape Seafoods, Ocean Venture Inc., Thunder Bay Inc., Superior Bait and Salt, My Lady Inc., Beaver Enterprises and Stinsons for their generous financial contributions. 

 

Appendix A: Abbreviations for spatial and temporal strata

 Appendix B: Length frequency charts by trip number

Literature Cited

 

Armstrong, M.P. and Steven X. Cadrin, Morphometric Variation among Spawning Groups of the Gulf of Maine-Georges Bank Herring Complex. Herring: Expectations for a New Millennium, Alaska Sea Grant College Program, AK-SG-01-04, 2001.

 

Creaser, E.P., D.A. Libby and G.D. Speirs. 1984. Seasonal movements of juvenile and adult herring, Clupea herengus, tagged along the Maine coast. J. Northw. Atl. Fish. Sci., 5:71-78.

 

NMFS, Commercial Fisheries Landings Statistics Web Page; http://www.st.nmfs.gov/st1/commercial/index.html. 2002.

 

NEFMC, Scientific and Statistical Committee Meeting Report. June 2003a.

 

NEFMC, The Role of Atlantic Herring (Clupea harengus) in the Northwest Atlantic Ecosystem. 2003b. 

 

NEFMC, Herring PDT/TC Report. May 2004.

 

Overholtz, W.J., The Gulf of Maine-Georges Bank Atlantic herring (Clupea harengus): spatial pattern analysis of the collapse and recovery of a large marine fish complex, Fisheries Research 57, 237-254, 2002.

 

Overholtz, W.J.; Jacobson, L.D.; Melvin, G.D.; Cieri, M; Power, M.; Libby, D.; Clark, K. 2004. Stock assessment of the Gulf of Maine – Georges Bank Atlantic herring complex, 2003.  Northeast Fish. Sci. Cent. Ref. Doc. 04-06; 290 p.

 

For a printed version of this report please click here - 2004 Report
You will need Adobe Acrobat Reader to view this