Comparison of Spawning Characteristics of Cod(Gadus morhua) Stocks in the North Atlantic

K. Brander1

Ministry of Agriculture, Fisheries and Food, Fisheries LaboratoryPakefield Road, Lowestoft, Suffolk, United Kingdom

Abstract

A fundamental problem in studying fish populations is to identify the processes whichgovern their dynamics and the appropriate scales (in time and space) at which to investigatethese processes. One possible way to identify the processes is by describing the life historycharacteristics and comparing those characteristics for different areas in which the speciesoccurs. Some preliminary results of a comparative study of cod (Gadus morhua) andhaddock (Melanogrammus aeglefinus (L.)) life histories and population dynamics, under-taken by scientists within ICES, are presented to stimulate further development of theinitiative.

This paper concentrates principally on the timing and location of cod spawning, andbriefly summarizes the information available for most of the geographic range of the speciesin the North Atlantic. It looks in greater detail at the area around the British Isles and at theScotian Shelf, in order to investigate variability in timing of spawning on a smaller scale.Some conclusions are drawn about the processes and scales which may be appropriate forfurther investigation.

Key words: Cod, Gadus morhua, distribution, life history characteristics, regional compari-son, spawning

NAFO Sci. Coun. Studies, 18: 13–20

some of the issues of methodology and samplingprocedure which arise are discussed here.

This paper will concentrate principally on thetiming and location of cod spawning, and will brieflysummarize the information available for most of thegeographic range of the species. It will look ingreater detail at the area around the British Islesand at the Scotian Shelf in order to investigatevariability in timing of spawning on a smaller scaleand the factors which may be responsible for ob-served differences in timing between differentspawning sites. Finally, some conclusions will bedrawn about the processes and scales which maybe appropriate for further investigation.

One of the characteristics of temperate fishspecies is their relatively short spawning period,whereas tropical species breed over a long periodof the year (Fig. 1). In a species such as cod in theNorth Atlantic, eggs are generally released over aperiod of a few weeks in spring and one may com-pare the timing of spawning and its duration (themean and standard deviation of the egg productioncurves) between different stocks, or within the samestock in different years. Possibly the best exampleof the latter type of comparison is for the Arcto-

Introduction

A fundamental problem in studying fish popula-tions is to identify the processes which govern theirdynamics and determining the appropriate scalesin time and space at which to investigate theseprocesses. For example, annual recruitment to somefish stocks is influenced by temperature; such arelationship might arise in several ways, but it isdifficult to distinguish which of the many possibleprocesses in fact operate and therefore what theappropriate measure of temperature should be.

One possible way to identify efficient processesis by description of the life history characteristics ofthe species concerned and comparison of thosecharacteristics for different areas in which the spe-cies occurs. This is one of the purposes of a com-parative study of cod (Gadus morhua) and had-dock (Melanogrammus aeglefinus (L.)) life historiesand population dynamics, being undertaken by agroup of scientists within ICES. I have been askedto prepare a paper for this symposium on behalf ofthat group in order to present some preliminaryresults and to stimulate further development of theinitiative. The paper is partly based on informationsupplied through the ICES comparative study and

1 On behalf of the ICES Study Group on Cod Stock Fluctuations.© Crown copyright 1993

14 Sci. Council Studies, No. 18, 1993

Fig. 1. Schematic graph of seasonal spawning charac-teristics.

Norwegian cod in the Vestfjord, where “spawningintensity” over an eight year period shows remark-able consistency from year to year (Fig. 2).

These estimates of “spawning intensity” arederived from egg surveys, which are probably themost accurate but the procedure is costly. Thereare alternative methods for estimating the spawningperiod, and one of these which has also been ap-plied to Arcto-Norwegian cod for the period since1929 is based on an index of the weight of roe. Thismethod has shown that over the period 1930–60,spawning tended to occur later by about 7–15 daysand the delay was ascribed to a decline in theaverage age of spawning fish over that period(Pedersen, 1984). For this stock there was no evi-dence that year-to-year variability in temperatureaffects the date of spawning, but this may be be-cause the vertical temperature differences in thespawning area are greater than the year-to-yeardifferences and the fish can select a preferredtemperature by vertical movement. These Arcto-Norwegian cod spawn at a temperature of 4° to 6°C,at the thermocline between the cold coastal waterand the warmer Atlantic water. The time of spawningat the main spawning site of the Arcto-Norwegiancod, at Lofoten, is about 2 weeks earlier than at thenorthernmost spawning field.

To what extent can the Arcto-Norwegian cod beused as a paradigm for the other cod stocks? Is itgenerally true that the timing of cod spawning at aparticular site is calendar fixed and does not re-spond to proximate environmental variability (e.g.annual temperature variation)? Is it generally truethat the location of spawning is fixed either geo-graphically or in relation to a hydrographic featureor discontinuity? Which factors govern the ob-served differences in timing between differentspawning sites? These are the sorts of questionswhich the comparison of spawning characteristics

1976

1977

1978

1979

1980

1981

1982

1985

1976–1985

2 000

1 000

0

2 000

1 000

0

2 000

1 000

0

2 000

1 000

0

2 000

1 000

0

2 000

1 000

0

2 000

1 000

0

2 000

1 000

0

2 000

1 000

0

Num

ber

of

egg

s p

er m

2

50 60 70 80 90 100 110 120 130Julian day

Fig. 2. Seasonal “spawning intensity” for the Arcto-Nor-wegian cod. Vertical bars represent mean dateof 50% spawning for the period 1976–83 (fromPedersen, 1984).

between different cod stocks is designed to an-swer. The development of generally applicablemodels of the relationship between cod population

15BRANDER: Comparison of Spawning Characteristics of Cod Stocks

dynamics and the physical environment (e.g. in thecontext of the ICES Cod and Climate program)depends partly on answers to questions of this kind.However, this paper is intended to have much lessscope, and will concentrate principally on the tim-ing and location of cod spawning, and will brieflysummarize the information available for most of thegeographic range of the species.

Pepin and Myers (1991) recently proposed thatvariability of recruitment may be a function of theduration of the pelagic phase and that this could betested among stocks of single species. This andsimilar hypotheses require information on the dura-tion of the pelagic period of the life history. Sincethe pelagic period lasts from the time of spawning tothe time of settlement, information on time of spawn-ing will be useful in this regard and information ontime of settlement will also be sought through theICES checklist.

Materials, Methods and ResultsThe ICES checklist of spawning characteristics

provides information on location and timing of codspawning from many published and unpublishedsources (Anon., 1990). The major spawning sites

Fig. 3. Distribution of cod stocks in the North Atlantic, with their principal spawning sitesand dates (month) of peak spawning (based on Anon., 1990).

and dates of peak spawning are shown in Fig. 3 and4. The most southerly stock in the Northeast Atlanticspawns in the English Channel (50°N) in January-February and the northernmost stock spawns aroundnorthern Norway (70°N) in April. The most southerlystock in the Northwest Atlantic spawns on GeorgesBank (42°N) during the first 3 months of the year.Further information on the duration of spawning indifferent areas and the depth at which it occurs isgiven in Table 1. Spawning takes place at depthsranging from 15 m to more than 800 m, but as can beseen from the gaps in Table 1, there is no informa-tion on depth of spawning for most areas.

The quality and kind of information available ontiming of spawning varies a great deal, even forsupposedly well studied areas such as the NorthSea. Figure 3 shows a scatter of points in the NorthSea, which presumably is intended to indicate wide-spread spawning. Since there has been no compre-hensive series of eggs surveys in the North Sea it isdifficult to judge the relative contribution of particu-lar areas to the total, but Fig. 5 summarizes avail-able information on spawning for most of the areasaround the British Isles (Graham, 1923; Heesen andRijnsdorp, 1989).

16 Sci. Council Studies, No. 18, 1993

Fig. 4. Dates of peak spawning (month) for cod stocks in the Northwest Atlantic.

In the English Channel the level of spawning islow, but the fact that it is the most southerly stock inthe Northeast Atlantic and one of the earliest tospawn gives it some interest. Spawning starts inJanuary at two locations close to the coast of France.Off the northeast coast of England, spawning be-gins in early-February about 40 miles offshore andit remains mainly offshore. Egg surveys of the BristolChannel in 1990 show a clearly defined spawning inspace and time and surveys in 1953, 1971, 1972and 1980 confirm that this is a regular area.

A detailed picture of the location and timing ofspawning on the Scotian Shelf was obtained fromichthyoplankton surveys carried out in 1979–81(Brander and Hurley, 1992; fig. 6). Here spawning iswidespread, but occurs mainly on the various off-shore banks. The timing is earliest at the southwest-ern end of the area and progressively later to thenortheast (Fig. 7).

Discussion

Several questions which were posed abovemight be answered by comparisons between codstocks. These are: is the location of spawning fixed?,

does it vary from year to year in response to environ-mental factors?, and, can we identify the factorsresponsible for variations in timing between areas?The material collected so far in the checklist exer-cise can assist in answering these questions, but itquickly becomes apparent that the questions couldbe answered more fully if they were posed as part ofthe checklist exercise from the outset. Therefore,the next stage of this exercise should perhaps be toapproach the contributors again with more specificinformation requirements.

There are three principal factors to considerregarding the question about causes of variation intiming between areas: latitude, temperature andproduction timing. Latitude per se (and its relationto day length) is not a dominant factor since thelatitudinal ranges are very different on the two sidesof the Atlantic, but within each area it may play arole. One contrary example seems to be the timingof spawning off Labrador, which is apparently ear-lier further north.

Spawning occurs over a fairly wide temperaturerange (perhaps 0° to 10°C, although it is mucheasier to tell the temperature at which eggs occur

17BRANDER: Comparison of Spawning Characteristics of Cod Stocks

TABLE 1. Timing and location of cod spawning throughout the North Atlan-tic, from the checklists of cod spawning characteristics.

Location Depth (m) Peak timing Range

Northwest Atlantic*5Z Feb–Mar Dec–May4X Apr Feb–May4VsW Apr–May Feb–Jun

Oct–Dec4T4Vn 35–90 late May–mid June May–Sep3Pn4RS Apr–May3Ps May–Jun3NO May–Jun2J3KL Apr–May------------------------------------------------------------------------------

Central AtlanticWest Greenland 600–800 Apr Mar–JulEast Greenland 170–400 Mar–JulIceland Apr Mar–MayFaroe Bank Mar–Apr Feb–MayFaroe Plateau 80–180 Mar Feb–May------------------------------------------------------------------------------

Northeast AtlanticWhite Sea 15-100 Apr Mar–MayNE Arctic 100–500 Apr Mar–AprNorway Coast 50 Feb–MaySkagerrak Feb–AprE. Baltic 60 Mar–AugW. Baltic 40–50 Feb–AprN. North Sea Apr Feb–AprC. North Sea end Feb Jan–AprS. North Sea mid Feb Jan–AprEnglish Channel early Feb Jan–MarBristol Channel late Mar Feb–AprIrish Sea Mar Mar–Apr

* NAFO Divisions and Subdivisions.

Fig. 5. Distribution and timing of cod spawning around the BritishIsles, with March surface isotherms (°C) for the North Sea.

6

45

5

6

4

58˚

6̊ 4̊ 2̊ 0̊ 2̊ 4̊ 6̊ 8̊

English Channel

BristolChannel

IrishSea

midMarch

late March

late March

March

February

January

early February

early February

mid February

February

NorthSea

March

March

7

56˚

54˚

52˚

50˚

48˚

18 Sci. Council Studies, No. 18, 1993

Fig. 6. Distribution of stage 1–3 cod eggs from Scotian Shelf Ichthyoplankton Surveys (SSIP) in 1979–81 (from Branderand Hurley, 1992).

than the temperature at which spawning occurs, inthe areas with strong vertical or horizontal tempera-ture gradients there may be considerable differ-ences). Within the North Sea, cod spawn at a timewhen temperature is stable and the differences intiming of about 1 month between the south and thenorth do not correspond to a difference in tempera-ture (Fig. 5).

It is difficult to define or measure productiontiming as it may affect cod spawning. In the NorthSea the Continuous Plankton Recorder shows adifference in timing of peak abundance of the zoop-lankton Calanus finmarchicus of 2 weeks betweenthe north and south, which is similar to the differ-ence in time of cod spawning. On the Scotian Shelfthe timing of peak abundance of Calanus sp. in

19BRANDER: Comparison of Spawning Characteristics of Cod Stocks

Fig. 7. Numbers of stage 1–3 cod eggs and Calanus finmarchicus per m2, averaged arithmetically in 20-day periodsfor Emerald, Sable, Banquereau and Misaine Banks. Data for the years 1979, 1980 and 1981 have beencombined. The averages of the dates of sampling within each 20-day period have been used in the plots, notthe midpoint. The arrowed date is the mean date of cod spawning.

TABLE 2. Proposed table structure for data on duration of the pelagic stage for cod eggs and larvae.

Duration Distance Final sizeLocation (d) (km) (mm)

Northwest Atlantic*

5Z4X4VsW

4T4Vn3Pn4RS3Ps3NO2J3KL-------------------------------------------------------------------------------------------------------------------

Central Atlantic

West GreenlandEast GreenlandIcelandFaroe BankFaroe Plateau-------------------------------------------------------------------------------------------------------------------

20 Sci. Council Studies, No. 18, 1993

TABLE 2. (Continued)

Duration Distance Final sizeLocation (d) (km) (mm)

Northeast Atlantic

White SeaNE ArcticNorway CoastSkagerrakE. BalticW. BalticN. North SeaC. North SeaS. North SeaEnglish ChannelBristol ChannelIrish Sea

* NAFO Divisions and Subdivisions.

spring is over a month later close to the LaurentianChannel than it is on Emerald Bank, and thesecoincide with the differences in timing of cod spawn-ing (Fig. 7). Production timing may therefore affectthe timing of cod spawning, but this can only beexamined by looking at the relationship on a suffi-ciently fine scale.

The process of bringing together information onNorth Atlantic cod for comparative purposes and formodelling population dynamics is an iterative one inwhich preliminary evaluation and interpretation sug-gest new questions and ways of defining and mea-suring things. So far the checklist exercise hashelped to retrieve some of the historic material. Theplan is to edit the contributions for a publication byICES in 1993. If the approach has any value beyondthis, it is probably as providing a means of bringingtogether information which can be used to testexisting hypotheses or models which arise fromprograms such as the ICES Cod and Climate. Forexample it would be very useful to complete Table 2in order to follow up Pepin and Myers (1991) sug-gestion that variability is a function of duration of thepelagic phase.

References

ANON. 1990. Report of the study group on cod stockfluctuations. Appendices III and IV. ICES C.M. Doc.,No. G:50 (mimeo).

BRANDER, K. M., and P. C. F. HURLEY. 1992. Distributionof early stage cod, haddock and witch eggs on theScotian Shelf: a reappraisal of the evidence on thecoupling of cod spawning and plankton production.Can. J. Fish. Aquat. Sci., 49: 238–251.

GRAHAM, M. 1923. The annual cycle in the life of themature cod in the North Sea. Fish. Invest. Lond., Ser.II, (VI No. 6): 77 p.

HEESEN, H. J. L., and A. D. RIJNSDORP. 1989. Investiga-tions on egg production and mortality of cod (Gadusmorhua L.) and plaice (Pleuronectes platessa L.) inthe southern and eastern North Sea in 1987 and 1988.ICES Rapp. Proc.-Verb., 191: 15–20.

PEDERSEN, T. 1984. Variation of peak spawning of Arcto-Norwegian cod (Gadus morhua L.) during the period1929–1982 based on indices from fishery statistics.In: The propagation of cod. E. Dahl, D. S. Danielssen,E. Moksness and P. Solemdal (eds.). Flodevigenrapportser, 1: 301–336.

PEPIN, P., and R. A. MYERS. 1991. Significance of eggand larval size to recruitment variability of temperatemarine fish. Can. J. Fish. Aquat. Sci., 48: 1820–1828.