NORTH EAST ATLANTIC - ATLANTICO NORDORIENTAL (Continued)

ON THE BIOLOGY AND CHANGES IN EXPLOITATION AND ABUNDANCE OF THE ICELANDIC CAPELIN

by

H. Vilhjálmsson

Marine Research Institute
Reykjavik, Iceland

Resumen

Se describe la biología y explotación del capelán que habita las aguas entre Islandia, Groenlandia y Jan Mayen. Se describen, asimismo, y se discuten los efectos aparentes de la pesca y las acciones de ordenación.

INTRODUCTION

The capelin, Mallotus villosus (Muller), is a boreoarctic, pelagic and schooling, salmonid species with a more or less circumpolar distribution. Although capelin seldom grow larger than 20 cm in size and are short lived (3-5 years) their biomass in certain areas may reach millions of tonnes.

Capelin is a highly significant item in the diet of many commercially important northern fish species. Although feeding in arctic waters it may transfer large amounts of biomass and energy to warmer areas, which it periodically visits for example, in order to spawn, thus becoming available as food for larger fish such as the cod.

The largest known stock of capelin inhabits the Barents Sea and adjacent waters while smaller but nevertheless important stocks are found off Iceland, Greenland, Newfoundland and Labrador as well as in the North Pacific.

During the last 2-3 decades capelin have become an important source of raw material for the production of meal and oil and lately on a smaller scale for human consumption. Because of its short life span and pelagic schooling nature capelin have fallen easy prey to modern fishing technology and stocks have collapsed because of overfishing.

The present paper deals with Icelandic stock of capelin which inhabits the area between Iceland, Greenland and Jan Mayen. The biology of this stock is outlined as well as its exploitation and the apparent effect of the fishery as well as management actions are described and discussed.

ICELANDIC CAPELIN

Spawning times and areas

The Icelandic capelin spawn on the seabed in shallow coastal waters mainly during the second half of March and the first half of April. Usually the spawners migrate clockwise from North to Southeast Iceland the first arriving at the southeast coast in late January or February. In some years spawners also arrive at the southwest coast of Iceland directly from the banks off the northwestern peninsula in February-March. In both cases spawning usually begins about 2-3 weeks after the fish entered the warm coastal waters.

The main spawning area extends from Southeast Iceland along the south coast to the Snaefellsnes peninsula in the west. The bulk of the south and west coast spawning is usually finished by mid-April.

Spawning also takes place off the north and northwest coasts of Iceland. This is a less important and later spawning which may last throughout May into June.

The spawning areas of the Icelandic capelin are shown in Figure 1.

Distribution of larvae

After hatching the larvae ascend towards the surface waters and are transported by currents to the west and north of Iceland. In some years a westward drift across the banks to the north of the Irminger Sea basin towards Greenland is observed. The system of surface ocean currents in the Iceland-Greenland-Jan Mayen area is shown in Figure 2.

The distribution and abundance of capelin larvae in the Iceland-Greenland area have been studied annually in August since 1970. A schematic representation of the results of these investigations is shown in Figure 1.

Fig. 1. Spawning areas (black) and general larval distribution in August (shaded) of the Icelandic capelin.

Distribution and migration of 1-3 group capelin

Through extensive acoustic surveying and sampling, first in connection with scouting for the fishery and later for stock assessment purposes, the migrations of the Icelandic capelin have been mapped and may be described as follows:

In winter juvenile 1-2 group capelin are most frequently found in waters over or near the continental shelf off North and Northwest Iceland, but sometimes also further east.

In spring a varying proportion of this stock component migrates out into deeper and colder waters to feed during the summer months. Thus, at that time of the year, it may also be recorded in deep waters between Iceland and Greenland and over the East Greenland plateau from the Scoresby Sound south to the Dohrn Bank. Juvenile capelin, probably belonging to the stock which spawns at Iceland, have also been recorded further west, particularly at the edge of the continental shelf off the Angmagsalik Bay.

In autumn and early winter the immatures retreat with the advancing cold waters from the north and overwinter off North-east and North Iceland or in the Denmark Strait area (NW-Iceland) as the case may be.

Fig. 2. Surface ocean currents in the Iceland-Greenland-Jan Mayen area (after Stefánsson 1962).

Maturing 2-3 group capelin exhibit a somewhat similar behaviour pattern but as a rule have a much wider distribution. Thus, most of these capelin migrate to the deep and colder regions in spring and early summer. In recent years the maturing stock has been divided in two parts during the feeding season which we may refer to as northern and southern components.

The southern component feeds in the Denmark Strait and off North Iceland and is often mixed with juveniles. The northern component, consisting of the oldest and largest fish, undertakes a long northward migration to the area between Greenland and Jan Mayen and to the north of Jan Mayen to 72-73°N. These capelin stay out of the distribution area of the juveniles for at least 2-3 months in summer and autumn.

Toward autumn both components of the maturing stock start migrating back south and together gradually assemble on the overwintering grounds off North or Northwest Iceland in October-November. From there the main spawning migration begins in December-January and follows the edge of the continental shelf off North and East Iceland until the fish meet the warm Atlantic water when they turn westwards towards the southeast coast. These movements of the maturing stock were very clearly demonstrated during a tagging experiment in 1978/79 (Figure 3).

Fig. 3. Migration routes of adult capelin as indicated by tag returns. Summer 1978 - winter 1979.

In those years when capelin arrive to spawn off West Iceland from the north and northwest, a part of the spawning stock stays in the overwintering area off Northwest Iceland until late February when its southward in migraton begins.

The distribution areas of juvenile and adult capelin are shown in Figure 4 and the general migration pattern of the maturing stock in Figure 5.

Fig. 4. A generalized representation of the distribution areas of juvenile and adult capelin in the feeding season.

Fig. 5. The general migration pattern of the maturing stock.

Age and mortality at spawning

Most of the Icelandic capelin spawn as 3 year olds. A varying proportion of each year class does, however, not mature until one year later. Thus, the spawning stock consists of two age groups where the ratio of 3 and 4 year old spawners varies somewhat depending on feeding conditions and relative year class strength. In later years the ratio of 4 year old fish has been unusally low which possibly is connected with the changes in fishing pattern and mortalities brought about by the recent advent of a summer and autumn fishing season.

On the other hand the survival rate of spawners is considered very low and second time spawners do not seem to contribute to the spawning stock. This is for example, borne out by the consistently low percentage of the older age group in the spawning stock even though a large year class has spawned as three year olds in the year before. Without risking oversimplification it seems that for practical purposes the survival may be set at zero.

THE FISHERY

The total annual catch from the Icelandic capelin stock by season and countries for the period 1964-82 is shown in Table 1. Prior to this period capelin belonging to the Icelandic stock were only caught for bait and the catch amounted to no more than a few hundred tonnes a year.

Table 1: The total annual and seasonal catch of capelin in the Iceland, E-Greenland, Jan Mayen area (in '000 tonnes).

In the mid-sixties Iceland started a capelin fishery for a reduction to meal and oil. This was a purse seine fishery and carried out by the local herring fleet outside the traditional herring season.

Until 1973 the capelin fishery was limited to coastal waters during the spawning season from late February to April. But with the depletion of the Atlantic-Scandian herring stocks in the late sixties and the North Sea herring in the seventies the capelin fishery gained momentum and importance.

By 1969 almost all the Icelandic herring fleet had become involved in the winter capelin fishery and it soon became obvious that processing facilities scattered along the N-Icelandic coastline, were a limiting factor. Thus, the catch did not exceed 200,000 tonnes during the 1969-71 period.

In the 1972 winter season a coordinating body, the Capelin Board, was set up in order to synchronize fishing and production. This was mainly done by dividing the coastline into price zones the boats getting higher prices for longer trips. This system had the immediate effect of increasing the catch by about 50 percent as compared to the average for the previous 3 years. The price zone system was operated until 1979 when it was discountinued, the board from then on mainly functioning as collector and distributor of information on catch statistics, available landing spaces and waiting time in the various ports where the factories were located.

As early as 1969 research had indicated that, at least in some years, capelin could easily be caught in January and February when the spawning migration still was in deep waters off East and Northeast Iceland. Such a fishery was attempted in 1970. Difficulties were encountered by the fishing fleet and another attempt was not made until 1973. This time success was immediate and the winter catch jumped to 440,000 tonnes.

With the exception of 1976 when the fleet went on a lengthy strike because of a price dispute the annual catch during the 6 year period 1973-1979 ranged between 440,000 and 575,000 tonnes. Since then the winter fishery has been subject to increasively severe restrictions ending in a total fishing ban in 1983. Apart from a limited Faroese participation in 1977-79 the winter fishery has been conducted by Iceland alone.

In 1975 an attempt was made to catch capelin on the feeding grounds north of Iceland in July and August. At that time there was much drift ice in the area and the experiment failed.

A summer fishery was again attempted in the following year and was an immediate success. The catch amounted to 115,000 tonnes in spite of the fact that most of the fishing fleet was occupied in the North Sea herring fishery that summer and autumn. During the 6 year period 1977-81 the average Icelandic summer and autumn capelin catch has been about 410,000 tonnes taken by about 50 vessels.

The number of Icelandic vessels participating in the capelin fishery is shown for the period 1973-82 in Table 2. It must be stressed, that in this case the number of fishing boats is no measure of effort. The increase in size of vessels and fishing gear and technological improvements in the acoustic search systems more than make up for the reduction in the number of vessels participating in the winter fishery during the above period.

In the summer of 1978 the Norwegian capelin fleet started a purse seine fishery on that part of the stock which migrates to feed in the area between Jan Mayen and Greenland The catch during that first season was 154,000 tonnes but has since been subject to restrictions agreed upon with Iceland which, for her part, has mainly reduced the winter catch as stated previously.

In 1980 and 1981 vessels from the Faroes and EEC countries, mainly Denmark, caught about 40,000 tonnes each year lending an international atmosphere to the capelin fishery in the Iceland-Greenland-Jan Mayen area.

Apart from 13,000 tonnes caught during the 1982 winter season there was a complete fishing ban on fishing from the Icelandic capelin stock in 1982 as well as during the 1983 winter season.

STOCK ABUNDANCE

Because of the capelin's short life span and spawning mortality large and sudden changes in stock abundance may be expected. For the same reasons researches have to resort to direct, fishery independent methods in order to assess stock size and changes therein.

Comparative measurements of the abundance of O-group capelin have been obtained in August annually since 1972 (Figure 6). This information has been collected by a combination of sampling by conventional and acoustic methods and is intended to serve as a rough index of recruitment. During the period 1972-75 the O-group index indicates a high level of recruitment followed by a sharp downward trend and a much lower and decreasing level during the last 5-6 years.

Fig. 6. O-group indices 1972-1982.

In the seventies several tagging experiments were carried out in order to evaluate the size of the Icelandic capelin stock (Vilhjálmsson and Reynisson, 1979). The marking was done with internal steel tags which were recovered by means of magnets in the processing plants ashore. Because of the small size of the fish the mortality of juvenile capelin proved very high and returns of tags from such taggings were extremely low.

The tagging of adult capelin was, on the other hand, successful and results of estimates of the 1979 spawning stock abundance from the returns were on a similar level as those obtained by the acoustic method. Although tagging apparently can give information on the size of the maturing stock, final results do not become available until during or after the spawning season and consequently are of little value for current management purposes.

Echo surveying of capelin at Iceland has been conducted since the mid-sixties. The main purpose used to be to chart the distribution and movements of the stock as well as to study the behaviour of the fish and so to help the fishing fleet in catching the capelin. With the rapid increase in catches following the advent of the multinational summer and autumn fishery the need for an accurate assessment of the abundance of the stock became of paramount importance.

Due to the information collected during the above scouting cruises it was possible to pinpoint the times of the year during which information on stock abundance most likely could be obtained by acoustic methods. The main requisitions were relatively limited area of distribution and minimum danger of interference by drift ice, weather, other scatterers and bottom echoes. For the maturing or fishable part of the stock the choice was the autumn period when the capelin were in the process of or had assembled on the wintering grounds off North and Northwest Iceland and also during the earlier stages of the spawning migration before the spawners entered the shallow waters at the southeast or southwest coasts of Iceland.

Since 1978 the stock has been surveyed jointly by Iceland and Norway in October, the Icelandic Marine Research Institute carrying out an additional survey in November 1981 when the October survey failed because of adverse ice conditions. In addition Iceland has carried out annual survey of the abundance of the same spawning stocks in January/ February 1979-1983 for comparison to the autumn estimates.

The basic technique was described by Nakken and Dommasnes (1975) and the variant adopted for the Icelandic capelin stock by Vihjálmsson et al. (1982). A summary of the technique is as follows:

Echo intensities from fish are registered continuously from the depth column along the ships' track and the values obtained integrated and recorded as average/nautical mile for each 5 nautical miles sailed. Trawling is undertaken when necessary in order to ensure adequate biological sampling and to check changes in the echo recordings. In all instances a Simrad 38 kHz acoustic system has been used.

The survey area was divided into suitable subareas depending on variations in the length composition of the capelin as observed from the trawl catches. The integrated echo intensity was converted to fish densities using the equation:

N = C × M × A

Where:

N is the total density in terms of numbers,
C is the number of capelin per unit area and dependent on mean length of the fish and instrument characteristics,
M is the mean integrated echo intensity for the relevant subarea (mm ref. 40 dB gain on echo integrator),
A is the subarea measured in square nautical miles.

The sum of the number of fish calculated for each subarea then gives the total stock size in number.

The reference C value had been previously established for the instrumentation of the Norwegian research vessel G.O. Sars and used in abundance surveys of the Barents Sea capelin. In the course of the surveys an intercalibration of vessels has generally been done in two different ways:

1.   Calibration with a standard reference target (Foote et al., 1981) simulating echo intensity over l nautical mile. This served as a check of the C value as well as giving an inter-ship relation irrespective of transducer directivity.

2.   Inter-ship calibration on a scattering layer of mostly O-group capelin and plankton thus including the directivity of the transducers.

The performance of the acoustic instrumentation of the vessels employed has always been remarkably similar and consistent. Examples from October 1982 are shown in Figures 7 and 8.

Fig. 7. Intercalibration on O-group capelin and plankton recordings. R.v. G.O. Sars and R.V. Bjarni Saemundsson, 21 October 1982.

Fig. 8. The regression line from the intercalibration shown in Figure 7.

Although it is known that the northern component of the stock feeding in the Jan Mayen-Greenland area does not always return to the wintering grounds until November the joint 2 ship autumn surveys have generally been scheduled for the first 3 weeks in October. The reason for this is of course the need for an assessment of the fishable stock abundance as early as possible. With the exceptions of an attempt during the period 25 September - 5 October 1979 as well as of the joint autumn survey in 1981, which had to be repeated in November because of adverse ice conditions, the October surveys have been successful. Examples of a recent survey grid and fish distribution are shown in Figures 9 and 10.

Fig. 9. Survey grid and trawl stations, 2-20 October 1982.

Fig. 10. The distribution and relative abudance of capelin, 2-20 October 1982

The Icelandic surveys in January-February have been carried out with 1-2 ships at the time when the spawning or fishable stock normally has a small distribution area, the location of which may vary considerably from one year to another. After having located the stock through rough scouting the actual assessment can, therefore, be carried out in a very short time period. However, weather conditions in the Icelandic area can often be difficult for acoustic surveying. When the stock has been located considerable time has on several occasions been spent in waiting for a suitable opportunity. The reward has been that acoustic assessments were carried out under good conditions and the results obtained credible.

The results of the acoustic estimates of the 1979-1983 spawning stocks are shown in figure 11. The points on the top line represent the results from the October- November surveys and those on the second line estimates of the abundance of the same spawning stocks obtained in January-February in the following year. When account has been taken of the catch and natural mortality during the intervening period there appears a small but consistent difference between the above sets of stock abundance estimates. On average for the 1979-1983 period this difference is in the order of 5-15 percent with the higher estimate on all occasions being obtained during the January-February surveys.

Fig. 11. Changes in abundance of the spawning stock of the Icelandic capelin as estimates by th acoustic method.

On several occasions it has proved possible to survey the spawning stock component 2-3 times within a very short time interval (10-30 days) as shown in Table 3. The difference within such series of estimates have usually been small and thus established confidence in the applicability of the method to the mature stock. When important deviations were observed they were immediately explainable. Cases in point are survey II in October 1978 when a short spell of windy weather lowered the abundance estimate and surveys I and II in January 1981 when cruise tracks were much too widely spaced with respect to the highly variable distribution density observed at the time.

The results of acoustic estimates of the abundance of the spawning or fishable stock of the Icelandic capelin are, therefore, considered accurate at least in a relative if not the absolute sense and show an alarming decline in stock size during the 1978/79-1982/83 period.

As yet it has proved difficult to assess the abundance of juvenile 1-2 group capelin. The reason for this is the frequent distribution of this stock component in or near areas periodically covered by drift ice. With the advancing of the ice border in autumn, part of the juvenile stock often becomes inaccessible and therefore not accounted for during the October surveys. This is indeed suggested by the highly variable occurrence of the juveniles in these surveys. A similar situation is encountered in winter. Recent surveying in late August has, however, brought promising results but further experiments are needed.

MANAGEMENT

Initially regulatory measures were mainly precautionary in nature. In Iceland there was a closed season in spring and summer from 1973 to 1978 lasting for 2-4 months. In 1979 the closed season was extended lasting until 20 August and in 1980 until 5 September. In 1981 Icelandic vessels were, however allowed to fish capelin from 10 August but until 15 September the feeding grounds of the juvenile 1-2 group fish west of 21°W as well as south of 68°N remained closed to them.

In 1975 a minimum landing size of 12 cm was introduced with a minimum mesh size of 19.6 mm. In order to facilitate the release of the small juvenile capelin Iceland increased the minimum mesh size to 21 mm in 1981.

Since most of the capelin spawn only once and die thereafter the main management objective is to prevent the spawning stock from being fished to the level of reduced recruitment, not to mention recruitment failure.

In 1979, when fishing was stopped in the 3rd week of March, about 600,000 tonnes were left to spawn according to acoustic estimates of stock size, when account had been taken of the fishery and the natural mortality rate.

In 1980, it was decided that while gaining further experience it would be inadvisable to reduce the spawning stock to more than 2/3 of the 1979 level or 400,000 tonnes. This has been the management objective since then.

It was soon recognized that overfishing could well have taken place already by the time when the results from the October abundance survey became available. The reason for this is the large fishing power of the multinational fleet during periods of easy availability of the capelin. Already in 1979 a precautionary TAC was, therefore, advised to be revised according to results from stock abundance surveys in autumn and winter. This advice was not accepted.

Instead during the 3 year period 1979-81, preliminary and somewhat arbitrary TAC's were set prior to the fishing season after bilateral negotiations between Iceland and Norway. These preliminary TAC's were then divided and allocated to individual vessels. Recommended TAC's were based on results of the acoustic stock abundance surveys carried out in October/November and January/February each season. By that time, however, the Norwegian vessels had always fished their quota and so had a large part of the Icelandic boats. The result was that it proved impossible to have the fishery stopped in time and in particular the 1981 and 1982 spawning stocks were reduced to extremely low levels as shown in Figure 11.

A summary of this unfortunate procedure as well as the actual catches taken from the 1980, 1981 and 1982 spawning stocks is given in the text table below.

1) The catch taken before 1 November. Zero TAC was then recommendedpending further survey results.

Although it had proved difficult to assess the abundance of juvenile capelin it was evident from acoustic as well as O-group surveys that the abundance of the 1980 year class, which was to spawn in 1983, was very low. The stock was considered to be in a grave condition even without any fishery taking place. A zero TAC was, therefore, recommended for the 1982-83 season pending results from the 1982 October abundance survey. This advice was accepted by all parties concerned, i.e. Iceland, Norway, the Faroes and EEC countries.

The results of the above survey corresponded to a spawning stock of about 190,000 tonnes in March-April 1983 and the fishing ban was extended to cover the whole season. As expected the 1983 January/February survey produced a slightly higher stock estimate or corresponding to about 220,000 tonnes of capelin spawning in March/April 1983. Estimates of spawning stock abundance as calculated from the January/February acoustic surveys for the period 1979-1983 are shown in the following text table.

The management objective of allowing 400,000 tonnes to spawn was, therefore, not reached except in 1979.

Concluding remarks

The Icelandic stock of capelin does not carry a long history of exploitation. As has been related the fishery began in a modest way in the mid sixties and did not expand seriously until in the early seventies. Then came a period of 5 years when the catch from the spawning stock was near the half million mark annually. In this period the fishery mainly took place during the 3 months preceding the spawning process and apparently did not produce excessive mortality rates.

When the summer and autumn fishery came into being the catch increased to over one million tonnes from 1976-1978, remained near that level for 2-3 years whereupon the stock collapsed.

Unfortunately information on the abundance of the maturing or fishable stock did not become available until in the autumn of 1978. Since then the abundance of each years' spawning stock (1979-1983) has been assessed twice, i.e. in autumn and in January/February in the following year, by acoustic methods. There is little doubt in the minds of the researchers that obtained these abundance estimates that they are reliable and give a correct picture of the development of the stock during the 5 year period which they cover.

Nevertheless, these abundance estimates have been heavily questioned most of the time by fishermen and authorities alike and advice on TAC's based on them not fully accepted. Restrictions upon the fishery have been too lenient and been imposed too late. The stock has been overfished.

There are several reasons for this:

1. General reluctance to face the economic consequences of suddenly having to limit catches to a fraction of what they were.
2. The almost unbelievably rapid decrease in stock abundance to a level that was little more than the recommended spawning stock size.
3. Before the fishing season opened the preliminary TAC, generally set far too high, was divided between the vessels participating in the fishery. The best part of the TAC had been taken before the October survey results became available while the Icelandic vessels, nevertheless, were in varying stages of filling their respective quotas. To have the fishery stopped in time was, therefore, difficult to say the least.
4. The occasional failure of some acoustic surveys, especially that from October 1981. Although the inadequacy of this survey was recognized at the time and a new survey executed immediately afterwards, its results were used to throw doubt on the applicability of the method by fishermen and authorities for a long time afterwards.
5. The nature of the capelin to aggregate and become available in areas of limited size the location of which may vary from year to year. The resulting concentration of the fishing effort gives the fishermen a false impression of abundance if their scouting effort is reduced from what it used to be. As the catch quotas and the extension of the closed season imposed in later years have indeed greatly lessened scouting activity, many of the fishermen have been under the impression that their own information on stock abundance is far better than it is in reality.
6. The improved acoustic gear, mounted in many vessels recently, also in most cases seems to have established an illusion of continued high abundance rather than helping in verifying the actual decline in stock size.

The faith in acoustics as a reliable tool of management is, however, improving. This is borne out by the fact that the authorities of all parties concerned accepted the advice of zero TAC for the 1982/83 season as described in a previous section.

In the absence of data on stock abundance prior to 1978 it is difficult to judge whether the recent decline in stock size is solely due to overfishing. Comparable data on the abundance of O-group capelin are available from 1972 onwards (Figure 6). These O-group indices point to a high level of recruitment until 1975/76 which is followed to a sharp downward trend and a much lower level since then.

The development coincides of course with the large increase in fishing effort brought about by the recent multinational summer and autumn fishery. It should, however, be pointed out that since 1976 large year classes of other fish species at the O-group stage have, as a rule, not been observed in the Icelandic area. Therefore, the decline in the abundance and recruitment of the capelin stock, brought about by overfishing, has probably been accelerated by a prolonged period of inhospitable environmental conditions.

REFERENCES

Anon. 1973-1981. Annual Reports of LODNUNEFND, the Icelandic Capelin Board.

Anon. 1982. Atlanto-Scandian Herring and Capelin Working Group Report. Part II. Copenhagen, 28- 30 October 1981. ICES, C.M. 1982/Assess:2.

Anon. 1982. Atlanto-Scandian Herring and Capelin Working Group Report. Part I. Copenhagen, 4-6 May 1982. ICES, C.M. 1982/Assess:12.

Anon. 1983. Report of the Atlanto-Scandian Herring and Capelin Working Group Report. Part I. Copenhagen, 27- 29 October 1982. ICES, C.M. 1983/Assess 4.

Foote, K.G., H.P. Knudsen, G. Vestnes, R. Brede and R.L. Nielsen. 1981. Improved calibration of hydroacoustic equipment with copper spheres. ICES, C.M. 1981/B:20.

Nakken, O. and A. Dommasnes. 1975. The application of an echo integration system in investigations on the stock strength of the Barents Sea capelin (Mallotus villosus, Müller) 1971-1974. ICES, C.M. 1975.

Stefánsson, U. 1962. North Icelandic waters. Rit Fiskideildar II, 8.

Vilhjálmsson, H. and P. Reynisson. 1979. Abundance estimates of the 1979 spawning stock of the Icelandic capelin. ICES, C.M. 1979/H:17.

Vilhjálmsson, H., P.Reynisson, J. Hamre and I. Rottingen. 191981. No.21.