REPORT OF THE FAO WORKING GROUP ON THE ASSESSMENT OF SMALL PELAGIC FISH OFF NORTHWEST AFRICA - BANJUL, REPUBLIC OF THE GAMBIA, 5-12 APRIL 2002

ISBN 92-5-004881-5

All rights reserved. Reproduction and dissemination of material in this information product for educational or other non-commercial purposes are authorized without any prior written permission from the copyright holders provided the source is fully acknowledged. Reproduction of material in this information product for resale or other commercial purposes is prohibited without written permission of the copyright holders. Applications for such permission should be addressed to the Chief, Publishing Management Service, Information Division, FAO, Viale delle Terme di Caracalla, 00100 Rome, Italy or by e-mail to copyright@fao.org

© FAO 2002

PREPARATION OF THIS DOCUMENT

A permanent FAO Working Group composed of scientists from the coastal States, and from countries or organizations that play an active role in Northwest African pelagic fisheries, was established in March 2001. The main objective of the Working Group is to improve the assessment of the small pelagic resources, as well as to advise options for management and exploitation of their fisheries and to ensure optimal sustainability of the resources for the economic benefit of Morocco, Mauritania, Gambia and Senegal.

The second meeting of the FAO Working Group on the Assessment of Small Pelagic Fish off Northwest Africa was held in Banjul, Republic of the Gambia, from 5 to 12 April 2002. Altogether 17 scientists from nine different countries participated.

A first editing of the report was made by all the participants of the working group. We are grateful to Stephen Cofield, Marie-Thérèse Magnan and Merete Tandstad for their assistance in the final editing of this document.

ABSTRACT

Distribution:
Participants in the Working Group/Participants au Groupe de travail
FAO Regional Fishery Officers/Fonctionnaires des pêches régionaux de la FAO
FAO Fisheries Department/Département des pêches de la FAO
NORAD
IMR
Ministery of Agriculture in Netherlands and RIVO Institute of Netherlands

1. INTRODUCTION

The second meeting of the FAO Working Group on the Assessment of Small Pelagic Fish off Northwest Africa was held in Banjul, Republic of the Gambia, from 5 to 12 April 2002. The overall objective of the working group was to contribute to the improved assessment of small pelagic resources in Northwest Africa and the analysis of fisheries management and exploitation options aimed at ensuring optimal and sustainable use of small pelagic fish resources for the benefit of coastal countries.

The species to be assessed by the group were: sardine (Sardina pilchardus), sardinellas (Sardinella aurita and S. maderensis), horse mackerels (Trachurus trecae and T. trachurus) and mackerel (Scomber japonicus), in the region between the southern border of Senegal and the northern border of Morocco.

The meeting was funded by project GCP/INT/730/NOR: International co-operation with the Nansen Programme, the Netherlands Institute for Fisheries Research (RIVO) and the Institute of Marine Research (IMR), Norway and organized by FAO.

Altogether 17 scientists from 9 different countries participated. The chairman of the group was Mr. Reidar Toresen, Norway.

The meeting was a follow-up of the meeting held in Nouadhibou, Mauritania, from 24 to 31 March 2001.

1.1 Terms of reference

The terms of reference for the group were:

1. Presentation of working papers on research activities
2. Presentation of reports on acoustic surveys 2001
3. Review of catch, effort and biological data from 2000 and 2001.
4. Report on the progress made on age readings of sardine in Morocco and effort studies in Mauritania
5. Updating the existing data base
6. Updating stock assessments for all species
7. Formulation of management advice
8. Co-ordination of biological research projects

1.2 Participants

1.3 Definition of working area

The working area for the working group is defined as the waters between the southern border of Senegal and the northern border of Morocco.

1.4 Structure of the report

A separate section is devoted to each of the main groups of species (sardine, sardinellas, horse mackerels and mackerel). For each of these, standardized information is given on stock identity, the fisheries, catch and effort, biological data, abundance indices, assessment, management recommendations and future research.

1.5 Course in VPA-methodology

During the first two days of the meeting, a course in VPA-methodology was held. The first day, Ad Corten lectured and guided the group through the basic principles of the Virtual Population Analysis (VPA). On the second day, Reidar Toresen lectured and guided in the application of the separable VPA and the Integrated Catch-at-age Analysis (ICA) (Patterson and Melvin, 1995). The applications based on Excel-sheets as presented in the FAO Fisheries Technical Paper 400 (Lassen and Medley, 2001), were shown and the principles were explained. The Excel-based ICA-version provided with the technical paper was adjusted so that it could be used for the stocks analyzed in the working group.

1.6 Assessment methods

Separable VPA

The separable VPA model (Pope and Sheperd, 1982) separates the exploitation rate from the selectivity pattern. The assumption is that the exploitation pattern remains constant over some significant period of time and that only the overall level of exploitation varies between years, so that:

F_ay = S_aE_y

The fishing mortality (F_ay) is separated into an age-dependent exploitation pattern (S_a) and an age-independent exploitation level (E_y). The catch data are used to estimate the S_a and E_y. The method is further described by Pope and Shepherd (1982). The basic principle of the estimation of the age and year effect is to estimate a theoretical catch (by year and age), and to minimise the difference between this catch and the observed catch by the best combination of S_a and E_y.

Integrated Catch Analysis (ICA)

The Integrated Catch Analysis (ICA) (Patterson and Melvin, 1995) is a method to make a model of the stock, which gives the best statistical fit to all available data. The method can make use of both age disaggregated and non disaggregated data. The ICA makes use of an object function, and minimise the difference between the catch at age given by the model and the observed catch at age. The method is described by Deriso et al. (1985) and Fournier and Archibald (1982). The flowchart for the Excel version of ICA as used in the working group is in Appendix I.

1.7 Age reading workshop

Sardine

An age-reading workshop was organized by the Atlantic Scientific Research Institute of Marine Fisheries and Oceanography (AtlantNIRO), Kaliningrad, Russia from 28 - 31 August 2001, and the report was presented by Ana Maria Caramelo.

The main objective of the workshop was to standardise age reading methods for the Northwest African sardine and to establish uniform age reading criteria.

Specifically, the workshop aimed to:

a) Familiarise the participants with the terminology used during the workshop, including the terms “accuracy”, “precision”, “bias”, “average age”, “modal age”, “agreement”, etc;

b) Discuss the results of the 2000 sardine otolith exchange with the otolith readers and provide suggestions on how to improve their age reading techniques;

c) Describe and discuss the otolith processing methods and the age estimation criteria used by the different fishery institutes;

d) Present a synopsis of the biology of the north-west African sardine focusing on aspects such as stock distribution, spawning period and juvenile distribution, which are relevant to age reading;

e) Present otoliths images from the Iberian Sardine agreed collection as outcome knowledge in order to promote discussion on otoliths interpretation;

f) Perform a reading based on a sub-sample from the exchange and discuss the results immediately after reading in order to correct the age interpretation of each otolith reader;

g) Prepare a manual on sardine age estimation that would provide guidelines to experienced and inexperienced otolith readers on how to read and interpret the rings in order to achieve high precision;

h) Provide recommendations on how the quality of the age readings can be improved;

i) Provide information on the bias, accuracy and precision of the age reading by age group for each reader and for the whole group.

Age readings were analysed with a spreadsheet developed by Guus Eltink (Eltink et al., 2000). The total percentage of agreement was 72.5% with a coefficient of variance (CV) of 41.4%. The high CVs obtained for some otoliths correspond to individuals aged between 0 and 1 year. These high values (greater than 100%) are not an indication of high variation in age determination. The percentage of agreement for each reader was presented and shows an overall agreement higher than 80%. For ages 0, 1 and 2 all readers estimate the same mean length at age of the workshop sample by reader. Some differences were observed for ages 3 to 5. The overall conclusions of the meeting could be summarised in the following sentences: the criteria adopted in this workshop must be applied to estimate the age of the sardine in the area and the reliability of the age reading should be taken into account. It is convenient that two readers from each country read the otoliths, discuss and agree with the age readings.

Management decisions are made on the basis of the knowledge of the current state of the stock in relation to defined reference points. So different age-length keys, applied to the same length catch matrix can provide a different understanding of the status of the stock and consequently lead to different management decisions. The report from the workshop will be published as FAO Fisheries Report. No.685.

Sardinella

A new exchange of otoliths of sardinella started in May 2002. The rules and the schedule are in Appendix II. A workshop on sardinella age readings will be held in the last quarter of 2002.

2. SARDINE

2.1 Stock identity

Following the procedure of previous meetings (FAO, 1985; 1990; 1997; 2001) the Working Group adopted the existence of three stocks:

• Northern stock (35^o45'-32^oN)
• Central stock (32^oN-26^oN) (Zone A+B)
• Southern stock (26^oN - southern limit of distribution) (Zone C)

2.2 The fisheries

Four fishing zones are distinguished in this report: the Northern zone, Zone A, B and C (Figure 2.1.1). The fleets operating in these fishing zones are very different and composed of traditional small purse seiners, industrial seiners, pelagic trawlers and factory vessels working in combination with seiners. For a detailed description of the fisheries in the four zones, see Chapter 2.2 of the 2001 Working Group Report (FAO, 2001).

2.3 Catch and effort data

The new data available for 2000 and 2001 are:

- Annual catches of the “Moroccan seiners”, the “Ukrainian pelagic trawlers and others”, “non-EU pelagic trawlers working in Mauritania”, and “EU pelagic trawlers”;

- Annual fishing effort of “Ukrainian pelagic trawlers and others”, “non-EU pelagic trawlers working in Mauritania”, and “EU pelagic trawlers”.

The series of total annual catches (tonnes) for the different fleets working in the northern zone, zone A, zone B and zone C have been updated for the years 2000-2001 (Table 2.3.1). The catch and effort in the different fleets and zones are presented in Table 2.3.2. The effort series have also been updated and are presented in Table 2.3.3. It should be noted that the fishing effort of Ukrainian and other pelagic trawlers presented in Table 2.3.3 refers only to the area between Cape Bojador and Cape Blanc. The same units of fishing effort for each type of fleet have been used as during the previous WG meeting (FAO, 2001).

In the northern zone the catches of sardine show a decline, and the catches in 2000 and 2001 are in the order of 5000 tonnes per year. It should be noted that this fishery has recorded important catches of other target species, and that the catch of anchovy reached a record of more than 30 000 tonnes in 2001.

In zone A, the catches of sardine show a declining trend since the early 1980s. A slight increase is noted since 1998. In 2000 and 2001 catches have been in the order of 50 000 tonnes a year. The trend in fishing effort shows a similar pattern.

In zone B, the catches of sardine are increasing. In the last two years the catches have exceeded the record of 1995. The activities of the seiners also show an increasing trend.

For zone C, the catch and effort of sardine in 2000 and 2001 have dropped after the termination of the activities by the Spanish and Russian fleets.

2.4 Biological data

Zones A and B

Length distributions by quarter and year, corresponding age-length keys and age compositions were provided to the meeting by INRH for the Moroccan catches in 1993-2000. Length distributions by quarter for the year 2001 are also available and age compositions of the catches.

Zone C

New data provided to the Working Group in 2002 are rather scarce and incomplete. They refer only to some of the fleets currently operating in the Zone and can be summarized as follows:

- Length composition of the catches made by Moroccan purse-seiners in January and February 2000 and in May, October and December 2001.

- Length composition (in fork length) of the catches made by the European Union pelagic trawlers in 1999, 2000 and 2001. The months sampled as well as the sample sizes are unknown.

- Length composition of the catches made by the Russian pelagic trawlers in January (2 samples) and December 2000 and in January (2 samples), February and March 2001.

- Age-length keys of the catches made by the Russian pelagic trawlers in January (2 samples) and December 2000 and in January (2 samples), February and March 2001. These data have also been provided combined in the form of quarterly (1st and 4th quarters of 2000 and 1st quarter of 2001) and annual (2000 and 2001) age-length keys.

- Mean length at age (cm) and mean weight at age (kg) from the catches made by the Russian pelagic trawlers in the 1st quarter of 2000 and 2001.

- Maturity ogive from the catches made by the Russian pelagic trawlers in 2000.

2.5 Abundance indices

2.5.1 Catch per unit of effort

The CPUE series of the major fleet components in Zone C are presented in Figure 2.5.1. CPUE series were calculated by dividing catches in tonnes by the nominal fishing effort expressed in fishing days. The evolution of the different series in time is rather heterogeneous with no clear trends.

2.5.2 Acoustic surveys

a) Northern zone (R/V Al Amir Moulay Abdallah)

An acoustic survey in this area was conducted in December 2001 by R/V “Al Amir Moulay Abdallah”. The results of this survey show the presence of sardine especially in the region off Larache and Kenitra. The biomass of sardine in the region between Tanger and Casablanca is estimated at 20 000 tonnes.

b) Zone A and B (Safi - Cape Bojador) (R/V Dr. Fridtjof Nansen)

Cape Bojador - Cape Juby^[1]

The biomass in this sector was estimated at 334 000 tonnes in November/December 2001, which is 48% below the estimates of the November/December 2000 and May/June 2001 surveys. The length distribution shows two modes, one at 10 cm and one at 19 cm. The number of fish was estimated at 9.5 billion in 2001, which is a reduction of almost 50% in comparison to 2000.

Cape Juby - Safi

In this sector the biomass of sardine in November/December 2001 was estimated at about 900 000 tonnes whereas an estimate of 800 000 tonnes was obtained during the May/June survey of that year. The stock consists mainly of sardines less than 16 cm, which constitute more than 70% of the stock in number in this sector.

Total biomass - Safi to Cape Bojador

The total biomass of sardine between Safi and Cape Bojador in November/December 2001 is estimated to be close to 1.2 million tonnes, which is 25% less than the estimates obtained for the same period in 2000 and May/June 2001 (1.5 million tonnes).

c) Zone C (R/V Dr. Fridtjof Nansen)

After the drop in 1997, the biomass of sardine in the southern zone gradually increases since 1998. It has increased from 1.2 million tonnes in 1998 to 3.5 million tonnes by the end of 2001. In the November/December 2001 the adults (longer than 16 cm) constituted more than 80% of the stock in numbers. The juvenile stock component, with a length of less than 16 cm, was relatively weak in comparison to the two previous years (November/December 1999 and 2000).

d) Zones A+B+C

It was concluded that the total stock of sardine in the area between Safi and Cape Timiris (Mauritania) in November/December 2001 (4.7 million tonnes) had remained at the same level as in the same period in 2000 (4.5 million tonnes). During the survey, sardines were encountered only in the area between Safi and Cape Corvero (21^o45'N). The length distribution was dominated by adults, which constituted more than 70% of the stock in numbers.

The acoustic surveys by R/V Atlantniro in Mauritanian waters in 2001 have not provided an estimate for sardine in this area.

2.6 Assessment

Considering the lack of new information concerning the separation of stocks, the WG followed the procedure of previous meetings (FAO 1990, FAO 1997, FAO 2001) and attempted to make separate assessments for the central stock (zone A and B) and the southern stock (zone C). The northern stock was not assessed, due to lack of data.

Central stock (zone A+B)

Input data

- Annual age compositions for the period 1976-2001. The series for 1976-1992 for all fleets was taken from the report of the WG meeting in 1997 in Casablanca. The series 1993-2000 for the Moroccan fleet was combined with a series for the Spanish fleet for the period 1993-1995, provided during the meeting. For the year 2001, the available length distributions were converted into age distributions using age-length keys of the previous year (Table 2.6.1a).

- The average weights by age in the catches are presented in Table 2.6.1b. They have been calculated using the number of observations in the series for 1983-1992 (FAO 1997) and 1993-2000 (Moroccan fleet) as weighting factors.

- The average weights by age in the stock are presented in Table 2.6.1b. They are based on data for the first quarter in the period 1993-2000.

- The same maturity ogive and natural mortality (M = 0.6 per year) were used as in FAO (1997) (Table 2.6.1c and 2.6.1d)

- As abundance indices, the numbers of fish >15cm estimated during acoustic survey of R/V Dr. Fridtjof Nansen were used (Table 2.6.1e).

Method

Models used were the separable Virtual Population Analysis (Pope 1977, 1979; Pope and Shepherd 1982; Stevens 1984), and the Excel version of Integrated Catch Analysis (ICA) (Lassen and Medley, 2001). The period 1995-2001 was used for the separable part of the VPA.

Results

The analysis of the time series 1976-2001 by separable VPA yielded unrealistic results, both in terms of fishing mortality and level of biomass. High residuals were found for certain age groups, which could indicate possible errors in the input data.

The application of ICA gave the following results:

- Spawning stock biomass increased from 1983 onwards to reach a maximum in 1986. It then declined gradually until 1992 when it dropped sharply.

- Fishing mortalities show a trend that is opposite to spawning stock biomass.

- The patterns of theoretical and observed catches are diametrically opposed.

On the basis of these results, the WG could not reach a conclusion concerning the state of the central stock.

Southern stock (Zone C)

As for Zones A+B the Working Group reviewed the available information for the stock in zone C (Table 2.6.3). It was noted that there were still important gaps in data from the past years and that the new data submitted at this year’s meeting were scarce and incomplete in terms of the type of information collected as well as in the representation of fisheries and fishing seasons (see Sections 2.3 and 2.4).

The last attempt to assess this stock was made in 2001 by applying the Separable VPA using the Lowestoft software (Darby and Flatman, 1994). Results from the analyses showed high and no randomly distributed residuals. Also the regression statistics results were not satisfactory. The Working Group in 2001 therefore concluded that the results of the analyses were not reliable, possibly due to anomalies in the input data.

Considering that there is no new information, the Working Group decided not to make a new attempt at the present meeting to assess the Sardine stock in Zone C.

2.7 Management recommendations

Because of uncertainties in the assessment, a prognosis has not been made and the results of the assessment should not be used for management purposes.

Results of acoustic surveys indicate an increase in stock size during the last years.

2.8 Future research

1. A biological sampling programme has to be established for all fisheries in the area. This sampling programme has to cover the entire fishing season and all fleets. At least 100 fish should be measured per sample.

2. Sampling for age composition should cover the entire length distribution. In each length class, ten otoliths should be taken.

3. All input data for the assessment models have to be ready on the first day of the WG meeting. The data series should include data for the previous year.

4. The members of the sardine sub-group should remain in contact by E-mail and exchange information on the collection of data in the period between meetings.

3. SARDINELLAS

3.1 Stock identity

No new studies on stock identity of sardinella have been carried out in the interim period since the last meeting of the Working Group in March 2000. Following the procedure for that meeting, the Working Group agreed on the existence of one stock for the two species of sardinellas (FAO, 2001). For more information on stock identity, the readers are referred to the 2001 Working Group Report (FAO, 2001).

3.2 The fisheries

Gambia

The sardinellas are not target species in The Gambia. The landings of sardinella in the artisanal fishery are very low and no major variations in landings have been observed. There are no industrial pelagic fishing vessels.

Senegal

The resources of coastal small pelagics are exploited by an artisanal fishery and an industrial one. The artisanal fishery uses motorised canoes and a variety of fishing gears. The industrial fleet consists mainly of trawlers and seiners, some of which are of foreign origin, which work under commercial contracts.

The most important landings of coastal pelagic fish are made by the artisanal fleet, which is still developing. The main fishing gears are purse seine, ringnet and beach seine.

A census of the canoe fishery, made in November 1993, October 1995 and October 2001 gave the following results:

The industrial fleet consists of small seiners (“sardiniers Dakarois”) with a low tonnage, and foreign vessels with high engine power. The number of small seiners has been declining for many years. In 2001 and 2002, only five units have taken part in the fishery. Foreign vessels constitute a more important, but irregular factor. In recent years, the large seiners and Russian trawlers have no longer worked in the Senegalese EEZ. The seiners disappeared in 1994 and the Russian trawlers operated until June 1999. At the moment, the Senegalese industrial fleet consists only of the small seiners based in Dakar.

Mauritania

The fishery statistics of the CNROP for the years 2000 and 2001 have not yet been published. However, preliminary statistics are presented. Data on catches and fishing effort are split into two categories: European Union and others.

The sardinellas in Mauritania are exploited mainly by pelagic trawlers from the European Union, particularly the Netherlands. In 2000, three large units have been added to the fleet, which resulted in a considerable increase in fishing effort by that fleet. With the exception of the new vessels, the entire fleet returns to European waters by the end of the year.

Several factors affected the sardinella fishery in 2001. The most important of these were hydrographic changes, an increase in fishing effort, and the introduction of new regulations.

- An increase in temperature has been observed during the last two years. This higher temperature is caused by a reduction of upwelling in July - October, which results in lower food abundance.

- A new regulation was introduced in Mauritania in 2001, which resulted in an extension of the fishing limits for pelagic trawlers from 12 miles to 13-25 miles off the coast. This measure is aimed at the protection of the artisanal fleet and the coastal fish resources. The measure probably reduces the possibility for pelagic trawlers to exploit the coastal concentrations of sardinella.

- A substantial increase in fishing effort has been noted in recent years.

The combined effect of these three factors has resulted in a significant reduction of the catches per unit of effort.

Morocco

The exploitation of sardinellas in the southern zone gained importance in the early 1990s. In the 1990s, this fishery was conducted mainly by pelagic trawlers that worked for Moroccan joint ventures, and trawlers operating under the fishery agreement between Morocco and the Russian Federation. The characteristics of these fleets and the conditions under which they operate are described in FAO (2001).

In 2000 and 2001, only trawlers working under joint ventures have operated in this area. The activities of the Russian fleet ceased at the end of 1999, following the expiry of the fishery agreement.

The main fishing zone for the pelagic trawlers is found between Dakhla and Cape Blanc. During the last five years (1997-2001), the number of vessels in this fleet has declined from 35 to 22. The highest catches of sardinella have been recorded from July to October in the region between Cape Barbas (22°N) and Cape Blanc (19°40°N)

3.3 Catch and effort data

The catches for the period 1990-2001 are presented in Tables 3.3.1a (S. aurita) and 3.3.1b (S. maderensis). Effort data are presented in Table 3.3.2. Figure 3.3.1 shows the trend in sardinella catches in the whole region combined.

Catches

Gambia

Gambia presented data on the annual catch of the two sardinella species for both the industrial and artisanal fishery for the years 1990-2001. The landings of the two species are still low. Although it is government policy to encourage the development of the pelagic fishery, industrial and artisanal fishers are yet to enter into this fishery.

Senegal

Senegal has data on the annual catches of the artisanal fishery from 1974 to 2000. For the purpose of the present meeting a preliminary estimate was made of the catches in 2001. Moreover, the annual catch series for the years 1997-1999 was corrected, using a better raising factor. Catches of the industrial fishery refer to the period 1966-2001.

Data collected on board Russian trawlers do not differentiate between Sardinella aurita and S. maderensis. These two species are recorded as one category “Sardinellas”. In order to obtain an estimate of the catch of each species, we have used starting from 1997 the percentages of these species in catches made by R/V Dr. Fridtjof Nansen during its acoustic surveys in the area. In earlier years, the split was based on observations made by scientists on board the vessels.

Mauritania

Catch and effort data are split in two groups: European Union and others.

The estimate of the industrial catches by month for the two species of sardinella combined was based on logbook data provided by the inspection service DSPCM^[2]. These data have been split into the two species using the results of the scientific observer programme on board EU pelagic trawlers in 2000 and 2001.

In recent years the artisanal fishery has shown a gradual development. Landings of sardinella by this fishery have been earlier estimated at 3000 tonnes/year. The present level of catches is not known, but it is assumed that these are above the earlier level, due to the licensing of Senegalese canoes to fish for small pelagics in the Mauritanian EEZ. Starting from 2000, a total of 250 Senegalese canoes have been authorised to fish in the Mauritanian EEZ.

Morocco - Zone C

The WG had available monthly catches of sardinella for the years 1993 to 2001 from the Russian fleet and the “Ukrainian and others” fleet. The split of catches into the two species (Sardinella aurita et S. maderensis) was based on sampling on board vessels north of Cape Blanc during 1995-1999. For the years 2000-2001 the split of catches by species was based on hauls made by R/V Dr. Fridtjof Nansen during acoustic surveys in the area.

Effort

Total effort by country and fleet is presented in Table 3.3.2 for the years 1990-2001.

In Gambia, landings of the sardinellas are considered as by-catches and therefore no effort is available for this fishery.

Senegal provided effort data for encircling nets (purse seines and ringnets) for the artisanal fishery (number of trips). The data are available starting from 1982 on a monthly and annual basis. Effort data for the industrial fishery are available on a monthly basis starting from 1976, and on an annual basis for the period 1962-2001 (fishing days).

In Mauritania, the number of countries involved in the fishery was 12 and 13 in 2000 and 2001 respectively. The number of vessels was 79 and 82 respectively. The increase in number of vessels is not reflected in the number of fishing days.

For the EU fleet, an index of total fishing effort has been estimated on the basis of the number of fishing days reported by the vessel in their logbooks. For each vessel, the index was corrected for motor power. It is assumed that fishing power is linearly related to engine power. Fishing effort is expressed in fishing days for a standard trawler of 10 000 HP. It appears that the fishing effort by the EU fleet has substantially increased in 2001 (Figure 3.3.2).

Using an index of CPUE for the Dutch fleet, the total effort exerted by other fleets has been estimated on the basis of their total catches. It appears that the effort on sardinella by the non-EU fleet has decreased during the latest years (Figure 3.3.2). This could reflect a change in fishing strategy, with the focus shifting to horse mackerel, mackerel, anchovy, sardine, etc. This is illustrated by the increased catches of anchovy during the last two years, which have reached a level of 104 000 tonnes.

Morocco presented effort data (fishing days) from 1993 to 2001 for the Russian fleet and the “Ukrainian and others” fleet in area C. As far as the Moroccan fleets are concerned, their fishing effort is mainly directed at sardine. This effort, therefore, was not used by the subgroup.

3.4 Biological data

Senegal

For each of the two species of sardinella, the CRODT provided monthly length frequency data for the catches of the artisanal fishery in the period 1986-2001 (except for 1992). For the round sardinella, effort data were provided for the industrial fishery in the period 1986 - 1993.

For the artisanal fishery, the length distributions have been derived from catches by seiners and ringnets. For the industrial fishery, the length distributions were taken from the catches taken by the purse seiners from Dakar. The data have been raised to the total catches.

Other biological information available for the area is found in the report of the FAO working group meeting in 1993 (FAO, 1994).

Mauritania

Several length distributions are available for the Dutch vessels in 1999-2001. The data are derived from the CNROP observer programme on board the Dutch vessels in 1999-2001.

Annual age and length distributions for Russian vessels are available from 1992-2001.

A comparison between length distributions from Russian and Dutch catches in 1999 and 2000 has shown that during these years the two fleets have fished the same concentrations of Sardinella aurita (Figure 3.4.1).

The length distribution of the Dutch catches in 2001 has been extrapolated to the total catch of the industrial fleet in that year. Subsequently, a Russian age-length key for 2000 has been used to convert the 2001 length distribution into an age composition.

Age composition Sardinella aurita for all areas combined

An attempt was made to calculate age compositions of Sardinella aurita for all areas combined for the period 1994-2001. For the Senegalese artisanal fishery, catches in numbers by length were converted into age groups by applying Russian age length keys. For years in which no adequate age/length keys were available for area 34.3.1, the combined age length key for area 34.3.1 + 34.1.3 was used. The age composition of the total industrial catch in area 34.3.1 + 34.1.3 was estimated by extrapolating the age composition of the Russian industrial catches. The results of this calculation are presented in Table 3.4.1.

3.5 Abundance indices

3.5.1 Catch per unit of effort

For Senegal, the development of the CPUE for Sardinella aurita and S. maderensis is presented for the artisanal and the industrial fishery in Figures 3.5.1a and 3.5.1b. The CPUE for the artisanal fishery is expressed in tonnes per trip, and the CPUE of the industrial fleet is expressed in tonnes per fishing day.

The CPUE for S. aurita, both in the artisanal and industrial fishery, shows a slightly increasing trend from 1999 onwards. This increase followed a decline in both CPUE series during the period 1996-1999.

For S. maderensis, the continuous decline of CPUE in the industrial fishery from 1995-1999 was followed by a slight increase in 2000, and a decline in 2001. The CPUE for S. maderensis in the artisanal fishery is more variable, with peaks in 1996 and 1999.

Combining the two species, it appears that both in the artisanal and the industrial fishery in Senegal the CPUE shows an increasing trend in the last two years (Figure 3.5.1c).

Another CPUE series was calculated for the Dutch fleet in Mauritania (Figure 3.5.1d). This figure refers to the combined catches of both species of sardinella. The CPUE is expressed in tonnes per day of a standard trawler of 10 000 HP. The results for the period 1996-2001 show a decline from 1998 to 2001. This decline could be partly due to a reduction of the stock (and to the increase in fishing effort in Mauritanian waters). For 2000 and 2001, the decline in CPUE is probably also related to the factors mentions earlier (high water temperatures and the resulting emigration of sardinella across the northern border and extension of fishing limits).

3.5.2 Acoustic surveys

General remarks concerning sardinella acoustic surveys

Series of acoustic surveys are available from surveys by the Norwegian R/V Dr. Fridtjof Nansen and the Russian R/V AtlantNiro and R/V Atlantida (Figure 3.5.2 (FAO, 2001). Since 1995, the Norwegian surveys have always been conducted in November-December. This is the time of the year when sardinella are migrating south. Although the distribution of the stock during the survey may change from year to year (depending on the timing of the southward migration), the surveys are supposed to provide reliable indices for the total stock in the area. So far, the results of the surveys have been expressed in biomass and in numbers per length group. No age compositions are yet available for the surveys

Results of acoustic surveys

The indices from surveys of the R/V Dr. Fridtjof Nansen (Figure 3.5.2) show the same trend for both species of sardinella, i.e. a decline from 1995 to 1997 and a recovery in 1999. In the last years (2000 and 2001), the index has levelled out for S. aurita and increased again for S. maderensis. The relative abundance of S. aurita and S. maderensis in the acoustic series shows that S. aurita dominated in all years except in 1995 and in 1998.

3.6 Assessment

Using information on length and age distribution, the WG has tried to apply a VPA on the sardinellas, but the results were considered not satisfactory.

Methods

The BIODYN model (Punt & Hilborn, 1996) has been used for the assessment (FAO, 2001). The parameters that gave the best fit were estimated for both species using the OBSTWO spreadsheet.

Input data

Although total catch data were available for both species of sardinella, very few data existed on effort that was directed specifically at these species. There is the Dutch fleet that targets sardinella, but the effort series is too short. Therefore, the WG proposed to use the CPUE of the artisanal fishery in Senegal as an index of abundance. In addition, the stock estimates by R/V Dr. Fridtjof Nansen were used for the period 1995-2001. Catch data used were those presented in Tables 3.3.1 and 3.3.2.

It is important to remember that the use of dynamic models requires the estimation of additional parameters, viz. R (intrinsic growth rate), K (mean equilibrium biomass of the unexploited population), and B_init (biomass before the start of the first recorded catch). If data are available from the start of the fishery onwards, it is reasonable to assume that B_init equals K. The ratio between CPUE and biomass, the constant q (catchability coefficient), is also estimated.

Results and discussion

A good simulation of the observed data series for Sardinella aurita and S. maderensis by the Schaefer model was obtained, except for the Nansen indices for S. maderensis (Figures 3.6.1a and b). The values estimated for the main parameters are:

Sardinella aurita (CPUE index artisanal fishery Senegal)

Sardinella aurita (Nansen abundance index)

Sardinella maderensis (CPUE index artisanal fishery Senegal)

where SSQ = sum of differences between logarithms of CPUE observed and estimated.

It is noted that for S. aurita the two indices give similar estimates for MSY, with an SSQ that is much lower in case of the artisanal CPUE series. The results show a fluctuating CPUE, with an increasing trend during the last two years for both series of indices (Figure 3.6.1a and 3.6.1b).

For S. maderensis the results of the model show a relative stable CPUE during 1990-2001, except for the years 1992-1993 and 1997-1998 (Figure 3.6.1c).

The results obtained provide useful information on the dynamics and possible states of the stock. On the basis of these results, a precautionary TAC of 600 000 for the two species combined would seem to be justified. This estimate is higher than the one given in last year’s report (500 000 tonnes). This seems to be consistent with the increasing trends in some CPUE series for certain fleets. However, considering the uncertainties in the results of dynamic models, the WG decided to retain the value of last year, i.e. 500 000 tonnes.

3.7 Environmental effects on the stock of Sardinella aurita

The analysis of historical data has shown that recruitment strength in Sardinella aurita depends on environmental factors. Especially high summer temperatures have a negative impact (Kudersky et al., 2000; Barkova et al., 2002). The hot summers of 2000 and 2001 could have had a negative effect on the recruitment produced in those years.

3.8 Management recommendations

In view of the uncertainty of results based on global models, the working group considers that the results obtained this year should not be used to revise the recommendation given last year. It is therefore recommended to set a precautionary TAC of 500 000 tonnes for both species combined in the total area. This figure corresponds approximately to the maximum level observed in the period 1990-2001.

It should be noted that the total catch in the region has dropped below 500 000 tonnes in the last three years, despite a substantial increase in fishing effort of the EU industrial fleet in Mauritania. Catches per unit of effort in the EU-fishery in Mauritania show a decline over the last 4 years, suggesting a reduction of the stock in this area. This information, combined with the prospects of poor recruitment, should be a good reason to set the TAC not higher than 500 000 tonnes.

3.9 Future research

1. Continue the surveys by R/V Dr. Fridtjof Nansen and the calibration exercises

2. Joint surveys with research vessels of the sub-region

3. Starting and exchange of otoliths and the organisation of an age reading workshop

4. Conduct studies aimed at finding a representative index of fishing effort on sardinellas

5. Stimulate and conduct studies on the application of analytical models

6. Starting a monitoring programme of the artisanal fishery for sardinella in Mauritania.

4. HORSE MACKEREL

In this section we deal mainly with two species of horse mackerel, Trachurus trachurus (Atlantic horse mackerel) and Trachurus trecae (Cunene horse mackerel) for which time series on catches and age composition are available. The false scad (Caranx rhonchus) is gaining importance in the commercial fisheries in recent years and also in the acoustic surveys. Therefore, we consider in this section also the distribution and biology of this species. Furthermore we present the catches of this species in the sub-region.

4.1 Stock identity

For a detailed description of the stock identity of the two species of Trachurus, the reader is referred to the report of the WG meeting in Nouadhibou in 2001 (FAO, 2001).

Distribution and biology

The false scad distribution is more coastal than that of the two Trachurus species. Its abundance increases from Cape Timiris to the south. The species also migrates along the coast, from Guinea to Mauritania, following the seasonal shifts of the intertropical front (Garcia, 1982). It is absent from the northern area during the entire first half of the year. In Senegal, the spawning seasons lasts from April to November. In Mauritania, spawning lasts from April to August (FA0, 1989). The spawning area is situated south of Cape Timiris. The behaviour of the false scad resembles that of the other species of horse mackerels, but is more of a predator. Its food consists for 70% of fish, anchovy, squid and shrimp (Chavance, 1991).

4.2 The fisheries

The fishery of horse mackerel in the northern CECAF zone is conducted mainly by foreign fleets. These are fleets from countries in the former USSR (Russia, Ukraine, Lithuania, Latvia and Estonia), from Europe (Netherlands, France, UK, Malta, Sweden, Poland and Cyprus) and of a number of other countries (Panama, Ghana, Marshall Islands, St Vincent and Grenadine etc.).

General characteristics of the fleets fishing for small pelagics in the northern CECAF zone

In Mauritania and in the south of Cape Bojador (Zone C), the fleets consist mainly of large trawlers. The characteristics of these vessels have been presented in the report of the WG on small pelagics for this area (FAO, 2001). These fleets consist of vessels from Eastern Europe that have worked in the area during the last four decades, and more recently, of vessels from the European Union. The industrial vessels from the coastal states have a limited capacity.

The main changes that occurred during the last two years are related to the withdrawal of the Russian fleet from zone C due to the expiry of the fishing agreement by the end of 1999. Furthermore, in the Mauritanian zone an increase of fishing effort is noted by Eastern European fleets on horse mackerel.

4.3 Catch and effort data

The series of catch data has been updated until 2001. Catch statistics for the three species of horse mackerel are presented by country in Table 4.3.1a,b,c for the period 1990-2001.

For Senegal, the catch figures of horse mackerel in the artisanal fishery in 2001 are provisional. For the industrial fishery, however, the final catch figures were available. It should be noted that this category has been negligible since the departure of the Russian fleet in June 1999.

In the Moroccan area, the Atlantic horse mackerel is taken as by-catch by the local seiners. The target species of these seiners is the sardine.

In Mauritania, data are available for the industrial fleet. For the artisanal fleet, however, no information is available at all.

Fishing effort, expressed in hours fishing, has been standardised on the basis of data collected in the Mauritanian zone (the criteria are formulated in Section 4.5.1). Unfortunately, it was not possible to construct one complete effort time series covering the whole time period using the same criteria and different models were applied for different periods.

To split the horse mackerel catches in 2000 in the Mauritanian zone by species, data from the sampling conducted on board the EU vessels were used. Russian data, which are generally more suitable for this split, could not be used because they did not distinguish the false scad. This species was recorded in large numbers during the acoustic surveys, and also in catches of the EU vessels.

For the Moroccan and Senegalese waters, the split was made using the proportions of each species in the catches made by R/V Dr. Fridtjof Nansen during the acoustic surveys.

4.4 Biological data

For the stock of Trachurus trachurus between 26°N and 10°N, and the stock of T. trecae between 9°N and 23°N biological data are presented in Appendix II - Russian data.

The age compositions of the catches of Trachurus trecae, calculated on the basis of Russian age-length-key data and corrected by catch figures reported by various countries, are presented in Table 4.4.1. Similar data for T. trachurus are presented in Table 4.4.2.

4.5 Abundance indices

4.5.1 Catch per unit of effort

The CPUE constitutes both the basis of many assessment methods and an important factor for the regulation of fishing activities. In the case of small pelagics, the calculation of an index on the basis of CPUE is strongly criticised, mainly because of the large variability of these resources, and the fact that the fish are not randomly distributed. The concentration of these fish in schools may keep the CPUE constant, even when the abundance declines.

There are six types of vessel: STM, RTMA, RTMS, BMRT, BAT and undetermined. Only vessels with a minimum of 40% horse mackerel in their catch are used to calculate fishing effort. To estimate these indices, a GLM model was applied, using the method of Robson (1966) and the vessel- and year-effect.

The first model of the type

CPUE ~ year + vessel type + area + month

allowed to demonstrate the following effects:

- a year effect that represents the annual variation in abundance and possibly the annual variation in fleet efficiency

- a month effect that represents the spatial variation in abundance and the availability of the fishable part of the population

- an area effect that represents the spatial variation in abundance and the availabity of the fish

- a not-significant vessel effect

The second model was:

CPUE ~ vessel type: year

The interaction (vessel type: year) is not significant. The fact that there is no interaction between the variable “vessel type” and the variable “year” means that the variations in CPUE are similar during the different years.

The CPUE for the false scad showed opposite trends to CPUE series of the Atlantic horse mackerel during the entire study period. The relationship between the series of the two Trachurus species is less clear.

For T. trecae a similar trend is seen in the CPUE data and the acoustic indices of R/V Dr. Fridtjof Nansen.

4.5.2 Acoustic surveys

The time series of acoustic stock estimates for horse mackerel by R/V Dr. Fridtjof Nansen was presented in the report of the last WG meeting (FAO, 2001). This series has been updated to include 2001 data (Tables 4.5.1 and 4.5.2).

In 2001, the Cunene horse mackerel (T. trecae) was almost absent in the Mauritanian zone during the survey by R/V Dr. Fridtjof Nansen (November- December), but its abundance has increased in Senegal/Gambia. The Russian research vessel (covering only the Mauritanian zone in August) estimated the biomass of this species at 25 00 tonnes.

A drastic reduction in biomass of the Atlantic horse mackerel (T. trachurus) was observed during the acoustic surveys of R/V Dr. Fridtjof Nansen. The Russian vessel, that conducted its survey during the warm season, did not report in 2001 any detections of this species.

The false scad shows an increasing trend according to the results of both vessels, both in the Mauritanian and Senegalese-Gambian zone.

4.6 Assessment

VPA

For exploratory purpose, a separable VPA was run for the two species (Trachurus trachurus and T. trecae). Thereafter, an Integrated Catch Analysis (ICA) was run for each of the species separately.

Data on age composition and mean weight-at-age of catches by year, used by the 1998 Mauritanian working group (FAO, 1989), were updated, the available series covering the period 1979-2000. For T. trachurus the catch-at-age series presented in last year's report was changed because the group believed that there are two stocks, one from zone A+B and northwards in Morocco, and another one being distributed from zone C southwards (Figure 2.1.1). The catch-at-age matrix from 1990 to 2000 was recalculated so that the total catch (in tonnes) was consistent with the total catch reported by each country (Table 4.3.1 a and b).

The following parameters were used for both Separable VPA and ICA runs:

• For the years 1979-2000, the age composition in numbers and the mean weight-at-age of the catches were used for the analysis. Age 8 was considered a + group
• The natural mortality coefficient used was 0.5 per year (constant for all ages)
• The mean weight at age in the stock and in the catch for the two species are given in the input tables
• The mean proportion mature at age is given in the input tables.

For the ICA run the following object function was minimized for T. trecae:

SSQ = S ln(C - C) + S ln(I_acoust - Î)

Where SSQ is sum of squares, C is observed catch at age, C is estimated catch at age, I is total estimated biomass from the autumn R/V Dr. F. Nansen (1995-2000) surveys of horse mackerel and Î is the estimated spawning stock biomass of the ICA model (see flow chart in Appendix III).

In addition the following object function was minimised for T. trachurus:

SSQ = S ln(C - S) + S ln(I_acoust - Î)+ S ln(I_CPUE - Î)

where I_CPUE is the abundance index for the years 1995-2000.

The results of the Separable VPA and ICA for T. trachurus are presented in Figures 4.6.1-4.6.2, and for T. trecae in Figure 4.6.3.

The results of the separable VPA’s show that there are relatively high residuals in the analyses, especially for T. trecae.

The ICA was run successfully for T. trachurus. For T. trecae, the ICA did not find a reasonable solution. This may be explained by the large discrepancy between the stock trajectory of the separable VPAs (of the catch at age analysis) and the stock trajectory of the R/V Dr. Fridtjof Nansen acoustic time series for this stock.

The results of the ICA for T. trachurus indicate that the stock has been through fluctuations during the last twenty years, and that the stock has been in a declining phase in recent years.

The acoustic time-series for T. trecae shows the opposite trend with an increasing trend in stock biomass, with ICA Model.

During the trial runs of the Excel-version of ICA, it was discovered that the program found several different solutions (minimum SSQs), with quite different SSB and F-levels. The present assessments must therefore be looked upon as preliminary and the stock and F-levels must not be used for management purpose. However, the trends in stock and F’s seem to be estimated well.

4.7 Management recommendations

The annual catches of the different species have fluctuated strongly over time, but all species show an increasing trend from 1990 to 2001. The CPUE data, calculated for vessels that targeted horse mackerel, show a declining trend. Moreover, the acoustic estimates of R/V Dr. Fridtjof Nansen show a decreasing trend for both species of Trachurus. This decline is partly compensated by the increase in biomass of false scad.

Considering the many uncertainties in the assessment of these stocks, a precautionary approach should be taken in management of the stocks. For this reason, the WG recommends a restriction of fishing effort to the current level.

4.8 Future research

With the exception of recommendations 3 and 4 in last year’s report, some new recommendations have not yet been implemented. Therefore, the WG reiterates them as follows:

1. At least monthly sampling on board industrial vessels to establish species composition, length and age distribution etc. This activity is the most important.

2. Starting an age reading programme for the three species.

3. A full utilisation of historic data series on catches, effort and biological parameters.

4. The acoustic surveys should be continued and the acoustic estimates should be split by age groups, if possible.

5. A regional programme should be established to assess the entire stock including the coastal component of the horse mackerels and other species of small pelagics.

6. Stock assessment methods should be developed to use environmental factors to forecast possible changes in stocks, and to use the results of these methods as an additional source of information.

5. CHUB MACKEREL

5.1 Stock identity

The southern mackerel (Scomber japonicus Houttuyn, 1782) is a neritic-oceanic species, which is distributed in the Eastern Atlantic.

In the Central East Atlantic it is found in waters between 15 and 400m depth. The adults occur mainly in waters of more than 80m depth. However, during the period of the seasonal migration and the presence of unfavourable conditions on the shelf, the concentrations of adult fish can move outside the shelf edge over water of more than 1000m depth.

The mackerel has a varied diet, consisting of more than 50 different species: Euphausids, Copepods, Cladocerea, and also young fish (horse mackerel, anchovy, sardine, hake, mackerel), Crusteaceans and also Polychaetea. The diet depends on the age of the fish and the season.

Spawning of mackerel occurs between December and April at temperatures of 16.5°-18.5°C in the area between Cape Blanc and Cape Bojador, and between Cape Juby and Cape Ghir.

For the assessment of the mackerel, the WG (FAO, 2001) has assumed that there are two stocks in the region from Morocco to Senegal:

• The “northern stock” distributed from Bojador to the north of Morocco;
• The “southern stock” distributed from Bojador to Senegal.

During this meeting, only the “southern stock” has been considered.

5.2 The fisheries

The northern fishery between Tanger and Cape Bojador is conducted only by coastal seiners from Morocco that target sardine.

The southern fishery between Bojador and Cape Vert is conducted exclusively by pelagic trawlers from the former USSR (Russian, Ukranian and others), that operate under the fisheries agreement or under joint ventures. These trawlers fish for mackerel in Moroccan and Mauritanian waters. In Senegal and Gambia, the mackerel is taken as by-catch. The characteristics of the fleet are described in the section on sardine in zone C (2.2), sardinellas (3.2) and horse mackerel (4.2).

5.3 Catch and effort data

A series of catch and effort data, updated until 2001, is presented in Tables 5.3.1 and 5.3.2.

In the Moroccan zone between Tanger and Bojador, mackerel is a by-catch in the fishery for sardine. The total catch in stock North during 2001 reached about 25 600 tonnes, compared with about 33 100 in 2000.

In Southern stock for the vessels working under joint ventures (from Ukraine and other countries) that continued to fish in 2000 and 2001, the catch was 90 500 and 65 000 tonnes respectively. Fishing effort by this fleet has decreased from 3900 fishing days in 2000 to 2200 days in 2001.

In Mauritania, the total catch of the trawlers was 65 000 and 60 000 tonnes in the years 2000 and 2001 respectively. Fishing effort remained constant during these last two years at a level around 9 400 fishing days.

In Senegal and Gambia, mackerel is exploited by the artisanal and industrial fleets. The total catch in this area in 2000 and 2001 was around 2 000 - 2 500 tonnes.

5.4 Biological data

A summary of biological data is presented in Appendix III.

5.5 Abundance indices

5.5.1 Catch per unit of effort

Last year report explains how the CPUE was calculated (FAO, 2001). The total catches, total effort in fishing days and CPUE in tonnes per fishing day from 1992 to 2001 are presented in Table 5.5.1.

5.5.2 Acoustic surveys

Series of acoustic surveys from 1994 to 2001 are available from the surveys by Russian R/V ATLANTNIRO and ATLANTIDA. In 2000 and 2001 the surveys were conducted mainly in summer in the Mauritanian area (Figure 5.5.2).

The indices of abundance obtained by the acoustic surveys were too small and did not cover the whole area for all the years, so they could be used only for comparative analysis.

5.6 Assessment

Methods

An Integrated Catch Analysis (ICA) was run for this species. The input data are given in Table 5.6.1.

Data on age composition and mean weight-at-age of catches by year, used by the 1998 working group, have been completed for recent years. The available series covers the period 1992- 2001.

The following parameters were used for the ICA run:

• For the years 1992- 2001, the age compositions in numbers and the mean weights-at-age of the catches were considered for the analysis. Age 6 was considered a + group
• The natural mortality coefficient used was 0.5 per year (constant for all ages)
• The mean weight at age in the stock and in the catch is given in the input tables
• The mean proportion mature at age is given in the input tables

Series of catches at age and the series of CPUE from 1992 to 2001 were used in the model. For the ICA run the following object function was minimised:

SSQ = SS(lnC_y^obs - lnC_y^est)² + lS [(ln F_y - ln(qE_y)]²

Where SSQ is sum of squares, C_y^obs is observed catch at age, C_y^est is estimated catch at age, q is the catchability coefficient, F_y is the coefficient fishing mortality, and E_y is the effort.

Fish aged 4 years were considered to be the youngest age group fully represented in catches. A linear regression (I=q * N) was used to estimate catchability coefficient for all age groups. A total of 14 iterations was sufficient to get an acceptable value of the object function (0.0141).

Results and discussion

The results of the ICA show that the total biomass reached a maximum in 1995 (1.1 million tonnes) and decreased from 1996 to 2001. The residuals obtained are generally satisfactory (Table 5.6.2). The distribution of fishing mortality at age shows variations between years according to the change in strategy of the pelagic trawlers fleet in Morocco and Mauritania.

There is no information on the size distribution of mackerel in catches between Dakhla and Cape Barbas, where small fish is normally very abundant and sometimes caught in large numbers.

Generally, the estimations by ICA were in good agreement with the results of Russian acoustic surveys, which gives credibility to the calculations (Figure 5.6.1). The use of oceanographic factors to forecast chub mackerel abundance should be complemented by recruitment surveys that provide actual indices of recruitment strength.

5.7 Management recommendations

The results of the ICA analysis indicate a decreasing trend in biomass. Although the results are uncertain, due to inconsistencies in the input data, the Working Group considers that there is reason for a precautionary approach, avoiding an increase in fishing effort.

5.8 Future research

For a better assessment of the mackerel stocks, the WG makes the following recommendations (in order of priority):

1. to establish an adequate sampling programme for all the countries fishing in the region;

2. to adopt the same age reading method for mackerel;

3. to conduct further studies on stock identity;

4. all acoustic surveys should give a separate estimate for mackerel;

5. it is recommended that the comparative studies of environment and biomass indices are presented at the next meeting.

6. GENERAL CONCLUSIONS

During this year's meeting the emphasis was on updating the data base as well as knowledge of assessment methods within the group. At the beginning of the meeting a two-days course in basic assessment methodology was held.

A number of preliminary assessments for various species were carried out, but with few reliable results. The reason for this was poor quality of the input data, and to some extent problems in using the existing models. During future meetings, this aspect of the work will require more attention. A Planning Group to discuss the data collection will be held in Senegal before the November-December Survey.

Although the group is not yet in a position to make short/long-term projections, it is expected that the pelagic stocks will decline further in case of an expansion of the (industrial) fishing effort in the area.

As a general recommendation the group therefore suggests to adopt a precautionary approach, and not to increase the combined catch of small pelagics above the average level attained during the most recent 5 years (1997-2001).

BACK COVER

The second meeting of the FAO Working Group on the Assessment of Small Pelagic Fish off Northwest Africa was held in Banjul, Republic of the Gambia, from 5 to 12 April 2002. During this year's meeting the emphasis was on updating the database as well as knowledge of assessment methods within the group. At the beginning of the meeting a two-days course in assessment methodology was held. A separate section is devoted to each of the main groups of species (sardine, sardinellas, horse mackerels and mackerel). For each of these, actualized information until 2001 is given on stock identity, the fisheries, catch and effort, biological data and abundance indices. A number of preliminary assessments for various species were carried out. Although the group is not yet in a position to make short/long-term projections, it is expected that the pelagic stocks will decline further in case of an expansion of the (industrial) fishing effort in the area. As a general recommendation the group therefore suggests that the fishing effort should be restricted to the current level.