Table of ContentsNext Page


Report of the FAO Working Group on the Assessment of Small Pelagic Fish off Northwest Africa - Saly, Senegal, 17 - 27 March 2004


PREPARATION OF THIS DOCUMENT

A permanent FAO Working Group composed of scientists from 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 give advice on 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 fourth meeting of the FAO Working Group on the Assessment of Small Pelagic Fish off Northwest Africa was held in Saly, Senegal, from 17 to 27 March 2004. A total of 15 scientists from seven countries and from FAO participated.

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

Distribution:

Participants in the Working Group
FAO Regional Fishery Officers
FAO Fisheries Department
Norwegian Agency for International Development
Institute of Marine Research (IMR), Norway
Ministry of Agriculture of the Netherlands
Netherlands Institute for Fisheries Research (RIVO)

1. INTRODUCTION

The fourth meeting of the FAO Working Group on the Assessment of Small Pelagic Fish off Northwest Africa was held in Saly, Senegal, from 17-27 March 2004. The overall objective of the Working Group was to assess the status of the small pelagic resources in Northwest Africa and recommend on 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 assessed by the group were: sardine (Sardina pilchardus), sardinellas (Sardinella aurita and Sardinella maderensis), horse mackerels (Trachurus trecae, Trachurus trachurus and Caranx rhonchus) 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. Fisheries Management and Marine Environment” and the Netherlands Institute for Fisheries Research (RIVO) and organized by FAO and the Oceanographic Research Centre in Dakar, Thiaroye, Senegal (CRODT).

Altogether 15 scientists from seven countries and FAO participated. The chairman of the group was Reidar Toresen, IMR, Norway.

1.1 Terms of reference

The terms of reference of the Working Group were:

1. Presentation of new data on catch, effort and sampling intensity by country; updating the existing data base.

2. Presentation of working papers on research activities; Review of research activities recommended by the Small Pelagic Fish Working Group in 2003 and carried out during 2003/2004.

3. Presentation of reports on R/V DR. FRIDTJOF NANSEN acoustic surveys in June and Oct/Nov/Dec 2003 and from the Research Vessels of the different countries.

4. Report on the progress made on age reading of sardine and sardinella in the region.

5. Carry out analyses of catch, effort and biological data for the period 1990-2003, if possible extending this to the period before 1990.

6. Update stock assessments for sardine, sardinellas, horse mackerels and mackerel.

7. Advise on short and long term management for each resource/stock.

8. Coordination of research projects on Small Pelagic Fish.

1.2 Participants

Pedro Barros

FAO

Ana Caramelo

FAO

Malika Chlaida

Morocco

Hamid Chfiri

Morocco

Ad Corten

The Netherlands

Elhabouz Hammou

Morocco

Aziza Lakhnigue

Morocco

Ebaye Mahmoud

Mauritania

Asberr Mendy

Gambia

Ahmedou Moustapha

Mauritania

Birane Samb

Senegal

Abdoulaye Sarre

Senegal

Ibrahima Sow

Senegal

Nikolay Timoshenko

Russia

Reidar Toresen (chairman)

Norway

Names and full addresses of all participants are given in Appendix I.

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, in Atlantic waters.

1.4 Structure of the report

The structure of the report is the same as that of the previous Working Group report (FAO, 2003). A separate section is devoted to each of the main groups of species (sardine, sardinellas, horse mackerels and chub mackerel). For each of these, standardized information is given on stock identity, fisheries, abundance indices, sampling, biological data, assessment, management recommendations and future research.

1.5 Age reading

Sardinellas

A second workshop on sardinella age reading was organized by FAO in co-operation with the Instituto Español de Oceanografía (IEO) in Tenerife/Canary Islands (Spain) from 8 to 12 December 2003 with funding by RIVO, The Netherlands and from project GCP/INT/730/NOR: “International Support to the Nansen Programme. Fisheries Management and Marine Environment”. In 2003, the exchange was done with otoliths of Sardinella aurita. The objectives of this workshop were to standardise the methodology criteria for age reading, to prepare a set of guidelines for Sardinella aurita age reading and to estimate an Age-Length Key (ALK) for the landings in 2003.


The second Sardinella aurita otolith exchange was initiated in October 2003 with two sets of otoliths from Mauritania and Senegal mounted on black plastic slides. These were circulated among 12 readers from the countries involved in the sardinella fisheries off Northwest Africa.

The Mauritania set consisted of 304 otoliths from commercial catches and 75 otoliths from the R/V DR. FRIDTJOF NANSEN survey, for a total of 379 otoliths. The Senegalese set consisted of 267 otoliths from commercial catches. Otoliths from the months February to September and a total length range from 12.5 cm to 39.3 cm were represented. The age range as determined by the 12 readers was 0-8 years.


The results of the exchange seem to indicate that the different readers do not apply the same criteria for the allocation of the rings and ages. Draft preliminary criteria were established in the Dakar workshop (February 2003), but the results show different interpretation of these criteria, since, for some of the readers, mean length of the fish does not increase with age.


The participants in this second workshop agreed a set of preliminary age-reading criteria that should be used for the interpretation of sardinella otoliths (Appendix II).

Workshop recommendations:

Sardine

In October 2003, an exchange of sardine otoliths was initiated, but reading is not yet completed.

Horse mackerel and mackerel

It was recommended not to initiate any age-reading activities for these species this year, but to concentrate efforts on sardine and sardinella. An exchange of horse mackerel and chub mackerel otoliths should be carried out in early 2005.

1.6 Planning Group for acoustic surveys

The meeting of the Planning Group for the coordination of acoustic surveys off Northwest Africa was held in Dakar, Senegal, 28-29 October, 2003. The meeting was the second of its kind, following a meeting in Dakar, Senegal from 26 to 28 October 2002. The general objective of the Planning Group is to coordinate the acoustic surveys in the region including intercalibration of research vessels and to act as a forum for discussion on issues important to acoustic surveys such as standardisation of methods, acoustic research and training.

The main conclusions of this Planning meeting were:

The countries in North West Africa have Research Vessels and they conducted acoustic surveys in 2003. These were better than in 2002 in terms of planning and data processing. The scientific capacity for running surveys in the Sub-Region was increased in collaboration with the Nansen programme, but an effort should be made to conduct studies in Target Strength for each main species of small pelagic fish, other studies for decreasing the noise in the acoustic recordings and to develop techniques for improved pelagic trawling.

The WG discussed the report from the Planning Group and concluded that to facilitate future co-operation on the monitoring and research of the pelagic stocks in the region, as much as possible of the process should be standardized (equipment, gear, methods, data storage, etc.).

1.7 Overview of landings

Total landings of the main small pelagic fish in the region for 2003 were estimated at 1.4 million tonnes, thus continuing the downward trend observed since 2001 (Figure 1.7.1a). In 1999, total landings were estimated to be around 1.3 million tonnes whilst attaining a peak of about 1.7 million tonnes in 2001. Overall, total landings of small pelagic fish in the region for the period 1990 to 2003 have fluctuated around 1.4 million tonnes.

Sardine (S. pilchardus) represents about 45 percent of the total landings of small pelagic fish in the region. Landings of sardine over the last five years ranged from 530 000 to nearly 790 000 tonnes with an average of about 670 000 tonnes. Total catches of sardine in the region have stabilized over the last two years (Figure 1.7.1a), at a level of about 700 000 tonnes. Catches of this species were also relatively stable from 1993 to 2000, with total annual landings around 600 000 tonnes.

The second most important small pelagic fish in the region is the round sardinella (S. aurita) constituting nearly 21 percent of total landings of small pelagic fish in the region. Total catches of round sardinella have fluctuated between 270 000 and 350 000 tonnes in the last five years with an average of about 310 000 tonnes. The overall trend is a relatively steady decrease in landings for this species since 1998 (Figure 1.7.1a). Total annual landings of flat sardinella (S. maderensis) have been significantly lower than those of round sardinella, with a last five year average of about 130 000 tonnes (1999-2003).

The Cunene horse mackerel (T. trecae) is the most important species in the horse mackerel sub-group. Average annual landings of Cunene horse mackerel over the last five years were estimated at about 144 000 tonnes while average landings of Atlantic horse mackerel (T. trachurus) were around 63 000 tonnes. Average landings of false scad (Caranx rhoncus) (1999-2003) were about 38 000 tonnes. Landings of the chub mackerel (Scomber japonicus) fluctuated between 130 000 and 190 000 tonnes with an estimated average of about 155 000 tonnes over the last five years.

Anchovy (Engraulis encrasicolus) and bonga (Ethmalosa fimbriata) have been considered very important small pelagic fish species in some countries in this region. Important landings of anchovy have been recorded in Mauritania while bonga constitutes the bulk of fish landings in The Gambia. In the later years, Senegal has also started recording significant landings of bonga. A brief evolution of the landings of these species is given in the country-specific review of landings. There is a need to include these species in the group of the main small pelagic fish to be assessed by the Working Group in the future.

Morocco

In Morocco, sardine dominates the landings of small pelagic fish. The landings of sardine during 2003 have remained approximately stable relative to 2002 (Figure 1.7.1b). In the overall series, the second most important species is chub mackerel S. japonicus, followed by Atlantic horse mackerel T. trachurus and round sardinella S. aurita. However, landings of all these species have been declining steeply since the late 1990s, and have kept approximately stable, at very low levels, from 2002 to 2003.

Landings of anchovy (Engraulis encrasicolus) have been recorded throughout the time-series of landings, but fluctuating appreciably. In 2003, landings of this species have not reached 17_000 tonnes.

Mauritania

Landings of small pelagic fish in Mauritania have shown important inter-annual fluctuations in the period from 1990 to 2003 (Figure 1.7.1c).

In 2003, an increase of the landings of the species with a preference for temperate waters (Sardina pilchardus, Trachurus trachurus, Scomber japonicus, Engraulis encrasicolus), accompanied by a decrease of the landings of the species associated with tropical waters (Sardinella sp, Trachurus trecae and Caranx rhonchus) was observed.

Despite this, sardinellas remain strongly dominant in the landings. The landings in decreasing order are Sardinella aurita (39 percent), Trachurus trecae (23 percent), Scomber japonicus (15 percent), Sardine pilchardus (6 percent), Trachurus trachurus (6 percent), Caranx rhonchus (6 percent) and Sardinella maderensis (4 percent).

Landings in 2003 decreased slightly (6 percent) for sardinellas, but markedly for T. trecae, passing from 149 000 tonnes in 2002 to 98 000 tonnes in 2003, and for Caranx rhonchus, decreasing from 66 000 tonnes in 2002 to 19 000 tonnes in 2003.

Anchovy (Engraulis encrasicolus) is targeted especially by the East European fleet. An increasing trend in the landings can be noted since 1997. During the last 6 years, landings of anchovy increased steadily, reaching 140 000 tonnes in 2003 (Figure 1.7.1c).

Senegal

The most important species in the landings of small pelagic fish off Senegal are, in decreasing order, sardinellas, horse mackerels and chub mackerel (Figure 1.7.1d). Landings are clearly dominated by sardinellas. During the period 1990-2003, sardinellas represent about 92 percent of the total landings of small pelagic fish in Senegal. Landings of horse mackerels are on average ca. 6 percent of total landings of small pelagic fish, and those of chub mackerel 2 percent.

Up to 1998, round sardinella was dominant in the landings, but since then flat sardinella has taken up the lead in the landings. Horse mackerels and chub mackerel are always bycatch species.

The average annual landings of bonga (E. fimbriata) during 1990-2003 were about 17 000 tonnes.

The Gambia

Bonga shad (Ethmalosa fimbriata) is the main target species in the small pelagic fish group in The Gambia. Landings of this species constitute over 70 percent of the total catches of all fish species of all groups. An annual average of around 19 000 tonnes of bonga was landed over the last five years. Although the landings of bonga over the period 1990 to 2002 have fluctuated from 8 000 tonnes in 1990 to over 22 000 tonnes in 1996 and about 19 000 tonnes in 2002, the general trend is that of an increase in the landings of this species.

Despite Government efforts to encourage the development of a fishery for the other small pelagic fish, small pelagic fish other than bonga are not target species in The Gambia since the cessation of operations of the Ghanaian fishing company, Seagull Coldstores. Other small pelagic fish such as sardinellas are mainly bycatch or alternate species in the absence of the main target species, bonga shad.

However, landings of other small pelagic fish are common and becoming significant (Figure 1.7.1e). The estimated average landings of other small pelagic fish over the last five years are 1_800 tonnes. Following the tripling of landings of these fish from 1_000 tonnes in 2001 to 3 000 tonnes in 2002, a 10 percent decrease in landings has been observed in 2003.

Of the sardinellas, flat sardinella (S. maderensis) is the most important species in terms of landings, and landings of this species were appreciably higher than those of round sardinella (S. aurita) for the past two years (Figure 1.7.1e). The annual landings of horse mackerels (Trachurus trecae, T. trachurus and C. rhonchus) and mackerel (Scomber japonicus) have been relatively low, with annual averages of 475 and 255 tonnes, respectively, for the past five years.

1.8 Hydrographical effects

The small pelagic fish species in West Africa can be divided into a group with a preference for temperate waters, and another group with a preference for tropical waters. The group of temperate species includes the pilchard (Sardina pilchardus), the chub mackerel (Scomber japonicus), the anchovy (Engraulis encrasicolus), and the Atlantic horse mackerel (Trachurus trachurus). The group of tropical species includes the sardinellas (Sardinella aurita, S. maderensis), the Cunene horse mackerel (Trachurus trecae), and the false scad (Decapterus rhonchus).

The distribution of each group of species will be determined by the distribution of the two different water masses we find in West Africa, the cold waters of the Canary Current system, and the tropical waters of the Guinea Current. The front between these two water masses shows seasonal north/south shifts between Senegal and Morocco. In the border area between the two water masses (Mauritania), the species composition will be strongly influenced by the seasonal shifts of the front, and also by inter-annual variations in the average position of the front. In addition to the shifts of the thermal front, the fish stocks will be influenced by variations in upwelling and plankton production.

The year 2003 in Mauritania was characterized by an increase of temperate species (sardine, Atlantic horse mackerel, anchovy), and a decrease of tropical species (sardinella, Cunene horse mackerel, false scad). This consistent change in both groups suggests a southward extension of cold-water masses in 2003. This change also seems to have affected fish distribution in Senegal, where sardine occurred at very southern positions (M’Bour), whereas sardinellas were relatively scarce.

It is recommended that more quantitative hydrographical data, including ship-based measurements and satellite measurements of Sea Surface Temperature (SST) and/or primary productivity, are made available before future meetings, so that the effect of environmental variations on fish abundance can be taken into consideration.

1.9 Overview of acoustic surveys results by R/V DR. FRIDTJOF NANSEN

The Norwegian research vessel, R/V DR. FRIDTJOF NANSEN has surveyed the sub-region during the period 1995-2003, carrying out acoustic surveys during the months October-December each year. In addition, in the period 2001-2003, the vessel carried out acoustic surveys covering the same area in May-July. The surveys aimed to map the distribution and estimate the abundance of the main small pelagic fish species, sardine (Sardina pilchardus), sardinellas (Sardinella aurita and Sardinella maderensis), horse mackerels (Trachurus trachurus and Trachurus trecae) and chub mackerel (Scomber japonicus). The distribution of other pelagic resources (other carangids, and anchovy) was also mapped and their abundance was estimated. The abundance estimates from the surveys are presented as numbers and biomass per length-group.

Figure 1.9.1a shows the estimated abundance for all the target species during the surveys in October-December, while Figure 1.9.1b shows the estimated abundance for the target species excluding sardine. For S. pilchardus, there has been an increase in the estimated biomass through the years 1997-2002, from a level of around 1 million tonnes to more than 6 million tonnes, with a relatively small decrease from 2002 to 2003. For Sardinella aurita, there has been an overall decreasing trend in the acoustic estimates from 2.1 million tonnes in 1999 to 1.3 million tonnes in 2003. For S. maderensis, the values have fluctuated between 1 and 1.5 million tonnes since 1989, and the 2003 estimate of 1.8 million tonnes was the highest on record. Of the horse mackerels, T. trecae has been the dominating species in the acoustic estimates, and its abundance was estimated as 800 000 tonnes in 1998. Since then, its biomass, as estimated from the acoustic surveys, has decreased, and it was estimated at 390_000 tonnes in 2003. The abundance of the other main horse mackerel species, T. trachurus, has fluctuated in the acoustic estimates, but its estimated biomass has shown an increasing trend from 2001 on, and was estimated at 320 000 tonnes in 2003. The estimated biomass of Scomber japonicus has shown an increasing trend, from the rather low level of 100 000 tonnes in 2000 to 550 000 tonnes in 2003.

Detailed estimates for the different species are given in the respective sections.

1.10 Quality of data and assessment methods

For the analysis of data, the group has the long-term aim to apply age-based analytical assessment methods to all the main stocks. These are the VPA-based methods like ICA, XSA, or others. However, to use such methods it is a requirement that catch statistics can be age-disaggregated with a high degree of consistency in the series, and that it is possible to follow the different year-classes age by age and year by year through the time series of catch data. For the main stocks to be analysed by the group, there are age-disaggregated data series. These data series are, however, not yet of sufficient quality to use analytical assessment methods. The reasons for this are problems with age reading, non-representative sampling of the catch (fishing fleets by quarter) and uncertainty in stock definition. The group aims to enhance the quality of these data series, encouraging developments to be done in all these fields, like arranging otolith age reading workshops, studies on stocks components or other. The quality of these data series may therefore change for the better in the future.

The quality of the age-disaggregated data series can be controlled by simple methods, such as the correlation between the number of fish in the catch at a certain age and the corresponding number of the same year-class the next year (the numbers at age 0 versus the numbers of the corresponding year-classes at age 1, and so on for all age groups). If the data series is consistent the correlation coefficient ® should be (as a guideline, on average) higher than 0.8. Data sets showing values of the correlation coefficients lower than 0.60 should not be used in the analysis. If the data are of poorer quality than this, methods not requiring age-disaggregated catch data, as e.g. surplus production models or length based models, should be used.

1.11 Methodology and software

After revision of the available data, the Working Group concluded that the only class of methods that could be applied to all stock groups were Logistic Production Models (Appendix III).

2. SARDINE

2.1 Stock identity

The previous Working Groups recommended that studies on sardine stock identity using, for example, molecular biology, should be carried out. Following these recommendations, a study on the genetic characterisation of the sardine stocks, using allozymic markers, was carried out by INRH (Chlaida, 2003). The preliminary results from this study indicate the possible presence of four distinct sardine populations. The first would be a homogeneous population from Tarfaya to Dakhla (28°N-24°N), the second a population off Safi (32°N) and the third a population straddling the previous two off Agadir (30°N). A fourth population has been identified off Larache (35°N). This population would be similar to the one off southern Portugal.

However, considering that these results are still preliminary, the WG decided that in the present assessment the stocks assumed in the previous Working Groups would still be used: the northern stock (35°45’ - 32°N); the central stock, Zone A+B (32°N - 26°N); and the southern stock, Zone C (26°N? the southern extent of the species distribution) - Figure 2.1.1.

2.2 Fisheries

Total catches

The data on catch have been updated for all three zones, to include the data from 2003 (Table 2.2.1a, b).

Recent developments by country

Recent changes in the various economic zones are described below.

Morocco

In 2003, the sardine stock in Zone A+B was exploited exclusively by the Moroccan fleet. This fleet comprises over 350 fishing units, all very similar (40-60 GRT, with the exception of three RSW (Refrigerated Salt Water) vessels, with a tonnage of about 1000 GRT. These three vessels operate only in Zone C.

The landings of sardine in Zone A have recovered, passing from 23 000 tonnes in 2002 to over 74 000 tonnes in 2003. Catches in this zone had decreased abruptly during the 1990s, to a very low level in 1996 of about 3 500 tonnes. This reduction in landings was due to a marked decrease in availability, measured as the ratio between positive (with non-zero catches) trips and the total number of trips. A slight increase was also noted in the northern zone, with reported catches reaching 20 000 tonnes. The opposite trend of a slight decrease was observed in Zone B. In Zone C, the total catch of sardine has reached 120 000 tonnes in 2003, a doubling from 2002 (Table 2.2.1a) (Figure 2.2.1a).

Mauritania

The landings of sardine off Cap Blanc are increasing, despite the fact that sardine is not a target species for the fleets fishing the Mauritanian waters.

Landings of this species in the Mauritanian zone have increased from 11 500 tonnes in 1996 to more than 76 000 tonnes in 2003.

Sardine is captured seasonally by the EU pelagic trawlers. An increase in the effort directed to sardine has been observed during the last few years, following the double effect of the increase in its abundance, shown by the acoustic surveys, and the decrease in the abundance of the traditional target species, sardinellas and horse mackerels.

Senegal

Occasional captures of sardine have also been reported off the Senegalese coast, an unusual situation that was recorded earlier, in 1994.

2.3 Abundance indices

2.3.1 Catch per unit of effort

The CPUE series have been updated for 2003. In general, the CPUE for Zone A+B shows an increasing trend, even though with some fluctuations (Figure 2.3.1a).

In Zone C, effort is defined in terms of fishing days for the RSW Moroccan vessels and for the industrial vessels working the Mauritanian zone.

No CPUEs data of the different fleets operating in Zone C are available in 2003.

2.3.2 Acoustic surveys

R/V DR. FRIDTJOF NANSEN

The last acoustic survey by R/V DR. FRIDTJOF NANSEN, in December 2003, has shown a slight decline of the estimated sardine biomass, especially for the Cap Cantin-Cap Timiris area (Figure 2.3.2).

National surveys

Mauritania

The acoustic survey carried out by R/V AL-AWAM in December 2003 has detected sardine concentrations as far south as 18°N. A very large biomass of sardine, exceeding four million tonnes, was estimated for the Mauritanian waters in December 2003.

2.4 Sampling of commercial fisheries

Morocco

The Moroccan Biological Sampling Programme has been enlarged in 2003 to cover two new landing sites (Sidi Ifni and Tarfaya). Sampling intensity has been increased by increasing the number of individuals sampled, both for length distributions and for collection of biological parameters and otoliths (Table 2.4.1).

Mauritania

In Mauritania, sampling of the commercial catches of sardine is done only on board of the vessels of the EU fishing fleet (IMROP observers).

Sampling on board is carried out all year round. These data are used to estimate the monthly species composition of the catches and the length distributions of the main species in the landings and the discards. Biological data of the main target species (sardine, sardinellas, horse mackerel and mackerel) is also recorded.

The sampling intensity of the EU fleet in Mauritania is 2.7 samples/1 000 tonnes catch. The average weight and size of the fish sampled are respectively 100 g and 19 cm.

2.5 Biological data

The biological data on landings of sardine from Zones A, B and C obtained from Moroccan samples are available for 2003. These data include length distributions, age-length keys and the length-weight relationship. Lengths reported are total length (TL) recorded to the ½ cm below.

The length distribution of the sardine landed in Zone A+B is bimodal, with modes at 16 cm and 22 cm. In Zone C, this distribution has a single mode at 22 cm. The absence of small individuals in the landings from this zone is probably due essentially to the range of operation of the fleet, which is limited to the coastal area. (Figures 2.5.1 a,b).

For Zone C, the annual length composition of the Mauritanian fleet landings is also available for 2003. Fork length (FL) has been measured to the nearest centimetre (FL, 1cm). In order to combine this length distribution with the Moroccan one, which is reported as total length, recorded to the ½ cm below (TL, 0.5 cm), the Mauritanian length-frequency distribution has been converted into the Moroccan format by using the conversion factor TL (0.5cm.inf) = 0.195598 + 1.111756*FL (1 cm.prox) (CECAF series/PACE 90/50).

The combined length distribution for Zone C is bimodal, with modes around 19 cm and 24.5 cm.

The age compositions of the landings were obtained applying the age-length keys calculated from the Moroccan data to the length compositions of Zone A+B and C (Table 2.5.1a,b).

The age-length relationships used for the calculation of average weights at age were also obtained from the sampling carried out in Moroccan landing sites.

Both the age composition and the average weight at age of the landings have been updated for the two zones (A+B and C) for 2003 (Table 2.5.2a,b).

Due to some difficulties in communication between the Working Group members, some of the biological data collected in Zone C, especially age-length keys, have not been utilized in work carried out during this meeting of the Working Group. Therefore, the Working Group suggested that these data should be updated for the 2005 meeting.

The length distribution estimated during the December survey of R/V DR. FRIDTJOF NANSEN is unimodal in the northern zone (Cap Cantin-Cap Juby), with a mode at 17 cm. In the southern zone, however, the length distribution is bimodal, dominated by adults (modes at 15 cm and 22 cm) in the Cap Juby-Cap Blanc area, and by young individuals (modes at 17 cm and 27 cm) in the Cap Blanc-Cap Timiris area (Figures 2.5.2 a,b,c,d).

2.6 Assessment

Data quality

In order to test the quality of the data available for the assessment, the sub-group carried out an exploratory analysis of the data, calculating the linear correlation between the estimated catches of each age-group and the corresponding number of the same year-class the following year. The results obtained (Figure 2.6.1) indicate that in Zone A+B no correlation can be detected between the estimated catches of supposedly the same cohort along life, with the exception of the correlation between age-groups 0 and 1.

For Zone C, the correlations are a little better than in Zone A+B, except for the age-group pairs 1-2 and 2-3 (Figure 2.6.2).

This lack of correlation, that indicates that cohorts cannot be reliably followed on the catch data, could be the consequence of several factors. These could include incorrect ageing, changes in the exploitation pattern (alternance of several fleets), inadequate sampling, specific targeting on a few length groups, more sought by the canneries, or incorrect stock delimitation. Irrespective of the source of uncertainty, however, the consequence is that it is not possible to rely on these data for an age-structured analytical assessment.

Therefore, the Working Group decided not to use the age-based analytical models for the assessment of this stock, at this stage.

Methods

The logistic surplus production model, implemented on an Excel Spreadsheet, was used. This model is further described in Annex III.

Data

The time series of sardine landings data for Zone C between 1995 and 2003 was used as input data for the model.

The abundance indices from the November-December acoustic surveys with R/V DR. FRIDTJOF NANSEN from the same set of years were used for fitting the model. Given that these surveys are carried out at the end of the year, it was considered that they represented the stock abundance the following year.

Most of the data series in the area showed some strong deviations from average expectations, that could not be explained from either the fishery or from average stock dynamics parameters. Current knowledge suggests that these changes are due to strong environmental forces. Accordingly, an environmental index has been incorporated into the model, allowing the stock to grow more or less than average depending on the state of the environment in each year. In the current formulation, and because complete data series on environmental factors were not available, only the exceptional years were marked with different than average environmental status. In this case, this meant considering the year 1996 as an exceptionally poor year for sardine growth and survival.

Results

The model managed to capture the main dynamic behaviour of the sardine stock in Zone C, with the introduction of a negative environmental effect in 1996, even in the presence of large fluctuations in the abundance indices.

The summary results indicate that current stock biomass is well above the biomass producing the maximum sustainable yield, and that current fishing mortality is below the sustainable fishing mortality at current biomass levels (Table 2.6.2).

Table 2.6.2:

Summary of the results of fitting the logistic production model with environmental covariates to the sardine catch and abundance index data


Stock

B/BMSY

Fcur/FSYCur

Sardine, Zone C

215%

81%

No reliable assessment could be carried out for sardine in Zone A+B.

Discussion

The data available indicate that catch levels were about stable during the whole period, even if the abundance indices from R/V DR. FRIDTJOF NANSEN surveys showed large changes, with an abrupt decrease in 1997 followed by a steady recovery until 2003.

This stability of catch rates with varying stock abundance has been documented for several schooling pelagic fish stocks, and it is considered to reflect the fact that the fleet always manages to locate the concentrations of pelagic fish, even when their total abundance is low.

The drastic decline in 1997 could not be explained solely by a high fishing pressure. Current views indicate that this decline in biomass was most probably due to an environmental effect occurring at the level of the region, with important effects on the survival, recruitment and movement of sardine. This effect was also noted on the values of Sea Surface Temperature (SST) in the whole region, and had an influence on other pelagic species like sardinella and horse mackerel.

Given the high sensitivity of this species to changes in environmental conditions, and the instability of the environment in the whole Canary Current Region, it is expected that the sardine stock may go through periods of contrasting production potential, shifting from periods of high productivity to other periods of low production capacity. The change from one condition to the other may be progressive as well as abrupt.

2.7 Management recommendations

The results from fitting the assessment model, even if they should be considered with caution, have given consistent indications on the current status of the sardine stock in Zone C. These results show that, at the current biomass levels, it should be possible to increase catches relative to the last few years. Given the high instability of the stock, however, clear in the drastic decline observed in 1997, a constant monitoring of the stock abundance and structure, by scientific surveys, and independent from catch data, should be ensured, to detect unanticipated changes that may require urgent management measures.

A system for quick adjustment of the fishing effort, that may react to the fluctuating abundance and productivity of the stock, should also be put in place.

This Working Group could not achieve reliable results on the assessment of the central (Zone A+B) stock. By a precautionary approach, however, it is recommended that catches of sardine in this area should not exceed the average catch levels observed during the last four years (ca. 500 000 tonnes).

2.8 Future research

The Working Group recommended that research in the following areas should be intensified:

1. the exchange of otoliths in the region, in order to improve the age reading of sardine;

2. age-reading on board, during the scientific surveys, to obtain age-disaggregated abundance indices;

3. continue the research on stock identity;

4. improve the sampling of catches in all fisheries targeting sardine;

5. continue the R/V DR FRIDTJOF NANSEN surveys, to improve and increase the time series of survey abundance indices, especially because in this fishery, fishery-based abundance indices are not reliable;

6. estimate the abundance by age group on the scientific surveys, by zone (A, B and C);

7. recalculate the R/V DR. FRIDTJOF NANSEN abundance indices by zone (A, B and C) for the whole time series;

3. SARDINELLAS

3.1 Stock identity

No new study on sardinella stock identity has been carried out since the last meeting of the Working Group. As in the previous Working Groups agreement was reached on the adoption of a single stock for both species of sardinella (FAO, 2001). For further information on stock identity refer to the report of the 2001 Working Group (FAO, 2001).

3.2 Fisheries

Sardinellas are fished in four major fisheries in the subregion: the industrial fishery in Morocco, the EU-industrial fishery in Mauritania, the non-EU fishery in Mauritania, and the artisanal fishery in Senegal.

Total catches

Catches by fleet and by country are presented in Table 3.2.1a for Sardinella aurita and in Table 3.2.1b for S. maderensis. Total catches for the whole region are presented graphically in Figures 3.2.1a and 3.2.1b.

The revised figure for the total catch of S. aurita in 2002 is 291 000 tonnes, which is 9 000 tonnes higher than the provisional estimate used last year. The preliminary figure for 2003 is 274 000 tonnes. Although this figure may still be revised, it seems that total catches of S. aurita have decreased from 2002 to 2003. The declining trend that started around 1998 has thus continued until now. The decline was noticed in all fleets, i.e. the Senegalese artisanal and industrial fleet, and also the industrial fleet in Mauritania.

For S. maderensis, the revised figure for 2002 is 148 000 tonnes, which is 22 000 tonnes higher than the provisional figure used last year. The revision was due mainly to an update of Senegalese catch statistics. The preliminary Figure for 2003 is 129 000 tonnes. This suggests that catches have declined from 2002. However, this conclusion may be changed after the final figure for 2003 has become available. So far, there does not appear to be a trend in total landings of S. maderensis in the whole region since 1996.

Total effort

Effort data for each zone are presented in Table 3.2.2. and Figures 3.2.2.a,b and c. No attempt has been made to separate effort on sardinella from effort directed at other species, and the data presented here represent the total effort exerted by the industrial fleets.

Recent developments by country

Recent changes in the various economic zones are described below.

Morocco

In Morocco, there was no directed fishery for sardinella in 2003. A small quantity of sardinella was taken as bycatch in the sardine fishery.

Mauritania

A decline was noted in catches of the industrial fleet in Mauritania. This decline was associated with a late arrival of the fish in Mauritania. Whereas the fishing season normally starts around April-May, the fishery in 2003 only started in June. The late arrival of the sardinella in Mauritania may have been related to low water temperatures during the first half of the year.

The percentage of S. maderensis in the landings of the EU-fleet declined to 4 percent, compared to 10 percent in the previous year. This decline may have been an effect of the late arrival of S. aurita in Mauritania. When the fishery finally started in June, it was directed at shoals consisting almost exclusively of S. aurita. In normal years, the sardinella fishery starts earlier, and takes a mixture of both species in the first months of the year.

More precise estimates have now become available for the artisanal landings in Nouakchott. During last year’s meeting, only an approximate estimate was available for the landings in 2002. Due to improved monitoring of the landings, an estimate of 15 000 tonnes is now available for landings in 2003.

A new development in Mauritania was the start of a purse seine fishery in the coastal zone off Nouadhibou. The two vessels engaged in this fishery catch a mixture of S. maderensis and S. aurita with a total length around 25 cm. These fish are available all year round in the coastal area.

Total fishing effort by the EU fleet has been estimated as in last year’s meeting by adjusting the number of fishing days for total horsepower of the vessel. The EU-fleet in Mauritania does not target only sardinella throughout the year. During the winter months (November-March) when sardinella is scarce in Mauritania, the fleet also targets sardine, horse mackerel and mackerel. The total effort of the EU fleet presented in Table 3.2.2, therefore, tends to over-estimate the effort directed at sardinella.

The same applies to the effort exerted by the non-EU industrial fleet. Most of these vessels target horse mackerel, mackerel, anchovy and sardine, and only some of them target sardinella. The effort figures presented for this fleet in Table 3.2.2, therefore, do not reflect effort directed at sardinella.

Senegal

In Senegal, there were no major changes in the fishery compared to 2002. As usual, the main catches were taken by the artisanal fishery, located mainly in the villages of M’Bour and Joal, south of Dakar.

A census of the number of canoes active in the artisanal fleet was made in April 2003. The results are compared with data for earlier years in Table 3.2.3.

Table 3.2.3: Results from the censuses of the artisanal fishing fleet of Senegal from 1993 to 2003

Number of canoes using different gears


1993 Sept

1995 Oct

1997 Sept

2001 Oct

2002 Oct

2003 Apr

Purse seine

344

294

394

476

393

395

Surrounding gillnet

72

89

184

101

67

137

Beach seine

91

95

177

85

85

85

The number of canoes using purse seines appears to have remained constant, whereas the number of canoes using surrounding gillnets has increased. The purse seine is used for both species of sardinella, whereas the surrounding gillnet is used exclusively for S. maderensis and Ethmalosa fimbriata.

A new development in the Senegalese artisanal fleet was the limitation of fishing effort by fisherman’s organizations. They restricted the number of fishing trips by purse seine canoes in order not to saturate the sardinella market.

A part of the Senegalese artisanal fleet operates in the southern part of Mauritania and land their catches in Saint Louis. The number of Senegalese canoes allowed to fish in Mauritania is 250. However, in practice this number is much lower. The total number of canoes using purse seines in Saint Louis in 2003 was 83. They fish sardinella only from November to June. During summer (July-October), the sardinella are out of reach because they have migrated further north into Mauritania. Total landings of sardinella in Saint Louis in 2002 were 16 000 tonnes of S. aurita and 6 000 tonnes of S. maderensis. The proportion of the catch coming from Mauritanian waters cannot be established.

The industrial fishery was conducted by four small trawlers, operating from Dakar. They contributed only one percent to the total sardinella catch in Senegal.

The Gambia

In The Gambia, catches were at the same level as in 2002 (around 2 000 tonnes for both species of sardinella combined). Although catches have increased in recent years, the total catch of sardinella taken in The Gambia is still less than 1 percent of the total regional catch. In The Gambia, there is no directed fishery for sardinella. The species is taken only as bycatch in the fishery for Ethmalosa fimbriata.

3.3 Abundance indices

3.3.1 Catch per unit of effort

Mauritania

For Mauritania, two series of CPUE are presented: one for the EU-fleet and one for the rest of the industrial fleet (Russia, Ukraine and others). Both series refer to the combined catch of S. aurita and S. maderensis as no distinction can be made between effort directed at either of the two species individually. In practice, however, about 90 percent of the catch consists of S. aurita. The CPUE calculated on the basis of the effort, therefore mainly refers to S. aurita.

The EU fleet in Mauritania targets primarily sardinella and switches to other species only when sardinella is not available in the area. The CPUE of sardinella in this fleet can thus be considered to given an indication of the availability of sardinella in Mauritania, in particular for S. aurita. Effort data for the EU fleet do not include the two Irish vessels, as no effort data for these vessels were available. For the remainder of the EU-fleet, fishing effort was adjusted for total horsepower of the vessel, using the factors presented in last year’s report, and expressed in standard fishing days of a 10 000 HP trawler.

The series for the EU fleet show a declining trend over the last five years. The CPUE for the EU fleet in 2003 remained at the low level of the previous year (Figure 3.3.1a).

The non-EU fleet targets horse mackerel and mackerel, and only takes sardinella when the other species are scarce. The CPUE for sardinella in this fleet has fluctuated around the same low level during the last five years.

Senegal

The CPUE series for S. aurita and S. maderensis are given only for the artisanal fishery of Senegal (Figure 3.3.1b). These series are expressed in tonnes per trip.

The highest CPUEs of S. aurita have been obtained during the periods 1992-1994 and 1996-1997. Since then, it has shown a decrease over the period 1996-1999, followed by a small increasing trend from 1999 to 2001 and a slight decline in 2002 and 2003.

For Sardinella maderensis, the yearly trends are more fluctuating with the highest peaks observed in 1996 and 1999. Relative stability can be seen over the last three years, even though an increase was observed in 2002.

3.3.2 Acoustic surveys

R/V DR. FRIDTJOF NANSEN

Morocco

The estimated biomass of sardinellas (Figure 3.3.2.a) shows strong fluctuations from 1995 to 2003. Peaks in 1996 and 2001 were followed by sharp decreases of 78 percent in 1997 and 52 percent in 2002. The biomass estimated for 2003 is slightly lower than that recorded in 2002.

The round sardinella (S. aurita) has always predominated over flat sardinella S. maderensis. In the 2003 survey, the estimated biomass of round sardinella represented 81 percent of the total estimated sardinella biomass.

Mauritania

The Norwegian research vessel covered the Mauritanian zone both in June and in November 2003. S. aurita accounted for 66 percent of the total biomass in the June survey, but for only 19 percent in the December survey.

The increase in the total abundance of sardinellas estimated in the December 2003 survey relative to the two previous years, was mainly due to a sharp increase in the abundance of flat sardinella S. maderensis (Figure 3.3.2b).

Senegal - The Gambia

The estimated biomasses show large fluctuations from one year to the next (Figure 3.3.2a). High biomasses were estimated in 1995, 1999 and 2002, while 1996, 1997 and 2000 have seen low estimated values for biomass of sardinellas. Flat sardinella always dominates the biomass of sardinellas.

Total sub-region

The estimated biomass for S. aurita and S. maderensis in the sub-region from 1995 to 2003 showed fluctuating trends with average values of 1.5 million tonnes and 1.3 million tonnes respectively (Figure 3.3.2d) in the November/December surveys. The lowest biomass for S. aurita was recorded in November/December 1998, peaking at over 2 million tonnes in 1999. Since then the biomass of S. aurita has been gradually declining to just over 1.2 million tonnes in November/December 2003. Unlike round sardinella, the flat sardinella stock has shown a more regular fluctuating pattern over the last five years but with a slight upward overall trend. The total biomass estimates in November/December for the two species combined show a similar trend to that of flat sardinella in the period 1999 to 2003 (Figure 3.3.2d).

National surveys

Senegal

The first survey with R/V ITAF DÈME was carried out from the 12-27 March, 2003. The period 14-21 March was devoted to the southern area while the northern area was covered during 21-27 March.

The results indicate a strong presence of sardinellas during this period. Total fish biomass (all species) was estimated as 1 014 000 tonnes, of which 30 percent was flat sardinella and 14 percent round sardinella. Concentrations of round sardinella were detected only in the northern area, in the southern area only a few traces were detected.

Mauritania

The Mauritanian research vessel R/V AL AWAM carried out three surveys during 2003. Two of the surveys (March and December) covered the entire Mauritanian zone. In June a 5-day survey was conducted in the northern part of the area. This survey was primarily intended for testing a new pelagic trawl. However, the results were also used to calculate fish abundance in the northern area. These surveys estimated the total biomass of sardinellas at respectively 324 000, 78 000 and 958 000 tonnes. The proportion of S. aurita in this biomass was estimated as 63 percent, 100 percent and 77 percent respectively.

During all these surveys, sardine was much more important than sardinella. This was a very unusual situation, indicating a possible ecological change in the area. The estimate for total sardinella in the December survey (958 000 tonnes) is comparable to the figure produced by R/V DR. FRIDTJOF NANSEN (1 021 000 tonnes). However, the trawl catches made by R/V AL AWAM contained a much higher percentage of S. aurita than the catches of R/V DR. FRIDTJOF NANSEN. The catch composition of R/V AL AWAM is in accordance with the catch composition of the EU commercial fleet. The difference in catch composition between the two vessels could be due to the fact that the pelagic trawl of “Al Awam” has been designed specifically for sardinella, and may be more efficient in taking S. aurita.

3.4 Sampling of commercial fisheries

Tables 3.4.1 and 3.4.2 present the sampling intensity for Sardinella aurita and Sardinella maderensis in 2003.

Mauritania

Length sampling in Mauritania is quite intense in the EU and the Russian fleets as some of the vessels in these fleets carry on board scientific observers. The average number of samples per 1 000 tonnes of Sardinella aurita is about 2.5. The number of fish sampled per 1 000 tonnes was estimated to be about 270. For Sardinella maderensis (Table 3.4.2), less sampling was undertaken than for S. aurita which is the main target species of the two fleets in Mauritania. Length samples were also taken on the artisanal catches of this species (Table 3.4.1).

Senegal

Sampling for the two species in Senegal was low compared to Mauritania. Out of 408 tonnes of total catch of Sardinella aurita in the industrial catch, 41 samples were taken (Table 3.4.1). Full information on sampling intensity for sardinella in the artisanal sub-sector during 2003 was not available at the time of the meeting, since not all samples had been reported.

The Gambia

For The Gambia, there is no sampling scheme in place for small pelagic fish as these are not the main target species of the fishery.

3.5 Biological data

Length distributions of commercial landings taken by the Dutch fleet in Mauritania in 2003 showed an unusual amount of relatively small fish (Figure 3.5.1g).

Whereas in previous years the landings always had an approximately unimodal distribution around 29-30 cm fork length, the length distribution in 2003 showed a sizeable amount of fish between 22-27 cm. In last year’s report mention was made of a strong cohort of young fish that was found in the acoustic surveys of R/V DR. FRIDTJOF NANSEN in June and November 2002. It is possible that the high percentage of small fish in the Dutch landings in 2003 represents the recruitment of the new year-class detected by R/V DR. FRIDTJOF NANSEN in 2002 (Figure 3.5.1a to f). It is the first time since observations on Dutch landings were started in 1999 that length distributions show evidence of the recruitment of a strong year-class.

The relative scarcity of fish > 30 cm (fork length) may have several explanations. It is possible that the new year-class was very strong and that the fleet directed its effort entirely at concentrations of this age group. Another explanation could be that the older component of the stock has declined in recent years, and that the fishery has now become dependent upon the incoming year-class.

3.6 Assessment

Data quality

In order to test the quality of the data available for the assessment, the sub-group carried out an exploratory analysis of the data. The results obtained (Figure 3.6.1) do not indicate a good correlation between the estimated catches of supposedly the same cohort along life.

Methods

The logistic surplus production model, implemented on an Excel Spreadsheet, was used. This model is further described in Appendix III.

Input Data

The model requires a complete time series of data on total catch, as well as an index of stock abundance.

The estimates of total catch obtained by adding the catch estimates for the different fleets and countries were used for the total catch series.

For the abundance index, two time series were used, the acoustic abundance indices from R/V DR. FRIDTJOF NANSEN (1995-2003), and the CPUE from the Senegalese artisanal fleet (1990-2003). This CPUE series was chosen because it is considered that it will track sardinella abundance better than the industrial series, and because it shows a better agreement to the survey abundance indices.

Given that the R/V DR. FRIDTJOF NANSEN surveys are carried out at the end of the year, the indices obtained were considered as representative of the stock abundance for the following year.

Separate data series were estimated, for Sardinella aurita, for Sardinella maderensis and for the two species pooled.

In the data series, the year of 1999 showed up as an exceptional year, with stock growth conditions that could not be explained by the average stock dynamics parameters. Further investigations have revealed that 1999 had shown the highest upwelling activity in the last 10 years. Accordingly, an index of environmental quality, denoting 1999 as an exceptional year, was also incorporated into the model.

Results

The model could be fitted to the series on Sardinella aurita and on the two species pooled using the survey abundance indices for calibration (Figures 3.6.2 and 3.6.3). No satisfactory results could be obtained for S. maderensis.

Sardinella aurita

For the data on S. aurita, the model indicates a steady decrease in abundance over the last few years. This is in agreement with the trends observed both in the catches and the surveys across the whole sub-region.

The model results indicate that the current biomass is slightly below the biomass producing the maximum sustainable yield, on average, and that current fishing mortality is about 20 percent above the sustainable fishing mortality at current biomass levels (Table 3.6.2).

Sardinella sp.

When the two species pooled are considered, the model still captures the decrease in abundance during the latter years (Figure 3.6.3).

In this case, however, it indicates that the current biomass of the two species combined is at the level of the biomass producing the maximum sustainable yield, and that current fishing mortality is similar to the sustainable one.

Table 3.6.2:

Summary of the results of fitting the Logistic Production model with environmental covariates to the sardinella catch and abundance index data


Stock

B/BMSY

Fcur/FSYCurB

Sardinella aurita

93%

122%

Sardinella sp.

100%

105%

Discussion

The fitted model has shown that the abundance indices from the R/V DR. FRIDTJOF NANSEN surveys allow one to obtain a more reliable assessment than fitting the fisheries-based abundance indices alone. This is to be expected for fisheries of schooling pelagic fish, since it has been shown by several studies that CPUE in these fisheries tends to be independent of stock abundance. The positive effect obtained by the inclusion in the model of an environmental parameter, based on the upwelling index, also shows the importance of environmental fluctuations for the dynamics of these stocks.

The better condition of the two-species stock, when compared to the S. aurita stock, is in line with the observation of a slight increase in abundance of S. maderensis coinciding with a decline of S. aurita during the last few years.

Overall, therefore, the results from the model are consistent with the other information available on the stocks and fisheries of sardinellas.

3.7 Management recommendations

The results obtained during this Working Group confirm the relevance of the management measures recommended in previous years, i.e. the precautionary approach keeping the catches at the level of the mean catch of the previous three years.

Given the results obtained this year, however, and in view of the continued decline in S. aurita abundance, evident in all data sources, the Working Group recommends the reduction of the fishing effort on sardinellas (especially the round sardinella) by 20 percent.

3.8 Future research

The Working Group recommended that research on the following areas should be intensified:

1. continue the R/V DR. FRIDTJOF NANSEN surveys and the inter-calibration exercises, in order to maintain and improve the time series of fishery-independent abundance estimates;

2. joint surveys between the vessels of the sub-region;

3. carry out studies between the sessions of the Working Group aimed at solving the problems encountered, particularly in relation to fishing effort and to the results obtained using the production models;

4. continue the otolith exchange programme for age reading;

5. encourage and carry out studies for the application of analytical age-based models, particularly on the age-structure of sardinella stocks;

6. improve the sampling for length frequencies of sardinellas from commercial catches;

7. study in greater detail the length frequencies in the acoustic surveys and in the commercial landings.

4. HORSE MACKEREL

The two main species of horse mackerel in the region are the Atlantic horse mackerel (Trachurus trachurus) and the Cunene horse mackerel (Trachurus trecae). Only catch data will be presented for false scad (Caranx rhonchus).

4.1 Stock identity

This section has already been described by previous Working Groups (FAO, 2001 and 2002). More stock identification studies should be undertaken in the future.

4.2 Fisheries

Total catches

The evolution of the total catches of the three species of horse mackerels is presented in Figure 4.2.1. The landings data for the period 1990-2003, disaggregated by country, are presented in Tables 4.2.1a,b and c.

Total landings of Trachurus trachurus increased slightly in 2003 relative to the landings recorded in 2002 for Zone B and for the Mauritanian zone. Total landings of this species increased from about 46 000 tonnes in 2002 to 54 000 tonnes in 2003.

Landings of Trachurus trecae have seen an opposite trend. They have declined from 154 000 tonnes in 2002 to approximately 100 000 tonnes in 2003. Practically the totality of the catches of this species is taken in the Mauritanian zone.

The landings of Caranx rhonchus have also declined, to return to the levels of 2001.

For the sub-region as a whole, a declining trend is thus observed for the two species with a tropical preference (Trachurus trecae and Caranx rhonchus), accompanied by an increase of the temperate species (Trachurus trachurus). The landings of anchovy reached their highest level (140 000 tonnes) in 2003 (Figure 4.2.2).

Fishing effort

Most of the fishing effort of the industrial fleet on horse mackerel is in the Mauritanian EEZ. The industrial fleets targeting horse mackerels are those of Eastern Europe, Belize, Saint Vincent and the Grenadines. The nominal fishing effort (in number of vessels and in number of fishing days) of this fleet declined in 2003 relative to 2002. It should be noted that in 2003, a part of the effort of the Eastern European fleets was directed towards anchovy, following a trend started around 1997 (Figure 4.2.2).

Recent developments by country

Morocco

In 2003, the landings of Atlantic horse mackerel Trachurus trachurus increased relative to 2002 in Zone North, A and B. These landings are considered as bycatch from the trawlers and coastal purse seiners. No landings from Zone C were reported.

Mauritania

Most of the landings of horse mackerels in the region are made in Mauritania. In 2003, a strong decrease in the landings of Trachurus trecae and Caranx rhonchus was recorded, accompanied by an increase of the landings of Trachurus trachurus.

Senegal

The landings of false scad Caranx rhonchus) have stabilized around 5 000-6 000 tonnes, while those of Trachurus trecae have increased markedly, from 800 tonnes in 2001 to 4 500 tonnes in 2002. The estimates for 2003 are around 2 000 tonnes.

The Gambia

In The Gambia, reported landings of Caranx rhonchus and Trachurus trecae increased until 2002, but estimates for 2003 indicate a slight decline in the landings for these species.

4.3 Abundance indices

4.3.1 Catch per unit of effort

Two series of CPUE were considered for each species of horse mackerel: one based on nominal effort of the non-EU industrial fleet in Mauritania, and one based on the directed effort by the Russian fleet in Mauritania. The latter index was calculated as the geometric mean catch per day during the months when the species was most abundant in the catch (November-April for Trachurus trachurus and May-October for Trachurus trecae). The results are presented in Figures 4.3.1a and 4.3.1b.

For Trachurus trachurus, both CPUE series show a gradual decline from 1994 to 2000/2001. Both series show an increase in the last two years, the increase being stronger in the nominal series than in the seasonal series.

For Trachurus trecae, the seasonal series fluctuates without trend between 1994 and 2003. The nominal series shows a period of reduced values from 1996-1999, and a period of higher values from 2000-2003.

The CPUE series for the EU fleet in Mauritania, presented in last year’s report, was discontinued this year. The EU fleet takes only small quantities of horse mackerel and the CPUE for horse mackerel in this fleet is not expected to reflect variations in stock size.

4.3.2 Acoustic surveys

R/V DR. FRIDTJOF NANSEN

The acoustic surveys by R/V DR. FRIDTJOF NANSEN in May-June and October-December 2003 provided new abundance estimates for horse mackerel in the region. Results from the October-December survey have been used to update the series that started in 1995. The May-June survey was only the second survey during this time of the year (after 2002), and the results cannot be used to study long-term trends.

The results for the October-December surveys are shown in Figures 4.3.2a,b,c,d. The estimated abundance for Trachurus trecae was lower than in 2002. The decline in 2003 was associated with the complete absence of this species in the Moroccan zone (Fig. 4.3.2a). In Mauritania and Senegal, the abundance in 2003 increased slightly in comparison to 2002. On a regional level, the index for this species during the last three years (2001-2003) has been considerably lower than in the previous period (1996-2000).

The acoustic abundance index for Trachurus trachurus in 2003 was slightly higher than in the previous years. This was associated with an increase of this species in Morocco. For the first time since 1997, the index for this species at the regional level was now at about the same level as the one for Trachurus trecae.

National surveys

Mauritania

During the acoustic survey carried out by R/V AL-AWAM in December 2003, the biomass of Trachurus trecae was estimated to be about 520 000 tonnes, mostly juveniles around 15 cm total length. The biomass of Caranx rhonchus was estimated to be around 363 000 tonnes, including juveniles around 16 cm and adults around 26 cm total length.

4.3.3 Comparison of CPUE and acoustic indices

For Trachurus trachurus, all series show a relatively low abundance in the period 1998-2001 and an increase during the last two years. It cannot be decided which series best reflects the actual stock abundance. The CPUE series for Mauritania are probably influenced by north/south displacements of the stock. Mauritania is the southern limit for the distribution of the species.

For Trachurus trecae, there is a disagreement between the three series of indices. The acoustic surveys show a decline in abundance for the last three years. The seasonal Russian CPUE series shows a rather stable stock situation, whereas the nominal series for the non-EU fleet shows an increased abundance in the last four years. On the basis of the data available, it is not possible to decide which of the two CPUE series best reflects the variations in stock size but may be the acoustic series that are independent from the fisheries.

4.4 Sampling of commercial fisheries

Most of the sampling of horse mackerel landings is done in the Mauritanian zone, where most of the catch of these species is taken (Tables 4.4.1, 4.4.2 and 4.4.3).

Sampling in the Russian fleet continues to be carried out throughout the year, but sampling intensity for the horse mackerel species was reduced. The three species Trachurus trachurus, Trachurus trecae and Caranx rhonchus have been sampled during respectively three, two and two quarters.

Sampling data for the Russian fleet were presented by quarter in 2003, while they had only been reported as global values in 2002 A reduction of the number of samples for T. trachurus has been noted in 2003.

Sampling intensity of Trachurus trachurus in the EU fleet has dropped from 15 per 1 000 tonnes in 2002 to 8 per 1 000 tonnes in 2003.

Sampling intensity for Caranx rhonchus has been kept at approximately the same level, varying little from 11 samples per 1 000 tonnes in 2002 to 9 in 2003.

4.5 Biological data

The length distributions for the two Trachurus species estimated during the surveys by R/V DR. FRIDTJOF NANSEN during the two periods (May-June and October-November) are presented in Figures 4.5.1a to d. The length distributions of both species show multiple modes. For Trachurus trecae, the modes observed in May-June and November-December are respectively 10, 14 and 20 cm, and 10, 19, 26 cm.

The length distributions from the catches of the EU fleet (Figure 4.5.2) show an important component of small Trachurus trecae. A large part of these small individuals is discarded.

Age-length keys for Trachurus trecae, Trachurus trachurus and Caranx rhonchus based on the sampling carried out on board the Russian fleet fishing off Mauritania were supplied to the WG.

4.6 Assessment

Data quality

In order to test the quality of the data available for the assessment, the sub-group carried out an exploratory analysis of the data, calculating the linear correlation between the estimated catches of each age-group and the corresponding number of the same year-class the following year. The results obtained (Table 4.6.1) indicate a very low correlation between the catches of supposedly the same cohort along life.

This lack of correlation, that indicates that cohorts cannot be reliably followed on the catch data, could be the consequence of several factors. These could include incorrect ageing, changes in the exploitation pattern (alternance of several fleets), inadequate sampling, specific targeting on a few length groups, or incorrect stock delimitation. Irrespective of the source of uncertainty, however, the consequence is that it is not possible to rely on these data for an age-structured analytical assessment.

Therefore, the Working Group decided not to use the age-based analytical models for the assessment of this stock, at this stage.

Table 4.6.1:

Values of the linear correlation coefficient between the estimated catches of consecutive ages of the same cohorts for horse mackerels


Ages

1-2

2-3

3-4

4-5

5-6

6-7

7-8

Species








Trachurus trachurus

0.7

0.14

0.17

0.1

0.63

0.86


Trachurus trecae

0.37

0.35

0.26

0.12

0.10

0.4

0.54

The logistic surplus production model, implemented on an Excel Spreadsheet, was used. This model is further described in Appendix III.

Input data

The data available on the Trachurus trachurus stock are not enough to run even the surplus production model. Furthermore, with the closure of Zone C to the activity of most fleets, it was considered that the relative exploitation pattern of this stock would have changed too much to allow the application of surplus production models. Therefore, only the Trachurus trecae stock was assessed by the Working Group.

As catch data, the Working Group used the series of total catch estimated by the Group.

For the abundance index, the WG used the series of acoustic abundance indices from R/V DR. FRIDTJOF NANSEN (1995-2003).

Given that the R/V DR. FRIDTJOF NANSEN surveys are carried out at the end of the year, the indices obtained were considered as representative of the stock abundance for the following year.

An environmental effect was introduced into the model, to account for non-average growth conditions for the stock in some years of the series, namely 2001.

Results

The fit of the model to the available data was considered satisfactory (Figure 4.6.1). It manages to follow the main trends in the abundance indices, reacting to the variations in catches.

The fitted model indicates that the stock fluctuated markedly during the period considered, but that it was not strongly overexploited during most of this same period.

The model results indicate that the current biomass is at the level of the biomass producing the maximum sustainable yield, on average, and that current fishing mortality is below the sustainable one at the current biomass levels (Table 4.6.2).

Table 4.6.2:

Summary of the current status of the stock and the fishery for Cunene horse mackerel Trachurus trecae


Stock

B/BMSY

Fcur/FSYCur

Trachurus trecae

122%

52%

One may thus consider that the adult fraction of the stock is fully exploited. However, there are still important uncertainties regarding, on the one hand the level of discards or the catch transformed into fish meal, and on the other, the changes in the relative exploitation pattern. Given these uncertainties, the results should be viewed with caution.

Discussion

The results of the assessments suggest that current fishing effort is below the level that could be sustained by the stock. However, there are reasons to consider these results with caution. First, the results of the assessment depend on the assumption of a constant exploitation pattern. If this has changed recently, for instance as a result of a concentration on younger age groups, the results of the assessment will not be reliable. The second reason for caution is the possible under-reporting of catches of small fish, due either to discarding or to processing into fishmeal. In this context, it is worth noting that the declared fishmeal production of 25 000 tonnes by non-EU vessels in Mauritania corresponds to a catch of approximately 100 000 tonnes fresh fish. The species composition of this catch is unknown, but it may contain a substantial amount of small horse mackerel. Commercial fishing vessels in 2003 reported substantial catches of small Trachurus trecae (14-20 cm).

4.7 Management recommendations

The results of the assessment model suggest that the stock of Cunene horse mackerel is moderately exploited. However, due to the limitations of the model used, uncertainties on the stock assessment and the diverse nature of the fisheries (especially the industrial fisheries), a prudent attitude should be observed in the management of these stocks. For this reason the Working Group decided to take a precautionary approach and to recommend not increasing fishing effort above the average level for the last five years.

The Working Group is concerned about the possible underreporting of catches especially of juveniles, and it recommends initiating sampling of discards and catches on board all vessels.

4.8 Future research

In order to reduce the uncertainties associated with the assessment, the Working Group decided to recommend that the following research themes should be pursued:

1. Improve the sampling of the catches in all fleet segments.

2. Implement a specific sampling programme for estimating the amount, the species composition and the length structure of the fish discarded or used for fishmeal in all fleets.

3. Monthly sampling of the main fleets for the determination of the species composition of the catch, composition by length and age, etc.

4. Establish a programme of age reading for the three species.

5. Analysis of the length frequencies of the commercial catches and the research surveys at the same time as the age reading to improve the understanding of the growth dynamics of these species.

6. Continued use of the historical series on catch, effort and biological parameters.

7. Establishment of a standard effort series by species, taking into account the limits of such an approach.

8. The acoustic surveys should be continued and the acoustic abundance estimates should be split by age groups.

9. Stock assessment methods integrating environmental factors should be developed, and research projects aiming at a better understanding of the effect of environmental changes on the dynamics of horse mackerel stocks should be encouraged.

5. CHUB MACKEREL

5.1 Stock identity

The distribution of chub mackerel (Scomber japonicus, Houttuyn 1782) is described in the reports of the previous Working Groups (FAO, 2001, 2002 and 2003).

Two chub mackerel stocks have often been considered in Northwest Africa, a northern stock from the north of Morocco to Cap Bojador and a southern stock from Cap Bojador to the south of Senegal.

In 2001 and 2002, the Working Group assessed only the southern stock. However, since its 2003 meeting, the WG considered that the uncertainty in stock identification and the migratory behaviour of chub mackerel made it more reliable to assess it as a single stock. Therefore, the assessment was based on one single stock.

5.2 Fisheries

The northern fishery between Tanger and Cap Bojador is conducted only by Moroccan vessels. The Russian trawlers ceased fishing in these areas in 1999, and the remaining foreign fleet by the end of 2001. This fleet is composed exclusively of coastal purse seiners targeting sardine. During the summer (May-August) this fleet also targets chub mackerel when this species becomes available in this area (mainly between Cap Juby and Cap Ghir) (Figure 2.1.1).

South of Cap Bojador, the fishery is conducted essentially by pelagic trawlers from various countries (Russia, Ukraine, EU and others) that operate under fisheries agreements or under joint ventures in the Mauritanian zone. Chub mackerel is normally not the main target species for these fleets. In Senegal and The Gambia, chub mackerel is taken as bycatch by the artisanal and industrial fleets.

Total catches

The evolution of landings of Scomber japonicus by country and fleet in the study area for the period 1990-2003 is presented in Table 5.2.1. This table has been revised relative to the last WG meeting, and some corrections have been made to the data from Mauritania and Senegal. Total catches of chub mackerel increased from 1991 to reach a peak of over 200 000 tonnes in 1997, with a downward trend since then, to reach a minimum of 136 000 tonnes in 2002, with a small increase in 2003 (Figure 5.2.1).

The trend in total catch mainly reflects the developments of the fisheries in Zone C, north of Cap Blanc which, for most years with the exception of 1992, 1996 and 2002, constitutes the bulk of the catch. The observed reduction of catch in Zone C, north of Cap Blanc since 1998 can be explained by the expiry of the fisheries agreement with Russia at the end of 1999 whereby the Russian trawlers operating under this agreement stopped their activities and by the fact that the trawlers operating under rental agreements or joint ventures (from Ukraine and other countries) stopped operating at the end of 2001. Hence fishing effort in this part of fluctuates between about 10 000 and 35 000 tonnes during the period 1990-2003. In 2003 the landings from this fishery increased approx. 60 percent to a total of 36 500 tonnes (Table 5.2.1).

The total landings of chub mackerel in Mauritania show an increasing trend from the early 1990s to 1996 when the catch peaked at around 100 000 tonnes. From 1996 to 1999 landings decreased, followed by an increase from 1999 to 2002. Since 2002, landings from Mauritania were the highest in the region, representing more than 70 percent of total landings. Two of the most important fleets operating in Mauritania are the Russian and EU fleet. The EU fleet operating in Mauritania targets mostly sardinella, but when sardinella is not available they catch more of the other small pelagic species such as chub mackerel and sardine. The Russian fleet does not target a specific species.

In Senegal the landings of chub mackerel ranged between 1 000-9 000 tonnes for the period 1990-2002, with a peak of 8 900 tonnes observed in 1999. The landings in The Gambia ranged from 50 to 350 tonnes.

Fishing effort

Fishing effort of the industrial fleet, measured in fishing days, peaked in 1998, with around 17 000 fishing days, and has decreased regularly since then, to reach around 10 000 fishing days in 2003. From 2002 to 2003 there was practically no change (Figure 5.2.2b).

The declared fishing effort of the coastal purse seiners, mostly on the northern fishery, however, has fluctuated appreciably from year to year, but with an overall increasing trend. The current level of about 50 000 trips is among the highest in the series, even though it suffered a reduction of almost 10 percent from the 2002 level in Zone A+B (Figure 5.2.2a).

Recent developments by country

Morocco

A marked increase in Moroccan landings of chub mackerel was observed in 2003, reaching a value of over 36 000 tonnes. This increase comes mostly from increased catches in Zone B, which nearly doubled relative to 2002.

Mauritania

Landings of chub mackerel in Mauritania for 2003 have been estimated at around 111 000 tonnes, an increase of about 5 percent relative to 2002 and the highest level in the time series.

The fleets fishing chub mackerel in Mauritania are mostly those from Russia, Cyprus and the EU. The highest catches of chub mackerel are reported during the warm season, coinciding with the peak abundance of sardinellas and horse mackerels. This indicates that chub mackerel is mostly a bycatch of the sardinella and horse mackerel fisheries.

Senegal

In Senegal, chub mackerel landings have increased markedly in 2002 to reach approximately the same level as in 1999, with around 9 000 tonnes. They decreased again in 2003, to a level of about 4 000 tonnes, closer to the average of the 1990-2000 period.

The Gambia

Reported landings of chub mackerel in The Gambia decreased by 15 percent in 2003 relative to 2002, to a level of 300 tonnes.

5.3 Abundance indices

5.3.1 Catch per unit of effort

CPUEs have been calculated from data from the Russian fleet, using only fishing RTMS days where more than 50 percent of the reported catch was chub mackerel (FAO, 2001). CPUE was standardized using the multiplicative model of Robson (1966) and Gavaris (1988). Fishing effort was standardized for the north and south stocks separately, and then applied to the total fishery.

The standardized CPUE, in tonnes/RTMS day, shows an increase in 2003 relative to 2002 (Figure 5.3.1). This increase may have been due to a reduction of the fishing effort, after the departure of a part of the pelagic trawling fleet from Mauritanian waters. Another explanation might be an actual increase in the biomass of chub mackerel in the study zone, as suggested by the acoustic surveys of R/V DR. FRIDTJOF NANSEN in November-December 2003.

5.3.2 Acoustic surveys

R/V DR. FRIDTJOF NANSEN

Starting in 1999, the abundance of this species was estimated during the acoustic surveys by R/V DR. FRIDTJOF NANSEN (Figure 5.3.2). In 2001 and 2002 chub mackerel biomass was estimated at around 300 000 tonnes. The main concentrations were registered between Cap Bojador and Cap Barbas. In 2003, two acoustic surveys were carried out by R/V DR. FRIDTJOF NANSEN, covering the area between Cap Cantin and Cap Blanc. In the first survey, carried out in May/June, chub mackerel biomass was estimated at 374 000 tonnes, while in the second survey, in November-December, this estimate was of 547 000 tonnes.

National surveys

The survey carried out in December 2003 by the R/V Al-AWAM in Mauritania has detected chub mackerel concentrations along the continental platform. The total biomass of this species in Mauritanian waters was estimated at 300 000 tonnes.

Other surveys

From 1994 to 2000, chub mackerel biomass in Moroccan and northern Mauritania waters was estimated to vary between 100 000 tonnes and 900 000 tonnes (AtlantNIRO surveys, FAO 2003).

The recruitment survey carried out by R/V AtlantNIRO during autumn 2003 detected the largest concentrations of chub mackerel juveniles between Cap Juby and Cap Barbas, as in May 1999.

5.4 Sampling of commercial fisheries

Table 5.4.1 shows the sampling intensity of chub mackerel off Northwest Africa in 2003.

In the northern fishery (Zone A+B), exploited by Moroccan purse seiners, 1 222 samples of mackerel were taken, with 40 individuals per sample.

In the Mauritanian fishery, length samples were taken by scientific observers on board the Russian and EU pelagic trawlers. The total number of samples taken was 399, with an average of 244 individuals measured in each sample. The 2003 Working Group recommended that at least one sample per 1 000 tonnes of chub mackerel landed should be taken. The present Working Group considered that the number of samples taken is adequate, but the number of individuals measured per sample is low, particularly for the Moroccan coastal seiners.

5.5 Biological data

The length distributions of landings from the northern and southern fisheries were analyzed for the period 1992-2003. The length distributions obtained in 2003 were compared with those from 2002 (Figure 5.5.1a,b,c,d).

The length distributions of the landings of chub mackerel from the Moroccan seiners in the northern (Zone A+B) fishery usually present three modes. In 2003, the proportion of small individuals (11-14 cm) in this area was relatively high.

The length distributions of landings of chub mackerel in the southern fishery, usually show one mode. The lengths of chub mackerel landed in this fishery have been in general larger than those from the northern zone.

The length distributions estimated for the stock during the acoustic surveys by R/V DR. FRIDTJOF NANSEN in May-June and November-December 2002 and 2003 are presented in Figure 5.5.2a,b,c,d.

In May-June 2002, four modes were estimated, while only three were detected in the November-December survey.

In May-June 2003 only two modes were detected. In November-December the same year, however, three were again found, albeit corresponding to smaller fish than in 2002. The number of juveniles estimated for 2003 was appreciably higher than in 2002, reinforcing the observations made from the landings of the Moroccan purse-seiners.

The estimated catch-at-age matrix (Table 5.5.1) was obtained applying the Russian Age-Length keys by quarter to the estimated length distribution of the landings from the northern and southern zones. The estimated catch at age for 2002 in the southern zone were corrected during the current WG meeting, to account for the corrections introduced to the data on chub mackerel landings in Mauritania.

5.6 Assessment

Data quality

In order to test the quality of the data available for the assessment, the sub-group carried out an exploratory analysis of the data, calculating the linear correlation between the estimated catches of each age-group and the corresponding number of the same year-class the following year. The results obtained (Table 5.6.1) indicate a very low correlation between the catches of supposedly the same cohort along life.

Table 5.6.1:

Values of the linear correlation coefficient between the estimated catches of consecutive ages of the same cohorts for chub mackerel


Age-group pair

1-2

2-3

3-4

4-5

Linear correlation coefficient

0.36

0.08

0.29

0.58

This lack of correlation, that indicates that cohorts cannot be reliably followed on the catch data, could be the consequence of several factors. These could include incorrect ageing, changes in the exploitation pattern (alternance of several fleets), inadequate sampling, specific targeting on a few length groups, or incorrect stock delimitation. Irrespective of the source of uncertainty, however, the consequence is that it is not possible to rely on these data for an age-structured analytical assessment.

Therefore, the Working Group decided not to use the age-based analytical models for the assessment of this stock, at this stage.

Methods

The logistic surplus production model, implemented on an Excel Spreadsheet, was used. This model is further described in Appendix III.

Despite the unreliability of the catch-at-age data, some WG members tried to apply the VPA model to the data. However, the results of this modelling exercise were not accepted by the Group.

Input data

As catch data, the Working Group used the series of total catch estimated by the Group. A time series from 1992 to 2003 was used.

The abundance index used was the standardized commercial CPUE series (tonnes/RTMS day, FAO 2001). It was necessary to use the commercial CPUE series, despite the known limitations of this approach for pelagic schooling fish, because the series of acoustic survey abundance indices from R/V DR. FRIDTJOF NANSEN only started in 1999, and was therefore too short for the needs of the model.

An environmental effect was introduced into the model, to account for non-average growth conditions for the stock in some years of the series, namely 1999 and 2000.

Results

The fit of the model to the available data was considered reasonable (Figure 5.6.1). It manages to follow the main trends in the abundance indices, even though there is little contrast in the time series of abundance indices, that reduces its reliability.

The model results indicate that the current biomass is above the level of the biomass producing the maximum sustainable yield, on average, but that current fishing mortality is slightly above, or at the same level, as the sustainable one at the current biomass levels (Table 5.6.2).

Table 5.6.2:

Summary of the current status of the stock and the fishery for chub mackerel


Stock

B/BMSY

Fcur/FSYCur

Scomber japonicus, Northwest Africa

153%

114%

One may thus consider that the adult fraction of the stock is fully exploited. However, the reliability of these results is hampered by the doubts around the abundance index used to fit the model, and by the possibility of marked changes in the relative exploitation pattern on the other. Given these uncertainties, the results should be viewed with caution.

Discussion

The lack of contrast in the time series of catch and abundance indices decreases appreciably the reliability of the results obtained in the modelling. This reliability is also affected by the use of the commercial CPUE data in the fitting, since it is well documented that CPUE from fisheries of pelagic schooling fish may often fail completely in tracking stock abundance. Therefore, great care must be exercized when interpreting these results. However, the modelling results indicate that the stock is currently not overexploited, and these results find some support in the observation that the abundance indices obtained from the acoustic surveys have shown a slight increasing trend.

5.7 Management recommendations

Given the results obtained in the assessment, and the uncertainties associated with it, the Working Group recommends taking a precautionary approach, and maintaining the effort level in this fishery below or at the average level recorded during the last 5 years.

5.8 Future research

1. Enhance length sampling of the catches, increasing the number of individuals measured per sample and ensuring coverage of all size intervals.

2. Broaden the sampling system to cover the catches of all fleets in every country in the region.

Standardize all length measurements, to measurements of total length (TL) recorded to the cm below.

4. Carry out biological studies of growth and reproduction, in order to obtain better information on the biological cycle of chub mackerel.

5. Encourage the collection and reading of otoliths, to estimate appropriate age-length keys for each fishing area.

6. GENERAL CONCLUSIONS

A summary of the assessments and management recommendations by the Working Group is presented below:

Stock

Assessment

Management recommendation

Sardine/Zone A+B

The model applied gave no reliable results

Do not increase catches above average level of last 5 years as a precautionary measure

Sardine/Zone C

Stock not fully exploited

Current fishing effort may be temporarily increased but should be adjusted to natural changes in the stock

Sardinella/all sub-region

Stock of S. aurita over-exploited; no reliable results for S. maderensis

Reduce fishing effort on sardinella by 20 percent, particularly in fleets targeting S. aurita

Horse mackerel/all sub-region

Stocks of Trachurus trecae probably fully exploited; no results for T. trachurus and Caranx rhonchus

Do not increase fishing effort above average level of last 5 years

Mackerel/all sub-region

Stock probably not fully exploited

Do not increase fishing effort above average level of last 5 years as a precautionary measure

In the absence of reliable length and/or age compositions, the Working Group used production models for all stocks. The disadvantage of these models is that they cannot follow individual age groups, and thus cannot simulate the effect of changes in relative exploitation patterns on the stock. In previous years, the Working Group used the models assuming average stock growth parameters for all years. This year it was realized that most stocks in the area had been influenced by abnormal hydrographical conditions in certain years. The fit of the production models was significantly improved by specifically including an index of environmental quality. However, the values used for this index were defined on a rather subjective basis, and a more objective procedure has to be developed in the future. For this purpose, more research on hydrographical variability in the region and its effects on stock dynamics should be encouraged and the results made available to the group.

In formulating the results of the assessments, the Working Group lacked a set of uniform reference points and management objectives for all stocks in the area. The Working Group agreed that next year a discussion should take place on the definition of appropriate reference points.

Although the amount of catch, effort and biological data available to the Working Group has increased in recent years, some deficiencies persist. The main deficiency remains reliable age data for most of the stocks. The study of age and growth, therefore, remains a priority for the Working Group. Other data deficiencies concern the species and length composition of the landings and discards of the industrial fleet in Mauritania and the length distribution of artisanal catches in Senegal and Mauritania.

7. FUTURE RESEARCH

The Working Group recommended that the following research areas be pursued in 2004/2005:

1. Adhere to recommendations concerning sampling given in 2003.

2. Adequate sampling of all industrial fleets in Mauritania including discards and catches used for fishmeal.

3. More sampling of artisanal catches in Senegal and Mauritania.

4. Improve age reading of sardinella.

5. More research on hydrographical variations and their effects on fish stocks.

6. Improve national acoustic surveys and do comparative surveys with R/V DR. FRIDTJOF NANSEN.

7. Study of length distributions of all target species in acoustic surveys and commercial catches.

8. Inclusion of environmental parameters in production models.

9. Refine reference points for fisheries management.

10. Improve definition of effort.

BACK COVER

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 give advice on 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 meeting was held in Saly, Senegal, from 17 to 27 March 2004. A total of 15 scientists from seven countries and from FAO participated. The results of the assessments indicate that the sardine stock in Zone C is not fully exploited and the Working Group hence noted that current fishing effort may be temporarily increased, but should be adjusted to natural changes in the stock. As regards the central stock of sardine, it is recommended to not increase catches above the average level of the last five years as a precautionary measure. For sardinella, the Working Group recommended to reduce fishing effort by 20 percent, particularly in fleets targeting the round sardinella. The Working Group also recommended not to increase fishing effort above the average level for the last five years for the Cunene horse mackerel and Chub mackerel.


Top of Page Next Page