2.1 Stock identity
2.2 The fisheries
2.3 Catch and effort data
2.4 Biological data
2.5 Abundance indices
2.7 Management recommendations
2.8 Future research
The sardine (Sardina pilchardus Walb., 1792) found off the Northwest African coast, mainly between Cape Spartel and Cape Blanc, is divided in four fishing zones:
(Sidi Ifni-Cape Bojador)
(Cape Bojador-the southern extent of the species)
Based on the results of these studies, at the time of the 1978 Working Group (FAO, 1979) the evaluation of the sardine in Zones A, B and C was carried out as if they were distinct stocks. After observations made on the migration of sardines from Zone A to Zone B however, evaluation was limited to two distinct stocks, Zones A+B and Zone C.
The Working Groups which followed (FAO 1985, 1990, 1997) chose to adopt three separate stocks:
The evaluation of sardine was made on two stocks: The Central Stock (Zones A+B) and the Southern Stock (Zone C) (Figure 2.1.1).
Recent studies on sardine distribution between the 35°N and 19°N parallels have shown the existence of two sardine populations (28°N-35°N) and (27°N-20°N). One study on sardine stock identity, based on methods such as genetics, is currently being carried out by the National Fisheries Research Institute of Morocco (INRH).
In the whole of the CECAF zone, four sardine fishing grounds have developed from the north to the south (Figure 2.1.1). The fleets operating in these fishing grounds are very diverse and composed of traditional small purse seiners, freezer-purse seiners, pelagic trawlers and mother ships with their purse seiners.
Northern Zone (Cape Spartel-Eljadida)
A Moroccan fleet composed of around a hundred traditional-type purse seiners (gross tonnage 40 tonnes and 250 HP) operates in this area, based for the most part in the ports of Larache and Casablanca, the landings from these two ports represent over 80% of the sardine catch for the whole Northern Zone. During the last ten years, whereas the fishing effort of the fleet has remained stable, sardine catches have decreased, passing from the 24 000 tonnes registered in 1993 to less than 5000 tonnes in 1999.
Zone A (Safi-Sidi Ifni)
This zone is fished exclusively by the Moroccan fleet. The number of purse seiners operating in this zone is around 150 during the fishing season (May-September). The fishing effort of these purse seiners which have the same characteristics as those that operate in the Northern Zone, is progressively decreasing (in terms of positive trips). The catches have also been declining since the beginning of the 90s.
Zone B (Sidi Ifni-Cape Bojador)
Since 1983 the Moroccan fishing fleet has become quite important in Zone B following the transfer of a part of the fleet that operated in Zone A to the new ports of Tan Tan, Laâyoune and Tarfaya, which were opened respectively in 1982, 1989 and 1994. From 1990 to 1999 the total number of active vessels in this zone was around 200, with a gross tonnage of between 50 and 55 tonnes and an HP of between 250 and 300. The average annual catch of the fleet was 350 000 tonnes.
The Spanish fleet, composed of purse seiners, has traditionally fished for sardine between the Straits of Gibraltar and Cape Juby. Part of this fleet, coming from peninsular Spain, fished in the zone north of Casablanca, while the other part of the fleet, based in the Canary Islands, fished in Zone B. In 1976 the total number of vessels was forty purse seiners with an average gross tonnage of 130 and around 400 HP engines. Since then there has been a progressive decline in the size of the fleet.
Zone C (Cape Bojador-the southern extent of the species)
Two types of fleet operate in Zone C: the purse seiners (Moroccan and Spanish) and the pelagic trawlers (Russian, Ukrainian and others).
The Moroccan fleet is composed of approximately ten purse seiners operating in the Dakhla zone. The Moroccan catch from this zone has been increasing since 1996.
In 1983 the fleet that began working in Zone C was made up of thirty vessels with a gross tonnage of 214 tonnes and 659 HP on average. Smaller vessels were replaced by larger ones, but always within the limits established by the fishing agreements signed by Morocco and the European Union. In 1995 the fleet had been extended by eleven vessels of which 70% were between 250 and 500 GRT.
Since 1996 the fleet of Spanish purse seiners has only fished in Zone C following a transfer of effort imposed by the last fishing agreement between Morocco and the EU which was signed in 1995. The conditions of this agreement made provision for an annual closure to fishing during February and March and that the fishing zone which had originally been established at two nautical miles from the coast be moved to fifteen miles from 1998.
In the Zone north of Cape Blanc, the number of Russian vessels operating within the framework of the Morocco-Russian fishing agreements was between 14 and 12 from 1995 to 1999 with a quota of between 200 000 tonnes and 90 000 tonnes.
During this period, catch and effort of the trawlers has undergone changes connected to the conditions established by the fishing agreement. After 1996, the sardine quota for the Russian vessels was fixed at 5% of the total annual catch. The Russian fleet during this period targeted mackerel, horse mackerel and other species.
In addition the annual closure to fishing between September and October was required to meet the conditions of the fishing agreement mentioned above. Furthermore, the fishing zone established beyond twelve nautical miles from the coast was changed to twenty-five miles.
The other pelagic trawler fleets (Ukrainian and others) operating on a charter basis, increased from fifteen vessels in 1995 to twenty-seven vessels in 1999. These charter vessels land their catch at the ports of Laâyoune and Agadir.
It should also be noted that it is forbidden to fish for pelagic fish along the coast for 15 nautical miles between 24°-25°N. This measure was put into place in 1998 in order to replenish sardine stocks.
In Mauritania, the Russian pelagic trawlers are geared towards horse mackerel, mackerel and sardinella fishing. The sardine catch in the Mauritanian EEZ is not important. Sardine is also caught by EU and other fleets south of Cape Blanc, but it is not the target species.
The total annual catches (tonnes) for the different fleets operating in the Northern Zone, Zone A, Zone B and Zone C in the period 1990-1999 are shown in Table 2.3.1 and the historical catches in Table 2.3.2. The effort series available to the sub-group are shown in Table 2.3.3.
The time series prepared for analysis were 1976-1999 for Zone A and Zone B and 1983-1999 for Zone C. The Northern Zone was not considered in this Working Group.
For the period up to and including 1995 the catch and effort data for Zones A and B as reported by the Working Group on sardine held in Casablanca in 1997, were used (FAO, 1997).
For the period 1996-1999, the catch and effort data from the Moroccan fleet fishing in Zones A and B were made available to the Working Group. The catches and effort were made available on a monthly and annual basis. The Moroccan effort is measured as the number of positive trips and the total number of trips. Only the effort in terms of the number of positive trips is shown in Table 2.3.3, as these are the data used in the assessments. Spain reported a small catch with corresponding effort in 1996, after this year the Spanish fleet stopped fishing in Zone B.
The trends in catches and effort are shown in Figures 2.3.1a and 2.3.1b. Figure 2.3.1a shows an increasing trend in the Moroccan catches in Zones A+B, from 1976 to 1999. Figure 2.3.1b shows fluctuations in the Moroccan effort for the time series.
For Zone C it was decided to maintain the catch data as reported by the 1997 Working Group for the period until 1995 (FAO, 1997).
For the period 1996-1999, the group had catch information available from Morocco, Mauritania, Spain and Russia. Catch information from the Moroccan Coastal purse seiners and the Spanish purse seiners were available on a monthly and annual basis. The information from the pelagic trawlers operating in this area is also available. These were divided into two fleet categories Ukraine and others and Russia.
The annual catches from the Mauritanian part of Zone C were available from the EU, Mauritanian and others and Russian fleets. The catch data from the Russian fleet were also available from official Russian statistics for the sub-region, corresponding to the CECAF statistical divisions 34.1.3 and 34.3.1.
The effort from the Spanish purse seiners operating in Zone C is available for 1983-1999 by month and year. The Moroccan data are not available. The Russian effort until 1995 is available from the 1997 Working Group Report. This effort was obtained by taking a sample of the Russian vessels operating in Zone C in the period between 1983 and 1995. The criterion used for selection was that the vessels daily catch of sardine should represent more than 75% of the total catch per day. The data available on catch and effort were then combined for different Russian vessel types, region (Southern Morocco and Mauritania), month and year. The effort series presented to the Working Group was standardised by using the multiplicatif model of Robson (1966) and Gavaris (1983). This effort is shown in table 13, page 17 of FAO (1997).
For the years 1996-1999, problems were encountered in updating the above effort series for the Russian fleet using the method described above. Since 1996, the Russian vessels have reduced their activities in the area substantially and at the same time the target species has changed so that it was not possible to find samples that concurred with the 75% criteria described above.
Effort from the pelagic trawlers (Ukraine and others and Russia) fishing in Zone C north of Cape Blanc expressed as fishing days for the period 1993-1999 was also presented to the group. The data from the pelagic trawlers are collected by Moroccan observers onboard the pelagic trawlers
Figures 2.3.2a and 2.3.2b show the trends in total catch and effort by fleet in Zone C respectively.
For more details on the preparation of the catch and effort series, please refer to the reports from the meetings in Casablanca and Tenerife in 2000 in the Supplement (FAO Fisheries Report No. 657 - Supplement).
Zones A and B
INRH/Morocco supplied length composition both monthly and annually for sardine landings in Zone A for 1976-1999. The age-length keys and length-weight relationships are available quarterly for 1976-1995. Age composition for the period 1976-1995 is available quarterly and annually. Data relative to reproduction (sex, maturity stages) are also available for the same period.
For Zone B, length composition is available monthly, quarterly and annually for the period 1988-1999. Biological data collection has been regular and systematic since 1993. It should be noted that the 1993-1995 series, used by the previous Working Group, held in Casablanca in 1997, has been revised taking into account actual catches and the corresponding fishing zone.
Quarterly and annual age compositions are available for the period 1976-1992. Data are not available for the period 1993-1999.
For Zone B, IEO/Spain provided the biological data (length-weight, age-length keys, sex ratio, etc.) for 1976-1996, both quarterly and annually (FAO Fisheries Report No. 657 - Supplement). Length and age compositions are available quarterly and annually for the period 1976-1994.
For the years 1996-1999, data are available on length composition for the Moroccan fleet catches.
For the Spanish fleet, length composition, catch-at-age and biological data relative to growth and reproduction, the age-length keys, length-weight relationship etc. are available quarterly and annually for the years 1983-1999.
For the Russian fleet, length composition and catch-at-age for the years 1995-1999 is available. The maturity ogives for the years 1997-1999 as well as biological data relative to growth and the length-weight relationship for the years 1995-1999 are also available.
2.5.1 Catch per unit of effort
2.5.2 Acoustic surveys
For Zone A+B, the annual abundance indices (CPUE) were calculated separately for the Moroccan and Spanish fleets for the years 1976-1999. For the Moroccan fleet, effort is expressed in terms of positive trips. For the Spanish fleet it is expressed as fishing days.
Figure 2.5.1 shows the trend of three different abundance indices in Zone A+B. The Moroccan and Spanish CPUE and biomass index of R/V Dr. Fridtjof Nansen do not show the same trend.
As for Zone C, the CPUE of the pelagic trawler fleet was calculated using the standardised fishing effort for the series 1983-1995, which was used at the last Working Group (FAO, 1997). For the Spanish fleet, the CPUE was calculated for the years 1983-1999, the fishing effort used is expressed in terms of fishing days.
Figure 2.5.2 shows the trend of Spanish and Russian CPUE and biomass index by R/V Dr. Fridtjof Nansen. Different trends can be observed.
R/V Dr. Fridtjof Nansen
Since 1995 this vessel has carried out surveys in the region on a regular basis during the winter season (December). The vessel also surveyed the region in 1982 (August and November), 1989 (September) and 1992 (February). The R/V Dr. Fridtjof Nansen only covers sardine in the area south of Safi.
Figure 2.5.3 shows the biomass of sardine for the time series 1995-2000. The biomass estimates are presented for two sub-regions , from Safi to Cape Juby and from Cape Juby to Cape Timiris (taking into account that sardine in some years occurs also in Mauritania). In the northern area (Safi to Cape Juby) no major changes seem to have occurred and the biomass shows a slight increasing trend over the time series.
In the area between Cape Timiris and Cape Juby a marked drop in the biomass estimate can be seen in 1997. Over the last three years however the stock seems to be increasing again.
R/V Atlantida and AtlantNiro
Russian research vessels have worked in the region between 17°N to 32°N for many years. In 1995 and 1998 two surveys were conducted, one in February-March and the other in June-September. In 1994, 1996 and 1999 summer surveys were conducted, while in 1997 a winter survey was carried out. The biomass estimates for sardine for the years 1994 - 1999 are shown in Figure 2.5.4a and Figure 2.5.4b shows the biomass estimates for the third quarter for the same period. For the Southern area, this figure shows an increase in biomass in 1995 followed by a drop in 1996 and a decreasing trend until 1999.
Only the length composition of the 1998 survey was available to the meeting. The age composition from the Russian surveys for the years 1994-1999 was made available, but only on the last day of the Working Group (Table 2.5.1). This table was hence not used in the assessments.
In previous Assessment Working Groups (FAO, 1990 and FAO, 1997) two assessments were made, one of the sardine in Zones A+B and a separate one for Zone C. The Northern Zone was not assessed in these working groups. During the present meeting the question of stock structure was discussed and a suggestion to consider the sardine as one entity was put forward at the meeting. However, as no new information has been presented to change the approach taken in previous stock assessment working groups, this Group decided that the main approach should remain the same as in the 1997 Working Group where the sardine was considered as 2 units (Zones A+B and Zone C).
The new data only allowed a new assessment of Zone C.
One type of model was applied, the Virtual Population Analysis (VPA) with the Lowestoft software (Darby & Flatman, 1994). The Virtual Population Analysis (VPA) is a method of estimating the fishing mortality coefficients, Fi, and the number of survivors, Ni, at the beginning of each age group, from the annual catches, in number, of a stock, over a period of years.
The methods applied were within the Lowestoft Virtual Population Analysis (VPA) assessment tools (Darby & Flatman, 1994).
Data Input Zone C
Catch-at-age in numbers and the mean weight-at-age
Table 2.6.1 shows the catch-at-age and the mean weight-at-age for the period 1983-1999 in Zone C.
For the years 1983-1995 the age composition in numbers and the mean weight-at-age from the 1997 Working Group were used (FAO, 1997). For 1996-1998, age-length keys from Spain and Russia were used to convert length composition to age composition. Age 7 was used as a plus group. A detailed explanation of these tables can be found in the Tenerife Report 2000 (FAO Fisheries Report No. 657 - Supplement).
For the years 1983-1995 the mean weight-at-age in the catch and in the stock from the 1997 Working Group were used (FAO, 1997). For the years 1996-1998 mean weight-at-age in the catches was calculated using weight-at-age data collected by Spain and Russia. For each age group in each year the mean weight-at-age was calculated as the mean weighted by the corresponding catch in numbers of the two countries. From 1996 to 1999 the values of weight-at-age in the stock were considered the same as for 1995. The Russian 1996 catch distribution was modified using the proportions of fish at each length from the age composition of Spain in the same year.
Abundance indices from acoustic surveys
Acoustic abundance estimates from R/V Dr. Fridtjof Nansen for sardine by 1cm length group were used for tuning the VPA for the period 1995-1999 (Table 2.6.2). The length composition was converted to age composition using a Spanish age-length key for the 4th quarter. Due to the great variation in the 1996-97 biomass indices, biomass estimates were smoothed using the moving average technique over three years. This technique allowed less noise in the tuning models. The length compositions were then raised to the smoothed biomass estimates.
The maturity ogive from Spain used in the 1997 working group was used in the assessments (Table 2.6.3).
The natural mortality value used was the same as for the last assessment (M=0.6 per year).
Other input parameters
The Separable VPA method (Pope, 1977, 1979; Pope and Shepherd, 1982; Stevens, 1984) as used to examine the total catch-at-age data before tuning. High residuals were found for some age groups and some years and this could indicate possible data anomalies.
A reference age (2-year) was chosen against which the selection values for other ages were scaled. The values for the fishing mortality on the reference age in the last year (terminal F) were 0.4 and 0.5 per year. Two terminal selection values, i.e. that for the oldest independent age in the data range (used for all years) were selected (S=1.0 and 1.2). High individual residuals between ages and between years were found and the selection pattern was not consistent. Despite these results, the Extended Survivors Analysis (XSA) (Shepherd, 1992; Pope, 1977, 1979; Pope and Shepherd, 1982; Stevens, 1984), an extension of Survivors Analysis (Doubleday, 1981), was applied using the biomass indices from R/V Dr. Fridtjof Nansen. XSA focuses on the relationship between catch per unit of effort and population abundance, allowing for the fitting of a more adequate model for the relationship between CPUE and year class strength at the youngest ages.
The catchability dependent on stock size was set for ages < 2. For other inputs the program defaults were chosen.
Tuning converged after 11 iterations. When applying the Spanish age length key to the R/V Dr. Fridtjof Nansen biomass estimates, the pattern of the Log catchability residuals presented in Figure 2.6.1, show that the residuals are not randomly distributed over the ages.
The standard error of the mean log catchability of the 3 to 6 age groups (catchability independent of year class strength) are bigger than 0.5, only age group 2 is consistent.
The regression statistics for the ages with catchability independent of year class strength, the Rs squared, are not significant (only 5 observations) and the overall regression standard error gave values too high for all age groups except for age group 2 and therefore the results of the analysis are considered unreliable. Figure 2.6.2 shows the recruits (0+1) in thousand individuals, total biomass (t) and Fbar (2-4) (year-1) estimates by method and catch (t) from 1983 to 1999. The trends in recruitment and total biomass show a decline from 1983 to 1996. However an increase is observed over the last 3 years as can also be seen in the R/V Dr. Fridtjof Nansen survey indices (section 2.5 in the present report).
The total biomass estimates obtained from the present assessment using XSA are lower than the values obtained in 1997 using the ICA method for the same period. The quality of this assessment cannot be compared directly with the quality of the 1997 assessment as reliable assessment results cannot be derived from the present data.
Various inaccuracies in the sardine data were detected:
- In some years the length composition of the catch for some fleets was based only on the sampling distribution of one month of that year which means that the sampling is not adequate.
- Age composition: only one fleet has an age-length key (ALK) for all years by quarter and for some fleets the length distributions are not well covered by applying these ALKs.
- Some fleets have changed their exploitation pattern from a fishery targeting sardine to a fishery directed at other small pelagic species in the same area (Section 2.5.1). This might explain the inconsistencies of the selection pattern found in the separable VPA.
- The year to year variation in the biomass estimates from the acoustic surveys (Section 2.5.2).
Due to uncertainties in the assessment, projections for sardine were not done and the assessment results should not be used for management purposes. However, acoustic abundance indices in recent years indicate an increase in the stock.
1. Each fishery should have a proper biological sampling programme established in the fishing zone, covering the whole fishing season and all fleets. At least a hundred fish should be measured in each sample.
2. The sampling programme for age reading should cover the whole length distribution. In each sample all the otholits should be taken.
3. An age reading workshop will be organised in September 2001 in AtlantNiro, Kaliningrad, Russia in order to standardise the methodology and to establish the same age reading criteria for the Northwest Africa sardine.
4. All the input data for the assessment models should be ready on the first day of each of the future Working Groups. The data series for assessment must be the prepared including the previous year.
5. The members of the Sardine Sub-Group should be in contact by e-mail, exchanging information about the data assessment during the interim period.