The chub mackerel (Scomber japonicus, Houttuyn 1782) is a neritic-oceanic species, which is distributed in the Eastern Atlantic.
In the study area it occurs in the catches in Morocco, Mauritania, Senegal and the Gambia, although it is more abundant in the northern part of the area. In Zone C, north of Cape Blanc and Mauritania the chub mackerel has been recorded from the coast to more than 300 m depth and the distribution is related to hydrological conditions (FAO 2001). It should be noted that it is difficult to find concentrations of this species during acoustic surveys.
In previous Working Groups (FAO 2001; FAO 2002), two stocks were assumed based on information from the fisheries. However, due to the uncertainty about the stock identity of this species, and taking into account its highly migratory nature this working group decided to attempt to assess the total fishery of this species as well.
In the following sections, the northern fishery refers to the fishery between the north of Morocco and Cape Bojador; and the southern fishery refers to the fishery from Cape Bojador to southern Senegal.
The northern fishery between Tanger and Cape Bojador is conducted by coastal seiners from Morocco that target sardine. During the summer (May-August) this fleet also targets chub mackerel when this species becomes available in this area (mainly between Cape Juby and Cape Ghir). South of Cape Bojador, the fishery is conducted by pelagic trawlers from various countries (Russian Federation, Ukraine, EU and others) that operate under fisheries agreements or under joint ventures. Chub mackerel is normally not the main target species for these fleets. In Senegal and Gambia, chub mackerel is taken as by-catch by the artisanal and industrial fleets.
The characteristics of the different fleets are described in Sections 2.2, 3.2 and 4.2.
Tables 5.2.1 and 5.2.2 show respectively the catch and effort of Scomber japonicus by country and fleet in the study area for the period 1990-2002. The effort series from Mauritania was revised as regards to last years report.
Figure 5.2.1 shows the total catch and the catch by country. The Figure shows that the total catch of Scomber japonicus increased from around 20 000 tonnes in 1991 to over 200 000 tonnes in 1997. Since 1997 a general decreasing trend can be observed with a total catch of around 100 000 tonnes in 2002. The trend in total catch mainly reflects the developments of the fisheries in Zone C, north of Cape 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 Cape Blanc since 1998 can be explained by the expiry of the fisheries agreement with the Russian Federation 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 Zone C was reduced to almost zero in 2002. The total catch from the northern fishery fluctuates between 10 000 and 30 000 during the period. In 2002 the catch from this fishery was about 22 700 tonnes, which is a slight decrease compared to the catch of about 25 600 tonnes in 2001 (Table 5.2.1)
The total catch of chub mackerel in Mauritania shows an increasing trend from the early 1990's to 1996 when the catch peaked at around 100 000 tonnes. From 1996 to 1999 the catch decreased, followed by an increase from 1999 to 2002. In 2002 the catch from Mauritania was the highest in the region. Two of the most important fleets operating in Mauritania are the Russian and EU fleet. The EU fleet operating in Mauritania targets 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. The observed increase in catch from 2001 to 2002 was probably due to some big catches of chub mackerel by the EU fleet in July and August 2002. These catches coincided with the biggest catches of sardinella. The Russian fleet fishing in Mauritania was also forced to turn to mackerel due to the lack of sardinella and horse mackerel.
In Senegal the catches 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 catches in the Gambia ranged from 50 to 350 tonnes.
Figure 5.2.2a and 5.2.2b show the nominal effort for the northern fishery (number of trips) and the southern fishery (fishing days) respectively. As mentioned previously, it should be noted that chub mackerel is not the main target species for these fleets.
The total catches, total effort in fishing days and CPUE in tonnes per fishing day from 1992 to 2002 are presented in Table 5.3.1. As regards the standardisation of effort, some small changes were made with respect to last year. For the standardisation, daily catches of more than 35 tonnes with chub mackerel content of more than 50 percent from the Russian fleet were used. Many of those were obtained by long tows. The standardized effort was obtained for the southern fishery and then estimated for the total fishery.
Figure 5.3.1 shows the standardized CPUE for chub mackerel.
The low CPUE observed in 2002 could be explained by the fact that the vessels in Mauritania were forced to turn to mackerel even though this species was also scattered (not dense) during periods of low availability of sardinella and horse mackerel. There is no reason to suggest that the observed reduction in CPUE is a result of a decline in biomass.
This species, known for its migratory nature, has only been partially covered by the acoustic surveys.
From 1994 to 2000, chub mackerel biomass in the Moroccan waters and northern Mauritania varied between 100 000 tonnes and 900 000 tonnes (AtlantNIRO surveys).
Figure 5.3.2 shows the time series of biomass estimates from the R/V DR. FRIDTJOF NANSEN surveys. In 2001 and 2002 the chub mackerel biomass was estimated at around 300 000 tonnes. The main concentrations were registered between Cape Bojador and Cape Barbas.
Table 5.4.1 shows the sampling intensity of chub mackerel off Northwest Africa in 2002.
In the zone north of the Cape Juby, exploited by Moroccan coastal purse seiners, the number of samples taken to evaluate the catch structure was 25-32 individuals measured per sample. These samples are too small to cover all the different lengths landed, in order to analyse the population dynamics.
For the Russian pelagic trawlers, the samples were made onboard commercial vessels. The most intensive observations were carried out in the third quarter of 2002. As most fish were young, the number of fish aged was relatively low compared with the number of fish measured.
The length distributions for the northern fishery were analysed for the period 1992-2002 as seen in Figures 5.5.1a-b. For most of the years, peaks can be observed every 2 cm. This is probably due to the fact that the fleet, for most of the year, targets sardine and that mackerel is only caught as by-catch, but during the summer the fleet also targets mackerel. Another reason could be the absence of a standardized way of measuring the fish i.e. small fish are grouped to ½ cm and larger fish to 1 cm. However as the noise can be reduced by grouping at age, these data were used to prepare the catch at age matrix.
In order to prepare the catch at age matrix for the total fishery, the Russian age length key from the southern fishery was applied to length distribution of the northern fishery. Some estimations were made for the gaps in the age length key (Tables 5.5.1a-f).
Figure 5.5.2 shows the length distributions from the R/V DR. FRIDTJOF NANSEN surveys in 2000, 2001 and 2002 (May-July and November-December). Two modes, one around 20-23 cm and one around 27-28 cm can be observed for the November-December surveys in 2000 and 2001. In 2002 two modes are also observed but with the first mode at around 18-20 cm and the second at 28-29 cm. Fish of smaller lengths were observed in the spring survey (May-July 2002).
The following relationship between the weight in grams and total length in mm was used W=0.0000003L3.585 (2002 data, Russian Federation).
Data Analysis
Figures 5.6.1 and 5.6.3 show the catches at age per year class for the southern and total fishery (fishery south + north) respectively for Scomber japonicus. Figures 5.6.2 and 5.6.4 show the log catch ratios by age.
Figure 5.6.1 shows a varying pattern with recruitment to the fishery at age 2 for year classes 92, 94 and 96, recruitment at age 3 for year classes 93 and 95 and at age 4 for the 91 year class. This may indicate noise in the data that might be caused by different availability of the age groups for the fishery in different years, wrong assumption of the stock unit, problems with age readings or problems in sampling. The same general pattern as for the southern fishery can be seen for the total fishery (Figure 5.6.3), although with slightly less noise in the data. For the total fishery the 1994 and 1996 year classes can be seen to follow the same pattern as can the 1993 and 1995 year classes. A Working Group held in Nouadhibou, Mauritania in December 2002 also highlighted the high variability of fishing patterns over the last years.
The log catch ratios for the southern region (Figure 5.6.2) show large inconsistencies in the catch matrix. The number at age in many years for an older age group of the same cohort is higher than the younger one. Excluding the log catch ratios for the younger ages (0-1 and 1-2), the log catch ratios for the total fishery were found to be somewhat better (Figure 5.6.4). From the figure it can be seen that the log catch ratios for the 2-3 group are consistent for the period 1996-1999 for the 4-5 group consistent estimates can be seen for the last three years.
It may be that the data reflect real ratios of the ages in the catches. However, the observed relations are not favourable to VPA based analysis methods, which require the stable availability of age groups.
Methods
In spite of the inconsistencies in the data, the Working Group decided to do exploratory analyses using Integrated Catch-at-Age Analysis (ICA) and Extended Survivors Analysis (XSA). The analyses were first made on the southern fishery and then on the total fishery.
In summary, the following input data were used for the ICA and XSA runs:
The age compositions in numbers and the mean weight-at-age of the catches for the years 1992-2002. Age 6 was considered as 6 + group.
The natural mortality coefficient used was 0.5 per year (constant for all ages)
The proportion of mature at age
The standardized CPUE was used as abundance index
The input data are given in Table 5.6.1.
Southern fishery
Data on age composition and mean weight-at-age of catches by year, used by the 2002 Working Group (FAO 2002), were updated to include 2002. The 2001 data were also revised to reflect the final statistics for that year. The available series covers the period 1992-2002.
The maturity ogive and natural mortality were assumed to be the same as in the last Working Group (FAO 2002).
Total fishery
The catch-at-age matrix for the northern fishery was prepared for the period 1992-2002 using the length compositions from the Moroccan purse-seiner fleet in Zones North, A and B, and the Russian age-length keys from CECAF Statistical Area 34.1.3 with the following modifications:
In general it was noted that the Moroccan catches included smaller individuals than the Russian catches and that therefore some length classes were not represented in the Russian keys. To overcome this problem an estimate was made for the % age distribution of the smaller fish. Estimations were also made for gaps in the age-length matrix.
The Russian age-length keys are prepared by quarter, but not all quarters are available for every year. Also the length composition from the fleet is not always available on a quarterly basis. Hence for the period 1992-2000 it was decided to prepare annual age-length keys. Due to the lack of an age-length key for 1993, the 1994 key was applied to that year.
For 2001 and 2002 the age compositions were prepared by quarter and then summed. As the 2001 age-length key was very poor (only one quarter sampled and with many gaps) it was decided to apply the 2002 key to that year.
The catch-at-age for the total fishery was obtained by adding the composition for the southern and northern fisheries.
As regards the mean-weight-at age and the maturity ogive, the same matrices as for the southern fishery were used including an estimation for Age 0. The estimation of weight for that group was made using a length-weight equation (see Section 5.5)
Results and discussion
Southern fishery
Analysing the results obtained by these two methods, it was found that the residuals of the ICA run were very high while those of the XSA were somewhat lower. However as the residuals were not found to be satisfactory, these results were not retained.
Total fishery
The results with XSA were obtained using the following settings:
catchability independent of age for ages more than 4;
power dependence between the abundance indices and estimated numbers by ages 0-3, linear dependence by ages 4-5.
The following setting was used for ICA:
down weight of 0 group in the catch at age matrix (to 0.1);
linear dependence between the abundance indices and the numbers by age groups were used.
Figure 5.6.5 shows the results from the assessments of Scomber japonicus using XSA and ICA. The figure shows that with XSA the residuals have a decreasing tendency by year whereas the residuals for ICA do not seem to have any tendency. However the total value of the residuals are somewhat higher for ICA than the XSA. The SSB and fishing mortality trajectory for the two methods are different in terms of level and trend for the SSB and F level. The inconsistencies in these results may be explained both by the insufficient data and by the different distribution of the fishery relating to age groups in different years. Further exploratory analysis and studies on stock definition should however be made.
As this was the first trial to assess all of the fisheries assuming one stock, and due to the uncertainties of the results, the Working Group is not in a position to give advice on the catch levels based on these assessments. As a precautionary approach, the Working Group recommend that the catch levels should not exceed the average recorded during the last 5 years.
