The significant impact of upwelling on sardinella stock recruitment and abundance has been demonstrated in many papers (ORSTOM/FRU, 1976; Binet, 1983; Eréon, 1986; Cury and Roy, 1987).
Models used for the “Fishery-Climate” relationship are a possible approach if upwelling indices are introduced in the global models. Gulland and Garcia (1984) discussed the implications of basing management decisions on a model fishery, and the models used in the study on fishery/climate relationship can be used in applying upwelling indices to global models (Fréon, 1986; Cury and Roy, 1987).
Sardinella aurita
Binet (1983) and Fréon (1986) have demonstrated the limitation resulting from the utilization of one single model for the interpretation of catch or CPUE fluctuations over a whole time series (1966–1986). It appears as if the responses of a stock to climatic fluctuations are different in the case of its collapse or of its abundance. This would suggest two situations, one of poor abundance and the other of prosperity, with different response to climatic fluctuations. Therefore it is difficult, using a single model, to demonstrate the dynamics of Sardinella aurita. Additional work is needed and proposed under the ORSTOM “Fishery-Climate” programme.
Sardinella maderensis
There is less difficulty with this species which did not experience any collapse.
To analyse CPUE trends in relation to fishing effort and upwelling indices, the model used had the following formula:
CPUE = e-Ao f (B upwi + C upwi - 1) + r
where: i = year index
i - 1 = previous year index
Ao, B and C = constants
r = residue
At the level of the working group the model was used to analyse the Sardinella maderensis CPUE trends of industrial seiners off Côte d'Ivoire. Upwelling indices were those of Ghana and the fishing effort was calculated on the basis of the number of days at the fishing grounds.
Observed CPUE values and model-adjusted CPUE values are presented in Figure 29, which shows the significant contribution of the two upwelling indices (years i and i-1), and the particularly important effect of the index of year i-1 which suggests an impact on recruitment. Also important is the impact of fishing effort on CPUE structure, which partly explains the increase in CPUE values during the recent period when fishing effort decreased (since 1973).
CONCLUSION: Application of the model in this way allows for a faithful simulation of the historical variations of the stock. If the yearly fishing effort n + l and the upwelling index n are known, it should be possible to make usable predictions.
REMARK: The suggestion to use Port-Bouët upwelling indices and to take into account the artisanal ringnet catches of Côte d'Ivoire (whose development started in 1982) and off Ghana, was accepted. Calculation of a theoretical effort was made by dividing values of these catches since 1982 by CPUE of seiners from Abidjan; a further adjustment of values observed was made using the model, but results achieved are not conclusive.
It has not been possible to set up an analytical model for either species, the reason being the lack of a complete table of frequency of weighted and extrapolated sizes, to establish demographic tables and to undertake cohort analysis. The low rate of sampling for the various fisheries under study did not allow any such exercise. However, this was done for the 1982 Ghanaian artisanal fisheries (Appendix 4) to serve as an example.
The group expressed concern in view of these deficiencies which perpetuate the previous unsatisfactory conditions. Actions should be taken so that a demographic table for the two species be available in the near future.
The current state of exploitation of S. aurita stocks is not precisely known; however, the following has been noted:
high catches in 1983, 1985 and 1986, with an all-time record in 1985;
good consecutive upwelling indices from 1979 to 1986;
low preliminary upwelling indices in 1987 whose climatology was comparable to that of 1968, which suggest a high probability of poor recruitment in 1987. This should be verified at the earliest;
high catches of the 0+ age class in 1986 by the Ghanaian artisanal fleet and sharp decrease in 1987 (Table 9, Appendix 7);
Côte d'Ivoire catches mainly of the 2+ age class in 1987.
On the basis of this information, the group undertook a survey (detailed in Appendix 5). It is expected that low catches shall be recorded in 1988 and that a risk, by no means negligible, exists of a collapse of S. aurita stocks in 1988.
Sardinella maderensis
Proper understanding of the state of the stock(s) is not readily achieved. It is acknowledged that the EEZ of Côte d'Ivoire corresponds more or less to one stock. The global predictive model related to upwelling indices and to total effort of Côte d'Ivoire suggests high abundance levels (CPUE) (Figure 27).
No conclusion has been reached with respect to the other fishing zones, particularly the Ghanaian EEZ. It should be recalled that one assumption is that S. maderensis displays better resistance than S. aurita to (adverse) environmental conditions (and to fishing effort ?).
Catches of anchovy within the zone under review are mainly by the Ghanaian artisanal fishery (ringnets and beach seines) east of Cape Three Points (Table 5). Over the period 1972–1986 the catches followed two major trends, an increase from 1972 to 1981, then a decrease.
Compared to S. aurita and S. maderensis, anchovy is dominant in the catches by the Ghanaian artisanal fishery. The fishing season is from September to December, just after the sardinella season. Fishing grounds are in the vicinity of Cape Saint Paul (Eastern Ghana). It is also noted that pole and line tuna vessels fish anchovy from the stock around Tema and use it as bait; it would appear that this stock is more readily accessible between September and December (Appendix 6).
Two sets of CPUE data on the Ghanaian artisanal fishery covering the period 1972–1986 were available to the working group, one concerning the beach seines, the other the Poli/Ali nets (Figure 28). These two CPUE values indicate moderate fluctuations with no plain trend. From the discussions it emerged that CPUEs are dependent on the results of the sardinella fishery, and thence on the economic importance of the anchovy fishery as against the sardinella fishery. This was particularly evident in 1981. Thus, effort is transferred to the anchovy when the sardinella fishery is unfavourable; a transfer which seriously compromises a comparative analysis of variations in abundance between the different pelagic species.
In general, the size of fish caught in Ghana ranges between 5 and 10 cm, but no regular size sampling is available. In Togo, the size of fish caught with Watsa nets (purse seines) varies 6 cm throughout the year (Faggianelli, 1984).
It is difficult, therefore, to reach any conclusion on the current state of the anchovy stock(s) in relation to the Ghana-Côte d'Ivoire upwelling. Considering the actual tonnage of landings and their resulting economic value, this calls for additional studies on the species and its comparison with other upwelling-affected small pelagic species.
Compared with the other species, the volume of catches was relatively limited except in the year 1986.
In Côte d'Ivoire, total disappearance of the stock occurred from 1973 to 1982 followed by steady recovery as from 1983.
In Ghana, the exceptionally high landings (13 000 tons) of 1986 were solely made up of catches by the artisanal fleet, and of individuals of the same size (9–10 cm) as Sardinella aurita; this suggests a recruitment peak although in 1987 no significant catches of mackerel were recorded. The query arises as to what happened to this recruitment class.
Knowing that the lifespan of Scomber japonicus is 5–6 years, the above information is surprising and does not lead to any explanation about the dynamics of the species. A longer time series is needed for more definite conclusions. It is therefore necessary to proceed with measurements of the size of mackerel when re-entering the fishery.
