If CPUE could be interpreted as indices of abundance of the round sardinella, then the following estimates on the state of the resource can be made:
The CPUE of the Ghanaian beach seine taken as indices of recruitment allow for a first estimation that the cohorts of S. aurita born in 1988 and 1989 were relatively substantial. Unfortunately, there were no similar data in côte d'Ivoire. The CPUE of the beach-seine in Benin did not contradict those in Ghana.
Concerning the major fisheries of S. aurita, the availability of the resources in 1988–89 did not follow the pattern observed from 1985 to 1987; and in côte d'Ivoire, as in Ghana, the CPUE of the different fisheries did not follow parallel evolutions in 1988. In 1989, all the CPUE returned to high levels and the abundance of S. aurita appeared strong.
Sardinella aurita
The working group attempted to reconstruct for 1988–89 in the whole region, and for 1990 in Côte d'Ivoire, a vector of weights by length class in numbers of individuals (Figures 12a–12d).
In Côte d'Ivoire, the vector obtained for the industrial catch could be considered less biased because of good sampling. The same applied for the artisanal fisheries, except for the fact that the beach-seine catches had not been sampled.
In Ghana, there were only a few length samples in 1988 and these were from the Greater Accra Region only. An extrapolation of the length samples to total catches of the country was done month by month, for the ali/poli/watsa nets on one hand and for the beach-seine on the other. In 1989, the same procedure was followed for the first six months, and for July-December the samples collected in each region were extrapolated to the monthly average catch of the region before being totalled, then extrapolating to the total monthly catch of the country when not all the regions possess samples. Concerning the beach-seines, for which there were no samples in Ghana, the following method was used: after evaluating the eventual negative impact that catches of juveniles could have on the stock, it was supposed, in the worst hypothesis, that all beach-seine catches of small-sized fishes were apparently the same as at Tabou, Côte d'Ivoire. The samples collected at Tabou were therefore, used for calculating the total individuals caught by the beach-seines in Ghana. This vector of weight can therefore only be utilized for measuring the impact of this gear in the total catch.
Given the nature of the methods used in the calculations and also because the sampling was not very representative (especially in Ghana), the working group cautioned that the conclusions drawn from the analysis of these catch vectors be considered with great circumspection. All the same, the following characteristics became clear:
the Ghana beach-seine seems to catch smaller quantities of juveniles compared to the poli net;
the catches of Ghanaian semi-industrial purse-seiners were also negligible compared to those of the canoes; and these were mostly mature individuals of over one-year-old;
in the same way, the Ivorian purse-seiners exploit only large-sized fish;
the catches of canoes, while convening all the age-classes showed a high exploitation of individuals aged less than one year. Of the three fishing gears, the poli net in particular exercised this mortality on the young, whilst the ali or watsa nets exploited mainly individuals longer than 15 cm (over one-year-old);
the working group also noted with interest, the U-shape form of the exploitation curve in 1988 and 1989, which could be explained as follows:
it is a consequence of the marked seasonality of the fisheries, and constitutes a “photograph” of the ages of the different cohorts ever since the season of high exploitation;
it originates from the phenomenon of mesh selection and/ or of catchability due to variable behaviour of the fish according to its size.
It is possible that the catches sustained at a high level in 1989, in spite of a weak upwelling were due to strong recruitment in 1988. The 1989 recruitment was very weak.
The working group was not able to draw from this analysis of the exploited lengths any firm conclusion on the state of the S. aurita stock because of the doubtful reliability of the data, as indicated above. From the analysis of the 1988–89 curve, it appeared likely that the recruitment of 1989 was less important than that of 1988. It seemed equally likely that there has been a certain stability in the exploitation pattern of the different gears.
Noting the progress made towards an analytical approach, the working group still recommended an intensification of the poli net sampling which could provide a good index of recruitment to complement that provided by the beach-seines.
Other species
Due to the lack of adequate data, the working group was unable to evaluate the state of stocks of the other species.