Calculations were made on an HP 9845B computer. The programmes were prepared by Fernandez (Instituto Español de Oceanografía, IEO) and Barbe (CECAF) for the Fox production models and the cohort analysis, and by Pereiro and Fernandez (IEO) for the generalized production model.
5.1.1.1 Catches
In accordance with the recommendations of the preceding Working Group, a separation trial of the declared catches in Division 34.1.3 to distinguish between the presumed stock units preserved to be from Dakhla and Cape Blanc was attempted using the data available for Spain and Japan (which were the two main fishing countries) and some hypotheses concerning other countries were advanced.
The Cape Blanc catches were estimated as follows:
1966–68: only Japan exploited Cape Blanc (data supplied by Hatanaka, 1979);
1970–74: the Japanese catch data were given by Hatanaka (1979). The Korean catches were distributed between Cape Blanc and the Dakhla area following the distribution scheme of Japanese catches between the two sectors for lack of a better hypothesis. The Group stressed the fact that with the withdrawal of Japan from this fishery the Korean catch data have become essential and that adequate knowledge of their distribution between the Dakhla and Cape Blanc sectors has become absolutely necessary (cf. Tables 12 to 14). It was also considered that 10 percent of the total catches of other countries (which represent very small quantities) in the CECAF Northern Region are captured at Cape Blanc (the hypothesis accepted by the preceding Working Group);
1975–77: one hundred percent of the catches of Mauritania, Japan and Korea were considered as coming from Cape Blanc. Hatanaka's data show that this hypothesis was valid for Japan. Moreover, ten percent of Spain's catches as well as 20 percent of those of other countries were considered as coming from Cape Blanc;
1978–79: one hundred percent of the catches of Japan, Korea and Mauritania and 20 percent of the catches of other countries come from Cape Blanc. Spain no longer exploited this sector at all;
1980: the hypotheses were the same as for the preceding period, but only 77 percent of Korean catches came from Cape Blanc (Korean data).
The series we constructed in this way are given in Table 14.
The catches fished in the Dakhla region were obtained by calculating the difference between total catch (CECAF data) in Division 34.1.3 and estimated catches for the Cape Blanc region.
5.1.1.2 Cpue
For the Cape Blanc stock Hatanaka (1979) supplied a series of cpue calculated by biological years and already used at the preceding meeting. Since the declarations published in the CECAF Statistical Bulletin were broken down by calendar years, the cpue calculated by Hatanaka were adjusted (by mobile average for two years).
As Japan had not fished in 1980, the cpue estimate was based on the average ratio between Japanese cpue and that of Korea in 1978.
| Japanese cpue Hatanaka (1979) Data of Working Group No. 2 | Korea cpue Shimura (1981) | Japan/Korea ratio | |
|---|---|---|---|
| 1978 | 360 | 571 | R = 0.63 |
| 1979 | 210 | 355 | R = 0.59 |
| 1980 | (435) | 714 | R = 0.61 |
The Spanish cpue was adopted for the Dakhla stock as in the preceding meetings (from 1967 to 1975 the cpue used was based on a combination of the Spanish and Japanese data). In the future the ISPM will try to supply the cpue data of the Moroccan fleet.
5.1.1.3 Effort
Effort was simply calculated dividing total catches by the available cpue.
5.1.1.4 Results of Models1
The evolution of catches, effort and cpue is given in Figures 5 and 6.
Dakhla: the fox model fits the data well (see Figure 7). The best adjustment is obtained by m = 0.2 (see Figure 8). The main results are shown in the following table.
| Model | fMSY | MSY | f1980/fMSY |
|---|---|---|---|
| Fox | 640 000 h | 74 000 t | 2.4 |
| m = 0.2 | 470 000 h | 56 000 t | 3.3 |
The production model for m = 0.2 indicates heavier overexploitation than the Fox model. In all cases the stock is heavily over-fished, confirming the analyses of the preceding Working Groups. Effort calculated for 1980 is two to three times higher than the average which would be produced by balanced maximum catch.
A rough comparison of the cpue of several Moroccan ships, Spanishcpues (of which that for 1981 is provisional) and biomass assessments made by the N/O IBN SINA indicated that abundance increased in 1981 from 20 percent (Moroccan data) to about 100 percent (Spanish data). The provisional results therefore agree as to the tendency but diverge on the extent of the change.
The apparent improvement would supposedly be confirmed by Figure 16 which shows the evolution of the indexes for recruitment in the Dakhla area and particularly of its apparent increase in 1980.
However, this improvement does not change the conclusions concerning over-fishing of the stock based on the production model (1980 data), since although the number of Spanish ships decreased in 1981 by 36 percent that of the Moroccans increased by 50 percent and the total effort applied to the stock in 1981 will not be really known until all the final statistics become available.
Cape Blanc: the results are as follows (see Figures 9 and 10):
| Models | MSY | fMSY | f1980/fMSY |
|---|---|---|---|
| m = 0a | 62 000 t | ∞ | - |
| Shaefer | 46 000 t | 156 000 t | 0.17 |
| Fox | 44 000 t | 192 000 h | 0.14 |
If the available catch statistics are correct this analysis indicates that there has been a sharp decline in the level of exploitation of this stock in recent years from a state of full exploitation to one of marked underexploitation. The considerable increase of the cpue argues in favour of this interpretation.
The divergence of the conclusions between the analyses of the Cape Blanc and Dakhla stocks tends to confirm the presence of two separate stocks.
5.1.2.1 Catches
It was assumed that the cuttlefish captured in Division 34.1.3 belonged to a single stock; the chronological series was then re-established from 1969 to 1980. It appears in Table 15.
5.1.2.2 Cpue and Effort
The Group used the cpue data of Spanish ships to estimate total efforts in the area. Table 15 gives the evolution of these cpues and of efforts calculated on this basis.
5.1.2.3 Results of Models
Division 34.1.3
The production models for different m values were calculated. The adjustments were not very satisfactory. The results selected are as follows (see Figures 11, 12 and 13).
| Models | fMSY | fMSY | f1980/fMSY |
|---|---|---|---|
| m = 0.1 | 1 806 000 h | 7 600 t | 0.25 |
| Fox | 352 000 h | 7 000 t | 1.28 |
| Shaefer | 317 000 h | 8 400 t | 1.42 |
Although the best adjustment may be obtained for m = 0.1, the results are meaningless in terms of rational exploitation since they imply that the maximum catch (which differs little from that of recent years) would be obtained with an effort equal to four times the 1980 level. The more conservative Shaefer model indicates that the stock is heavily over-exploited.
The Fox model leads to a similar diagnosis.
However these results must be considered with caution since the adjustment is unsatisfactory and the lack of detailed data has led the Working Group to analyse the Cape Blanc and Dakhla concentrations together. Bearing in mind the widely divergent evolution of fishing effort in these two sectors (see discussion on octopus), this joint analysis may not be justified and it is more probable that, similar to the findings for the octopus fisheries, the Dakhla region is exploited more intensively than that of Cape Blanc.
The preceding Working Group has tried to make a yield per recruit analysis of the Cape Blanc stock using per fish mortalities varying according to age calculated by cohort analysis. The present data do not yet justify the use of this method (which is still the most satisfactory in theory) for the Dakhla sector because the catch per age vectors for some major fleets are not known.
Therefore, since time was lacking, the Working Group confined itself to examining the most probable situation of the fishery for this sector on the assumption of a constant F based on size at first capture. The analysis employed estimates of Z and mesh selectivity calculated in 1980 and 1981 based on the IBN SINA research campaigns. The results were calculated on an assumption of an M equal to 0.5 (the value accepted as being most likely) and that of a higher M equal to 1.0.
A range of sizes at first capture between 4 and 8 cm was used because of the continuing uncertainty concerning real age at first capture. Nevertheless the 4 cm value seems the most probable if one considers the Spanish trawlers' catch vectors by size category.
The following results were obtained:
| 1980 | 1981 | ||||
|---|---|---|---|---|---|
| Data | M | 0.5 | 1.0 | 0.5 | 1.0 |
| Z | 2.30 | 2.30 | 1.84 | 1.84 | |
| F/M | 3.60 | 1.30 | 2.68 | 0.84 | |
| M/K | 1.50 | 2.50 | 1.5 | 2.5 | |
| Results read on Beverton and Holt yield tables | Y/Ropt | 55 | 19.9 | 54 | 18.1 |
| Y/R for tc = 8 | 30 | 17.1 | 28 | 16.7 | |
| 19 | 12.1 | 15 | 13.2 | ||
| Y/R (tc = 8 to tcopt) | + 83% | + 16% | + 92% | 7% | |
| Y/R (tc = 4 to tcopt) | + 189% | + 64% | + 16% | 37% | |
| tcopt = | 17 cm | 11 cm | 17 cm | 12 cm | |
In the most probable hypotheses for the Working Party of M = 0.5 and tc = 4 cm (with the 60 mm mesh size and present fishing strategies), the stock appeared to be heavily exploited since size at first capture was well below the size at first capture likely to maximize yield per recruit. In fact, the estimated percentages of potential growth of yield per recruit can be biased, by the probable variations of F with age which are omitted in the model, particularly for juveniles at recruitment and for older individuals. (However, cohort analysis by the Jones method (see below) suggests a relative constancy of F during the entire exploitation phase). However, the Group drew attention to the fact that this analysis is still very rough since it does not take into account the fact that females disappear from the fishery much earlier than males. This kind of bias can lead to underestimation of recruitment per recruit and size at theoretical optimum first catch and consequently the conclusions have only scarce value. The Group recommended that a similar study by sex be undertaken.
It seems that, at least for the Dakhla stock which is heavily over-fished in terms of the production model, yield per recruit might be increased significantly by a reduction of per fish mortality exercised on young individuals (increase of tc): this result can be obtained either by increasing mesh size or by a reduction of fishing effort in the juvenile concentration areas. Additional analyses of selectivity of mesh sizes larger than 60 mm and studies of the space-time distribution of the youngest octopus concentrations and an analytical study by sex would be needed before the Working Group could make specific recommendations concerning an optimum fishing strategy.
An analysis of annual size frequency distributions by the Jones method was conducted on catches per commercial categories of the Spanish fleet from 1976 to 1980. These results were calculated taking two F/M values corresponding to the latest age. The results are given in the following table:
| Size in cm | Final F/Z = 0.8 | Final F/Z = 0.6 | ||||||
|---|---|---|---|---|---|---|---|---|
| 1976 | 1977 | 1978 | 1979 | 1976 | 1977 | 1978 | 1979 | |
| 2–5 | 0.74 | 0.82 | 0.76 | 0.71 | 0.74 | 0.82 | 0.76 | 0.71 |
| 5–6 | 0.89 | 0.89 | 0.82 | 0.81 | 0.89 | 0.88 | 0.82 | 0.80 |
| 6–8 | 0.86 | 0.85 | 0.84 | 0.82 | 0.86 | 0.85 | 0.83 | 0.82 |
| 8–11 | 0.73 | 0.70 | 0.68 | 0.71 | 0.72 | 0.79 | 0.67 | 0.70 |
| 11–13 | 0.75 | 0.74 | 0.72 | 0.76 | 0.74 | 0.72 | 0.70 | 0.75 |
| 13–15 | 0.87 | 0.86 | 0.86 | 0.87 | 0.85 | 0.84 | 0.84 | 0.86 |
| 15–17 | 0.80 | 0.80 | 0.80 | 0.80 | 0.60 | 0.60 | 0.60 | 0.60 |
 | 2.46 | 2.46 | 1.97 | 1.98 | 2.33 | 2.33 | 1.83 | 1.86 |
These results suggested that the exploitation rate according to age was relatively constant. Moreover we observe that in the Spanish fishery from 1976 to 1979 the 2–5 cm (that is, less than the theoretical tc of the mesh used) commercial class seemed to be fully recruited and retained by the meshes in use since its fishing rate was closely comparable to that of older individuals.
The validity of the conclusions obtained by this method is limited in fact by a number of problems particularly because of the idea of the pseudo-cohort which implies that the fishery is in equilibrium whereas the recruitment and exploitation rate are variable according to years.) The Working Group therefore recommended that the analysis be conducted according to the traditional cohort analysis method using the catch per age vectors of all the fisheries as soon as they can be obtained.
It can be said at this time that the per fish mortality according to age scheme obtained in the present study on the Dakhla stock is very different from that obtained for the Japanese Cape Blanc fishery from 1969 to 1971 by the preceding Working Group in October 1980; in fact the difference in the calculation methods makes a quantitative comparison of these results difficult. The pattern of the F vectors according to age for these two stocks is shown in Figure 14. Recruitment would appear to be late in the Cape Blanc fishery and the exploitation rate would show marked seasonal variations (resulting in variations according to age). On the other hand at Dakhla recruitment seems to be early and the exploitation rate much more stable. The marked difference in the rates of the F vectors probably mainly accounts for the differences in the conclusion of the yield per recruit analyses which used these data.
The Group attempted to make a preliminary analysis of the data on the biological cycle and recruitment.
The biological cycle of the Cape Blanc octopus stock was reconstructed (Figure 15) based on the available data in Hatanaka's study (1979) on spawning seasons, recruitment and growth. The cohort spawned in springtime (May-June) would be recruited in November (or pass to the adult stage) and would participate massively in the springtime spawning of the following year. The cohort spawned in autumn (September-October) would be recruited in May (consequently its slower growth could be explained by winter temperatures) only for the first spawning, but would not spawn massively until the following autumn (the size interval in which maturity takes place was indicated). From this we can deduce the existence of a main spawning season in May-June (the two new cohorts spawn) and of a secondary season in September-October (only one of the cohorts spawns). The main springtime spawning would produce the main autumn recruitment.
5.3.2.1 Dakhla
The Spanish cpue (commercial category No. 8) from 1977 to 1980 in spring (April-May) and autumn (September-November) made it possible to calculate the recruitment indexes whose evolution is shown in Figure 16.
5.3.2.2 Cape Blanc
The seasonal variations of cpue given by Hatanaka (1979) indicate a sharp increase of yields in August. If we consider that this increase reflects recruitment, the year-to-year evolution of the cpue in August would indicate evolution of the spring recruitment.
Figure 17 shows that recruitment at Cape Blanc would have decreased significantly from 1967 to 1975.
Figure 18 gives an entirely new approach to this problem of cpue in August; in the figure cpue for n + 1 year (recruitment index) is related to average annual cpue for the preceding year (parent stock index).
There is a relationship between these two data but the Group did not have time to discuss its validity.
The Group stressed that these results were given only as an indication and recommended that a more thorough analysis be performed on this point before final conclusions were to be drawn.
