7. STOCK ASSESSMENT

7.1 Catch and effort analysis
7.2 Biomass estimates by acoustic surveys
7.3 Cohort analysis

7.1 Catch and effort analysis

Data on catch by fishing zone are given in Table 11. Data on fishing effort are available for the Moroccan fleet fishing in Zone A, the Spanish fleet fishing in Zone B, and the Polish, Soviet and Bermuda fleets fishing in Zone C.

The group decided to attempt to make a catch and effort analysis for the fishery in each zone separately, which analysis would provide information on the state of exploitation and the potential yield by zone if each zone contains its independent sardine stocks, and for various combinations of zones for which the Group felt that the sardine in these zones might belong to the same stock.

7.1.1 Zone A

Fishing in this zone was nearly exclusively by the Moroccan fleet. It should be noted that the catches included for this area are different from those in Table 3 of the Report of the Third Session of the CECAF Working Party on Resource Evaluation, due to the exclusion of catches made in the area of Casablanca and further north, because of the indications that the fish in that area may belong to a separate stock. The effort has been calculated for the fleets of Safi, Essaouira and Agadir, and is expressed as the sum of the products of the number of trips and the gross tonnage of all vessels in the fleet.

Table 12 gives the Moroccan catch and effort data for this zone. Data on the Soviet catches for 1978 have not been included, in the absence of detailed statistics on the distribution of these catches between Zones A and B. As it is understood that a large percentage of the Soviet catches has been obtained from Zone B, for the purpose of the present analysis they have been completely included in the statistics of the latter zone.

Table 6

NOMINAL CATCHES (t) OF SARDINA PILCHARDUS IN THE NORTHERN SECTOR OF THE CECAF AREA, DIVISIONS 34.1.1; 34.1.2; 34.1.3 and 34.3.1

Country	1965	1966	1967	1968	1969	1970	1971	1972	1973	1974	1975	1976	1977	1978
Bulgaria								1 215	1 862	4 417	9 764	13 845	30 105	0
France	5 531	4 400	11 589	8 313	10 352	11 583	11 300	4 574	5 310	4 486	6 910	2 080	1 498	0
Greece													1 000	1 000
Morocco	150 354	244 627	203 496	162 566	168 373	163 028	173 296	194 823	334 309	209 144	157 167	209 423	91 375	94 737
Norway										11 469	182	0	0	0
Mauritania													250	300
Poland						6	315	3 120	15 293	23 685	76 720	106 119	136 804	2 633
Portugal								4	2				2 000	2 000
Romania										21 275	34 939	17 641	26 771	0
Spain1	25 000	25 000	30 000	30 000	30 000	30 000	30 000	20 000	10 000	5 000	-	-	-	-
Spain2	11 569	13 979	19 530	26 532	28 146	41 427	46 711	44 520	63 500	79 899	65 740	114 950	131 9205	138 4435
German Dem. Rep.										1 561	1 913	7 410	13 159
Bermuda									1 7343 15 7884	25 795	182	270	47 279	47 0004
USSR					80 100	88 700	123 600	136 000	204 800	342 200	357 200	350 000	396 857	142 9006
Total	192 454	293 006	264 615	227 411	316 971	334 744	385 222	404 256	636 810 650 864	728 931	710 717	821 743	839 826	429 0136

¹ Estimated catches made by boats based in continental Spain harbours (Cadiz, Huelva, Algeciras, Barbate) operating in the northern part of Division 34.1.1

² Catches made by Spanish boats fishing in Zone B

³ Minimum estimation

⁴ Maximum estimation

⁵ Estimated

⁶ Data-incomplete

Table 7

NOMINAL CATCHES (t) OF SARDINA PILCHARDUS IN THE NORTHERN SECTOR OF CECAF, DIVISION 34.1.1

Country	1965	1966	1967	1968	1969	1970	1971	1972	1973	1974	1975	1976	1977	1978
Bulgaria												716
France	5 531	4 400	11 589	8 313	10 352	11 583	11 300	4 574	5 310	4 486	6 910	2 080	1 498
Morocco	150 354	244 627	203 490	162 566	168 373	163 028	173 296	194 823	334 309	209 144	157 167	209 428	91 375	94 737
Poland								52	530	3 923	7 096	5 689
Spain1	25 000	25 000	30 000	30 000	30 000	30 000	30 000	20 000	10 000	5 000
Spain2	9 069	16 479	17 030	24 032	25 646	38 927	44 211	42 020	60 500	76 000	63 500	112 450	129 420	135 943
USSR													1 0843	32 0004
Total	189 954	290 505	262 115	224 911	234 371	243 538	258 807	261 469	410 649	298 553	234 673	330 363	223 377	262 680

¹ Estimated catches made by boats based in Cadiz, Huelva, Algeciras, Barbate operating in the northern part of Division 34.1.1

² Catches of Spanish vessels in Zone B. Catches of 1977 and 1978 estimated

³ Estimate by Soviet scientists (3 months in 1977)

⁴ Estimate by Soviet scientists

Table 8

NOMINAL CATCHES (t) OF SARDINA PILCHARDUS IN NORTHERN SECTOR OF CECAF, DIVISION 34.1.2

Countries	1965	1966	1967	1968	1969	1970	1971	1972	1973	1974	1975	1976	1977	1978
Bulgaria
France
Greece
Morocco
Norway
Poland.
Portugal								4	2
Romania
Spain1
Spain2	2 300	2 500	2 500	2 500	2 500	2 500	2 500	2 500	3 000	3 899	2 240	2 500	2 500	2 500
German Dem. Rep.
Bermuda
USSR													38 409
Total	2 300	2 500	2 500	2 500	2 500	2 500	2 500	2 504	3 002	3 899	2 240	2 500	40 909	2 500

¹ Continental
² Canary Islands catches estimated in the periods 1965-73 and 1976-78

Table 9

NOMINAL CATCHES (t) OF SARDINA PILCHARDUS IN NORTHERN SECTOR OF CECAF, DIVISION 34.1.3

Country	1965	1966	1967	1968	1969	1970	1971	1972	1973	1974	1975	1976	1977	1978
Bulgaria								1 215	1 862	4 417	9 764	13 129	30 105
France
Greece													1 000	1 000
Morocco
Norway										10 906	182	0	0	0
Poland						6	315	3 068	14 763	19 762	69 624	100 430	136 804	2 633
Portugal													2 000	2 000
Mauritania													250	300
Romania										21 275	34 939	17 641	26 771
Spain1
Spain2
German Dem. Rep.										1 561	1 913	7 410	13 159	-
Bermuda									1 7343 15 7884	25 795	182	270	47 279	47 0005
USSR					80 100	88 700	123 600	136 000	204 800	342 200	357 200	350 000	316 654	110 900
Total					80 100	88 706	123 916	140 283	223 159 237 213	425 916	473 804	488 880	574 022	163 833

¹ Continental
² Canary Islands
³ Minimum estimate
⁴ Maximum estimate
⁵ Estimate
- No data available

Table 10

SOVIET CATCHES OF S. PILCHARDUS IN DIVISION 34.3.1 DURING 1977

CECAF Division	Catches (t)
34.3.1	40 710

CATCHES (in t) OF S. PILCHARDUS IN CECAF AREA (DIVISIONS 34.1.1, 34.1.2, 34.1.3) BY FISHING ZONE

Year	Zone A1	Zone B2	Zone C3	Others4	Total
1965	148 496	14 600		29 385	192 454
1966	237 657	20 879		34 370	293 006
1967	196 597	28 619		39 399	264 615
1968	161 683	32 345		33 383	227 411
1969	164 101	35 999	80 100	36 772	316 972
1970	161 443	50 510	38 706	34 085	334 744
1971	165 998	55 511	123 915	39 798	385 222
1972	173 514	46 646	140 283	43 809	404 252
1973	325 193	66 340	223 1595 237 2135	22 116	636 8105 650 8645
1974	204 181	84 409	425 916	13 861	728 367
1975	147 588	77 506	473 804	11 819	710 717
1976	176 564	120 219	488 880	36 080	821 743
1977	84 410	132 002	574 872	9 465	799 879
1978	91 399	167 943	163 833	5 838	429 013

¹ Total catch of Moroccan ports (Safi, Essaouira, Agadir)

² Spanish fleet from Canary Islands, French and Polish catches in 34.1.1 + USSR (1977-78)

³ Bulgaria, Norway, Poland, Romania, German Democratic Republic, Bermuda, USSR (until 1977), Greece, Mauritania, Portugal

⁴ Spanish fleet, Moroccan (northern part of 34.1.1), Canary fleet fishing in 34.1.2, and Soviet catches in 34.1.1 except 1977 and 1978

⁵ Two figures due to minimum and maximum estimates of Bermuda catches

Table 12

CATCH, EFFORT AND CATCH PER UNIT OF EFFORT IN ZONE A (SEE TEXT)

Year	Total catch ('000 t)	cpue	Total effort	Effort (average of 2 years)
1962	120	236	508	-
1963	119	123	963	735.5
1964	128	122	1 045	1 004
1965	148	161	922	983.5
1966	238	232	1 023	972.5
1967	197	213	921	972
1968	162	177	913	917
1969	164	187	875	894
1970	161	151	1 071	973
1071	166	125	1 329	1 200
1072	174	111	1 558	1 443.5
1973	325	171	1 899	1 728.5
1974	204	114	1 797	1 848
1975	148	84	1 761	1 779
1976	177	92	1 929	1 845
1977	84	45	1 867	1 898
1978	91	48	1 900	1 883.6

The fishing effort has remained fairly constant in the period 1963-1970, then increased rapidly to reach nearly the double of the previous level in 1973, after which it remained again fairly constant with limited fluctuations. The catch per unit of effort, after having fluctuated until 1966, has shown a nearly continuous decline thereafter, reaching in 1977/78 a level of less than one quarter of the 1966/67 level.

The continuous decline in the catch per unit of effort since 1973, in a period when the effort estimate remained constant, suggests that the stock has been affected not only by fishing in Zone A, but also by other factors which could be either fishing in other zones, in which case Zone A does not contain a separate stock, or changes in environmental factors affecting the distribution or abundance.

7.1.2 Zone B

At the first session of the Group, the Spanish scientists were asked to collect information on the size of the fishing vessels in order to correct the estimate of effort in days fishing for increases in vessel fishing capacity.

It was now reported (Appendix 5) that a factor which takes account of both gross tonnage and hold capacity gives the best correlation between vessel characteristics and catch of vessels fishing in the same period. Data were therefore provided on total catch in Zone B, catch per unit of effort expressed as:

of the Spanish fleet, and the estimated fishing effort expressed in these units (Table 13).

Table 13

TOTAL CATCH (t) EFFORT AND CATCH PER UNIT OF EFFORT IN FISHING ZONE B (SEE TEXT)

Year	Catch (t)	Effort (units; see text)	cpue	Average effort (2 years)
1965	14 600	1 849	7 896	-
1966	20 879	1 958	10 663	1 903
1967	28 619	3 092	9 256	2 525
1968	32 345	3 290	9 830	3 191
1969	35 999	3 979	9 047	3 635
1970	50 510	13 813	3 658	8 896
1971	55 511	13 170	4 212	13 492
1972	46 647	5 696	8 190	9 433
1973	66 340	8 529	7 779	7 113
1974	84 409	7 120	11 855	7 825
1975	77 506	14 921	5 195	11 020
1976	120 219	20 390	5 896	17 656
1977	132 002	23 862	5 532	22 126
1978	167 943	46 088	3 644	34 975

The data now show a continuous increase in estimated fishing effort, and a decline in catch per unit of effort from 1965 till the early seventies from about 8 000 to 4 000 whereas thereafter, the fishing effort continued to increase; the catch per unit of effort did not decrease further, but rather showed signs of an increase to slightly above 5 000. This coincided with the unusual decline in the Moroccan catch per unit of effort in the same period. This would appear to strengthen one of the alternative conclusions reached in Section 7.1.2, that part of the sardine stock previously fished by Morocco may in the early seventies have changed its distribution or migration and now is caught more in Zone B. This suggests that, Zones A and B, at least partly, contain the same stock which in recent years changed the magnitude of its seasonal northward displacement. The alternative hypothesis, which looks less likely on the basis of evidence available, would be that the environmental conditions have changed in such a way to cause a decline in stock A and a simultaneous increase in an independent stock in Zone B.

7.1.3 Zone C

Various series of catch per unit of effort data in Zone C were available to the meeting, from Poland, USSR and Bermuda. The USSR data showed large, irregular variations in the figures. This was related with the introduction of pelagic trawls during the period for which data were available and with the year-to-year changes in fishing strategy of the USSR fleet. These data could, therefore, not be used for catch/effort analysis. Similar observations would probably apply to the greatly varying catch per unit of effort data of the Bermuda fleet. Polish data on the catch per hour fishing of the B 18 trawlers gave a much more consistent picture (Table 14).

Table 14

CATCHES PER UNIT OF EFFORT OF THE POLISH, USSR AND BERMUDA FLEETS IN ZONE C

Year	Poland	USSR		BERMUDA
	Catch per hour fishing of B 18 trawlers	Catch/day BMRT vessels	Catch/day RTMA vessels	Catch/day
1972		1.1	6.1
1973	7.8	3.6	5.9	10.2 (or 1.1)
1974	7.5	3.1	25.9	25.6
1975	6.2	10.1	21.5	0.3
1976	5.4	26.5	27.0	0.4
1977	5.5	12.1	18.1	42.5

Table 15 shows an increase in total catch from 1973 to 1976, after which the catch remained about stable until 1977, then probably falling greatly in 1978 with the departure of the Polish fleet. The estimated total effort increased until 1976 and decreased somewhat in 1977, and the catch per unit of effort showed a decline until 1976. If Zone C can be taken to contain a separate stock of sardine, Figure 2 shows the corresponding relationships between effort, catch per unit of effort and total catch.

The figure indicates that the total catch in 1975-77 was close to the maximum potential yield of this stock of about 520 000 t, although the effort was still some 30 percent below that giving the maximum yield. It should be noted however that this analysis has been based on data for six years only, in three of which the effort was nearly the same, and may therefore not be very reliable.

Table 15

TOTAL CATCH IN 1 000 t, CATCH PER UNIT EFFORT (CATCH IN t/h FISHING OF POLISH B 18 TRAWLERS) AND ESTIMATED TOTAL EFFORT (EXPRESSED AS NUMBER OF HOURS FISHING OF POLISH B 18 TRAWLERS REQUIRED TO OBTAIN THE TOTAL CATCH) IN ZONE C

Year	Total catch ('000 t)	Catch per unit of effort	Estimated total/effort	Effort averaged over two years
1973	223.2	7.8	28 615
1974	425.9	7.5	56 787	42 701
1975	473.8	6.2	76 419	66 603
1976	488.9	5.4	90 537	83 478
1977	574.9	5.5	104 522	97 530
19781	163.8	5.4	30 399	67 431

¹ incomplete data

7.1.4 Combination of fishing zones

The conclusion reached above on the likelihood of a shift of the population of the fish from Zone A southward and hence a change in the distribution or migration pattern of the sardine, makes it invalid to relate the catch per unit of effort in Zones A and B with the fishing effort in these zones separately, because the fish in Zone B would at least partly belong to the same stock as the fish previously caught in Zone A. For a proper analysis a catch/effort study would be needed for Zones A and B together. Such analysis, however, would require estimates for the successive years of an index which is proportional to the fish abundance in the two zones together and neither the catch per unit effort in Zone A nor that in Zone B fulfils that requirement. The Group considered possible ways to combine the catch per unit effort data in the two zones, but concluded that the basic data for calculating a meaningful combined estimate, such as information on the relative fishing power of an average boat fishing in Zone A with respect to that of an average boat fishing in Zone B, were lacking and that hence no catch/effort analysis for Zones A and B together could be carried out.

For the same reason, it was not possible to carry out a catch/effort analysis for Zones A, B and C together, which would be necessary if the sardine in all three zones belonged to the same stock.

7.2 Biomass estimates by acoustic surveys

An acoustic survey carried out in November 1974 between Point Stafford (27°40' N) and Cape Cantin (32°35' N) gave an estimate of biomass in that area of about 1.4 million t (Johannesson, Villegas and Lamboeuf, 1975). Lamboeuf (1977) estimated the amount of pelagic fish other than sardine which may have been included in the 1974 survey results and arrived at an estimate for sardine, 1.1 million t. A second survey carried out in January 1977 (Lamboeuf, 1977) gave an estimate of the biomass between Cape Cantin and Oued Drâa of 325 000 t and between Oued Drâa and 27°N of 715 000 t, total 1.04 million t in an area slightly larger than that covered in 1974. Finally, a survey carried out by USSR in cooperation with Morocco in, the area between 35°40' N and 28° N (see Appendix 8) in December 1978-January 1979 gave an estimate for the sardine biomass of 1.53 million t, substantially larger than the previous estimates. Important concentrations were found all along the coast between Casablanca and Cape Juby. The total biomass of all species was according to provisional calculations estimated at 2.3 million t, but the analysis is not yet completed.

The calibration of the echointegrator readings was during the first two surveys based on measurements with live fish in a cage, whereas during the Soviet survey the calibration was based on measurements with one dead fish fixed in the beam of the echosounder.

The Group considered that the differences in the results of the surveys cannot be taken as an indication of an increase in the stocks in this region, because they may be due to the differences in the methods used during the surveys.

7.3 Cohort analysis

7.3.1 Successive attempts

As mentioned in the report of the first session, the availability of data on age composition provided an important improvement of the data base for population studies on the sardine resource, as compared with the 1976 CECAF Working Party meeting. Uniformity of and agreement on the age determination of the sardine had been obtained between the scientists concerned by previous exchange of age reading attended by scientists from Poland, Spain and Morocco, held in Tenerife in 1978 (COPACE/Tech/78/8).

The data available to the meeting in Casablanca included:

(a) catch of S. pilchardus by the Moroccan fleet in Zone A, expressed in numbers of each age group by three-monthly periods, from the third quarter of 1973 up to the last quarter of 1978

(b) catch of S. pilchardus by the Spanish fleet in Zone B, expressed in numbers of each age group by three-monthly periods, from the third quarter of 1975 up to the last quarter of 1978

(c) total annual catch by age group in numbers in Zone C, based on Polish age-composition data, for the years 1972-1977. In addition, Soviet scientists presented to the meeting in Dakar their total annual catches in numbers by age groups for the Zone C and the years 1972-1978.

Zones A and B together
Zone C separately
Zones A, B and C together

Because, as it is said in the introduction, the Group was unable to achieve the computations, it recommended that the Moroccan scientists continue and complete the cohort analysis initiated during the Casablanca meeting along the lines mutually agreed at that time. The numbers of the Ad hoc Group who attended the Fourth Session of the CECAF Working Party on Resource Evaluation met in Dakar to review the computations done by the Moroccan scientists.

The Moroccan research workers had prepared for the Dakar meeting separate cohort analyses for Zones A, B and C. These had to be recalculated, however, because corrections had been introduced into the Soviet catch statistics and the demographic composition of the Polish catches. The age composition of the Soviet catches in 1977 and 1978 was also submitted.

In this connection Mr. Overko stated that the USSR had been collecting catch statistics broken down by the species and divisions recommended in the CECAF statistical system only since 1978. For the preceding years statistics were available only for larger species groups and CECAF sectors. The Soviet research workers had endeavoured to break them down by species (for example, sardine and sardinella) and in accordance with the CECAF statistical divisions on the basis of their knowledge of the probable composition of the catches, etc. This explained why corrections had had to be made in the statistics submitted by the Soviet research workers at the previous meetings. In future the corrections should be minimum.

The cohort analyses were therefore recalculated in conformity with the recommendations adopted at the Casablanca meeting. Generally speaking, the analyses remain fairly imprecise, owing to the small number of age classes (from 5 to 6) which does not make it possible for the estimates, particularly of F, to converge and become stabilized sufficiently early during the life of the cohort. Besides, the short duration of the observations (1966-1978 for fishery in Zones A + B. 1972-1977 for that in Zone C, and hence 1976-1977 the only years in which data are available for the whole sector) has limited analysis of the effects on stock and catches of changes in the rate of exploitation, and of variations in recruitment. Finally, the uncertainty which subsists regarding annual catches, and the divergences that exist in the age determination, directly affect the confidence that can be placed on the results of the cohort analyses.

7.3.2 Zones A + B

The Spanish and Moroccan data on the demographic composition of the annual catches in the two Zones A and B were regrouped. Since the limited number of years (3) for which the necessary data were available made it impossible to follow the evolution over time of the dynamic parameters, these data were combined by taking the mean of the catches, per age, made in 1976, 1977 and 1978. This gave a standard cohort for which the characteristics and behaviour were studied for the following combinations:

0.1 < F < 0.6 and M = 0.4, 0.6 and 0.8

The estimates of biomasses and corresponding fishing mortality are given in Table 16. These results are also shown in Figure 3 (for natural mortality M = 0.4 and 0.6).

Comparison of the F vectors shows that the combinations M = 0.4/terminal F = 0.3 and, to a lesser extent, M = 0.6 and terminal F = 0.2 appear to be the most probable. With M = 0.8 almost all the possible terminal F values seem excessive. For the two combinations mentioned above, the F vector is more or less stabilized between the ages of 2 and 5 years; all the other F vectors converge towards the corresponding F values.

Comparison of the biomass estimates obtained from these cohort analyses with the evaluations made by acoustic surveys in January 1977 and January 1979 also provides indications of the most probable values of the two mortality coefficients during the period 1976/78. According to the surveys, the biomass for the whole of the two Zones A + B consisted of between 1 million and 1.5 million t. The cohort analysis provides estimates of this order for combinations:

M = 0.4 and 0.1 < F < 0.2 on the one hand and
M = 0.6 and 0.4 < F < 0.6 on the other (Table 16).

Another cohort analysis used the Spanish and Moroccan catch statistics, by age and quarter (from the third quarter of 1975 to the last quarter of 1978). The calculations were made by the Pope method. All the combinations of M and terminal F explored show a decline in the biomass between 1976 and 1978 (Table 17, Fig. 4). For the most probable available estimates (M = 0.4; F = 0.4), the decline is more than 20 percent. The cpue relating to the Spanish and Moroccan, fleets show an even greater decline in the biomass (Table 18).

Finally, the Group studied the level of stock exploitation in Zones A + B in terms of production per recruit (i.e., assuming independent recruitment of the level of exploitation) and on the assumption, obviously, that the concentrations fished in Zones A and B constitute a distinct stock. Figure 5 shows the production curve per recruit corresponding to the Reverton and Holt model. The values used for the parameters correspond to the estimates relating to the Moroccan fishery; they are indicated in Figure 5. On the basis of the most probable F vector given in Figure 3, it was assumed that F equalled 0.35 and that the average age at recruitment was one year. According to this analysis and insofar as the initial hypotheses (distinct stock, stability of recruitment over time and in terms of the biomass) prove to be correct it can be concluded that at present the stock is moderately exploited. A certain increase in production could therefore be expected from an intensification of fishing. However, the maximum increase possible is probably considerably lower than that shown in Figure 5, since if the rate of exploitation reaches high levels a recruitment deficit probably occurs.

Table 16

STANDARD COHORT (1976 + 1977 + 1978) DEDUCED FROM THE CATCHES IN ZONES A + B: F ESTIMATES CORRESPONDING TO DIFFERENT COMBINATIONS OF M AND TERMINAL F; CORRESPONDING BIOMASSES

Initial assumptions		Ages						Biomasses 103 t
M	Fterm	0	I	II	III	IV	V
0.4	0.1	9.04	0.10	0.23	0.20	0.20	0.1	1 346
	0.2	0.05	0.12	0.30	0.29	0.31	0.2	1 024
	0.3	0.06	0.13	0.34	0.33	0.38	0.3	916
	0.4	0.06	0.14	0.36	0.36	0.43	0.4	863
	0.5	0.06	0.15	0.38	0.38	0.47	0.5	830
	0.6	0.06	0.15	0.39	0.40	0.50	0.6	826
0.6	0.1	0.02	0.05	0.12	0.12	0.13	0.1	2 739
	0.2	0.02	0.07	0.18	0.18	0.22	0.2	1 897
	0.3	0.03	0.08	0.21	0.22	0.28	0.3	1 616
	0.4	0.03	0.08	0.23	0.25	0.33	0.4	1 476
	0.5	0.03	0.09	0.25	0.27	0.37	0.5	1 391
	0.6	0.03	0.09	0.26	0.29	0.40	0.6	1 336
0.8	0.1	0.01	0.02	0.07	0.07	0.09	0.1	5 935
	0.2	0.01	0.03	0.10	0.12	0.16	0.2	3 818
	0.3	0.01	0.04	0.13	0.15	0.22	0.3	2 871
	0.4	0.01	0.04	0.14	0.18	0.26	0.4	2 752
	0.5	0.01	0.04	0.16	0.19	0.29	0.5	2 512
	0.6	0.02	0.05	0.17	0.21	0.32	0.6	2 407

ZONES A + B: ESTIMATED BIOMASSES (IN '000 t) CORRESPONDING TO VARIOUS POSSIBLE VALUES OF M AND F_TERM (COHORT ANALYSES, POPE METHOD, THREE-MONTHLY PERIOD)

Quarterly mortality rate	1975 3rd quarter	1976 1st quarter	1977 1st quarter	1978 1st quarter
M = 0.15 Fterm = 0.10	3.147	3.968	3.033	2.804
M = 0.15 Fterm = 0.15	2.462	3.083	2.330	2.058
M = 0.15 Fterm = 0.20	2.095	2.601	1.933	1.775
M = 0.10 Fterm = 0.10	1.601	2.057	1.673	1.597
M = 0.10 Fterm = 0.15	1.300	1.651	1.312	1.203
M = 0.10 Fterm = 0.20	1.143	1.440	1.124	1.022

EVOLUTION OF THE SARDINE STOCK BIOMASS IN ZONES A + B AS IT EMERGES FROM THE COHORT ANALYSIS AND FROM COMPARISON OF THE CATCHES PER UNIT OF EFFORT
(IN PERCENTAGE OF THE 1976 VALUE)

Biomass obtained by:

Year	1. Cohort analysis (M = 0.4; F = 0.4)	2. Spanish cpue	3. Moroccan cpue (catch/day × GRT)
1976	100%	100%	100%
1977	81%	94%	49%
1978	78%	62%	52%

Fig. 3 Standard cohort (1976 + 1977 + 1978 catches) deduced from the catches in Zones A + B: F vectors as a function of age for different assumptions of terminal F and M (M = .4)

Fig. 3 Standard cohort (1976 + 1977 + 1978 catches) deduced from the catches in Zones A + B: F vectors as a function of age for different assumptions of terminal F and M (M = .6)

Fig. 3 Standard cohort (1976 + 1977 + 1978 catches) deduced from the catches in Zones A + B: F vectors as a function of age for different assumptions of terminal F and M (M = .8)

Fig. 4 Zones A + B: Evolution of the biomass (in millions of t), as deduced by cohort analysis for various assumptions of M and terminal F

Fig. 5 Application of Beverton and Holt model to sardine concentrations in Zones A + B: production curves per recruit and present level of exploitation

When one varies the average age of first catch (t_c), one notes that although heavy fishing of individuals less than one year old should be avoided, it does not seem that any great benefit is to be expected from postponing the start of exploitation to a later age.

7.3.3 Zone C

For this zone the annual catches by age group are available for the Soviet and Polish fleets. The total annual catches were estimated separately by age, on the basis of these two series of data. The Polish data led, on the whole, to more homogeneous results than the Soviet data. The Soviet research workers put forward as a possible explanation the fact that their fishery was more versatile as regards the target species. It was decided to carry out the analyses without combining the Soviet and Polish data since, owing to the size of the Soviet catches (from 60 to 90 percent of the total), the Soviet data would have dominated the Polish ones. The following analysis therefore relates only to the Polish data, extrapolated to the total catches (Table 19). The group considered it desirable, however, that calculations be carried out in the near future using the verified Soviet data and according to the procedure followed in processing the Polish data.

Table 19

SARDINE IN ZONE C: TOTAL ANNUAL CATCH BY AGE GROUP (IN MILLIONS), AS DEDUCED FROM SAMPLING AND FROM THE POLISH AGE DETERMINATIONS EXTRAPOLATED TO ALL OF THE CATCHES

Year	I	II	III	IV	V	VI
1972	194.9	977.5	805.2	208.0	19.2	-
1973	230.6	1 394.8	1 260.9	448.4	36.6	3.4
1974	470.4	1 810.5	3 261.2	604.0	73.7	-
1975	388.3	699.0	2 962.1	1 583.3	349.4	73.1
1976	87.1	977.4	2 693.4	1 959.1	356.0	30.2
1977	429.5	1 392.6	3 272.2	1 806.8	433.3	150.5
	28.1	51.5	78.2	93.1	109.9	125.7

On the basis of the results obtained for Zones A + B, the value M = 0.4 was adopted in processing the data for Zone C.

In order to study the possible value of terminal F, two standard cohorts were considered separately: one consisting of the mean of the numbers caught in 1975/76/77, a period corresponding to a well established and, according to the production model data (Table 16), relatively stabilized fishery; and the other of the mean of the numbers caught in 1972/73/74, as corresponding to a period of development of the fishery, concomitant with the appearance of large recruitments to the population (Table 21). As for the cohort analysis relating to Zones A + B, an attempt was made to determine the most probable F terminal by analysing the deformations of the F vector age for various values of terminal F selected at random.

The crude results, given in Table 20 and Figures 6a and 6b, give rise to three observations:

- F apparently increases constantly with age, i.e., without reaching any obvious plateau; however, it should be noted that the major uncertainty refers to the older age only and that the F estimates rapidly converge for young ages;

- Each cohort starts to be effectively exploited later (1-2 years) than in Zones A + B; the Soviet data (Figure 7) agree with these first two observations;

- Finally, the terminal F values shown as the most probable by this analysis are extremely high (3.0 < F < 3.5 for ages 5 and 6 years) for the 1972/73/74 "cohort" corresponding to the period when the fishery was starting.

Table 20

SARDINE IN ZONE C: F VECTORS AS A FUNCTION OF AGE FOR TWO STANDARD COHORTS (1972/74 AND 1975/77) AND VARIOUS POSSIBLE VALUES OF TERMINAL F; CORRESPONDING BIOMASSES (COHORT ANALYSIS BASED ON THE POLISH DATA EXTRAPOLATED TO THE TOTAL CATCHES)

a) standard cohort type 1972/73/74

	Fterm	Ages						Biomasses ('000 t)
		0	I	II	III	IV	V
M = 0.4	1	0.04	0.32	1.24	1.98	3.18	1.0	896
	1.5	0.04	0.32	1.24	1.98	3.27	1.5	896
	2	0.04	0.32	1.24	1.99	3.33	2.0	896
	2.5	0.04	0.32	1.24	1.99	3.36	2.5	896
	3	0.04	0.32	1.24	1.99	3.38	3.0	896
	3.5	0.04	0.32	1.24	1.99	3.40	3.5	896
	4	0.04	0.32	1.24	1.99	3.41	4.0	896

b) standard cohort type 1975/76/77

	Fterm	Ages						Biomasses ('000 t)
		0	I	II	III	IV	V
M = 0.4	0.3	0.02	0.09	0.55	1.08	0.94	0.3	2 407
	0.5	0.02	0.10	0.57	1.14	1.10	0.5	2 351
	0.7	0.02	0.10	0.57	1.27	1.20	0.7	2 327
	0.9	0.02	0.10	0.58	1.29	1.26	0.9	2 314
	1.1	0.02	0.10	0.58	1.30	1.30	1.1	2 306
	1.3	0.02	0.10	0.58	1.21	1.33	1.3	2 300
	1.5	0.02	0.10	0.58	1.21	1.35	1.5	2 296

Fig. 6 Sardine in Zone C (Polish data): F vectors as a function of age for M = 0.4 and various assumptions of terminal F

(a) - standard cohort 1972/73/74

(b) - standard cohort 1975/76/77

Table 21

SARDINE IN ZONE C: POPULATION NUMBERS (MILLION t), FISHING MORTALITY RATE AND ANNUAL BIOMASSES ('000 t) (COHORT ANALYSIS, POLISH DATA EXTRAPOLATED TO THE TOTAL CATCHES)

		Ages						Biomasses ('000 t)
		I	II	III	IV	V	VI
1972	a	21 701.4	6 749.6	2 324.5	322.2	29.1	19.5	1 175
	b	0.01	0.20	0.55	1.54	1.64	0
1973	a	22 372.5	14 387.3	3 724.1	898.9	46.4	3.8	1 750
	b	0.01	0.13	0.63	0.94	3.33	3.5
1974	a	17 168.0	14 807.9	8 502.1	1 464.0	235.4	1.1	2 072
	b	0.03	0.16	0.63	0.70	0.48	0
1975	a	14 833.9	11 122.9	8 447.7	3 029.1	486.8	97.5	1 998
	b	0.03	0.08	0.56	1.02	2.05	1.2
1976	a	10 486.6	9 625.6	6 883.8	3 234.8	734.2	40.3	1 716
	b	0.01	0.15	0.65	1.35	0.90	1.2
1977	a	26 299.5	6 963.0	5 652.0	2 409.1	564.4	200.7	1 852
	b	0.02	0.10	0.55	1.2	1.2	1.2

a: population numbers in millions
b: fishing mortality rate

The table of F vectors and population numbers (Table 21) shows that the fishery has developed since the early 1970s, i.e., since the appearance of the first heavy recruitments. This observation is not surprising, since there were already sizable fishing fleets on the spot, engaged in exploiting other coastal pelagic stocks which had started to be fished several years before. Although the cohort analysis is likely to accentuate this phenomenon, it is to be noted also that the very high terminal F values apply only to the weak cohorts born before high recruitment occurred. If these F values are correct, it implies that the older age groups are more vulnerable the smaller their numbers, i.e., that fishing mortality is inversely dependent also on biomass. Examples of such a relationship between q and B are sufficiently numerous among coastal pelagic species to prevent a priori rejection of such a possibility.

Finally it should be noted that the analysis of the cohort formed by taking the mean of the numbers per age/groups captured in 1972, 1973 and 1974 contains a bias resulting from the fact that recruitment varied greatly during the life of this fictitious cohort.

In the absence of information making it possible to choose the most probable terminal F values, the final cohort analysis was therefore carried out taking 1.2 as the value of terminal F for the years 1975 to 1977, on the basis of these observations. For 1972 to 1974, the period for which there is the greatest uncertainty, the calculations were repeated by taking first a terminal F value of 3.5 for the three years, then the following low values in principle more acceptable: 0.01 in 1974 (total catch nil), 1.0 in 1973 (i.e., a value comparable to that adopted for the period 1975-77) and 0.01 for 1972 (total catch nil). These new values of terminal F do not, in fact, significantly affect the annual biomass estimates (the differences concern thousands of tons). This is because the modification of terminal F for the years 1972 to 1974 affects only the biomass of the very last age groups (IV to VI), for which recruitment is low and numbers very reduced. For the terminal F values of the cohorts fished in 1977, the F vector values shown in Figure 6 were adopted (terminal F = 1.2).

Estimates of fishing mortality and population numbers by age and year as well as of the corresponding annual biomasses obtained by cohort analysis, are given in Table 21. Recruitment started to increase in 1971 (age 0). Although the 1975 class was slightly lower, the sizable catches of one-year-olds in 1977 suggest that recruitment in 1976 was high. This observation needs, however, to be confirmed. To the extent that recruitment has been exceptional only since 1972, according to the series of observations available, it is important to monitor the future developments in recruitment. There are no grounds for affirming, in fact, that recruitment may not return one day to the level of the years preceding 1972. It is therefore important that measures be taken very quickly to bring catches back to a level compatible with the size of the stock.

Between 1972 and 1977 the biomass, as shown by these data, after reaching a maximum of about 2 million t, remained high (1.85 million t in 1977 - Figure 8). Comparison of these biomass estimates - bearing in mind that they represent total biomasses, not exploited biomasses - with present catches (575 000 in 1977) makes it possible to assume that the stock must now be approaching full exploitation. It must be emphasized, however, that this conclusion is only tentative, in the sense that it is reliable only to the extent to which the cohort analysis is reliable. In addition, it is valid only if the initial assumptions are correct, i.e., the existence of a distinct stock in Zone C - which has not yet been proved - and the continuation of the high recruitment observed during the last years.

In order to obtain a better idea of the present rate of exploitation of "stock C", the production per recruit was calculated, as for Zones A + B, taking as the value of F the mean value, weighted proportionally to catches, of age groups III, IV and V fished in 1975 and 1976. This evaluation can only be very approximate, insofar as the F vector is, as has been said, far from being stabilized after recruitment. Several values of t_c were used in order to take account of differences in the demographic composition of the Polish and Soviet catches (Figure 9). Despite its limitations this treatment leads to conclusions comparable to those drawn from the application of a production model to the Polish catches per unit of effort (Figure 3). These conclusions are decidedly more pessimistic as regards the state of the stock than those reached by following the same method as for Zones A + B (Figure 5). An improvement in the state of the stock should, however, have followed the departure of the Polish fleet in 1978.

In order to take into account the non-stability of the F vector after recruitment of the "stock", this production analysis by recruit should be redone using a Ricker-type model. Owing to lack of time and processing facilities, it was not possible to carry out these calculations during the Dakar meeting.

Fig. 7 Sardine in Zone C: comparison of the F vectors/age deduced from the Polish and Soviet data (average vectors for the period 1972 to 1977 reduced to the same value for age group III)

Fig. 8 Sardine in Zone C: evolution of the biomass according to the results of the cohort analysis

7.3.4 Zones A + B + C

The same calculations were carried out again for the three Zones A + B + C taken together. The catches in Zones A + B (obtained from the Spanish and Moroccan age compositions) were added to the catches in Zone C previously used (Polish data), for the years 1976 and 1977. A standard cohort was then established taking the mean of the numbers per age group fished during these two years.

The F vectors were calculated for values of M = 0.4 and 0.5 and of terminal F between 0.3 and 1.6 (Figure 9 and Table 2). The most probable vectors, according to the criteria adopted in Sections 7.3.2 and 7.3.3, correspond to M = 0.4 and a terminal F of the order of 1.0-1.2.

This F vector is very comparable to that corresponding to Zone C (Figures 6a and 6b). This similarity is explained by the predominance of the catches made in Zone C. If the concentrations fished in Zones A + B + C can be considered as belonging to the same stock, the conclusions relating to Zone C, and particularly those concerning its probable rate of exploitation, should be more or less applicable to A + B + C as a whole.

Table 22

STANDARD COHORT (1976 + 1977 CATCHES) BASED ON THE CATCHES IN THE WHOLE OF ZONES A + B + C: ESTIMATED F VALUES CORRESPONDING TO DIFFERENT POSSIBLE VALUES OF TERMINAL F AND CORRESPONDING BIOMASSES

	Fterm	Ages						Biomasses ('000 t)
		0	I	II	III	IV	V+
M = 0.4	0.3	0.01	0.05	0.17	0.86	0.77	0.3	3 861
	0.4	0.01	0.05	0.18	0.49	0.86	0.4	3 692
	0.5	0.01	0.06	0.18	0.51	0.92	0.5	3 590
	0.6	0.01	0.06	0.18	0.52	0.97	0.6	3 523
	0.8	0.01	0.06	0.19	0.53	1.03	0.8	3 438
	1.0	0.01	0.06	0.19	0.54	1.08	1.0	3 388
	1.2	0.01	0.06	0.19	0.55	1.11	1.2	3 354
	1.4	0.01	0.06	0.19	0.55	1.14	1.4	3 330
	1.6	0.01	0.06	0.19	0.56	1.16	1.6	3 312

Fig. 9 Application of a Beverton and Holt model to sardine concentrations in Zone C: production curve per recruit for different ages of first catch and values of M

(1) fishing mortality estimated on the basis of the cohort analysis (1975-76; age groups III-V)