3.1 The Estimates of the Demersal Resources
3.2 Comparisons with other Estimates of the Demersal Resources
3.3 The Catches of Demersal Fish
3.4 Estimates of Yield of the Demersal Resources
Biomass estimates from the trawl survey programme, expressed as mean densities, are shown in Annex 4, Tables A1 -A6. These estimates concern demersal fish only, as small pelagic fish, sometimes present in demersal catches, have been excluded from the calculations. The precision of the estimates is low as few hauls have been carried out in each region in the course of each survey, this gives a high standard error in the data. A random very high or low catch can severely bias the estimates when the hauls are few. This effect is smaller when the material is not split between samples north and south of Ras Hafun, represented by tables A5 and A6, as negative and positive sampling errors tend to balance with a higher number of samples. Therefore total estimates give higher precision and are thus more reliable than the regional ones.
The tables in Annex 4 show that the mean density of demersal species for the shelf between Alula and Ras Mabber are 36.2 and 13.6 t/nm2 for the first and second survey respectively. The low last estimate is probably due to the upwelling and migration in relation to this, as already mentioned in the section concerning the small pelagic fish. The first figure of 36.2 is probably more representative for the normal standing stock.
The area south of Ras Mabber has a lower density than the northern part of the shelf. The two southern estimates are 12 and 4.6 t/nm2 for the first and second survey, while the corresponding figures for the northern part are 63.4 and 18.1.
Considerable parts of the demersal biomass are fish of minor commercial value. If one restricts the analysis to high-market-value fish, the average densities drops to 50% or less of the total demersal biomass. We have considered the following fish groups to be of high market value:
|
Groupers, |
(Serranidae) |
|
Emperors |
(Lethrinidae) |
|
Seabreams |
(Sparidae) |
|
Grunts |
(Pomadasyidae) |
|
Snappers |
(Lutjanidae) |
|
Jacks |
(Carangidae) |
|
Cobia |
(Rachycentron canadus) |
|
Barracudas |
(Sphyraenidae) |
|
Croakers |
(Sciaenidae) |
|
Tunalike fish |
(Sarda orientalis,
Scomberomorus) |
|
Survey/area |
number of hauls |
tonnes/nm2 |
% of tot.dem. fish |
|
|
First survey |
||||
|
|
North of Ras Hafun |
8 |
32.2 |
51 |
|
|
South of Ras Hafun |
9 |
5.2 |
43 |
|
|
Both |
17 |
19.0 |
52 |
|
Second survey |
||||
|
|
North of Ras Hafun |
14 |
8.1 |
45 |
|
|
South of Ras Hafun |
7 |
1.1 |
24 |
|
|
Both |
21 |
5.4 |
39 |
Applying the above given estimated densities on the areas concerned the following absolute biomass estimates of commercial fish are obtained:
|
Survey/area |
Estimate |
|
|
|
1000 tonnes |
|
|
First survey |
||
|
|
North of Ras Hafun |
66 |
|
|
South of Ras Hafun |
8 |
|
|
Both |
74 |
|
Second survey |
||
|
|
North of Ras Hafun |
16 |
|
|
South of Ras Hafun |
2 |
|
|
Both |
18 |
All calculations on the demersal stocks are based on a catchability coefficient q = 1.0, that is that all fish within the path between the wings of the trawl are assumed caught. In the literature, coefficients between 0.5 and 1.0 are often used. Applying 0.5 on our catches would raise the estimates to the double of given above. Our estimates thus represents conservative estimates within the range of catchability coefficients normally used.
The 1975-76 R/V “Dr. Fridtjof Nansen” surveys, which were pure acoustic surveys in design, do not give accurate information on the demersal resources. This is partly due to limitations in the acoustic methods itself, but also due to the design and scale of these surveys.
Since the early sixties various surveys and fishing activities have been carried out off NE-Somalia. Most of these activities bear more the characteristics of experimenting fishing surveys or regular fishing activities, than surveys to estimate the abundance of the resources. The only exception is perhaps a Spanish survey with a trawler in 1981. Although the objective was mainly experimental fishing, the vessel was staffed with fishery biologists and a report was issued with information on fish distribution, catch records, and estimates of biomass based on the swept area method. Only the catches of the commercial species were recorded. Based on their findings, a biomass estimate of abt 31 thousand tonnes of commercial fish for the whole east coast was obtained. This is based on a catchability coefficient of q = 1.0, the same as used in our calculations. The shelf of the whole east coast is abt. 9500 nm2, and the total biomass estimate thus corresponds to an average density of 3.3 t/nm2.
Both “Dr. Fridtjof Nansen” surveys come out with considerably higher figures, 19 and 5.4 t/nm2 for the first and the second survey respectively. However as these surveys are restricted to the high-productive areas of the Somali upwelling, they are not necessarily in conflict with the Spanish ones.
The estimated mean densities by species, given in Annex 4 can easily be converted back to the catch rates, on which the calculations are based. For the trawl used on “Dr. Fridtjof Nansen”, which is abt. 18-20 m between the wings, the catch rate by hour trawling at 3 knots is found by multiplying the given mean density (t/nm2) with a factor of 30, giving the catch rate in kg/hour. Note also the left part of the table, giving information on the catch distribution and number of hauls with the species concerned in the catch. If the estimate is based on a single high catch of a species, the precision of this estimate is very low and must be used accordingly.
The method of Beddington and Cooke applied in 2.3.3 to assess the yield of the small pelagic fishes, cannot easily be applied to the demersal stock. The demersal biomass is composed of a variety of species with different sizes, growth patterns, natural mortalities and ages at recruitment. Nevertheless, as the much used Gulland’s formula (Y = 0.5M B0.) seems to overestimate the yield in most instances, some corrections should be made.
Pauly (1980) has listed growth parameters and natural mortalities for 175 fish species. Of these, 34 species are from waters with temperature higher than 20°C and represent families which are common in Somali waters. Table 3 shows the range and the mean of the natural mortality (M) of these species grouped into two classes; a) less than 50 cm maximum length, and b) more than 50 cm maximum length. The means of the M are 0.74 and 1.18 respectively. We therefore suggest M = 0.8 as a first approximation of the natural mortality in Somali waters.
Table 3 Classification in economical classes of estimated biomass based on species composition in the catches and their market value.
|
|
Thousand tonnes |
|
Class 1 (<5 Dirham/Kg) |
104 |
|
Class 2 (5-9 Dirham/Kg) |
120 |
|
Class 3 (10-15 Dirham/Kg) |
114 |
|
Class 4 (>15 Dirham/Kg) |
7 |
|
Total |
345 |
Y = 0.4 M * B0
With a natural mortality M = 0.8 and a standing stock of 74 thousand tonnes of commercial valuable fish, the annual yield is estimated to 24 thousand tonnes from the NE-Somali shelf. Due to the limited number of stations in our trawl survey, and to the fact that one big catch raises the average catch considerably, we feel that until more data becomes available, a safer limit would be 20 thousand tonnes for commercial demersal fish.
