6. Summary

The Institute of Marine Research, Bergen, operating the Research Vessel "Or Fridtjof Nansen" made a survey in 1975-76 of the fishery resources of the North-west Arabian Sea, from near Mogadiscu on the Somali coast, through the Gulfs of Aden and Oman, to the Iran/Pakistan border. The survey was accomplished by six cruises which effected 5 complete traverses, or coverages, of the region.

The work of each cruise was described in a Cruise Report issued soon after completion of the cruise (Cruises 1 and 2 were reported together), and a final report was made. These documents were transmitted to FAO and were discussed at a workshop, in Karachi, Pakistan in January 1978, "on the fishery resources of the north Arabian Sea". From those discussions emerged agreement that still more information could be drawn from the data obtained in the course of the survey and that perhaps a more accurate estimate might be made of biomass. Further analysis of the data was therefore undertaken and is reported here.

For the sake of completeness, so that this document should be quite self-contained, much material has been taken from the previous reports.

A brief account is given of the conduct of the survey. Some notable events which affected operations are mentioned. The equipment and methods are described. The execution of the work is described briefly and a tentative assessment is made of its efficiency. Data processing methods are described in detail.

A brief summary is given of meteorological, biological and oceanographic Information available before the survey which served as background to the plans for it.

The principal results of the survey are measurements of biomass observed in each of the several sectors of the region and, from these measurements, estimates of catches which might be expected. These results are given below as lower and upper values.

Biomass:

Sector	Demersal	Pelagic	Both
		thousand tonnes
1 South Somalia	392 - 822	953 - 1,214	1,258 - 1,527
2 North Somalia	67 - 128	128 - 225	194 - 307
3 Socotra	55 - 116	112 - 224	190 - 315
4 South Yemen	164 - 238	376 - 495	528 - 645
5 South Oman	98 - 156	730 - 903	846 - 1,017
6 North Oman	94 - 108	123 - 142	150 - 317
7 Iran	106 - 139	113 - 161	140 - 214
Totals	976 - 1,707	2,535 - 3,364	3,306 - 4,342

Prospective Catch:

Sector	Demersal	Pelagic	Demersal plus pelagic
			Sum	Direct
		(thousand tonnes)
1 South Somalia	69 - 144	333 - 425	402 - 569	385 - 534
2 North Somalia	12 - 22	45 - 79	57 - 101	34 - 107
3 Socotra	10 - 20	39 - 78	59 - 98	33 - 110
4 South Yemen	34 - 42	132 - 173	166 - 215	92 - 225
5 South Oman	17 - 27	256 - 316	273 - 343	148 - 360
6 North Oman	15 - 19	43 - 50	58 - 69	26 - 111
7 Iran	19 - 24	40 - 56	59 - 80	25 - 75
Totals	176 - 298	888 -1177	1064 -1475	743 -1522

Where "Sum" columns are obtained by addition of the preceding columns, the "Direct" columns result from application of the lower values of B_max (combined estimates) on page 37 in the expression Y_dem = 0.175 B_max, and the higher values of B_max on page 37 in the expression Y_pel = 0.35 B_max.

The procedures by which the biomass values were calculated are described in detail with special reference to particular circumstances of this survey and features of the stocks. This account of the procedures shows clearly the nature of the results obtained and gives a basis for judging the confidence that can be placed in them.

Two major difficulties were encountered in applying these procedures. The first concerned the setting of a value for the factor for conversion of a unit of integrator recording into a biomass value. The value of the factor depends, for any set of integrator recordings, upon the size and other characteristics of the individuals of the target stocks at the time of observation; these characteristics are primarly particular to each species although the biomass of species similar in form, size and other characteristics can be estimated from a single factor. In this connection the information from fishing and surface observations and from biological examination of samples from catches is of critical importance and has been discussed in detail in the report. The conclusion must be drawn that while the information as to stock composition obtained by these methods is fully in conformity with what was to have been expected from existing knowledge of the biota of this region, the frequency of sampling and intensity of analysis were not such as to yield good estimates of the conversion factor. Conservative values of this factor have therefore been applied and in general it is thought that the estimates are more likely to be below the true values than to be above them.

The other difficulty arose in seeking an estimate of annual mean biomass, which is the stock characteristic from which potential yield is properly to be calculated. For this purpose it is necessary to have a reliable picture of the annual cycle of events in a stock, with such increase in abundance as there may be from recruitment of juveniles and from their rapid growth and such decline in abundance as mortality causes. The effect of these processes may be augmented or nullified by migration into and out of a particular area. Some evidence on these matters was obtained in the survey and is discussed in the report, but it is inadequate, and the timing of observations in each sector was such that a true measure of the annual cycle is not available, and therefore mean annual biomass cannot be calculated from the data. A rough estimate can be made, however, on the principle that the annual mean is unlikely to be less than 50% of the maximum observed and more probably is close to 70% of that maximum. For these reasons the above measurements of biomass are the maximum observed values, presented as lower and upper values calculated from the confidence limits of the mean integrator recording of each sector.

The estimates of likely catch level have been obtained from the measurements of biomass by application of the familiar formula:

C = 0.5 × M × B_max

Perhaps the most important outcome of this further examination of the results of the survey has been the light it has thrown on the numerous technical and methodological problems of this kind of work. Even if the estimates here of biomass and yield potentials are no great improvement on the estimates in the Final Report, their presentation with upper and lower values derived from an estimate of confidence limits (even if admittedly the calculation of those limits is not very satisfactory) suggests the bounds within which planning decisions ought, at present, to be set. Moreover, the structure of the argument leading to these estimates is described so that as new information becomes available the estimates can be adjusted with respect to particular components of the system; which means that adjustment of the estimates need not await some repetition of work such as was done on this survey but should be achieved by systematic testing of the assumptions described in this report and by intensive work designed expressly to measure the effects of the causes discussed in Section 3.3 and to obtain a better measure of the conversion factor C.

Given the magnitude of the task undertaken for the survey -- to traverse some 44,000 nm² of continental shelf five times in less than two years, given also the large number of species and the relative scarcity of information wit regard to the display of almost all of these, allowing for weather conditions which made work impossible in at least one month and brought some difficulties at other times, and recognising the inevitability of contingencies and of human error, it is obvious that the results would be marked by a number of important uncertainties. In particular it could be expected that the results would have to be with regard to aggregates of species and could not provide a precise and detailed account of any one species.

On these terms, however, the survey was a successful exploration which has characterised the distribution and main compositional features of the fish stocks and has provided fair approximations to the annual mean biomass and yield potential of those stocks. Admittedly those estimates are low in comparison with others which have been made; indeed, we are convinced that significant upward revision of them will be made when the stocks in very shallow waters and in the 3 m bottom stratum can be assessed. The degree to which the estimates might be increased after such a revision, or might have to be increased for other reasons, it is a matter on which nothing can be said at this stage. However we believe that industrial development based on the present estimates could be planned safely and would soon yield information from which to improve the estimates.