Catch rates in relation to density estimates obtained by echointegration

During trawling echo integration was run in a depth interval corresponding to the vertical opening of the trawl. From these values the average fish density (g/m³) at the trawl opening was estimated. These estimates are plotted against the obtained catch pr m³ filtered sea water in Fig. 13. Table 10 shows the average "trawl efficiency" (the ratio between catch/m³ and density) for each layer. The average "trawl efficiency" seems to be considerably higher for the D I layer than for any of the other layers. This difference is significant, while the difference between D II, N I and N II are not significant.

Fig. 13. Catch rate (grams per m³ filtered seawater) in relation to fish density (g/m ³) estimated acoustically in various layers. A: Gulf of Aden and Gulf of Oman July-August 1979.

Fig. 13. Catch rate (grams per m³ filtered seawater) in relation to fish density (g/m ³) estimated acoustically in various layers. A: Gulf of Aden and Gulf of Oman July-August 1979. B: Gulf of Oman January-February 1981.

Table 10. Average "trawl efficiency" [ratio between catch (gram) pr m³ and density (g/m³)] for various layers during the two cruises.

Kind of layer	D I		D II		N I		N II
Year	1979	1981	1979	1981	1979	1981	1979	1981
Average trawl efficiency	0.51	0.77	0.11	0.05	0.07	0.17	0.61	0.07
Standard deviation	0.38	0.51	0.08	0.10	0.05	0.20	0.61	0.06
No. of stations	6	23	16	11	18	17	4	7

The low "trawl efficiency" in the N I, N II and D II-layers may give reasons to suspect that the fish in these layers were mixed up with plankton giving high contribution to the integrator values. The N I-layer was in fact mixed up with some plankton which remained as a separate layer in the upper 100 m during day time. However, the echo-contribution from this plankton layer usually represented just one to ten percent of the total integrator values during day time in the Gulf of Oman and up to 50% in the Gulf of Aden. Juday net hauls did not give much plankton in any of the layers.

In 1981 some underwater photos were taken in connection with trawl hauls and light experiments. Fish was recognized on most of the pictures in all layers, while plankton organisms were rather scarce. Salps and jellyfish were the only recognizable plankton on the pictures (Some faint traces could be shrimps). Table 11 shows the average number of fishes and salps per picture. About 30 pictures were taken at each station.

Table 11. Average number of fish and salps per picture taken in the scattering layers in the Gulf of Oman, January 1981.

Kind of layer	Nearest Trawl st. no.	Average number per picture		Remarks
		fish	salps + jellyfish
D I	47	13	0.02
D I	51	2	0.1	Camera not in the middle of the layer
D II	52	7	0.4
D II	57	9	0.1
N I	48	66	0.3	Fish gathered in the light
N I	48	3	1.0	Camera below the most dense layer
N I	59	5	0

A cage calibration of salps indicated that the average target strength is in the order of -50 dB per kg salps, which means that some hundred kilos of salps are needed to give the same echo as one kg of fish.

The conclusions is that the echo contribution from plankton was small compared to the contribution from fish in all layers. Therefore the plankton is not responsible for the observed differences in "trawl efficiency". There has to be a real difference in catchability.

The D I layer was composed of schools in a presumably lighted zone, while the other layers in darkness, had a rather homogeneous density. This difference in light conditions may cause a difference in the fish reaction towards the trawl. Probably the fish in the D I layer are able to see the trawl. The fish has not enough mobility to avoid the trawl opening, but they may be able to avoid coming in contact with the walls of the foremost part of the trawl. This means that they may be herded towards the centre of the trawl, thus being more easily caught. The fish in the other layers are probably not able to see the trawl and may therefore have a less directional reaction, so that much of the fish may be filtered through the meshes in the foremost part of the trawl. The average "trawl efficiency" for the D II, N I and N II layers together is 0.13, which seems to be reasonable for small, passive fishes, when the shape and mesh size of the trawl are taken into consideration (Fig. 13).