ANNEX 2. Details of the acoustic equipment and settings used.

During both cruises a 38 kHz scientific sounder was connected to two echo integrators. The settings used were as follows:

TVG and receiver gain	20logR-20dB
Recorder gain	8
Pulse length	0.6 msec
Bandwidth Transducer	3 kHz (1 kHz in 1979) 8° x 8°, ceramic
10 log Y	-21 dB
Integrator gain	20 dB
Integrator recorder mode	x 10
Integrator threshold	1 (~0.01Vp)

At these settings calibration were performed in January 1981 both with a -29 dB steel sphere (130 mm in diameter) and a -35 dB cuprous sphere (50 mm diameter). The calibrations gave a sum of source level and voltage respons of 138 dB, at 20 m. depth. The integrated echo intensity from the steel sphere showed 1.4 dB drop when applying 1 kHz bandwidth instead of 3 kHz. Repeated hydrophone calibrations from August 1979 to August 1980 have shown a stable SL of 233±1 dB//1µPa ref. 1 m.

The receiver gain and time varied gain (TVG) have, however, been changed (in April 1980) as Figure 26 shows.

According to Fisher and Simmons (1977) and Foote (1981) the hydrographic conditions in the area gives a sound absorption of about 7 dB/km at 38 kHz. Based on this an ideal TVG is shown in Figure 26. This shows that the TVG applied in 1979 was too high while the one applied in 1981 was too low. In 1981 corrections were made according to the depth of the recordings, but no corrections were made in 1979. This is believed to be the main reason why the integrator values obtained during night tended to be higher than during day in 1979. (In 1981 the opposite tendency was found among the uncorrected values).

For the depth intervals where most of the myctophid fish recordings occurred the difference between August 1979 and January 1981 is about 7 dB.

The factors applied for converting echo-intensities to fish densities were

C79 = 0.1-L	tonnes/n.mile²	per mm	per n.mile
C81 = 0.6.L	“	“	“

for 1979 and 1981 respectively.

(L = standard length, cm).

These are based on an average target strength of -10logL -22 dB per kg fish. This is compareable to the value of -28.4 dB/kg for 5.2 cm fish applied by FAO (1981), when the backscattering cross section per kg fish is assumed to decrease linearily with fish length. During other cruises with "Dr. Fridtjof Nansen" in 1979 a conversion factor of 0.2-L tonnes/n.mile was applied. This value was, however, assumed to represent a cod type fish which is more heavy bodied (NAKKEN & AUNG 1980).

Threshold settings

In the Gulf of Aden the fish tended to be very scattered vertically. This distribution resulted in weak echoes over a great depth interval. This situation requires high signal/noise ratio. Figure 27 shows the smallest recordable target strength for single fishes as function of depth, when the lowest threshold not including noise is applied. At the standard setting (20 logR-20 dB) the threshold setting was limited by electronical noise (0.005 Volt) which corresponds to a TS min =-48 dB at 500 m depth. While applying the 20 logR - 0 dB setting, all acoustical signals at calibrated output were raised by 20 dB while the electronical noise remained constant. The threshold setting was thus limited by the acoustic noise corresponding to a TS min =-59 dB at 500 m depth. This latter setting was applied during the last week of the cruise in the Gulf of Aden in 1979. It was found to give 20-30% higher integrator values for the most scattered recordings at about 350 m depth. However, it was not applied during the survey in 1981 because signals from more dense recordings in the upper layer easily run into saturation in the receiver when the 20 logR -0 dB setting is used. This is shown by the low TSmax in Figure 27. At the most applied settings max (20 logR -20 dB and ca 0.01 Volt threshold) proper integration is made of targets of less than -30 dB target strength at 10 m depth or more than -42 dB at 500 m when referred to single fishes. This TSmin =-42 dB corresponds to a minimum volume back scattering strength of -72 dB/m³ or a "threshold density" of about 0.1 grams/m³ when referred to multiple scatterers.

Figure 26 Time varied gain functions:

I - Function applied in 1979
II- Function applied in 1981
III- Ideal function according to hydrographic conditions.

Figure 27 Maximum and minimum target strength (for single fishes) required for proper echo integration. A.: 20 logR - 20 dB mode

Figure 27 Maximum and minimum target strength (for single fishes) required for proper echo integration. B: 20 logR - 0 dB mode

I: Maximum target strength (limited by receiver saturation)
II: Minimum target strength to avoid masking by electrical noise (0.005 Vp at calibrated output)
III. Minimum target strength to avoid masking by "normal" acoustic noise.
IV: Minimum TSmin when using 0.01 Vp threshold.