Although the acoustic properties of the DSL have frequently been studied and “scattering strength of water column” has been measured (e.g. HALL 1971, 1973), few attempts have been made to use these data for estimating biomass of mesopelagic fish. BAIRD et al. (1974) measured volume-reverberation of a DSL in the Cariaco Trench and got estimates of density of fish in reasonable agreement with estimates based on catch rates. McCARTNEY (1976) working off Western Africa concluded, however, that a calibrated sounder in the range 10 - 30 kHz could be a useful tool, but “the records can be little more than a guide to net sampling programmes”.
Several factors make the Arabian Sea better suited for abundance estimation of mesopelagic fish with acoustic methods than most other areas. Firstly, most of the biomass ascribed to mesopelagic fish was distributed in layers above 400 meters, which makes the signal/noise level favourable and which is within the TVG range of the equipment used. Secondly, there were few other organisms as euphausiids, sergestid prawns or siphonophores in the DSLs. Thirdly, all the fish species contributing significantly to the biomass had gas-filled swim bladders, making them good acoustic targets.
The shallow position of the DSLs probably is related to the high production and therefore low transparency of the water (DICKSON 1972). Farther south, where the production is lower, the DSLs also have a deeper position (BRADBURY et al. 1970).
Although the trawls had no closing device, the acoustic netzonde made it possible to see whether the trawl caught the organisms in the DSL and whether the catches from deep layers were contaminated from more shallow ones. The identification of the DSL organisms seems therefore reliable. During daytime the catches from the DSLs usually contained myctophids with only small contributions from other groups. On cruise 1, however, some large catches of Synagrops sp. showed that this fish contributed significantly to the DSL in the Gulf of Aden, and on cruise 4 a station south-east of the Kuria Muria Island yielded mostly Champsodon sp. But the occurrence of large quantities of these fishes seemed to be restricted both in time and in room. Sometimes the catches from the deepest DSL gave various Gonostomatidae, Sternoptychidae, Astronesthidae and other deep sea families, but generally they were of minor importance. Invertebrates, which are sometimes supposed to make up an important part of the DSLs were seldom caught in large quantities. The Same conclusion was drawn by KINZER (1969) working in the north-eastern Arabian Sea.
During nighttime it was more difficult to distinguish the mesopelagic fish from plankton organisms. To solve the problem, the surface plankton was supposed to give constant echo both day and night. Therefore, when other factors were similar, the integrated echo intensity from plankton during the day was subtracted from the night recordings. Composition of the trawl catches and the relation between the recordings on the 38 kHz and the 120 kHz echo sounder were also taken into consideration. The similarity in the echo abundance of mesopelagic fish obtained during daytime and during nighttime seems to indicate that the method used did not introduce serious bias.
The transformation of integrated echo intensities to fish biomass is a difficult point. There are many studies of acoustic properties of myctophids and other small fish (e.g. McCARTNEY & STUBBS 1970, SHEARER 1970, DALEN, RAKNES & RØTTINGEN 1976, McCARTNEY 1976, NAKKEN & OLSEN 1977). These studies have shown that the density coefficient C becomes less dependent on the species and on the tilt angle as the fish length decreases towards the wavelength. It is also known that fish with swim bladders gives resonance at frequencies lying between approximately
where l is fish length in centimeters and D is depth in meters. When a 38 kHz sounder is applied and the depth is less than about 400 m, a fish must therefore be smaller than about 1.5 cm to give resonance. But there is doubt about what will happen when fish length approaches wavelength. For 38 kHz wavelength is about 4 cm, and most of the fish considered in the present study is therefore in the critical zone. All calculations are, however, based on the assumption that the relation C = constant lb is applicable to all the length groups considered.
In the Gulf of Oman the acoustic measurements indicated a density of 25-63 Benthosema pterotum per m2 surface area. Supposing that they are distributed in two DSLs with a total depth range of 100 m, this corresponds to about 0.3 to 0.6 fish per m3 in these layers. The density in the upper layer during night may be of the same order of magnitude.
The krill trawl used has an opening of about 320 m2. If the whole opening is catching myctophids with 100 per cent efficiency, st. Nos. 449 and 450 both taken during night (Table 5) would indicate a density of 0.5 and 1.3 g m3 or 0.6 and 1.6 fish per m3 respectively. A station (No. 419) taken during daytime in the upper DSL gave 6.3 g m3 corresponding to about 8 fish per m3 filtered water. These figures are underestimates, as the efficiency of the gear is obviously less than 100 per cent. These stations were, however, taken in areas where the density was higher than the mean one for the whole Gulf.
Various estimates of population densities in DSLs have been published, all giving much lower values than those obtained in the present study. JOHNSON et al. (1954) found about one fish per 1000 m3 of water. Based on catch rates BAIRD et al. (1974) estimated the density of Diaphus taaningi in the Cariaco Trench to about 2 fish per 1000 m3. Based on acoustic measurements they got estimates varying from 13 to 130 fish per 1000 m3. CLARKE (1973) studying myctophids in the Hawaiian area got about 0.55 fish/m2 based on catch rates.
From the size data and the counting of otolith zones it is tentatively concluded that the two most important species, Benthosema pterotum and B. fibulatum have a life cycle of one year or shorter. Few studies of tropical myctophids have been carried out, but BAIRD et al. (1974) concluded that Diaphus taaningi reaching a size of about 40 mm probably had a one year life cycle. LEGANG (1967) drew the same conclusion for Notolychnus valdiviae reaching about 30 mm. Boreal species as Stenobrachius leucopsaurus (SMOKER & PEARCY 1970) and Benthosema glaciale (HALLIDAY 1970, GJØSÆTER 1973) reach about 32 mm after one year, Myctophum affine reach about 36 mm (ODATE 1966) while Notoscopelus kroeyeri seem to grow about 80-90 mm during its first year of life (GJØSÆTER in prep.). The growth rates assumed for the Benthosema species therefore do not seem unreasonable. Consequently, the yearly production of these species is as high as, or higher than their standing stock.
From the figures given by CUSHING (1973) the mean primary production in the area covered by “Dr. Fridtjof Nansen” is about 220 gCm-1 180 day-1 in the SW monsoon and 50 gCm-1 180 day-1 in the NE monsoon. These values can be converted to gram wetweight using the factor 0.065 (see CUSHING 1971), and a primary production about 4.2 kgm-2 year-1 is found. The area studied is about 1.7 x 1012 m2 and the primary production is therefore 7.1 x 109 ton year-1. An assumed mean production of mesopelagic fish about 1 x 10 ton year-1 represents therefore between 1 and 2% of the primary production.
CUSHING (1973) also presented estimates of secondary production, and using his figures the secondary production in the area studied is about 1 x 109 ton year-1 or 0.6 kgm-2 year-1. Thus, the production of mesopelagic fish is about 10% of the secondary production. It seems, therefore, that if an ecological efficiency of 10% at each trophical level is assumed, the mesopelagic fish is utilizing the entire secondary production.
It has been shown in other areas too that the production of mesopelagic fish is higher than should be expected from the primary production figures (CLARKE 1973). This may partly be explained by higher efficiency than 10% in oceanic waters (e.g. GRAZE 1970) or by production by bacterioplankton (VINOGRADOV 1973).
In the northern part of the Arabian Sea myctophids were often observed at very low oxygen concentrations. The same was observed by KINZER (1969). From studies in Californian waters DUNLAP (1970) concluded that there was no general relationship between oxyclines and DSL. BAIRD et al. (1974) found Diaphus taaningi in water with oxygen concentrations about 0.35 ml/I in the Cariaco Trench.
KINZER (1969) observed full stomachs with contents showing only slight traces of digestion in Benthosema pterotum in the oxygen minimum in the Arabian Sea, and concluded that they feed on copepods in this layer. He also found a few Diaphus spp. which were all empty. BAIRD et al. (1975) concluded that D. taaningi from the Cariaco Trench feed little, if at all, during daytime. HOLTON (1969), who studied feeding of Triphoturus mexicanus, found mainly empty stomachs during the day and he wrote that “it is possible that this fish does not continue digestion of the food it has consumed in the surface waters, but regurgitates the undigested portion while descending in order to reduce metabolic oxygen needs while residing in oxygen minimum waters”. The present data show that the Benthosema species do not regurgitate their food when descending. It is not clear, however, whether the presence of little digested food sometimes found in their stomachs indicates that they stop digestion to save oxygen or they feed during day in the oxygen minium zone.
