Size distributions of the catches of each species were obtained from the trawl catches.
The echo-sounder results are given in measurements in decibels of the target strengths of the fish. The target strength of a fish largely depends on the size and shape of its swim bladder, which is the main reflective surface. The target strength will therefore vary between species and also with the angle of the swim bladder, and so with the fish's behaviour. Fish near the surface which dive as the boat approaches will present a smaller swim bladder area in the horizontal plane and so will appear as smaller fish.
Only the deep-water trawl catches and acoustic survey data have been used in order to minimize bias from the scaring of fish by the boat towing the transducer. In order to reduce net selection as much as possible, Commercial Vendace Trawl and Young Fish Trawl catches have been combined, and the percentage length distributions, to the nearest cm below, are given in Figures 6, 7 and 8.
It will be seen in Figure 6 that in the length frequency diagram from the trawl data there are four modes, at 6 cm, 8 cm, 14 cm and 22 cm. These correspond to the catches of ruffe, smelt, vendace and whitefish. In the frequency distribution of target strengths, the distribution also appears to be derived from four modes, at -53 dB and -54 dB, -49 dB and -50 dB, from -43 dB to -46 dB and at -37 dB and -40 dB, with -51 dB and -52 dB frequency derived from, overlap of the lower two modes. Lindem (personal communication) has suggested that the empirical formula
| T.S. = 20 log10 Lm - 27 | (1) |
gives an approximate relation between target strength (T.S.) and the fish length in meters (Lm). Since the relation of swim bladder size and shape to fish length will vary with different species, it is not surprising that a general formula will not give perfect fit to lengths from a mixed catch. With the material from Konnevesi it appears that a reasonable relationship may be obtained by associating the whitefish at 22 cm with -40 dB, the vendace at 14 cm with -45 dB and the smelt at 8 cm with -50 dB. Scales have been chosen and fitted i n t o Figures 6, 7 and 8 so that the target strength distributions agree approximately with the associations given above. When this has been done, it is found that the values are close to those obtained from equation (1). The fish with target strengths of -54 dB and -53 dB are probably 0-group smelt which were too small to be retained in the trawl (a very few of these fish were found in the trawl catches). Errors from the echo-sounder are most likely to occur at the lower size range and the modes of small fish are less likely to be quantitatively correct.
The figures not only indicate approximately the same positions of the modes from trawl and acoustic data, but also the frequencies agree well.
A programme that takes into account the multiple echoes when fish are closely packed, showed a total density of 6 239 fish/hectare in the trawling area on the night of 6 September after the trawl sampling had been completed.
It seems reasonable to conclude that, bearing in mind that net selection will allow the smaller fish to escape through the meshes and net avoidance will give biased results for larger fish, the acoustic method and the trawl catches give size frequency results which are in remarkable agreement.
