In order to determine the most efficient sampling method of Anguilla japonica leptocephali, the diel change of vertical distribution of leptocephali and sampling efficiency of gears were examined in a fixed water mass marked by drifting buoy with drag in the Western Pacific in September 1986. Forty-min horizontal tows by 3 m IKMT were conducted by night and day at 7 towing depth layers between 10 and 400 m deep. Number of leptocephali taken by night was 2.6 times larger (60 fish) than those sampled during daytime (23 fish). Most of leptocephali (85 %) of night sample were taken at shallow layers, 10, 20, 50 and 80 m deep, while no leptocephali occurred at layers shallower than 80 m by day. All of 10 leptocephali of A. japonica were also taken at shallow layers between 10 and 50 m deep by night. Upper limit of daytime sample and the layers of maximum catch by night roughly coincided with a sharp thermocline observed at 70–80 m deep. A. japonica leptocephali seem to be obtained most efficiently by a horizontal night tow at layer just up to a thermocline. Comparison of sampling efficiency between IKMT (8.7 m2 mouth opening, 0.5 mm mesh) and Hexagon Net (10.4 m2 mouth, 1–8 mm mesh) showed that Hexagon Net caused a bias in sampling leptocephali smaller than 25 mm TL, while IKMT collected fewer leptocephali exceeding 40 mm TL than Hexagon Net.
All Science Journal Classification (ASJC) codes