axis of the mesh is parallel to the long axis of the 

 net. The ratio of total filtering area to area of 

 mouth aperture (area of nylon thread included in 

 the filtering area) is about 7.6:1. The detachable 

 cod end is made of 56XXX nylon or silk grit 

 gauze, with mesh apertures 0.31 mm. wide; the 

 choice of this material was dictated purely by 

 convenience (the same type of cod end was rou- 

 tinely used on the same cruises for collecting 

 zooplankton). A larger mesh might have served 

 equally well. Lengths of 6-mm. galvanized-iron 

 welded-link chain are attached at one end to the 

 four corners of tlie square frame and at the other 

 end (in pairs, right side and left side) to the ends 

 of a 1.5-m. galvanized-iron bar of 25-mm. 

 diameter. Two similar chain bridles from the ends 

 of the bar are attached to a swivel at the end of 

 the 10-mm. towing wire rope. The arrangement 

 and approximate lengths of the chain bridles are 

 shown in figure 3. One sinker was attached to 

 each bottom comer of the square frame; two 

 45-kg. cylindrical iron weights were used during 

 cruise TO-58-1, but these were replaced by two 

 20-kg. bronze streamlined depressors during the 

 other cruises. 



When the 45-kg. weights were used, the haul- 

 ing operation was as follows : 4.50 m. of the 10-mm. 

 wire rope were paid out at 25 m./min. and then 

 immediately retrieved at 10 m./min., all at a ship 

 speed of 5 knots. The haul, thus, took about 63 

 min. (in practice there was some variation from 

 the desired winching sfjeeds — see table 1). It was 

 found, by attaching a bathythermograph to the 

 frame of the net, that the maximum depth ranged 

 from 72 to 115 m. and averaged 90.2 m. (12 hauls) . 

 Wlien the 20-kg. depressors were used, 350 m. of 

 wire were paid out and retrieved at the same winch 

 and .ship speeds, for a total hauling time of about 

 49 min. ; the maximum depth ranged from 73 to 

 160 m. and averaged 95.5 m. (48 hauls). Because 

 the difference between the means is small, the two 

 types of operation have be«n considered com- 

 parable except for the time; the second type has 

 been continued routinely, and the catches from 

 both types of hauls have been considered to rep- 

 resent broadly the quantity and quality of 

 micronekton present at night that can be caught 

 by the net in the upper 90 to 95 m. of ocean. The 

 actual depths, where obtained by bathythermo- 

 graph, are listed in table 1. The depressors were, 



as expected, more effective in sinking the net than 

 the much heavier cylindrical weights. A sampling 

 depth of 90 to 95 m. would be expected with the 

 Isaacs-Kidd midwater trawl also, if used on 350 m. 

 of cable at a towing speed of 5 knots, according 

 to tests made by Aron, Ahlstrom, Bary, Be, and 

 Clarke (1965). 



No attempt was made to use a flowmeter in the 

 net routinely, because it was thought that it might 

 prevent the entry of some of the micronekton. 

 Flowmeters were used twice to estimate the filtra- 

 tion coefficient of the net by making successive 

 tows in a precisely identical manner, and of pre- 

 cisely the same duration, first with a flowmeter 

 in the mouth of a fully rigged net, and then with 

 the net removed and only flowmeter, frame, 

 bridles, and depressors remaining. The ratio of 

 flowmeter readings was 0.757 on the fii-st trial and 

 0.738 on the second trial (3 years later). The net, 

 used as described, apparently filters 74 to 76 per- 

 cent of the water available (product of net-mouth 

 area in square meters and distance towed in 

 meters). Because the two estimates of the filtra- 

 tion coefficient are so close, the one first obtained, 

 which had already been used to calculate volume 

 of water strained on many of the hauls, was used 

 for all. Water was available for filtration at the 

 rate of 1,000 m.^ every 2.79 min. at a speed of 5 

 knots; the rate of actual filtration was estimated 

 by the coefficient 0.757 to be 1,000 m.' every 3.69 

 min. By dividing the measured vohune by the total 

 number of minutes for the haul and multiplying 

 by 3.7, the micronekton was standardized to vol- 

 ume per 1,000 m.' of water strained (see table 1). 



The most important feature of this net is the 

 mesh size, which retains most of the micronekton 

 and releases most of the zooplankton, and, more 

 importantly, is uniform throughout the net. This 

 miiformity gives the standardized volumes a much 

 more precise biological meaning than would be 

 possible for similar measurements with a mixed- 

 mesh net ; some parts of a mixed-mesh net permit 

 more escapement of organisms than other parts. 

 Aron (1959, 1962a, 1962b) admitted this problem 

 in relation to the Isaacs-Kidd midwater trawl, 

 which has larger meshes near the mouth than near 

 the cod end (as do most other nets and trawls 

 liitherto used for caA«hing micronekton and 

 zooplankton). King and Iversen (1962), who used 

 the Isaacs-Kidd and other trawls, observed that 



78 



U.S. FISH AND WILDLIFE SERVICE 



