vessels, approximately 250 m apart, steamed at 

 2 knots. Concurrent surface and oblique tows 

 were made by both vessels at six stations in 

 daylight between 1300 and 1700 h local time; the 

 start of tows on each vessel was co-ordinated by 

 radio and visual communication. The distance 

 separating the two vessels during the tows was 

 determined by radar. 



The plankton samples taken by the 70 cm net 

 were preserved in 95% ethanol and those from 

 the 2 m net in 10% formalin. Samples were 

 sorted, and tuna larvae identified to species 

 (where possible) and measured (standard length) 

 in the respective laboratories. At FSFRL the 

 samples were sorted in petri dishes with the 

 unaided eye, whereas at CSIRO samples were 

 sorted in a rotatable sorting ring under a dissect- 

 ing microscope with dark-field illumination. Both 

 laboratories used a dissecting microscope to aid 

 in identification of larvae. 



The relative efficiency of the 70 cm and 2 m 

 nets was described by ratios of the mean catch 

 per unit volume for each method of deployment 

 (surface and obhque tows) assuming that small 

 catches are more prone to error than large 

 catches (Snedecor and Cochran 1967). The 

 significance of differences in the number of tuna 

 larvae per unit volume collected by the 2 m and 

 70 cm net in surface and oblique tows was 

 tested by two-factor ANOVA of split-plot 

 design (Winer 1971). The data were log (x -I- 1) 

 transformed to make variances homogeneous 

 (Cochran's test; C = 0.47; df = 4,5; P > 0.05). 

 Differences in the mean lengths of larvae 

 caught by net type and tow type were analyzed 

 by the Kruskal-WalHs test (corrected for ties), 

 and multiple comparisons (with unequal sample 

 sizes) were made using the Dunn test (Zar 

 1984). 



Results 



The stations were at approximately 2.4 km 

 intervals and the vessels were, on average, 

 260 m apart during tows. The range of volumes 

 filtered by net and tow type was 505-560 m^ for 

 a 70 cm net in oblique tow; 244-307 m^ for a 70 

 cm net in surface tow; 4,600-5,650 m^ for a 2 m 

 net in oblique tow; and 3,600 m'^ for a 2 m net in 

 surface tow. Mean displacement volume of 

 plankton caught by the 70 cm net was higher in 

 oblique tows (27.3 mL/1,000 m^) than in surface 

 tows (4.4 mL/1,000 m'*). Both surface and 

 oblique tows with the 70 cm nets caught a total 

 of 283 tuna larvae whereas the 2 m nets caught 



123 (Table 2). The species composition of the 

 sorted and identified catches was similar except 

 for an unidentified component in the CSIRO 

 catches. This was due to a more conservative 

 approach to identification by the less experi- 

 enced CSIRO staff. The species composition of 

 the FSFRL samples suggests that most CSIRO 

 Thunnus spp. were probably Thunnus 

 alalunga. 



Table 2. — Numbers of tuna larvae taken in 2 m net hauls 

 (surface and oblique) by FRV Shoyo Maru and in 70 cm net 

 hauls (surface and oblique) by FRV Soela. 



'Small larvae 2.1-3.8 mm SL but usually < 3.3 mm SL not tiaving 

 the pigment pattern needed to distinguisti to species. 



The mean catch in the 70 cm net in oblique and 

 surface tows was 16.5 and 144.7 tuna larvae/ 

 1,000 m^, respectively. Catches in the 2 m net in 

 oblique and surface tows were much lower, 1.4 

 and 3.7 tuna larvae/1,000 m^, respectively. The 

 ratio of the mean catches (70 cm net/2 m net) was 

 11.8 (SE ± 3.5) for oblique tows and 39.6 (SE ± 

 10.7) for surface tows. Catches in the 70 cm net 

 were significantly higher than in the 2 m net {F 

 = 140.9; df = 1,5; P < 0.001) and catches in 

 surface tows were significantly higher than 

 oblique tows (F = 11.3; df = 1,10; P < 0.01). The 

 difference in catch between surface and oblique 

 tows was mainly attributable to the 70 cm net; 

 hence, there was a significant interaction be- 

 tween net type and tow depth (F = 6.1; df = 

 1,10; P< 0.05) (Fig. 2). 



There were significant differences in the mean 

 lengths of larvae caught by net and tow type 

 (Kruskal-Wallis test; H = 21.74; df = 3; P < 

 0.001). The mean length of tuna larvae caught in 

 the 2 m oblique tows (4.03 mm) was significantly 

 greater (Dunn test; P < 0.05) than those caught 

 in 2 m surface tows (3.60 mm), 70 cm oblique 

 tows (3.56 mm), and 70 cm surface tows (3.65 

 mm) (Fig. 3). This was due to a lower proportion 

 of small larvae (and the absence of larvae less 



978 



