332 
PACIFIC SCIENCE, Vol. V, October, 1951 
VARIATION IN CATCH WITH ORDER 
(direction) of haul 
Following Survey 1, Surveys \a and \h were 
undertaken to investigate in more detail the 
obviously large differences between A and B 
hauls at the same station, and also to study 
variation in egg and larva distribution with 
depth. To illustrate the differences between 
A and B hauls it might be noted from Tables 
3 and 4 that at Station 4, the 100 cm. net 
caught 1,930 eggs in the A haul (one direc- 
tion) and only 956 eggs in the B haul (oppo- 
site direction); similarly at the same station, 
the 22 cm. net took 244 eggs in the A haul 
and 193 eggs in the B haul. It is the lack of 
consistency in these differences from station 
to station which produces the large and sig- 
nificant SXH interaction demonstrated in the 
preceding section. The differences, which in 
many cases depart significantly from an ex- 
pected difference of 0 (using a Chi-square test) 
may be due to differences in volume of water 
strained in A and B hauls, to non-random 
distribution of the eggs, or to both factors. 
Surveys \a and \h were both undertaken 
in the vicinity of Station 4, which had yielded 
relatively large numbers of eggs in Survey 1. 
Each A and B haul was of exactly 10 minutes’ 
duration. Two identical 22 cm. nets were 
towed simultaneously, one just below the 
surface and the other at a depth. 
In Survey \a an attempt was made to run 
four replicate hauls (I to IV), each replicate 
including an A and a B haul, and each haul 
including two depths, thus yielding 16 
samples. Each A haul was in a southwesterly 
direction, with the wind, whereas each B haul 
was in a northeasterly direction, against the 
wind. The distance travelled in A hauls, about 
3,000 feet, was greater than in B hauls, about 
2,400 feet, due to the greater speed attained 
when travelling with the wind. In the A hauls 
it was necessary to diverge from the usual 
Station 4 course to miss the reefs off the 
southeast side of Coconut Island. Before 
starting the B hauls, the ship was run along 
the east shore of the island in order to assume 
Fig. 2. A portion of the southern sector of Kaneohe i| 
Bay in the vicinity of Stations 4 and 5, showing the ,1 
location of A and B hauls in Survey la, and indicating Ij 
the presumed distribution of the eggs (broken iso- I 
lines). ■ 
the usual Station 4 course, as illustrated in f 
Figure 2. The resulting data are shown in ' 
Table 8, and the analysis of the data, after 
transformation to logarithms, is shown in 
Table 9. In the latter, the data were adjusted 
to an average distance of 2,700 feet in an at- 
tempt to overcome one source of variation. 
This would correct for differences in volume 
of water strained in A and B hauls only if 
there were no change in the efficiency of the | 
nets with change in speed. 
The adjusted egg catches are as follows: 
REPLICATES A B 
surface deep surface deep 
1 148 130 79 37 
II 520 456 29 35 
III 323 302 106 64 
IV 306 254 56 46 
As shown by the analysis (Table 9), there 
are no significant differences between the 
(geometric) mean catches for replicates (87, 
125, 160, and 119, respectively). However, 
there is a highly significant difference be- 
tween the means for haul order or direction 
(A, 277; B, 52), despite the correction for dis- 
tance travelled. It is most unlikely that this is 
due to a change in net efficiency with speed, 
as the A hauls, at the faster speed, had the 
higher mean counts; efficiency would be ex- 
pected to decrease with increase in speed. 
The analysis also shows a significant RXH 
