44 
PACIFIC SCIENCE, Vol. XX, January 1966 
been combined for this and all subsequent parts 
of the analysis. The different parts of Table 2, 
representing three cruises made in the Northern 
Hemisphere autumn, have been similarly com- 
bined. All statistics were calculated with the 
logarithms of the measurements; micronekton 
data in Table ID were first raised by 0.1 to 
eliminate zeros. 
The two-variable regressions given below are 
structural regressions fitted, estimated, and in- 
vestigated by the methods of Bartlett (1949). 
These methods are appropriate when each vari- 
able is subject to error and when the functional 
relationship, rather than a predictive one, is of 
interest. They include the estimation of con- 
fidence limits and tests of linearity. In these 
regressions C, Z, M, and H are logarithms of 
standing crops of chlorophyll a, zooplankton, 
carnivorous micronekton, and copepods, respec- 
tively. The grouping of points for Bartlett’s 
method was done on the basis of C values for 
each regression except (6), where it was done 
on the basis of Z values. The assumption of a 
linear relationship among points was validated 
at the 5% level of probability in each regres- 
sion except ( 8 ) ; in ( 8 ) departure from linearity 
was significant at the 5% level but not at the 
2.5% level 
According to Bartlett (1949), the joint con- 
fidence region for the parameters of slope and 
position of the above-mentioned type of regres- 
sion, i.e., the region in which the true rela- 
tionship is likely to He, is elliptical For the 
regressions of this paper for which such regions 
were of Interest, it was found that the sides of 
the ellipses were nearly straight over most of 
the range of the observations, and they were ac- 
cordingly drawn as straight lines in Figures 3, 4, 
and 5. 
The positions of the regressions are of little 
biological interest or significance, except In a 
few Instances mentioned in the discussion, for 
such reasons as: differences in amount of inter- 
stitial water included In Z, M, and H; the omis- 
sion of some material of the trophic level that it 
was desired to measure, and/or the inclusion of 
some material of another trophic level, in Z, M, 
and H; and differences in the ways in which 
some crops were measured on certain cruises, 
namely C with respect to depth In Table 1, C, 
and Z with respect to time of day in Table 1, C 
and in Table 2. However, the slopes of the re- 
gressions may be considered significant biologi- 
cally, because the measurements of a given crop 
for a given cruise are probably all comparable, 
methodologically. 
Chlorophyll a, Zooplankton, and 
Carnivorous Micronekton 
The largest set of data (n = 36) is for the 
combined two cruises .of the Northern Hemi- 
sphere spring (Table 1, A and B) . This set is 
henceforth called AB-36-Z. Figure 2 shows 
where the observations were made. 
TABLE 3 
Total and Partial Correlation Coefficients Among Logged Standing Crop Data for Station- 
pairs in Table 1 and Stations in Table 2f 
SOURCE OF DATA 
n 
fern 
f am 
f es.m 
fern . 2 
Xzm.c 
Table 1, A and B 
36 
+0.702 ** 
+0.615** 
+0.626** 
+0.515** 
+ 0.345* 
+0.317 
n 
Ditto, selected data 
11 
+0.966** 
+0.860** 
+0.840** 
+0.880** 
+0.348 
+0.068 
n 
Table 1, C 
18 
+0.718* * 
+0.314 
+0.365 
+0.682** 
+0.078 
+0.207 
Table 1, D 
14 
+0.376 
+0.137 
+0.781** 
+0.435 
-0.271 
+0.794* 
# 
Table 1, E 
13 
+0.592* 
Table 2 
35 
+0.124 
f ch 
fern 
film 
fch.ra 
fcm.li 
fhm.c 
Table 1, A and B 
36 
+0.613** 
+0.615** 
+0.602** 
+0.386* 
+0.390* 
+0.361* 
n 
Ditto, selected data 
8 
+0.929** 
+0.874** 
+0.849** 
+0.728 
+0.440 
+0.202 
c 4 ] 
[q Series AB-36-Z Series AB-ll-Z [3] Series AB-36-H [*) Series AB-3-H 
f The letters c, z, h, and m indicate chlorophyll a, zooplankton, copepods, and carnivorous micronekton respectively, * and 
** indicate significance at 5 and 1 % levels of probability, and n is number of station-pairs or stations; see text for particu- 
lars of "selected data.” 
