52 
PACIFIC SCIENCE, Vol. XX, January 1966 
in the region of higher-than-average chlorophyll 
a (Fig. 2) and account for half the pairs in this 
region; whereas, of the other 25 station-pairs, 
16 are in the region of lower-than-average 
chlorophyll a and account for nearly all the 
station-pairs in that region. This is interest- 
ing because the high-chlorophyll area along the 
central American coast is a fairly eutrophic area; 
it is comparable in surface chlorophyll a and 
surface productivity with the region off southern 
California (Holmes, 1958). It has been sup- 
posed that steady-state conditions occur (a) in 
oligotrophic tropical ocean areas, but (b) not 
in eutrophic tropical ocean areas ( Cushing, 
1959^) • The data of this paper oppose (b) , and 
this point is further discussed below. They 
cannot confidently be said to oppose ( a ) because 
other relationships consistent with steady-state 
conditions, besides those represented by the re- 
gressions in Figures 3 and 4, could exist for 
certain groups of station-pairs. In this connec- 
tion, Table 5 shows that 4 adjacent pairs (B3, 
4, 5 and 7, all at the mouth of the Gulf of Cali- 
fornia; see Fig. 2) have high values of Z and 
M, whereas 5 other adjacent pairs farther west 
(A 1-4 and B2) have low values of Z and M. 
Figures 3 and 4 show that the range of C is 
about the same for both groups. The difference 
between these groups might be an effect of 
year (the A data are from 1958, the B data from 
I960) rather than of area. The matter warrants 
further attention when more data are available. 
Station-pair A4 was off Clarion Island. This 
area was visited on another cruise (Island Cur- 
rent Survey) in May 1957, and the oceano- 
graphic data then obtained were presented by 
Bennett and Schaefer (I960). Chlorophyll a in 
a water column 0-80 m averaged 18.0 mg/m 2 
for 10 offshore stations, and zooplankton (in 
hauls made like those described in this paper, 
at various times of day) averaged 16 ml/10 3 m 8 
for 10 offshore stations. The point correspond- 
ing to these measurements would fall well below 
the confidence region, and not far from A 1-4, 
in Figure 3- 
The productivity data of Table 4 are not ob- 
viously helpful in explaining the different kinds 
of deviations from regression shown in Table 5. 
For instance, a structural regression, fitted to the 
data of Table 4 and the corresponding chloro- 
phyll a data from Table 1, A, showed that the 
most deviant noon stations were 15 and 76, cor- 
responding to station-pairs A 5 and A21; their 
productivities were respectively very low and 
very high for the chlorophyll a, but they do not 
help to interpret the particular combinations of 
C, Z, and M observed. 
Biological Significance of Regression Statistics 
As mentioned above, the regression (slope) 
coefficients are likely to have biological signifi- 
cance. Those involving Z or H on C are of 
special interest because of the many previous 
studies of relationships between phytoplankton 
and zooplankton crops. In regression (3 ) , of Z 
on C, for the AB-ll-Z set of station-pairs with 
all standing crops consistent with steady-state 
conditions, the point estimate of the coefficient 
is 0.619 with 95% confidence limits 0.506 to 
0.810. In (9) , H on C, for the AB-8-H set with 
data less certainly consistent with steady-state 
conditions, the corresponding figures are 0.517 
and -0.244 to 0.744. Similar point estimates and 
confidence limits are available for three other 
significant regressions of Z or H on C (equa- 
tions ( 1 ) , ( 5 ) , and ( 8 ) ) in which steady-state 
conditions could not be demonstrated; they are 
respectively 0.634 (0.428 to 0.850), 0.702 
(0.342 to 1.060), and 0.652 (0.394 to 0.920). 
Equation ( 1 ) , reworked for the 27 station-pairs 
of Table 1, A which were all occupied in 1958, 
gave a point estimate 0.829 and limits 0.508 to 
1.150. It may be concluded that the standing 
crop of herbivores generally varies with some 
power less than 1.0 of the standing crop of 
phytoplankton when there is a significant rela- 
tionship between them, whether or not steady- 
state conditions prevail among all standing 
crops. This implies that herbivores utilize phyto- 
plankton with increasing inefficiency as standing 
crop of the latter increases, which is consistent 
with the observations of some workers (Cush- 
ing, 1959 a,b\ Beklemishev, 1962; and references 
cited there) on excessive feeding by herbi- 
vorous copepods: when phytoplankton is abun- 
dant the herbivores may kill more of it than 
they assimilate. 
In significant regressions of M on Z or H the 
coefficients are evidently much closer to 1.0 than 
