Standing Crops and Trophic Levels — Blackburn 
51 
the correlation coefficient of logarithms of 
chlorophyll a and productivity is +0.532, sig- 
nificant at the 5% level. Estimates of water- 
column productivity for noon stations of Table 
1, B were made using a simulated in situ "deck” 
incubator (Blackburn et al., 1962), and to a 
depth of only 40 m; they are not significantly 
correlated with chlorophyll a at 0-100 m (r = 
+ 0.461) but the data (not given here) are 
available for only 8 stations. It may be said that 
the available productivity data do not deny the 
existence of steady-state conditions for the 
station-pairs of the smaller group-— at least for 
those listed in Table 1, A, which are all located 
south of 12 °N. 
Figure 2 shows that 8 of the 11 and 6 of the 
8 station-pairs occur in the region bounded by 
12 °N, 95 ° W, 5°N, and the American coast. 
In this region there are only 2 station-pairs, A13 
and A 15, for which the standing crop data do 
not agree closely with the regressions describing 
at least one of the above-mentioned sets of data 
(AB-ll-Z or AB-8-H). The data for A15 do 
not deviate much from the regressions (Figs. 3, 
4, and 5). Those for A13 deviate considerably; 
this is not surprising because A13 is in an up- 
welling area, the Costa Rica Dome (Wyrtki, 
1964), where a steady state would not be 
expected. 
It is clear that station-pairs fulfilling steady- 
state conditions were largely confined to and 
characteristic of a particular ocean area, in the 
northern spring of 1958. It is more likely that 
ocean conditions permitting the existence of a 
steady state would be distributed in this way, 
than that they would occur in a number of scat- 
tered areas. This is discussed below. 
It is of interest to classify the other station- 
pairs for which the data deviate from the re- 
gressions. For this purpose the C, Z, and M 
observations were used, and a station-pair was 
considered deviant if it fell clearly outside the 
confidence region of either regression (3) or 
(4), i.e., outside dotted lines in Figure 3 or 4. 
The deviations were classified as to whether 
Z and M values were above, within, or below 
the confidence limits of the regression on C. 
The classification of the 36 pairs of Table 1, 
A and B, including the 11 nondeviant ones, is 
in Table 5. 
Table 5 shows that all but one of the possible 
combinations of values of Z and M (compared 
with C) were encountered. The most common 
combination was the one discussed above, with 
both Z and M in confidence limits ( AB-ll-Z) ; 
the two next most frequent combinations were 
those with Z and M both high, and both low, 
compared with C; other combinations occurred 
much less frequently. The station-pairs are all 
identified in Figure 2. 
Table 5 also shows that 9 of the 11 station- 
pairs consistent with steady-state conditions ure 
TABLE 5 
Classification of Station-pairs of Table 1 , A and B (Series AB-36), According to Whether 
Z and M Values Fall Above, Within, or Below the Confidence Limits of Their Regressions, 
(3) and (4) , on C* 
z 
M 
PAIRS WITH C> MEAN 
PAIRS WITH C< MEAN 
NO. PAIRS 
Within 
Within 
[All-12, A14, A16-A20] B9 
A7, B6 
11 
Above 
Above 
A13, A24, A26 
A8 [B3-5, B7] 
8 
Below 
Below 
A15, A21 
[Al-4, B2] 
7 
Within 
Above 
A23 
A10, A25 
3 
Within 
Below 
B8 
A6, B1 
3 
Above 
Within 
A9 
1 
Below 
Within 
A5 
1 
Above 
Below 
0 
Below 
Above 
All, All 
2 
TOTAL 
18 
18 
36 
* Station-pairs with C values above and below the mean are distinguished. Groups of geographically adjacent station-pairs 
within each class are bracketed. 
