138 



FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 



and temperature at the 100-meter level independ- 

 ent of variation with inorganic phosphate. A 

 multiple regression analysis (following Snedecor 

 1946, pp. 340-373) was therefore carried out em- 

 ploying cruise-5 data, treating zooplankton vol- 

 umes (cc./m. 3 ) as the Y variate, inorganic phos- 

 phates (i^g at./L) at 100 meters as the X t variate, 

 and temperatures (°C.) at 100 meters as the X 2 

 variate. 



The following multiple regression equation was 

 obtained : 



Y= 0.0484X! + 0.0023X-, - 0.044969. 



The first partial regression coefficient (&n. 2 = 

 0.0484; t= 5.232, P<0.001, at 40 degrees of free- 

 dom) and the first partial correlation coefficient 

 (r ri .2 = 0.638; F<0.01) are highly significant, 

 showing that there is a positive relation between 

 zooplankton volumes and inorganic phosphates, in- 

 dependent of the variation with temperature, at 

 the 100-meter level. This conclusion was expected 

 from the previous correlation analysis. The sec- 

 ond partial regression coefficient (5r 2 .i = 0.0023; 

 2=2.987, P<0.01, at 40 degrees of freedom) and 

 the second partial correlation coefficient (»'y 2 -i = 

 0.423; P<0.01) are also highly significant, show- 

 ing that there is also a positive relation between 

 zooplankton volumes and temperature, independ- 

 ent of variation with inorganic phosphate, at the 

 100-meter level. This conclusion is of interest 

 since it indicates a much higher degree of covaria- 

 tion between zooplankton and temperature, than 

 was revealed by the previous correlation analysis. 



We have shown that the abundance of zooplank- 

 ton in the central Pacific is correlated with inor- 

 ganic phosphates, temperature, dissolved oxygen, 

 and thermocline depth. We do not believe, 

 though, that the abundance of zooplankton is in 

 any way limited by the conditions of temperature, 

 oxygen, or thermocline depth prevailing through- 

 out the region. The relationships between zoo- 

 plankton and these three environmental factors, as 

 indicated by the statistical correlations, are most 

 likely independent variations due to common 

 causes. Zooplankton is more directly linked with 

 phosphate, however, through the phosphate— > 

 phytoplankton— ^zooplankton relation. There- 

 fore, the causal agent for these several variations 

 examined is directly or indirectly the Equatorial 

 Current system, principally the divergence at the 

 Equator which brings to the surface water that is 



relatively low in temperature, low in oxygen con- 

 tent, but high in chemical nutrients, and the con- 

 vergence to the northward which results in a 

 deepening of the thermocline and a possible con- 

 centration of the plankton. 



Other investigators have found essentially simi- 

 lar correlations between plankton and these en- 

 vironmental factors. Marshall and Orr (1927) 

 observed in the Clyde Sea area that where animal 

 life was rich, phosphates were high, but dissolved 

 oxygen and pH were low. Hardy and Gunther 

 (1935) found in the Pacific that numbers of zoo- 

 plankton were positively correlated with phos- 

 phate values. Jesperson (1935) states that there 

 is a direct relation between quantities of nitrate 

 and phosphate and macroplankton. Leavitt 

 (1938) found in the Atlantic Basin a correlation 

 between temperature, salinity, and density, and 

 the vertical distribution of zooplankton; a nega- 

 tive correlation was found between oxygen and 

 zooplankton. Graham (1941) reports that differ- 

 ences in productivity in the Pacific, as measured by 

 plankton dry weight, are correlated with the con- 

 centration of phosphate. 



COMPARISON WITH ZOOPLANKTON ABUNDANCE 

 OF OTHER REGIONS 



To facilitate the comparison of our data with 

 that of other regions, we have calculated average 

 numbers and volumes (table 11) which are repre- 

 sentative of the quantity of zooplankton in differ- 

 ent latitudinal zones of the central Pacific. It is 

 a difficult task, however, to reduce to comparable 

 terms plankton data which have been obtained by 

 different investigators, using different methods, at 

 different seasons, and with results expressed in dif- 

 ferent units. In surveying a considerable fraction 

 of the great bulk of available literature on quan- 

 titative plankton sampling, we found only a few 

 reports that have enough in common with our own 

 work to permit a comparison of the results. 



In Graham's report (1941) on plankton col- 

 lected by the Carnegie along a series of stations 

 extending from San Francisco to Samoa, he states, 

 "In the open Pacific Ocean from September to 

 November 1929, there was a greater production of 

 total plankton in the tropics between latitudes 

 20° N. and 11° S. than between latitudes 20° and 

 34° N." The richest collection of this section (fig. 

 10) was taken at 13° N. latitude, but was consid- 

 ered atypical since it consisted largely of salps. 



