for possible relationships between the geograph- 

 ical distributions of the taxonomic categories, 

 these data were subjected to a principal compo- 

 nents analysis. 



Figure 2 is a graph of the first latent vector 

 plotted against the second. The graph has a point 

 for each taxonomic category, and the disposition of 

 points represents in a spatial form the relation- 

 ships between the geographical distributions of 

 the taxonomic categories with respect to the first 

 two components which, in this case, account for 

 61% of the variability of the original geographical 

 distributions. The interrelationships are probably 

 best regarded in the form of a more or less circular 

 sequence; only the point for Medusae falls well off 

 the sequence. 



Figure 3 shows charts of the first two compo- 

 nents. The first component shows a very clear 

 north to south, alongshore gradient; and the 

 second shows an equally clear inshore to offshore 

 gradient, indicating that the sequence of cate- 

 gories in Figure 2 runs from categories with 

 northern distributions (Siphonophora to Radio- 

 laria) to inshore distributions (Euphausiacea to 

 Cladocera) to southern and inshore distributions 

 (Larvacea to Mysidacea) to offshore distributions 

 (Heteropoda to Ostracoda). Figure 4 shows the 



(ONSHORE) +0.5 



CRST 



-0.4-1- „CHET COPD 



CLAD. 



+ 0.3- - 



LARV 



DECP 



MYSD 

 (SOUTH) 

 -04 -03 -02 -0.1 



+ 0.2-- 



+ 0.I-- 



H 



-+- 



■+■ 



-+- 



HETP 

 PTER 



-0.2- - 



(OFFSHORE) -04-L 



EUPH 



(NORTH) 

 RADL 

 + 0.1 +0.2 +0.3 * 



-+- 



-+- 



MEDS 



1 amph* ¥ i 



CTEN 

 THAL 



SIPH 



3- • 



OSTR 



FIGURE 2. — A plot of the first vector against the second vector 

 derived from a principal components analysis of the geographi- 

 cal distributions of the taxa. A key to the abbreviations of the 

 names of the categories is given in Table 1. 



FIRST COMPONENT 



FISHERY BULLETIN: VOL. 75, NO. 2 



SECOND COMPONENT 



FIGURE 3. — Charts of the first and second components derived 

 from a principal components analysis of the geographical dis- 

 tributions of the taxa. 



distributions of the taxonomic categories ar- 

 ranged in this sequence. They are based on 

 averages of the transformed data, for each zone, 

 for each quarterly cruise for the period 1955-59, 

 excluding zones for which fewer than five stations 

 were occupied. These distributions show varia- 

 bility other than that involved in their relation- 

 ships with the first two components; nevertheless, 

 the north to inshore to south to offshore sequence 

 can be seen fairly clearly. Heteropods and 

 Pteropods are firmly placed in the sequence of 

 taxonomic categories in the vector plot in Figure 

 2. They have, however, fairly low values compared 

 with the other categories, and only parts of their 

 distributions conform with the south to offshore 

 transition indicated by their position in the vector 

 plot. The distribution of Medusae (Figure 4) can be 

 seen to include areas of relatively high biomass 

 both in the north and in the south, and clearly it 

 does not fit into the sequence of the other 

 categories. 



It is obviously unrealistic to attempt to classify 

 the internally diverse taxonomic categories used 

 here in terms of geographical distribution types 

 such as Brinton (1962) found for Euphausiacea. 

 Brinton found that the alongshore axis of the 

 California Current in the CalCOFI survey area 

 was characterized by transitions from "subarctic" 

 species in the north to "transition" species in the 

 region between lat. 30° and 40°N to "equatorial" 

 species in the south. "Central" species occurred 

 offshore and some "boundary" species occurred 

 inshore in the area. McGowan (1971) has shown 



360 



