118 



Fishery Bulletin 103(1) 



Oomo D P au 



Outjer 

 Oog\ 



Aroc' 



. 4hep~ 



, Mame 



Inner 



"D 



3 

 CO 



Oomi 

 Omi' 

 Mid /Wura 



B 



<rOo§p/ 



"^Mame 



Inner 



c 



3 

 3 



CD 



< 



Outer 



Oomo* 



Dbau* 

 Ecro' 



Outer 



Inner 



Omar/ 

 Ahep • fWame 

 Lxan 



Mid 



"n 



0) 



D 



zBoce 

 Oomp* 

 ?Dpau 



Mid 



;. 



Cspi 



F 1/Bf 



'B(yr 



CA 1 



Figure 5 



Correspondence analysis (CA) ordinations (portraying the first 

 and second dimension scores) of the larval fish community data 

 showing species in each season: (A) spring, (B) summer, (C) fall, 

 and (D) winter. A larval fish assemblage was associated with 

 each cross-shelf station group. Each station group is outlined 

 and labeled as in Figure 4. The dashed lines intersect at the 

 origin of the plot. Analyses were conducted by using larval 

 concentration data only. Refer to table 2 for definitions of larval 

 taxa codes. Three larval fish assemblages were defined based 

 on species association with station groups (see table 5). 



found primarily at outer-shelf stations (Fig. 6). Auxis 

 rochei and Bothus ocellatuslrobinsi [where the slash (/) 

 means "B. ocellatus and B. robinsi" or one of these spe- 

 cies] represented the outer-shelf assemblage (Table 5). 

 The region of the shelf with the highest species rich- 

 ness depended on the inclusion of rare taxa and season. 

 With the exception of fall, species richness was highest 

 in the mid-shelf group when only abundant taxa were 

 included in analyses (Table 5, Fig. 7A). When rare taxa 

 were included (the 1% data set), species richness was 

 highest in the mid-shelf group during spring and sum- 

 mer and highest in the outer-shelf group during fall 

 and winter (Fig. 7B). 



Relationship among cross-shelf patterns in 

 the larval fish community, larval assemblages, 

 and environmental variables 



Five environmental variables were correlated to the cross- 

 shelf pattern in station groups and larval assemblages. 

 Water density, salinity, temperature, depth, and strati- 

 fication of the water column had a significant relation 

 to the structure of larval assemblages and the grouping 

 of stations in the CCA (P<0.05 for each variable, Monte 

 Carlo permutation test; Table 6). The species-environment 

 correlation for the first two axes of the ordination was 

 greater that 0.79, indicating a strong association between 

 the environment and larval assemblages (Table 6). 

 Although the portrayal of station groups and larval 

 assemblages in ordination space was not identical when 

 environmental data were included (compare Figs. 4 and 5 

 to 8), the cross-shelf pattern in station groups and larval 

 assemblages was maintained (Fig. 8). 



The first CCA dimension, in all seasons, was most 

 highly influenced by the depth, temperature, salinity, and 

 density of the water (Fig. 8). In spring, summer, and win- 

 ter, the mid- and outer-shelf stations were aligned along 

 CCA 1 and separated from the inner-shelf stations along 

 this gradient (Fig. 8). Similarly, in fall, the three station 

 groups were arranged separately along this gradient 

 with the mid-shelf groups intermediate to the inner- and 

 outer-shelf stations. Thus, the separation between inner- 

 shelf and mid- and outer-shelf stations is related to a 

 gradient in depth, temperature, salinity, and density. 



The second dimension separated outer-shelf stations 

 from inner- and mid-shelf station groups. In spring and 

 summer, the second dimension (CCA 2) was clearly influ- 

 enced by stratification (Fig. 8). The outer-shelf stations 

 experienced a higher degree of stratification, separating 

 them from the inner- and mid-shelf stations. During fall 

 and winter, stratification still impacted the second di- 

 mension, but less dramatically. In summary, outer-shelf 

 stations were distinguished from mid- and inner-shelf 

 stations by increased stratification of the water. 



Relation between larval assemblages and 

 water mass distributions 



When hydrographic variables were combined to define 

 water mass, a possible explanation for the cross-shelf 



