FISHERY BULLETIN: VOL 86, NO. 1 



Table 2. — Catch comparisons of 50-0 m and 100-0 m Hensen net tows taken at lat. 28 -38 S, long. 71 -74'W during 

 3-17 December 1969. Abundances expressed as mean and standard deviations of numbers per 10 m2, F = percent 

 frequency of occurrence in samples Depth-related catch differences are shown as ratios of 50:100 m abundance 

 estimates, species percentage contribution to total identified larvae, PL and OL percentage contribution to total larvae, and 

 taxonomic diversity (mean numbers per tow and total numbers of taxa). N = number of samples. PL = larvae of pelagic 

 species; OL = other larval taxa. 



canus which was 2.4 x more abundant and 1.5 x 

 more frequent in 100 m tows. Because of the large 

 catch variability, none of the species abundance 

 differences nor the abundance differences of the 

 PL, OL, and total larvae are significant (Z tests, 

 P's all >0.10). Additionally, species abundances 

 within all positive tows from the two sampling 

 depths are not significantly different (Mann 

 Whitney U tests, P > 0.10 in all cases). 



The overall species composition of 50 and 100 m 

 tows was similar. Despite greater proportions of 

 PL in 50 m tows, the PSI value from comparisons 

 of total species lists was high (87.7). Species per- 

 centage contribution within the OL fraction of the 

 two tow types was also quite similar (94.5). Spe- 

 cies abundance rankings within the two total lar- 

 val data sets are significantly correlated 

 (p = +0.80; P < 0.01 ). Species diversity estimates 

 (total numbers of taxa and mean numbers of taxa 

 per tow) are also similar. 



From these comparisons it is apparent that the 



PL predominantly occur within the upper 50 m. 

 Similar shallow (e.g., <50 m ) distributions have 

 been described for dominant PL species off of Peru 

 (anchoveta, sardine, and hake; Sameoto 1982). 

 The generally lower 100 m abundance estimates 

 of these species is puzzling, but suggests possi- 

 bly shorter sampling time and/or less efficient 

 sampling within the upper 50 m of these tows. 

 Higher catch frequency and abundance of T. 

 mexicanus in 100 m tows suggest that large pro- 

 portions (e.g., 30-60%) of these larvae are at 50- 

 100 m. 



As a result of these catch differences we suggest 

 caution in making direct numerical comparisons 

 between the 1973 and 1983 vs. earlier data sets. 

 Although the overall compositions and abun- 

 dance relations should not be markedly altered, 

 some accommodation should be allowed for the 

 percentage contributions and across-year abun- 

 dance ranks of PL species (especially Trachurus) 

 and Triphoturus mexicanus. 



