LOEB and ROJAS: ICHTHYOPLANKTON COiMPOSITION AND ABUNDANCE 



Table 1. — Cruises yielding samples used for examination of ichthyoplankton abundance and composition variations 

 oft northern Chile Only samples from 1 8 -24 S are used for interannual comparisons. Data from two 1 968 cruises are 

 combined; data from cruise 71(4)69CD are used for analysis of sampling depth-related catch differ- 

 ences. N = number of samples used in ichthyoplankton analyses Tow types: H = Hensen net; WP2 = WP2 net; 

 V = vertical. 



aged 28.3 m'^ yielding a raw count to numbers per 

 10 m'- conversion factor of 35; this 179f increase in 

 conversion factor is associated with a 12*7^ de- 

 crease in mouth opening of WP2 vs. Hensen nets. 

 Larval fish diversity is expressed as total num- 

 bers of taxa per sampling period and mean num- 

 bers of taxa per tow. 



Various parametric and nonparametric tests 

 were used for statistical analyses. Differences in 

 mean abundances are tested with 2-tailed Z tests 

 and Mann Whitney U tests (Dixon and Massey 

 1969; Conover 19711. Similarity of abundance 

 ranks within data sets are tested with Kendall's 

 concordance iW) test (Tate and Clelland 1969) 

 and Spearman's rho (p) correlation test (Conover 

 1971). Significant values resulting from these 

 tests are indicated, but due to multiple testing 

 these values should be used only as indicators of 

 the relative strengths of relationships. Percent 

 similarity indices (PSI; Whittaker 1975) are used 

 for comparisons of species percentage composi- 

 tion. Because PSIs are strongly influenced by 

 abundant species, we apply these tests to the OL 

 fraction as well as to total larvae. We define as 

 "high" all PSI values >80, as "moderate" PSI's 

 >65 and <80, and as "low" values <65. 



SAMPLING CONSIDERATIONS 



Sampling Depth Differences 



The 100 m sampling depths in 1973 and 1983 

 potentially effect direct comparisons of abun- 

 dance estimates and species composition in these 

 vs. earlier data sets owing to individual species' 

 depth distributions. Evaluation of depth-related 

 sampling differences is possible through a com- 

 parison of data obtained from coincidental 50-0 

 m and 100-0 m Hensen net samples taken during 

 December 1969 at 28°-38"S (Table 1). These data 

 indicate that abundance estimates and percent- 

 age contribution of the PL to total larvae were 

 higher in 50 m tows while those of the OL were 

 higher in 100 m tows (Table 2). Four of five PL 

 species caught were more abundant and frequent 

 in the 50 m samples. The greatest differences 

 were for Trachurus murphyi, which was 5x more 

 abundant and 3.6 x more frequent, and Merluc- 

 cius gayi , which was present only in 50 m sam- 

 ples. Abundance estimates were higher for 11 of 

 16 OL taxa in 100 m tows although, with a few 

 exceptions, catch frequencies were similar. The 

 greatest difference was for Triphoturus mexi- 



