FISHERY BULLETIN: VOL. 86. NO 1 



due to the increased numbers of Group II species 

 and decreased numbers of L. parvicauda . The 

 1970-83 averaged abundance of L. parvicauda is 

 significantly lower (Z test; P < 0.01) than that of 

 1964-69. Triphoturus mexicanus averaged abun- 

 dance (both adjusted and unadjusted values) 

 is similar (P > 0.05) between the two time peri- 

 ods. 



DISCUSSION 



The northern Chilean ichthyoplankton data set 

 is obviously weakened by lack of information 

 from the 1974-82 period; this missing informa- 

 tion is critical for an appreciation of the temporal 

 extent and relative constancy of the apparent 

 ichthyoplankton composition change in 1970-73 

 vs. earlier years. This data set also suffers from 

 limited seasonal coverage which prohibits exami- 

 nation of between-year variations in spawning 

 time and intensity as a cause of interannual 

 abundance fluctuations and apparent composi- 

 tion change. However, the existing data set does 

 provide coherent coverage over varied hydro- 

 graphic conditions between 1964 and 1973 and is 

 sufficient to test for correlations with short-term 

 (e.g., year to year) fluctuations in hydrographic 

 conditions. 



The large interannual changes in abundance 

 and composition of the northern Chilean ichthy- 

 oplankton can to a certain extent be related to 

 interannual changes of hydrographic conditions 

 in the Humboldt Current. This has been demon- 

 strated through correlations of ranked tempera- 

 ture and salinity values and abundances of 

 coastal species, sardine, and Lampanyctus parvi- 

 cauda (Tables 9, 10). The temperature and salin- 

 ity values used in these correlation tests 

 represent ambient conditions during the July- 

 September spawning period and therefore 

 possibly reflect only conditions affecting egg and 

 early larval (e.g., to stages capable of substantial 

 net avoidance) survival. These values do not nec- 

 essarily reflect longer term conditions affecting 

 abundance, distributions, and fecundities of adult 

 populations or later larval survival and recruit- 

 ment. However, there is a generally good corre- 

 spondence between these values and reported 

 longer term hydrographic conditions in the Hum- 

 boldt Current over the 19-yr timespan (e.g., 

 Table 8; Robles et al. 1976; Bernal et al. 1983; 

 Guillen 1983; Bakun 1987). 



Despite significant correlations between abun- 

 dances of some ichthyoplankton components and 



temperature and salinity values, there is no ap- 

 parent consistency of total larval or OL species 

 percentage compositions during years of 

 "similar" hydrographic conditions (Table 11). 

 More coherent patterns emerge from consider- 

 ations of the 1964-69 and 1970-73 data sets 

 (Table 12). This chronological separation is also 

 supported by the ranked abundance patterns of 

 the various species groups and pairs {Table 7). 



Among the least confusing across-year abun- 

 dance patterns demonstrated by the ichthy- 

 oplankton are 1) generally greater abundance 

 of Group II species after 1969, 2) greatest abun- 

 dance of Pair I species after 1970, and 3) pre- 

 dominantly higher abundances of one of the Pair 

 II species prior to 1970 (Table 7). Associated with 

 the Group II and Pair II abundance patterns are 

 large shifts in their relative proportions 

 (Table 13). 



The shift from relatively large percentage con- 

 tributions by Lampanyctus parvicauda and 

 Triphoturus mexicanus to larger proportions of 

 Diogenichthys spp., Bathylagus nigrigenys , and 

 Vinciguerria lucetia after 1969 is notable. The 

 abundances of these mesopelagic species, unlike 

 those of anchoveta and sardine, are not directly 

 influenced by man's fishing activities and so may 

 be interpreted as indicators of environmental 

 change. Furthermore, the timing of these species 

 absolute and relative abundance changes pre- 

 ceded by several years the dramatic changes in 

 anchoveta and sardine stocks off of northern 

 Chile (Fig. 4) and so cannot be directly related to 

 biological consequences of change in the domi- 

 nant pelagic schooling fish stocks. 



Although fragmentary, there is evidence for a 

 change in zooplankton biomass values off north- 

 ern Chile (18°-24°S) occurring in 1969 (Fig. 5) 

 which, like OL percentage composition, suggests 

 a possible environmental change. Time series 

 analysis of quarterly zooplankton biomass values 

 during 1964-73 indicate generally lower biomass 

 during 1969-73 relative to the 1964-68 period. 

 As with total larval abundance (Table 9), these 

 zooplankton biomass variations do not appear to 

 be related to warm year-cold year events (Bernal 

 et al. 1983). 



The changes in OL composition and zooplank- 

 ton biomass suggest that there was subtle but 

 large-scale (low-frequency) environmental tran- 

 sition occurring in the 1969—70 period. Various 

 indications of environmental change occurring 

 about this time are present in long-term physical 

 data bases from Chile and Peru. Predominantly 



18 



