706 



Fishery Bulletin 98(4) 



Results 



First increment deposition 



Only thirteen larvae from the rearing experiments 

 in the laboratory were analyzed for determination of 

 age at the first increment deposition, and because of 

 the low number of larvae, the data from the treat- 

 ments were analyzed together These preliminary 

 experiments showed that no lai-vae formed the first 

 ring on the otoliths at hatching (two lai-vae were ana- 

 lyzed at hatching). Although the first ring on the oto- 

 liths was first observed in a larva at end of the first 

 larval day after hatching (two larvae were analyzed), 

 most larvae formed their first ring during the third 

 day after hatching (four out of six larvae, 3-days-old 

 or older). The larva analyzed at the end of the fourth 

 day had two rings deposited; therefore it formed its 

 first ring during the third day. At the end of the third 

 day after hatching, all larvae had deposited their first 

 ring. The mean diameter of the otolith focus for those 

 larvae hatched in our experiments was 10.23 (SD=1.25) 

 Jim and the mean diameter of the otolith with their first 

 increment formed was 12.83 (SD=1..56) pm. 



The relation between the number of rings and otolith 

 diameters from anchoveta larvae collected in the wild 

 was well described by a linear regression (r2=0.90; /!=112; 

 P<0.001)(Fig. 2). From this relationship, a lai-val ancho- 

 veta with one ring should have an estimated otolith diam- 

 eter of 12.6 micrometers, which is close to the mean otolith 

 diameter measured from larvae reared in our laboratory 

 (mean=12.83 pm, SD=1.56). A power model was used to 

 describe the relation between otolith diameter and larval 

 length (corrected for shrinkage after Theilacker 1980) of 

 wild lai-vae (;-2=0.79; ?? = 112; P<0.001i (Fig. 3). According 

 to these two relationships, a larva with one increment 

 (12.6-pm otolith diameter) should have a larval length of 

 5.6 mm. 



Larval growth of larvae caught in the wild 



Because deposition of the first nng in most larvae occurred 

 during the third day after hatching, age of lai-vae caught in 

 the wild was estimated by adding two days to the number 

 of rings in their otoliths. The standard lengths of larvae 

 used in the growth models varied between 5.68 mm and 

 20.74 mm, and their age ranged from 3 to 35 days. 



The three gi-owth models used to describe anchovy lai-val 

 growth during the winter of 1995 were highly significant 

 (P<0.001) (Fig. 4, Table 1). The linear model yielded a daily 

 growth rate of 0.47 mm/d (n=112 larvae; r2=0.82); the Gom- 

 pertz model yielded a mean daily growth rate of 0.50 mm/d 

 (n=\12 larvae; r'--'=0.84); and the von Bertalanffy model 

 yielded a mean daily growth rate of 0.48 mm/d (n=l 12 larvae; 

 r2=0.84). Estimated larval length at the end of the third day 

 after hatching (day of the first ring deposition) varied from 

 7.81 mm with the linear model to 7.17 mm with th(> Gomp- 

 ertz model and 6.13 mm with the von Bertalanfl'y model. 



The results of the analyses of growth rates between 

 groups of cohorts spawned early versus late in the winter 



season varied according to the models used to describe 

 growth. The lai-val growth rates determined from the 

 linear model revealed that the group of daily cohorts 

 spawned earlier in the season grew slower (1 Jul-17 

 Aug: 0.40 mm/d; ;!=63) than the group of lai-vae spawned 

 later in the season (18 Aug-11 Sep: 0.57 mm/d; n=49) 

 (ANCOVA, F=26.2, « = 112, P<0.001; Zar, 1984). For both 

 groups of cohorts, the linear regressions used to describe 

 lai-val growth were significant (P<0.001) and explained 

 over 78% and 749; of the variance in the respective data 

 sets. A comparison of lan,'al growth rates between periods 

 using the Gompertz and von Bertalanffy models, how- 

 ever, showed no differences between larvae spawned early 

 versus lar\'ae spawned later in winter (Gompertz: P=1.08. 

 /( = 112, P>0.05; von Bertalanffy: F=0.82, « = 112, P>0.05; 

 analysis of residual sum of squares, ARSS; Chen et al., 

 1992). 



