812 



Fishery Bulletin 101(4) 



Regular 

 rings 

 (141) 



Central 

 rings 

 (14) 



Nucleus 



B 



Central 



Regular 

 nngs 



1 1 null I I I I 



1 1 mill I I I I 



1 1 iiiiii I I I I 



1 1 mill I I I I 



I I I I I I I I I I I I I I 



I I I I I I I I I I I I I I 



II I I I I I I I I I I I I 

 II I I I I I I I I I I I I 



Nucleus «r 



Cut plane 



Figure 3 



(A) Photomicrograph in optical microscopy of the posterior radius of a polished otolith 

 from a 180-day-old juvenile white mullet iMugil curema). Different sequences of rings are 

 indicated which result from a horizontal plane not aligned with the changing growth plane 

 of the otolith. Numbers of rings in each sequence are shown in parentheses. (Bi Schematic 

 presentation of the changing growth plane of the otolith, the cut plane, and the resultant 

 increment distortion (black rectangles represent the distance between adjacent rings). 



less consistent in the anterior field, otolith counts were 

 always made along the posterior radius of the sagittae. We 

 also measured the area of the nucleus, which represents 

 the prehatch zone. All measurements were made under a 

 microscope which was connected to an image analyzer and 

 computer. 



To evaluate the error in counting growth increments 

 on otoliths, one reader made 7 independent counts of the 

 number of growth increments on otoliths obtained from 

 11 juveniles representing the size-range of sampled fish. 

 The number of growth increments ranged from 63 to 289 

 (mean=147) and the mean coefficient of variation was 

 8.71% (SD=2.13%) (Fig.4). We therefore considered an 

 error of approximately 10% for the counts of growth incre- 

 ments. In applying the technique to the subsamples of the 

 different cohorts sampled in the lagoon, at least two counts 

 were made for each otolith. All counts were made by the 

 same person. 



To evaluate if otolith growth increments were formed 

 daily, we read the otoliths of juveniles sampled on succes- 



sive sampling dates and compared the average increase 

 in the number of otolith increments to the number of days 

 between samplings (Struhsaker and Uchiyama, 1976; Jor- 

 dan, 1993; Jenkins and May, 1994). Birthdate was obtained 

 by subtracting the number of daily growth rings on otoliths 

 from the date of capture. We used the hatching mark as 

 defined for M. cephalus (Radtke, 1984) and M. iso-iuy {Li 

 et al., 1993) to locate the hatching mark on the otoliths of 

 the white mullet. 



Knowing that white mullet embryos hatch from 24 to 

 40 hours after fertilization (Anderson, 1957; Houde et al., 

 1976), we back-calculated hatching dates of recruits to es- 

 timate when successful spawning occurred. We examined 

 the relationship between the spawning dates of recruits 

 and an index of the intensity of upwelling. In calculat- 

 ing this index, we determined wind stress based on data 

 fnmi Fundacion La Salle, Margarita Island, and Cumana 

 Airport meteorological stations. The upwelling index (UI) 

 was based on Bowden's (.1983) theoretical calculations as 

 follows: 



