Wexler et al Temporal variation in larval growth of Thunnus albacares in the Panama Bight 



w 



 June 1990 (n=25) A September 1991 (n=43) 

 V'=2.98e0i06X ?= 3,66 eO 061 'f 



r2 = 0.86 r2 = 0.59 



A July 1 991 (n = 1 3) o June-July 1 992 (n = 22) 



D August 1997 (n=69) 



/= 4,44 e 063X 



r2 = 0.68 



Age (d) 



Figure 4 



Exponential relationships between standard length and age in days estimated from 

 otolith increment counts of larval yellowfin tuna tThunnus albacares) collected 

 during June 1990, September 1991, and August 1997. Data are also presented 

 for yellowfin tuna larvae collected during July 1991 and June-July 1992, but 

 growth models were not fitted to the limited range of data. 



1991 (collection group II) and between those of 1992 and 

 1997 (collection groups IV and V). 



The length-at-age data were used to examine and 

 compare growth relationships of all yellowfin tuna lar- 

 vae collected in June 1990, September 1991, and August 

 1997 (collection groups I, III, and V, respectively). An 

 exponential model provided the best fit to each of the 

 three groups of data (Fig. 4). The variances were homo- 

 geneous after log transformation of the length data for 

 each group, and the slopes were compared. The slope 

 and the average growth rate obtained through differ- 

 entiation of the exponential equation of the 1990 data 

 (1.28 mm/d, SE = 0.134) were significantly greater than 

 those of the September 1991 (0.60 mm/d, SE =0.033) and 

 1997 (0.71 mm/d, SE = 0.038) data (ANCOVA, P<0.0001, 

 Tukey multiple comparison test). The elevations (i.e., 

 adjusted means or intercepts) of the 1991 and 1997 

 data were significantly different (P<0.0001) and indi- 

 cated that the mean length-at-age was significantly 

 smaller for larvae of the September 1991 collection 

 group (Fig. 4). 



Similar results were obtained for otolith growth rates 

 (based on the exponential relationships between otolith 

 diameter and age) for the three years (Fig. 5). The 

 slopes were compared after log transformation of the 

 otolith data for each group. The growth rate of the 1990 

 data was significantly faster than that of the 1991 and 

 1997 data (ANCOVA, P<0.0001, Tukey multiple com- 



parison test), and the elevations were different between 

 the 1991 and 1997 data, indicating that otoliths were 

 significantly smaller for larvae of the September 1991 

 collection group (Fig. 5). 



A comparison of the linear relationships between oto- 

 lith diameter and SL (ANCOVA, P<0.0001, Tukey mul- 

 tiple comparison test; Fig. 6) revealed that the otoliths 

 of the 1997 group were larger and grew significantly 

 faster in relation to fish size than those of the 1990 

 and 1991 groups. Otoliths of the fastest (1990 group) 

 and slowest (1991 group) growth periods were grow- 

 ing at the same rate in relation to fish length, but the 

 slower-growing group had significantly larger otoliths 

 in relation to fish size than those of the faster-growing 

 group (P<0.0001; Fig. 6). 



Standing stocks of ichthyoplankton 



Ichthyoplankton and physical parameters were mea- 

 sured at four stations along the P and M transects on 

 two sampling days in June 1990 (16 tows total, each 

 to 50 m), at the MSB station on the M transect on six 

 sampling days in June and July 1991 (12 tows each depth 

 strata 1-4), at the MSB station on three sampling days 

 in August 1991 (six tows each strata 1-4 and four tows 

 each strata 1-3), and at the MSB station on two sam- 

 pling days in June and July 1992 (six tows each strata 

 1-3) (Table 3, Fig. 1). The sampling days lay within the 



