34 



Fishery Bulletin 105(1) 



o 

 O 



length (mm) length (mm) 



Figure 2 



(A) Observed standard length of delta smelt iHypomesus transpacificicus) plotted 

 against ages with respective linear regression line; (B) observed otolith radius 

 plotted against age sampled with respective linear regression line; (C) known-age 

 plotted against observed standard length with linear regression line; (D) observed 

 otolith radius plotted against observed standard length with linear regression line. 

 Ontogenetic stages are represented by different symbols: black circles represent 

 yolksac-stage larvae (•), gray circles represent feeding larvae (•), gray triangles 

 represent finfold-stage larvae (A), black triangles represent flexion-stage larvae 

 (▲I. open circles represent postflexion-stage larvae (O) and black squares represent 

 juvenile-stage fish (■). 



Table 3 



Comparison of the allometric slope (6) values (log {otolith 

 radius) = bx\og (standard length). P values are from Stu- 

 dent's ^-statistic. 



VI 



radius-on-age was 0.885 (P= 0.001) (Fig. 3A). Comparison 

 of the residuals of otolith radius-on-length and age-on- 

 length demonstrated that there was no significant growth 

 rate effect on the OS-FS relationship (Fig. 3B). The Pear- 

 son correlation coefficient was 0.068 (P=0.244). 



A simple log transformation of an allometic model 

 demonstrated ontogenetic differences in the slopes of 

 the OS-FS relationship (Fig. 4). Slopes were signifi- 

 cantly reduced for the newly hatched yolksac stage 

 (I) and the larval flexion stage (IV). However, stage-I 

 larvae were the same age; therefore variability due to 

 ontogeny may be difficult to evaluate. Variability in fish 

 length for stage-IV larvae may reflect subtle ontogenetic 

 differences in caudal fin flexing and thus variation in 

 the OS-FS relationship (Table 3). 



