COLLINS ET AL.: AGE AND GROWTH OF KING MACKEREL 



counted with good agreement between readers. The 

 regression of mean ring count (i?) on SL is R = 0.11 

 + 1.56 (SL); n = 29; r^ = 0.73 (siginificant at P < 

 0.001). That r- is not higher is attributed to coarse 

 length measurements (nearest mm). If the rings are 

 daily, the regression of SL on R (SL = 1.15 + 

 0A7(R)) indicates a growth rate of 0.47 mm SL/d 

 for early larvae (Fig. 1). 



Presumed daily ring counts were obtained for 54 

 (78%) of 69 YOY king mackerel 79-320 mm FL. A 

 strong correlation was found for the regression of 

 mean ring count (R) on FL (R = 2.0 + 0.32(FL); 

 n = 54; r- = 0.92, significant at P < 0.001). If 

 these rings are actually daily, the regression of FL 

 oni?(FL = 7.25 + 2.91(i?)) suggests that a growth 

 rate of 2.9 mm/d occurs at 30-100 days of age (Fig. 

 2). Attempts to produce evidence for the daily 

 nature of these rings by measuring diel variation in 

 marginal increments using SEM were not success- 

 ful, perhaps due to inadequate specimen prepara- 

 tion. Rings were normally visible on portions of the 

 lapilli, but we could not consistently read increments 

 near the margin. 



Two readers agreed on annual ring counts for 77% 

 of all whole sagittae and 70% for fish >850 mm FL, 

 resulting in 15 age (= number of rings) classes. Ex- 

 amination of sections made in the four planes veri- 

 fied that sections perpendicular to the long axis of 

 the sagitta were most legible, and no evidence for 

 splitting of rings was found. Agreement on read- 



ing sections was greater than that for whole sagit- 

 tae, with counts verified on 90% of all sections and 

 96% from fish >850 mm FL. The oldest fish aged 

 from sections was age 21. Agreement between the 

 two techniques was but 47% among fish on which 

 both whole sagittae and sections were used, and the 

 ages were significantly different (t test for paired 

 observations: P < 0.001). Counts were very similar 

 for the first three to five age classes, but sections 

 from older fish commonly showed one or more rings 

 not detected on whole sagittae and the difference 

 increased with age. The two procedures differ at an 

 earlier age for males than for females (Fig. 3). 



The correlations of fish length with otolith radius 

 were significant {P < 0.001 for all) for whole and 

 sectioned sagittae of males, females, and sexes com- 

 bined (Table 1). Plots of focus-ring measurements 

 from sections for successive age groups through age 

 5 show that the distribution was unimodal for each 

 increment, that distances to the rings varied little 

 with age, and that overlap increased with age (Fig. 

 4). The pattern for whole sagittae was not quite as 

 well defined (Fig. 5). Back-calculated lengths at ages 

 from whole and sectioned otoliths agree well with 

 observed lengths, especially among (younger) age 

 groups with large sample sizes (Tables 2-7). Annual 

 growth increments from whole and sectioned oto- 

 liths were generally higher for females than males, 

 especially during the first few years of life. Lengths 

 at age determined from whole otoliths were con- 



(0 9 



o 



z 



U- 7 



o 



m 



13 

 Z 



1 1 1 



3 5 



STANDARD LENGTH (mm) 



Figure 1.— Regression of number of presumed daily rings on standard length 

 of larval king mackerel. 



51 



