SIX and HORTON: ANALYSIS OF AGE DETERMINATION METHODS 



Age composition data were described graphi- 

 cally by FISHPLOT, a computer plotting routine 

 based on the method of Hubbs and Hubbs (1953). 

 Survival estimates were obtained by the 

 Chapman-Robson (Robson and Chapman 1961) 

 and the catch curve (Ricker 1975) methods. The 

 age-length relationship of yellowtail rockfish was 

 described by the equation L - cA h , where L = 

 length (centimeters), A = estimated age (years), 

 and c and b are constants. The age-length 

 relationships for canary and black rockfish were 

 described by the von Bertalanffy growth-in- 

 length equation with the computer program BGC- 

 2 (Abramson 1965) using the method of least 

 squares weighted according to sample size (Tom- 

 linson and Abramson 1961). 



A total of 71 young unsexed black rockfish, 

 mostly young-of-the-year, were used in the age- 

 length analysis. Their corresponding lengths 

 were applied to both males and females, with the 

 assumption that there were little or no sexual 

 differences in length at these younger ages. The 

 assumption was based on the fact that growth 

 curves for male and female Pacific ocean perch, 

 obtained by Westrheim (1973) for fish from 

 Oregon to British Columbia and by Gunderson 

 (1974) for Washington samples, were nearly 

 identical at ages less than 6 yr. 



RESULTS AND DISCUSSION 



Suitability of Structures for 

 Age Determination 



Only 5 of 25 anatomical structures sampled 

 were suitable for estimation of age. These were 

 the anal fin pterygiophore, opercle, otolith, scale, 

 and vertebra. The criterion used to determine 

 suitability for aging was the presence of enumer- 

 able growth zones. Based on examination of a 

 limited sample, most structures did not satisfy 

 this criterion because: 1) they were not suf- 

 ficiently calcified to reveal distinct growth zones, 

 or 2) calcification was evident but growth zones 

 were not discernible (Table 1). The above five 

 structures were examined further to determine 

 whether successive, independent estimates of age 

 were consistent. 



Consistency of Readings 



Percent agreement between two independent 

 counts (readings) of assumed annuli by the same 



person on anal fin pterygiophores, opercles, 

 otoliths, scales, and vertebral centra sampled 

 from the same yellowtail, canary, or black 

 rockfish is presented in Table 2. Exact agreement 

 ±1 assumed annulus is also given. Agreement 

 was low for all structures and species except oto- 

 liths of canary rockfish. Agreement between oto- 

 lith readings for yellowtail and canary rockfishes 

 was superior to agreement between readings of 

 other structures, with 71 and 97% agreement ±1 

 assumed annulus, respectively. For the sample of 

 black rockfish, otoliths and opercles were equally 

 readable with 74 and 75% agreement ± 1 assumed 

 annulus, respectively. 



Means of the two readings of the five structures 

 agreed fairly well for black rockfish, indicating 

 that counts of assumed annuli on the structures 

 were similar. Means were not similar for these 

 structures from yellowtail and canary rockfishes. 



A number of samples of each structure were not 

 read due to crystallization and breakage of 

 otoliths, regeneration of scales, and poor calcifi- 

 cation of opercles and pterygiophores. Throughout 

 the entire study at least one of the two otoliths was 

 partially or completely crystallized in 23 of 1,116 

 (2.1%) yellowtail rockfish, 27 of 666 (4.1%) canary 

 rockfish, and 29 of 302 (9.6%) black rockfish. 

 There were more readable vertebral centra and 

 otoliths than any of the other structures. Many 



TABLE 2. — Estimations of age, number of readable structures, 

 and percent agreement of two independent readings of five 

 structures sampled from 35 yellowtail rockfish, canary rockfish, 

 and black rockfish landed off Newport, Oreg., 1974-75. 



407 



