444 



FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 



We had an opportunity to examine critically the 

 nature of the accessory annulus while doino: some 

 scale back-calculations for another study. We 

 examined 513 scales collected in Februai-y and 

 March, 1956 and 1957, from Geoi-ges Bank had- 

 dock, 3 to 6 years old. The scales had been col- 

 lected as part of the routine port sampling pro- 

 gram and represented a typical sample of scales 

 used in our age determination procedure. In the 

 sample of 513 scales, 245 (47.8 percent) showed an 

 accessory annulus. A random subsample of 38 

 scales was taken for a more detailed study of the 

 accessory annulus. Back calculations were made 

 on the subsample to obtain these data: (1) Li — 

 calculated fish length at the end of the first year ; 

 (2) La — calculated fish length when the accessory 

 annulus first began to fonii; (3) Lb — calculated 

 fish length when the accessory annulus was com- 

 plete; (4) Ca — number of circuli from the focus to 

 the fii-st narrow circulus in the accessory annulus; 

 and (5) Cb — number of circuli from the focus to 

 the last narrow circulus in the accessory annulus 

 (see table 2). 



The accessory annulus starts to form when the 

 haddock attain an average length of 115 mm. and 

 is complete when the haddock attain an average 

 length of 139 mm. The accessory annulus consists 

 of about seven closely spaced circuli within a zone 

 of wideh' spaced circuli and includes the 10th 

 through the 16th circulus from the focus. 

 Broadly spaced circuli, denoting rapid growth, 

 are found on the scale after the accessory mark 

 and are in turn followed by a zone of narrow 

 circuli. This zone of narrow circuli is the first 

 true annulus and is considered complete on 

 Februaiy 1, altliough a few additional narrow 

 circuli are formed in Februaiy and Alarch. The 

 accessory annulus usually is well separated from 

 the first true annulus and with training and expe- 

 rience can be readily recognized by the scale 

 reader. 



Another sotirce of error may arise from mechan- 

 ical damage to the scale. The exposed posterior 

 edges of some scales, particularly from fish more 

 than 7 years old are ragged and appear to be 

 eroded, suggesting that some annuli may be 



Table 2. — Calculated fish length-scale length relationship 

 for accessory mark on Georges Bank haddock scales 



Confirmation of Ages from Length Distributions 



Comparing tlie ages read from scales with the 

 modal peaks in a length distribution of a sample 

 of fish (the Petersen method) helps to confirm 

 the scale method and particularly helps to estab- 

 lish the time of fonnation of the first true annulus. 

 Haddock scale readings were thus confirmed by 

 Duff (1916) for haddock from St. Andrews, by 

 Saetersdal (1953) for Norwegian haddock, and by 

 Schuck and Arnold (1951) for haddock from 

 Browns and Georges Banks. 



Figure 3 shows mean lengths of young-of-the- 

 year haddock collected by us in 1957 on Georges 

 Bank and in the Gulf of Maine. The tenn 

 "young-of-the-year haddock" includes those fish 

 that have not yet completed the first year of life. 

 As explained previously, the year is considered 

 complete on February 1. Small haddock, such 

 as are repi'esented in these data, are difficult to 

 collect in sufficient quantities at one time for 

 detailed length-distribution analysis. Therefore, 

 the various collections were incorporated into one 

 series of data in table 3. 



Collections of larval fish, made aboard the 

 Albatross III in April 1957, contained haddock 



