FISHERY BULLETIN: VOL. 80. NO. 2 



complete age distribution of 3-increment larvae 

 could be determined. If the range of ages of 3- 

 increment larvae is similar to the 2-increment 

 larvae, then the estimated age of 3-increment 

 larvae would range from 16 to 28 d with a mid- 

 date of 22 d from hatch. 



Otolith Growth 



The growth and morphology of young herring 

 otoliths has been described previously by Hempel 

 (1959) for specimens ranging in length (total) 

 from 25 to 130 mm collected in the German 

 Bight, by Watson (1964) for Maine herring of 85- 

 285 mm TL, and by Messieh (1975) for Bay of 

 Fundy herring of 32-118 mm TL. Here we de- 

 scribe the growth and morphology of herring 

 otoliths (sagitta) in relation to head length for 

 Gulf of Maine-Georges Bank larvae ranging in 

 size from 5.7 mm (hatching) to 35 mm SL (prior 

 to metamorphosis). 



The shape of the larval herring otolith at 

 hatching is essentially spherical having a slight 

 convex distal side and a flat proximal side with 

 three or four furrows radiating from a distinc- 

 tive central core called a primordium (see Fig. 

 2). A slight protuberance is apparent on the 

 anterior edge of the otolith from larvae starting 

 at about 20 mm SL which develops into the adult 

 rostrum. With further elongation along the 

 anterior-posterior axis, the otoliths become gen- 

 erally pear-shaped at metamorphosis (45 mm 

 TL) and attain the typical shape of adult herring 

 otoliths by 75 mm (Messieh 1975). The mean 

 diameter of the nucleus, defined here as the size 

 of the otolith at hatching prior to increment 

 deposition, is 22.5 ^m (1.1 /xm SD) based on the 

 laboratory-reared larvae. The otolith increases 

 exponentially in length (anterior-posterior axis) 

 to a mean size of 456 /um at 35 mm SL based on 

 the composite field data. Successive dark and 

 light layers are deposited around the nucleus as 

 the otolith grows. A single growth increment 

 comprised of a dark plus light band is generally 

 presumed to represent 1 d. The otolith nucleus of 

 the field-caught larvae is readily discernible as 

 its margin is usually darkened to form a nuclear 

 check (see Fig. 2). In some otoliths an additional 

 increment was seen inside of the check. Messieh 

 and Moore 9 also have observed 1 or 2, and some- 

 times up to 5, faint increments inside the nuclear 



check of otoliths from herring larvae collected in 

 the Gulf of St. Lawrence. However, no nuclear 

 check was evident in otoliths of the laboratory- 

 reared larvae and no increments were observed 

 within the defined nucleus. Nuclear diameters of 

 the field-caught larvae all fell within the 95% 

 confidence limit of the mean nuclear diameter 

 determined from the laboratory-reared larvae. 

 Immediately surrounding the nucleus of the 

 field-caught larvae, the first 3-9 growth incre- 

 ments appear to be less well defined than suc- 

 ceeding increments, i.e., lower optical density 

 and thinner in width. Distinctive, darker than 

 normal growth layers were noted across an oto- 

 lith transect but they did not suggest any pattern 

 or complex periodicity as observed by Pannella 

 (1971), nor was there any evidence of subdaily 

 rings as observed for some species by Taubert 

 and Coble (1977), Brothers (1981), and Brothers 

 and McFarland (1981). Scanning electron 

 microscopy techniques will be necessary to re- 

 solve the presence or absence of faint incre- 

 ments. 



The thickness of successive growth increments 

 was measured on otoliths from three field- 

 caught larvae along a posterior radius starting 

 from the nucleus edge (Fig. 4). Measurements 

 were made only through the penultimate incre- 

 ment in each case as the marginal increment was 

 still in the process of formation and could not 

 always be read clearly. Increment thickness 

 ranged from about 0.6 to 2.4 ^m along the radii. 

 All three otoliths show the same general pattern 

 of increment thickness up to about 23 increments 

 where the first 7-10 increments are relatively 

 thin (0.8-1.5 jum) and increase to near maximum 

 thickness (2.3 yum) by about 75 increments. The 

 thickness of the first 3 increments from the lab- 

 oratory-reared larvae was consistently 0.8-1.0 

 /xm, which compares closely with the initial in- 

 crement thickness of the field-caught larvae. 

 Otolith B tends to suggest a prolonged period of 

 relatively thick increments before thinner ones 

 start to be formed, whereas otolith C appears to 

 form thin increments immediately after maxi- 

 mum increment thickness at around 15 incre- 

 ments. The form of the otolith growth curves can 

 be seen more readily in Figure 5 where otolith 

 radii are plotted against the number of incre- 

 ments for the three larvae. These curves suggest 

 that otolith growth is initially slow, increases 



9 S. N. Messieh and D. S. Moore, Marine Fish Division, Fish- 

 eries and Oceans, Bedford Institute of Oceanography, Dart- 



mouth, N.S., Canada, B2Y 4A2, pers. commun. August-Sep- 

 tember 1981. 



192 



