FISHERY BULLETIN: VOL. 85, NO. 2 



enhances contrast, but does not allow the specimen 

 to be tilted. With SEI, tilting can increase increment 

 relief and visibility. 



Increments deposited during starvation were only 

 0.5 ^m in width, too closely spaced to be discerned 

 with the light microscope (Fig. 3). Additionally, the 

 material that is deposited appears to be more 

 homogenous in density, probably containing a lower 

 amount of matrix. When etched, less material was 

 dissolved in the area corresponding to starvation 

 periods. This resulted in a higher area of relief, form- 

 ing a broad ridgelike structure, subdivided into finer 

 increments. The etching properties were, therefore, 

 different compared to the same area in the otolith 

 of a fed larvae. This ridgelike structure consistent- 

 ly indicated periods of starvation during the first 

 2 weeks of life. Ridges were not apparent for older 

 larvae starved for shorter time intervals (Fig. 4). 



DISCUSSION 



Estimation of age obtained using the light micro- 

 scope was not always accurate. When larvae grew 

 well, the light microscope gave correct age esti- 

 mates. However, otoliths of larvae reared under sub- 

 optimal feeding conditions gave underestimates of 



true age. Age estimates were more accurate using 

 SEM, and starvation episodes were easier to recog- 

 nize in the otoliths. 



Light microscopy has been routinely used to esti- 

 mate age in field-captured larval fish (Jones 1985). 

 Only a few investigators have employed SEM. Al- 

 though SEM improves the accuracy of age esti- 

 mates, it is more costly, requires more precise prep- 

 aration, and is more time consuming. However, for 

 larvae as resistant to starvation as striped bass, 

 SEM verification of age estimation obtained with 

 the light microscope is necessary. In our view, inves- 

 tigators using increments to estimate growth should 

 check their results from the light microscope with 

 SEM studies. SEM analysis could be performed on 

 a randomly chosen subsample of otoliths. If prob- 

 lems were uncovered, a more extensive analysis 

 using SEM could be undertaken. Checks on a ran- 

 dom sample using SEM are particularly important 

 for field studies where application of the otolith 

 increment technique to estimate field growth is 

 relatively new. 



It could be argued that larvae which undergo 

 periods of starvation are more vulnerable to preda- 

 tion and may occur only infrequently in samples. 

 Although this is quite likely, it is precisely during 





Figure 3.— SEM photomicrograph of an otolith of a larva starved for the first 15 days 

 of life then fed ad libitum. A ridge, indicated by the circle, develops as the result of 

 etching the otolith with 0.02 N HCl. This ridge corresponds to the period of starvation. 

 Increment width during starvation is approximately 5 \m\. Legends in the micrographs 

 indicate 1) length of scale bars in ^^m, 2) accelerating voltage KV, 3) mm working 

 distance, 4) coded photo number. 



176 



