Prince et al.: Otolith analysis of Makaira nigricans age and growth 



455 



Discussion 



Limitation of the ageing method 



Within the age range of 9 to 495 days, reliability of the 

 ageing method decreased progressively with increas- 

 ing age. An upper limit for optical resolution of primary 

 increments occurred at abody length of about 212cm. 

 Beyond this length, depositions of individual incre- 

 ments were too close together to distinguish, and we 

 did not feel that accurate counts could be made. There- 

 fore, based on otolith appearance and increment spac- 

 ing at the margin, we did not apply the otolith micro- 

 structure method to fish larger than 212 cm. While this 

 limit is arbitrary and perhaps conservative, the prac- 

 tical application of this technique for Atlantic blue 

 marlin will certainly be restricted to no more than the 

 first 2 years of life. 



Because of the morphology of blue marlin otoliths, 

 the SEM technique is only practical for transverse sec- 

 tions. Although the microstructural record could some- 

 times be read for several hundred days longer than 

 with a light microscope preparation, other limitations, 

 such as irregularity in the ventral lobe, made this a 

 difficult and impractical approach (see Fig. 10). The 

 dorsal lobe has a very regular and easily read early- 

 growth record (up to about 1.4 years), but has a very 

 uneven margin. Incremental growth is greatly com- 

 pressed and probably interrupted to varying degrees, 

 depending on which particular radius is intersected by 

 the section. Furthermore, the extremely small size of 

 blue marlin otoliths (i.e., a maximum dorsal lobe radius 

 of about 2 mm), combined with a probable maximum 

 age exceeding 25 years (Hill et al. 1989), means that 

 increments will average about 0.2^m if a complete 

 record is present. Counting such fine increments re- 

 quires optimal sectioning, polishing, and etching and 

 an otolith crystalline structure that will allow such fine 

 structures to be seen. Increments of this width have 

 been reported (Radtke 1984, Brothers and Mathews 

 1986); however, it is not clear whether such fine incre- 

 ments can always be seen. 



The problem of increment resolution in larger/older 

 blue marlin may have resulted in an underestimate of 

 age and overestimated growth rate. Campana and 

 Neilson (1985) state that apparent non-daily increment 

 formation reported for some species (Geffen 1982) 

 could be due to limited instrument resolution, as well 

 as variable otolith preparation and retarded growth due 

 to environmental conditions during deposition. We 

 minimized these potential effects by rejecting hard-to- 

 analyze samples and by using increment counts on 

 lateral or sagittal views instead of transverse sections. 

 The SEM examination of sagittal and transverse sec- 



tions also helped confirm that errors of underestimat- 

 ing increments were minimized using the light micro- 

 scope counting procedure. In addition, the chi-square 

 contingency table analyses confirmed that significant 

 errors in undercounting increments did not occur, 

 assuming that the qualitative description of spawning 

 in the literature is accurate. 



Under optimal conditions, increments separated by 

 less than 0.2^m cannot be resolved with the light 

 microscope. In practice, our observation is that with 

 moderately thick sections and the complex three- 

 dimensional morphology of istiophorid otoliths, the 

 resolution limit is two or three times this value. Thus, 

 otolith growth zonations narrower than about 0.5^m 

 will be underrepresented in the increment count. Since 

 all increments in blue marlin sagittae were much larger 

 than 0.5^m (for fish < 1.4 years of age), we feel this 

 problem did not affect the results of this study. Al- 

 though subdaily increments were commonly observed 

 (Fig. 2), these features were easily distinguished from 

 the primary increments (Figs. 2,3) and thus overesti- 

 mation of total increment count was probably not a 

 source of error in this study. 



Periodicity of increment formation 



Back -calculation of spawning dates and descriptions of 

 otolith microstructure are inherently less desirable for 

 determining the periodicity of increment formation 

 than direct approaches such as rearing fish of known 

 age in captivity or chemical labeling (e.g., oxytetra- 

 cycline, Geffen 1987). Geffen (1987) reviewed seven 

 methods of age validation and rated each method ac- 

 cording to levels of reliability for providing evidence 

 of daily ring deposition. The otolith microstructure 

 approach was characterized as providing a medium 

 level of reliability, allowing only limited inferences for 

 validation of daily ring deposition. Conversely, Geffen 

 (1987) rated the estimation of hatching dates as a 

 medium-to-highly-reliable method for evaluating the 

 strength of assuming daily ring formation in teleost 

 otoliths. Therefore, the results of our study should be 

 tempered accordingly. 



Otolith microstructure Our examination of micro- 

 structural features of blue marlin otoliths identifies and 

 characterizes the primordium, core, primary incre- 

 ments, subdaily increments, and increment spacing. 

 These microstructural features in larval blue marlin 

 sagittae were indistinguishable from the same char- 

 acteristics described in the sagittae of related teleost 

 species, some of which had definitive age validation 

 based on rearing experiments. 



