FISHERY BULLETIN: VOL. 81, NO. 3 



same postrostral growth axis. As many readings as 

 possible were made from each preparation (x = 3 1 .9) , 

 subject to the constraint that counts be made only in 

 regions where increments were clear. 



Data were summarized for each specimen by com- 

 puting the ratios of segment length in pirn to the num- 

 ber of increments included at each specific segment 

 examined in the preparation. These figures repre- 

 sent the average increment width at given points in 

 the otolith. Under the assumption that one increment 

 forms each day, it is possible to approximate the 

 otolith growth rate on a size-specific basis as 

 follows: 



A2 



A increments 



AL 



At 



d2 

 dt 



where AS is the change in length (ju.m) of the otolith 

 over a given subregion, and At is the change in time 

 (days). The finite approximation to the differential 

 was always computed on a sufficiently small scale of 

 time and length. 



Given the series of ordered pairs, 



d2 

 dt 



at otolith 



length 2, it is possible to evaluate the functional 



d2 

 relationship between these quantities, , = f(2), 



and to estimate the parameters of the function using 

 regression analysis. The regression equation can 

 then be solved by separation of variables and treated 

 as a definite integral, bounded by the focus and the 

 total length (L) of the otolith at death, providing an 

 explicit estimate of age in days (T). That is: 



T 



J dt 



-i 



/ \f(2)\ d2 



1 = 



1=0 



(1) 



Size at Maturity 



Samples of opakapaka gonads were obtained from 

 throughout the Hawaiian Archipelago. Sampling was 

 haphazard with respect to both time and location. 

 Gonads were preserved in 10% Formalin until ex- 

 amined, when they were weighed to the nearest gram 

 after excess moisture had been removed by blotting 

 with a paper towel. Sex was determined by micro- 

 scopic examination of smeared gonadal tissues, and 

 ovaries were staged according to the following 

 classification: 



Stage I. Inactive ovary 



a) egg diameters 75-100 /xm 



b) transparent primary oocytes in ovary 



c) nucleii faintly visible 



d) eggs ovoid 



Stage II. Developing ovary 



a) egg diameters 100-325 /xm 



b) some opaque eggs in ovary 



c) nucleii clearly visible 



d) eggs wedge-shaped 

 Stage HI. Gravid/spawning ovary 



a) egg diameters 325-850 /xm 



b) presence of oil globules 



c) evidence of yolk granules 



d) eggs ovoid 



RESULTS 



Typical looking increments are found in the sagittae 

 of opakapaka (Fig. 3). In this SEM photograph the 

 distinct incremental growth of the otolith is readily 

 apparent, with a structure similar to that described in 

 previous descriptions of daily increments. Discern- 

 ible under high magnification is a more deeply 

 etched, discontinuous zone which transects the 

 radial growth of aragonite crystals found in the in- 

 cremental zone (Hickling 1931; Pannella 1971, 1974, 

 1980; Blacker 1975; Brothers et al. 1976; Timola 

 1977; Dunkelberger et al. 1980; Mugiya et al. 1981; 

 Tanaka et al. 1981; Watabe et al. 1982; Radtke foot- 

 note 3). 



Tetracycline Validation 



Photomicrographs of a sagittal section through the 

 otolith of one of the three experimental fish injected 

 with oxytetracycline are presented in Figure 4. In the 

 upper photograph the specimen was illuminated with 

 ultraviolet light, clearly showing the fluorescing tet- 

 racycline checkmark. The lower photograph shows 

 the same specimen illuminated with visible light. A 

 discontinuity coincident with the tetracycline label is 

 evident. We interpret this latter checkmark to be due 

 to exposure to acetazolamide and handling trauma. 

 The close coincidence of the two marks allowed us to 

 positively identify the experimentally induced 

 checkmark in all six specimens. 



The results of marginal increment counts from 

 these fish are summarized in Table 1. Several sec- 

 tions were taken from each otolith, although not all 

 preparations were readable. Furthermore, because 

 the quality of all sections was poor, multiple counts of 

 each section were made. The figures under the head- 

 ing "Number of marginal increments" represent the 

 range of counts. 



It is evident from these data that the number of in- 

 crements deposited on the otoliths after marking was 

 quite similar to the number of days elapsed after the 

 fish were marked. Since these fish were stressed dur- 



526 



