456 



Fishery Bulletin 89(3). 1991 



Back-calculated spawning dates There are several 

 potential sources of error in the back-calculated spawn- 

 ing dates reported in this study. Since istiophorids 

 have never been reared in captivity, we had no basis 

 for correcting for the time of initial growth-increment 

 formation in blue marlin sagittae. However, increment 

 deposition has generally been reported to start within 

 the first week after hatching in most teleosts, and in 

 the majority of studies the first ring usually forms dur- 

 ing the first 3 days after hatching (Brothers 1979, 

 Brothers et al. 1983, Radtke 1983b). Therefore, this 

 type of error probably did not bias our estimates of 

 spawning dates by more than 7 days. In addition, we 

 feel that the precautions taken to minimize underesti- 

 mates of increment counts (discussed earlier) avoided 

 major errors of this type. 



Mahon and Mahon (1986) summarized existing data 

 on spawning of blue marlin in the northwestern Atlan- 

 tic (Erdman 1968, Eschmeyer and Bullis 1968, Yeo 

 1978) and reported that the spawning season occurs 

 from April through November. Peak spawning is 

 thought to occur primarily in midsummer (Erdman 

 1968), but a smaller peak in the fall has also been 

 reported (Yeo 1978). 



Our data on back-calculated spawning dates (Fig. 5A) 

 agree with the qualitative data on spawning season 

 reported by Mahon and Mahon (1986) and the peaks 

 of spawning documented by Erdman (1968) and Yeo 

 (1978). Back-calculated spawning dates are based on 

 a wide range of fish age and dates of capture (Fig. 5B). 

 As shown in Figure 6, periodicities other than 1 allocate 

 substantial numbers of back-calculated spawning dates 

 and peak spawning to months outside the reported 

 spawning season. The results of the chi-square con- 

 tingency table analyses and the observations of otolith 

 microstructure are both consistent with the hypothesis 

 that the increments in sagittae from young Atlantic 

 blue marlin are formed once each day. 



Precision 



Blue marlin are considered to be a long-lived species 

 (Hill et al. 1989) and thus potentially have many age- 

 classes in the fishery. The APE method of evaluating 

 the precision of a set of age determinations, described 

 by Beamish and Fournier (1981), is not independent 

 of age and thus appeared well-suited for blue marlin. 

 Mean APE values (ranging from 0.3 to 4.0%) for blue 

 marlin, in the age range 21-495 days, are well within 

 the range published for other species (Table 2) using 

 annual or monthly ageing methods. The APE values 

 for these ageing techniques are not directly comparable 

 to daily ageing methods (i.e., errors in daily increment 

 methods obviously have a smaller effect on age estima- 

 tion than errors from annual ageing methods). Never- 



theless, the overall mean APE, 1.6% (as well as the 

 range in APE values), indicates that the otolith micro- 

 structure method applied to young blue marlin clearly 

 meets the requirement (<10%) described by Powers 

 (1983) for an acceptable level of precision for use of an 

 vageing technique in stock assessment. 



Growth 



The maximum absolute growth in length (1.66 cm/day 

 at 39 cm LJFL) we report for young Atlantic blue 

 marlin exceeds that estimated from length frequencies 

 by de Sylva (1957) for Atlantic sailfish Istiophorus 

 platypterus for the second month of life (1.10 cm/day 

 for the length range 18-51 cm total length). De Sylva 

 (1957) estimates that 6-month-old (180-day) sailfish at- 

 tain a modal total length of 142.2cm (~113.9cm LJFL), 

 while we found that blue marlin reach the same size 

 in about 130 days. Also, while blue marlin and sailfish 

 are almost the same size at the end of the first month, 

 our average relative growth rate (5.7%) computed for 

 the same size range as sailfish (RGR = 3.9%, 18-51 cm 

 total length or about 10-38 cm LJFL) is nearly 1.5 

 times larger during the second month. 



Growth rates are very rarely constant for extended 

 periods of time during early life cycles, i.e., periods of 

 rapid growth are usually followed by periods of declin- 

 ing growth giving rise to the S-shaped or asymptotic 

 growth curves. Thus, it is almost certain that growth 

 rates exceed the first 100-day average of 1 cm per day 

 somewhere during this period. Both the magnitude and 

 location of the estimated maximum depend to some 

 extent upon the validity of the choice of the growth 

 equation. 



Our data suggest that blue marlin is one of the most 

 rapidly growing teleosts in terms of absolute growth 

 rates, but that larval, juvenile, and young adult/adult 

 growth are not particularly exceptional measured on a 

 relative scale. For example, maximum growth of the 

 common dolphin Coryphaena hippurus does not exceed 

 0.5 cm/day for the first year of life (Pew 1957; C. Brown- 

 ell, The Oceanic Inst., P.O. Box 25280, Honolulu, HI 

 96825, pers. commun., 6 Sept. 1988), but the species 

 attains a maximum length of 1.5 m, compared with 

 4.5m for blue marlin, one of the largest North Atlan- 

 tic teleosts (Norman and Fraser 1948). Conversely, At- 

 lantic bluefin tuna attain a maximum weight similar 

 to Atlantic blue marlin (over 454.5kg), yet the max- 

 imum growth rate of bluefin tuna for the first year is 

 similar to dolphin and varies from 0.1 to 0.6cm/day 

 (Brothers et al. 1983). Similarly, as shown in Table 3, 

 the relative growth rate (17%) for 1.5-cm LJFL blue 

 marlin postlarvae is only slightly above that (13%) re- 

 ported by Hunter and Kimbrell (1980) for Pacific mack- 

 erel Scomber japonicus postlarvae averaging 1.5cm SL. 



