FISHERY BULLETIN: VOL. 87, NO. 1 



Validation of Ageing Procedures 



Larvae of M. novaezelandiae were reared in cap- 

 tivity to determine the age at which otoliths form. 

 Fertile eggs were obtained by stripping running ripe 

 males and females immediately after their capture 

 by trawl on the spawning grounds. Eggs were incu- 

 bated in 1 L plastic jars, which were filled with sea- 

 water, and placed in a seawater bath. Initial incu- 

 bation temperatures ranged from 14° to 18°C 

 (sea-surface temperature was approximately 14°C). 

 Upon return to the laboratory, eggs were trans- 

 ferred to 2 L glass jars and placed in an aquarium 

 maintained at a constant temperature of 14 °C 

 (±0.2°C). Incubation jars were not aerated, and no 

 attempt was made to feed the larvae. 



Under light microscopy, otoliths were first ap- 

 parent in M. novaezelandiae embryos 10 hours prior 

 to hatching. At hatching, the sagittae and lapilli 

 were developed and conspicuous. The asterisci were 

 first apparent 3-4 days after hatching. Otoliths from 

 newly hatched larvae characteristically have a con- 

 spicuous dark and broad band close to their edge, 

 which is apparently laid down at hatching. Scanning 

 electron microscopy indicated the otolith within this 

 hatching mark consisted of a spherical primordium 

 surrounded by an area with little conspicuous struc- 

 ture. The radius of the hatching mark varied be- 

 tween specimens but did not differ significantly 

 between reared larvae (x = 6.9 txm, range = 6.7- 

 8.5 iim, n = 17) and wild-caught larvae (x = 7.5 

 \xm, range = 5.3-9.7 yon, w = 17) (P > 0.1, two tailed 

 <-test). 



All wild-caught larvae had a second exceptional- 

 ly dense and very conspicuous band. The radius of 

 this band varied from 10.2 to 16.4 j^m (x = 13.2 

 fim, n = 17), i.e., approximately 5-7 \im outside the 

 hatching mark. Although the otoliths of reared lar- 

 vae reached sizes close to this (maximum radius = 

 13.1 ymi), this distinctive band was not evident in 

 their otoliths. As the largest of these larvae had fully 

 ossified jaws and well-developed guts and had all but 

 exhausted their yolk reserves, the second major 

 band in the otoliths may have formed close to or 

 coincident with first feeding. The microstructure of 

 the otolith differed markedly inside and outside of 

 the "first-feeding band". Within its radius, there 

 was little evidence of consistent structuring (other 

 than the hatching mark); beyond the first-feeding 

 mark, increments were unambiguous, increasing in 

 width exponentially. As we saw no indication that 

 any structure prior to the first-feeding mark formed 

 daily, otolithic age for the larvae examined is de- 

 fined as the number of increments external to this 



feeding mark. This age is used in analysis of growth 

 and advection patterns, unless otherwise indicated. 



Based on the observed incubation time (55-60 

 hours) and the observed time required for reared 

 larvae held under temperature conditions similar to 

 those during the spawning season to develop to a 

 stage where feeding was possible (6 days) (Bruce 

 1988), the total age of larval blue grenadier can be 

 estimated as otolithic age + 6 days, with a probable 

 error of about + 2 days. Hence, date at first-feed- 

 ing for a particular larva was calculated as date of 

 collection less otolithic age, and date of spawning 

 was date of collection less total age. In general, the 

 development of otolith structure prior to first-feed- 

 ing of larvae in M. novaezelandiae is remarkably 

 similar to that of other gadoids (Radtke and Wai- 

 wood 1980; Bolz and Lough 1983; Dale 1984), as is 

 the proposed time frame. 



The hypothesis that increments in the otolith are 

 formed daily was tested by following cohorts of in- 

 dividuals and determining whether the change in the 

 number of increments matched the known sampling 

 interval (Campana and Neilson 1985; Jones 1986). 

 Larvae were sampled within 0.5 km of a drogue 

 deployed near the spawning grounds (see descrip- 

 tion of drogue below). Larvae from three plankton 

 tows made near the onset of a 26 h period (0521- 

 0649) were compared with those from two tows 

 made close to its end (0628-0701). The respective 

 samples were pooled because the number of larvae 

 caught in each tow was small. Mean sampling in- 

 terval between the first and last set of tows was 24.6 

 hours (1.025 days). Size-frequency distributions of 

 larvae collected are given in Figure 2A. Modal 

 analysis (means and variances unconstrained) for the 

 first sample set indicated the presence of two nor- 

 mally distributed populations, with means at 3.61 

 and 4.66 mm SL (SE = 0.06 and 0.05, respective- 

 ly); analysis of the size-frequency distributions of lar- 

 vae collected approximately a day later also in- 

 dicates two means, at 4.12 and 4.91 mm SL (SE = 

 0.29 and 0.07, respectively). The smaller of these 

 two means is poorly defined statistically, however. 

 Re-analysis with the added constraint that larvae 

 grew at the same rate across the size range of the 

 two means (which is a reasonable approximation for 

 such small larvae— see below and Figure 9) indicated 

 means for the first set of samples at 3.62 and 4.65 

 mm SL, and for the second at 3.86 (which is within 

 one SE of the unconstrained mean) and 4.89 mm 

 SL, which fitted closely observed distributions. The 

 average difference in larval sizes between the first 

 and second set of samples (i.e., mean growth for the 

 24.6 h period) was 0.24 mm SL. The number of 



32 



