FISHERY BULLETIN: VOL. 87, NO. 1 



depth, time of day, and the interaction term were 

 all highly significant (Fg ig = 22.6, Fj ,e = 13.5, and 

 7^3 le = 9.6, respectively, P < 0.01 in all cases). Dif- 

 ferences between replicate samples were small, ac- 

 counting for only 12.7% of the variance, despite the 

 same patch of water not being sampled each time. 

 Although larvae of M. novaezelandiae were col- 

 lected at stations all along the western, southern, 

 and southeastern coasts of Tasmania, the age and 

 size-frequency distributions of these larvae differed 

 conspicuously between collecting sites. In 1984, lar- 

 vae younger than 5 d postfirst-feeding were caught 

 only on transects 5 (9% of total) and 6 (91%). This 

 is consistent with earlier suggestions (Wilson 1981, 

 1982) that the area along or on the continental shelf 

 between Sandy Cape (transect 5) and Cape Sorell 

 (transect 6) is the primary spawning area for M. 

 novaezelandiae in Australian coastal waters. The 



ages of the larvae caught at transect 5 varied wide- 

 ly. From transect 6 south and east along the coast, 

 the ages of larvae collected increased consistently 

 with increasing distance from the spawning area 

 (Fig. 6A). Differences between transects in age 

 distributions of larvae are highly significant (i^s no 

 = 38.8, P « 0.01), as is the correlation between 

 age and distance ( = number of transects, based on 

 the equal spacing of transects along the coast) from 

 transect 5 (r = 0.64, P « 0.01). The latter corre- 

 lation was also significant for each of the 1984 

 spawning season cruises individually, except the last 

 one (September), when all larvae collected were 

 relatively old. Differences between transects in the 

 sizes of larvae caught paralleled differences in ages 

 (differences between transects, i^5 no = 27.9, P « 

 0.01), v/ith the largest larvae collected farthest along 

 the coast from the spawning area (at transect 9) (cor- 



C 

 ■3 



60, 



50. 

 40. 

 30. 

 20. 

 10 

 0. 



03 



O 



a 

 IE 



S 



4 5 6 7 8 9 1 



Transect Number 



Figure 6.— Ages of larvae caught at each transect, pooled across sampling periods, for 

 1984 and 1985. Vertical bars indicate means ± 1 SD; SE are in all cases <3 days. Dif- 

 ferences in larval ages across transects are highly significant for both years, as are the 

 correlations between age and distance from transect 5. 



36 



