THRESHER ET AL.: LARVAE OF GADOID, MACRURONUS NOVAEZELANDIAE 



relation between distance from transect 5 and size, 

 r = 0.63, P « 0.01). 



The relationships between sampling site and lar- 

 val ages and sizes in 1985 were similar to those in 

 1984, though apparently complicated by several fac- 

 tors. As in 1984, differences between transects were 

 highly significant (F9355 = 14.2, P « 0.01, and 

 J^9 355 = 12.38, P « 0.01, for age and size, respec- 

 tively), as were the correlations between both vari- 

 ables and distance from the midwest coast (r = 0.33, 

 P < 0.01, and r = 0.36, P < 0.01, for age and size, 

 respectively) (Fig. 6B). At transects 5 and 6, 95% 

 of larvae aged <5 d postfirst-feeding were caught. 

 However, some larvae <5 d postfirst-feeding were 

 also collected at transects 4 and 7, and a few larvae 

 <15 d postfirst-feeding were found off the north- 

 eastern coast of Tasmania (near transect 1). The age 

 of the latter is much less than would be expected 

 based on northward advection from the known 

 spawning area (Gunn et al. in press), which suggests 

 strongly the presence of a second spawning area, 

 involving few adults, of the northeastern coast. The 

 occurrence of these larvae confounds a general 

 relationship between distance from the west coast 

 and the ages and sizes of larvae caught. Hence, 

 although for each cruise in 1985 larvae consistent- 

 ly increased in age and size with increasing distance 

 from transects 5 and 6, there was a broad range of 

 larval ages at each transect, relatively old larvae at 

 several transects, and young larvae on the east 

 coast. 



Larval Advection 



On the basis of the distribution of larvae of dif- 

 ferent ages and sizes around Tasmania, we hypoth- 

 esized that most larvae were being carried passive- 

 ly by a longshore current southwards around the 

 coast from the primary spawning area off the west 

 coast. The drift card returns, the movement of the 

 drogue deployed on the west coast, and the distribu- 

 tion of surface isotherms are generally consistent 

 with this hypothesis. 



Most drift card returns were from sites southeast 

 along the coast from the release points, including 

 all of those from the first series (July 1985) (Fig. 7A, 

 B). The drogue, deployed at the southern point at 

 the same time drift cards were released, also drifted 

 longshore and to the south (straight-line distance of 

 11.8 km in 26 hours). Movement of the drogue was 

 conspicuously related to wind speed and direction, 

 varying from nil at slack winds (<9 km/h) (as 

 measured by shipboard anemometer) to slightly >1 

 km/h for a 9 h period when wind speed averaged 



approximately 55 km/h from the northwest (350° 

 magnetic). For the second release series, drift cards 

 returned shortly after being released on 11 August 

 1985 at the northern site were predominantly from 

 points inshore (east) and slightly north of the release 

 point (Fig. 7C). Of the 43 cards returned from this 

 release, only two were found south of the release 

 point; four, found on mainland Australia, had been 

 transported north more than 150 km. In contrast, 

 southeasterly transport was indicated by the cards 

 released at the southern point on 12 August; only 

 three of 30 returns were from sites north of the 

 release point (Fig. 7D). One of these cards was found 

 on South Arm Beach (southeastern Tasmania) on 

 27 August 1985. It had drifted slightly over 350 km 

 in 15 days. Additional returns from this release in- 

 cluded three cards from New Zealand, one from 

 Flinders Island (northeast of Tasmania), and one 

 from the southeastern coast of mainland Australia. 

 All of these were found several months after being 

 released. 



The distribution of surface isotherms also suggests 

 the presence of a southward flowing current along 

 the west coast of Tasmania during the spawning 

 period of M. novaezelandiae. In both years of the 

 study, west coast temperature plots in late autxmin 

 and early winter were dominated by a tongue of 

 water, 1°-2°C warmer than the surrounding water, 

 that extended southwards along the coast, becom- 

 ing narrower and cooler to the south (Fig. 8). This 

 tongue of warm water, oriented parallel to the coast, 

 was observed on all winter cruises. Satellite imagery 

 has since documented it to be a regular seasonal 

 feature off the west coast of Tasmania (C. Nilsson, 

 in prep.). 



Growth 



Otolithic age was determined for 116 larvae in 

 1984 and 365 larvae in 1985. Growth trajectories 

 (length-at-otolithic age) for M. novaezelandiae lar- 

 vae were log linear for both years (Fig. 9), account- 

 ing for 96% of the variance in length at age in 1984 

 and 84% of the variance in 1985. Residuals exhibit 

 no conspicuous systematic deviation from linearity 

 in either year and no marked increase in variance 

 with age. Hence pooled data indicate consistent ex- 

 ponential growth through at least the first 50 days 

 of larval life, with no indication that the rate of 

 growth declined late in larval life. The slope of the 

 semilog regression was steeper, albeit only slight- 

 ly, in 1984 than in 1985 (0.043 vs. 0.039, respective- 

 ly, ANCOVA Fi,47; = 2.56, P < 0.001), which sug- 

 gests that growth was more rapid in 1984. 



37 



