FISHERY BULLETIN: VOL. 80, NO. 3 



rents between March and August, the main 

 period of postlarval recruitment in Shark Bay, 

 Western Australia. Penn proposed that during a 

 larval life lasting 2-4 wk the larvae would be dis- 

 placed 30-80 km before becoming postlarvae at 

 which time they would migrate more actively 

 using a tidal cue. In the present study the 12:12 

 larva that behaved in a manner dictated by 

 Penn's (1975) scheme seldom travelled in the 

 same direction or distance as the median larva 

 (Figs. 7, 8, 9). This is largely because these larvae 

 did not utilize the full water column over all their 

 larval life even under ideal conditions and that 

 subtle changes in environmental factors changed 

 the patterns of vertical distribution of all larval 

 stages quite markedly. 



With more detailed information on larval dis- 

 tribution and its causes, realistic dispersal dis- 

 tances and directions are even more dependent 

 upon detailed knowledge of current speeds and 

 direction at depth. In the present study there 

 were shortcomings in assessing both short-term 

 and long-term in situ current regimes. The mi- 

 gratory current meter used gave some indication 

 of variation of speed and duration of currents but 

 is probably biased by sampling technique and in- 

 fluenced, at times, by short-term wind events 

 and short-term nontidal pertubations. Further- 

 more there were errors introduced by deploying 

 the meter from a moored vessel. A certain 

 amount of oscillation in current direction ap- 

 peared during periods of low current velocity 

 when the ship would swing on its anchor. This 

 was overcome to some extent by using both a bow 

 and stern anchor. The most serious shortcoming, 

 however, is the robustness of the extrapolation 

 from the 24-h record to the entire larval life of 

 about 14 d. At best this would only be an approxi- 

 mation of the distance and direction of advection 

 of median larva. 



The distances estimated by extrapolation (ca. 

 70-100 km) are large enough to transport the lar- 

 vae from the farthest known offshore commer- 

 cial concentrations of adult penaeid shrimp in 

 the Gulf of Carpentaria (Lucas et al. 1979) to 

 their nearshore and estuarine nursery grounds. 

 This range of advection is greater than that esti- 

 mated by Penn (1975) even though in this study 

 only a 2-wk larval life was used as opposed to a 

 4-wk period in his. The 2-wk period seems more 

 realistic at ambient Gulf of Carpentaria temper- 

 atures (unpubl. data; Cook and Murphy 1969; 

 Mock and Murphy 1971). This period still allows 

 for one to two postlarval intermolts before settle- 



ment. It is suggested that the advection of the 

 larvae limits the offshore adult distribution of 

 those species of penaeid shrimp in the Gulf of 

 Carpentaria that rely on nearshore or estuarine 

 habitats for juvenile nursery grounds. There are 

 no apparent depth or substrate limitations that 

 would restrict the adult distribution to a coastal 

 zone <100 km wide. 



Whether or not differential larval advection 

 can account for the large-scale temporal and spa- 

 tial patterns of postlarval recruitment seen in 

 the Gulf of Carpentaria (Staples 1979) is still 

 open to speculation. There is some evidence from 

 this study of offshore transport of larvae in 

 March (Fig. 9d) at a time when large numbers of 

 ripe female shrimp are present in the commer- 

 cial fishery in the southeastern corner of the Gulf 

 of Carpentaria. It may be this offshore advection 

 of larvae that explains the subsequent low level 

 of postlarval recruitment in the following weeks 

 into the coastal rivers of this region of the Gulf of 

 Carpentaria (Staples 1980). The short-term sam- 

 pling of currents in the present study is not capa- 

 ble of annotating the long-term seasonal effects 

 of tropical wind regimes and the long-term pro- 

 gression of tidal phase. Long-term monitoring 

 and modelling of tidal and wind driven currents 

 (Church and Forbes 1981; Forbes and Church in 

 press) in the Gulf of Carpentaria has recently 

 been completed and will be used to overcome this 

 shortcoming of the present study. It is also likely 

 that a better understanding of short-term mete- 

 orological events at critical times in the larval 

 life history would help explain the year-to-year 

 variation seen in the strength of postlarval re- 

 cruitment in individual rivers around the gulf 

 (Staples 1979) and the subsequent commercial 

 catch (Lucas et al. 1979). 



ACKNOWLEDGMENTS 



C. J. Jackson and R. C. Pendrey assisted with 

 all aspects of sampling, curation, sorting, and 

 data analysis. M. Beamish, H. Keag, G. Peate, 

 and D. Savage sorted the larvae from the plank- 

 ton samples. J. Church, G. Cresswell, and A. 

 Forbes helped with the current meter analysis 

 and interpretation. J. Kerr provided assistance 

 with the statistical analysis. J. Church, W. Dall, 

 A. Forbes, D. Staples, and P. Hindley gave valu- 

 able criticism of the study and critically read the 

 manuscript. N. Hall prepared the manuscript 

 for publication. I would also like to thank the cap- 

 tains and crews of the RV Kalinda, FV Judy B, 



552 



