Abstract. — We compared sam- 

 pling performance of four nets and 

 two aggregation devices for larval 

 and pelagic juvenile coral-reef fishes. 

 The six sampling devices were de- 

 ployed simultaneously over three 

 nights near a coral reef at Lizard 

 Island, northern Great Barrier Reef, 

 Australia. The resulting 83 samples 

 captured 57,701 larval and pelagic 

 juvenile fishes of 70 families (exclud- 

 ing clupeoids which were not con- 

 sidered in this analysis). The bongo 

 net took the most families, and the 

 light-trap the fewest. In all meth- 

 ods, a few families dominated the 

 catch. Dominance was least in the 

 Tucker trawl catches and greatest 

 in light-trap catches, where poma- 

 centrids constituted 939c of the 

 catch. Composition of catches was 

 similar for the four nets. Catches 

 from the light-trap were markedly- 

 different from those taken by net; 

 catches taken by light-seine showed 

 similarities to those taken by both 

 net and light-trap. For four abun- 

 dant families (Apogonide, Gobiidae, 

 Lutjanidae, Pomacentridaei. the 

 bongo net gave the overall highest 

 density estimates, although those 

 from purse-seine were frequently 

 equivalent to bongo-net estimates. 

 The Tucker trawl provided the low- 

 est density estimates in most cases. 

 Catches of bongo, neuston, and seine 

 nets were similar in size structure 

 and were dominated by small lar- 

 vae; overall, however, bongo nets col- 

 lected the greatest size-range of 

 fishes. The Tucker trawl did not col- 

 lect small larvae well nor did it col- 

 lect significantly greater densities of 

 large larvae and pelagic juveniles 

 than the bongo net. Fishes collected 

 by aggregation devices were gener- 

 ally larger than those taken by net, 

 and light-traps caught very few fish 

 <5mm. Light-traps collected greater 

 numbers of large pomacentrids 

 (>6mm) than other methods. In an 

 extended sampling period of five 

 nights, both aggregation devices 

 showed obvious peaks in the den- 

 sity of large pelagic pomacentrids 

 and mullids; these patterns were not 

 detected by the nets. 



A comparison of towed nets, purse 

 seine, and light-aggregation devices 

 for sampling larvae and pelagic 

 juveniles of coral reef fishes* 



John H. Choat 



Department of Marine Biology, James Cook University of North Queensland 

 Townsville, Queensland 4811. Australia 



Peter J. Doherty 



School of Australian Environmental Studies, Griffith University 

 Nathan, Queensland 4111, Australia 



Present address Australian Institute of Marine Science. PMB No 3, Townsville M.C.. 

 Queensland 4810. Australia 



Brigid A. Kerrigan 



Department of Marine Biology, James Cook University of North Queensland, 

 Townsville. Queensland 4811, Australia 



Jeffrey M. Leis 



The Australian Museum, PO. BoxA285 

 Sydney South, NSW 2000, Australia 



Manuscript accepted 24 February 1993. 

 Fishery Bulletin, U.S. 91:195-209 ( 1993). 



Almost all species of marine teleost 

 fishes have a pelagic phase in the 

 early part of their life history (Moser 

 et al. 1984). Size, morphology, and 

 behavior of larval and pelagic juve- 

 nile phases vary greatly (Moser 

 1981), and this makes accurate sam- 

 pling of these fishes problematical 

 (Murphy & Clutter 1972, Frank 1988, 

 Suthers & Frank 1989, Brander & 

 Thompson 1989). The problem is ex- 

 aggerated in tropical waters due to 

 high taxonomic and developmental 

 diversity and the presence of many 

 demersal species with extended pe- 

 lagic phases (Leis & Rennis 1983, 

 Leis & Trnski 1989, Leis 1991b). 

 Studies of the pelagic phase can pro- 

 vide important information on popu- 

 lation biology of reef fishes. Despite 

 its brevity, the high mortality and 

 dispersion characteristic of this phase 

 can have important demographic con- 

 sequences for many species (Victor 

 1986). There is now a widespread in- 

 terest in the process of recruitment 

 in coral reef fishes (Doherty & Wil- 



liams 1988, Warner & Hughes 1989), 

 and sampling techniques which cover 

 the full size-range of the pelagic 

 phase are needed. 



A number of different methods are 

 available to sample this complex as- 

 semblage of early-life-history stages, 

 including towed nets, purse-seines, 

 and various types of aggregation de- 

 vices which attract fish into collec- 

 tion sites or traps. These methods 

 differ in their method of deployment 

 and capture, and each has its own 

 set of advantages and disadvantages. 

 All have biases in number, identity, 

 and sizes of pelagic fishes collected 

 (Clutter & Anraku 1968, Clarke 1983 

 and 1991). For the pelagic phase of 

 reef fishes, there have been few at- 

 tempts to evaluate the relative bias 

 of different sampling methods. Re- 

 cent studies have provided informa- 

 tion on the comparative performance 



Contribution of the Lizard Island Research 

 Station. Authorship alphabetical. 



195 



