LEIS ET AL.: DISTRIBUTION AND ABUNDANCE OF BILLFISH LARVE 



considered safe in the lagoon and close to the 

 windward face of the reef. Because of the great 

 variations in bottom topography in the latter 

 area, the net actually hit bottom upon occa- 

 sion. Further offshore, bongo net tows were done 

 with a standard amount of wire out which en- 

 sured a maximum sampling depth in excess of 

 100 m. 



Position fixing in the Coral Sea was by radar 

 reflection of the waves breaking on the reef crest 

 when close to the reef. This meant that actual 

 distance off the reef varied somewhat ( — 100 m) 

 depending on sea state and tide. Four cruises 

 were made to investigate horizontal distribution: 

 1) 2-5 November 1984, 2) 17 and 20-22 November 

 1984, 3) 30 January-2 February 1985, and 4) 9-13 

 February 1985. On each cruise, six samples were 

 taken between Lizard Island and the outer reef on 

 one day (Fig. 1), and three days were spent in the 

 Coral Sea running a transect each day and start- 

 ing at opposite ends of the transect on alternate 

 days. On each transect, two randomly located 

 samples were taken in each of five offshore blocks 

 defined by distance (nmi) from the outer reef crest 

 (Fig. 1): A, 0-0.25 nmi; B, 0.25-1.0 nmi; C, 1.0-3.0 

 nmi; D, 3.0-6.0 nmi; E, 6.0-10.0 nmi. Therefore, 

 six samples were taken in each block on each 

 cruise. The three transects on a cruise were each 

 centered off a different reef (i.e., either of Day, 

 Carter, Yonge, No Name, or Number 10 Ribbon 

 Reefs). Bad weather and high volumes of floating 

 pumice precluded the routine use of the neuston 

 net. There were some variations in this plan 

 owing to weather or equipment problems, the 

 most serious of which was missing 4 of 6 samples 

 in block A on the second cruise. Larvae from other 

 samples taken with similar methods in November 

 1983 were also included where appropriate. 

 Funding limitations prevented processing of sam- 

 ples from block D. 



The vertical distribution samples were taken in 

 the lee of Carter Reef primarily in the Great Bar- 

 rier Reef Lagoon, but partially in the pass to the 

 north of Carter Reef (Fig. 1). Samples were taken 

 in sets; each set consisted of a neuston tow and 3 

 bongo net (0-6 m, 6-13 m, and 13-20 m) tows. The 

 0-6 m stratum was sampled in the undisturbed 

 water flowing between the hulls of the catama- 

 ran. In February-March 1983, 22 such sets were 

 taken, 8 each in morning and afternoon and 6 at 

 night. 



Additional samples from within the Great Bar- 

 rier Reef Lagoon as reported by Leis and Gold- 

 man (1984, 1987) and Leis (1986) were used for 



seasonality information. Samples in the Lagoon 

 were taken in all months but May, June, August, 

 September, and December. Samples were taken 

 in the Coral Sea in October, November, January, 

 and February. 



Oblique bongo net tows typically filtered 1,000- 

 1,500 m^ and horizontally stratified tows filtered 

 400 m'^. Neuston tows typically travelled 1,200- 

 2,000 m. All nets were carefully washed after 

 each tow and the sample preserved in 5-10% 

 seawater-formalin. 



In the laboratory, samples were sorted using a 

 dissection microscope (~10x) and all larvae re- 

 moved. Samples from both sides of the bongo net 

 were fully sorted except for the Great Barrier 

 Reef Lagoon samples from February-March 1983 

 and November 1984 when only side was sorted 

 because of high plankton volume. Larvae were 

 placed in 70% ethanol prior to measurement. 

 Identification of larvae followed Ueyangi (1963, 

 1974a, b). Larvae were measured using an eye- 

 piece micrometer of a dissection microscope to the 

 nearest 0.1 mm. Notochord length and standard 

 length were measured for preflexion and postflex- 

 ion larvae, respectively (Leis and Rennis 1983). 

 Larvae from these samples are deposited in the 

 Australian Museum, Sydney. 



Numbers of larvae per sample were converted 

 to numbers per volume (concentration) and num- 

 bers per area (abundance) using standard meth- 

 ods (Leis 1986). In analysis of vertical distribu- 

 tion data, only positive sets (i.e., those in which at 

 least one larva was captured) were considered. 

 Statistical methods followed Conover (1971) and 

 Zar (1974). References to the Student-Newman- 

 Keuls (SNK) test refer to the version based on 

 ranks (Zar 1974). 



RESULTS 



Identification 



We captured larvae of black marlin, Makaira 

 indica; blue marlin, Makaira mazara; striped 

 marlin, Tetrapturus audax; and Indo-Pacific sail- 

 fish, Istiophorus platypterus . The larvae here 

 identified as black marlin correspond to the "non- 

 pigmented" sailfish of Ueyanagi (1974a, b). This 

 type of istiophorid larva has been captured only 

 in the seas off northern Australia and the south- 

 ern portion of the Indonesia-New Guinea 

 archipelago (Ueyanagi 1974a, b). In our Coral Sea 

 samples these larvae were found almost exclu- 

 sively during November, the time when large 



759 



