406 
Fishery Bulletin 106(4) 
habitats. Near-reef concentrations were determined by 
using short (<15 min.) horizontal tows at or near the 
surface (Miller, 1979; Leis et ah, 1991), whereas oceanic 
sampling in the region has generally involved tows 
of longer duration (>30 min.), with nets towed either 
horizontally at the surface or obliquely from depths 
reaching 200 m (Matsumoto, 1958; Nakamura and Mat- 
sumoto, 1966). Because distributions of tuna larvae are 
vertically stratified, and Thunnus spp. larvae are rarely 
found at depths greater than 50 m (Davis et ah, 1990b; 
Boehlert and Mundy, 1994), oblique tows to depths in 
excess of 100 m most likely underestimate concentra- 
tions in shallower strata. Because of this situation, it 
is appropriate to use the measure of abundance (i.e., 
the number of larvae below a standard surface area of 
water) rather than concentration (numbers of larvae per 
standard volume) to compare numbers of larvae among 
sites where tows were taken to different depths. 
Abundances of tuna larvae were determined at both 
near-reef (2 km offshore) and oceanic sites (~30 km off- 
shore) around the Hawaiian island of Oahu by Boehlert 
and Mundy (1994) by oblique sampling to 200 m depth. 
The abundance of Thunnus spp. larvae decreased with 
increasing distance offshore, but only on the leeward 
side of Oahu Island. Interestingly, the abundance of K. 
pelamis larvae actually increased with distance from the 
reef on both sides of the island (Boehlert and Mundy, 
1994), indicating the possibility of important taxa-spe- 
cific differences in near-reef distributions of tuna larvae. 
The possibility that tuna larvae are generally con- 
centrated near islands and reefs, and not in the open 
ocean, has important implications for the protection of 
these vulnerable life stages. Large numbers of tuna lar- 
vae near shore have been found in only two regions in 
the central Pacific Ocean, and only near oceanic islands; 
therefore further investigation of the generality of this 
phenomenon is required. Considering that patches of 
highly concentrated (10,945 larvae/500 m 3 ) T. maccoyii 
(southern bluefin tuna) larvae in the North East Indian 
Ocean were only 5-15 km in diameter (Davis et al., 
1990a), further investigation of fine-scale patterns of 
larval tuna distribution is also warranted. 
The aim of the present study was to investigate the 
near-reef abundance and on-offshore distributions of 
tuna larvae in the Coral Sea, near the Great Barrier 
Reef, Australia, over a fine (1-10 km) scale. Although 
the sampling design was originally intended for inves- 
tigation of distributions of reef fish larvae (Leis, 1986; 
Leis and Reader, 1991), the sampling scale and grada- 
tion of habitat from near-reef to oceanic in the offshore 
direction also made it appropriate for investigating the 
near-reef distributions of tuna larvae. 
Materials and methods 
Study area and experimental design 
Larval fish samples were collected in an area of the 
Coral Sea between Lizard Island (14°30'S, 145°27'E) 
and 19 km seaward of the outer ribbon reefs of the Great 
Barrier Reef, Australia (Fig. 1). Lizard Island is situated 
approximately halfway across the Great Barrier Reef 
Lagoon (hereafter “lagoon”) where water depths range 
from 25 to 40 m. The outer reefs lie on the continental 
shelf break, beyond which depth increases rapidly reach- 
ing 2000 m within 12 km. There is an abrupt change 
from shallow, protected waters of the lagoon to oceanic 
conditions in the offshore direction. Winds were usu- 
ally from the E to SE during the study period (common 
for this region); therefore the near-reef waters that we 
sampled were on the windward side of the outer reefs. 
Four cruises were conducted to investigate the hori- 
zontal distribution of fish larvae: 1) 2-5 November 1984 
(early November cruise), 2) 17 and 20-22 November 
1984 (late November cruise), 3) 30 January-2 Febru- 
ary 1985 (early February cruise), and 4) 9-13 February 
1985 (late February cruise). On each cruise, six samples 
were taken in each of five on-offshore blocks defined by 
distance (nautical miles, nmi) from the outer reef crest: 
A) 0-0.25 nmi (0-0.46 km), B) 0.25-1.0 nmi (0.46-1.85 
km), C) 1. 0-3.0 nmi (1.85-5.56 km), D) 3. 0-6.0 nmi 
(5.56-11.1 km), and E) 6.0-10.0 nmi (11.1-18.5 km, 
Fig. 1). On each cruise, six samples were also taken in 
the lagoon between Lizard Island and the outer reefs 
(Fig. 1). Samples were taken over four consecutive days 
on each cruise. Offshore transects were planned to be 
conducted over three days, and two samples to be taken 
in each block at randomly chosen distances from the 
reef on each day. All samples in the lagoon were taken 
on the same day. Transects were centered on a different 
reef each day and were started from opposite ends on 
alternate days. Distance from the reef was determined 
by radar reflection off the waves breaking on the outer 
reef crest. Because of this method of measurement, ac- 
tual distance from the reef varied by approximately 100 
m depending on the tide and sea state. Because of bad 
weather on the second cruise, two days elapsed between 
the sampling of lagoon and offshore waters, and only 
two samples could be taken in block A. Offshore tran- 
sects were conducted over four days during the fourth 
cruise because of mechanical problems. 
Sampling procedure 
Quantitative, double-oblique plankton tows were made 
from a 14-m catamaran with a bongo net (cylinder-cone 
mesh design) with 0.85 m mouth diameter and 0.5-mm 
mesh. The net was towed at approximately 1 m/s and 
was fitted with both a calibrated mechanical flowme- 
ter and a calibrated mechanical depth and distance 
recorder. Tows usually filtered 1000-2000 m 3 of water 
with a mean volume (and standard deviation) of 1554 
(585), 1644 (629), 1637 (306), and 1348 (278) m 3 for each 
cruise, respectively. All tows were completed during 
daylight, between one hour after sunrise and one hour 
before sunset. Tows were taken to a target depth of 200 
m on the first cruise and to 120 m thereafter, except in 
the lagoon and block A where they were taken as close 
to the bottom as considered safe. The net hit bottom 
