392 
Aggr e g a ti° ns of fish do occur over rocky 
mounds on a much smaller scale in shallower 
water. Immature Dascyllus albisella do this over 
individual coral heads and seek concealment in 
the branches of the coral when alarmed. How- 
ever, it is unlikely that the very large aggrega- 
tions of Chaetodon miliaris use for shelter the 
features of bottom topography above which 
they aggregate. This statement would also apply 
to Chromis verater. 
Fish concentrations on or over banks and in 
the vicinity of oceanic islands have long been 
noted by fishermen. Tuna fishermen in the 
eastern tropical Pacific have found concentra- 
tions of tuna in the vicinity of offshore banks 
regularly enough to make such topographic 
features of special interest (Bennett and Schae- 
fer, 1965). 
Four hypotheses have been offered to explain 
the apparent greater abundance of marine life 
about such topographic features, three of which 
would apply to banks as well as islands. The 
margins of continents, under many circum- 
stances, would have similar effects on the abun- 
dance of marine life. The four hypotheses, which 
are not mutually exclusive, are as follows: 
(a) nutrients from land runoff (Gran, 1931); 
(b) vertical movement of water transporting 
nutrients into the euphotic zone (Moore, 1949) ; 
(c) increased productivity through the growth 
of benthic algae in relatively shallow depths 
(Sargent and Austin, 1949); (d) the trapping 
of deep scattering layer organisms (Isaacs and 
Schwartzlose, 1965). 
A discussion of the first three of these possi- 
ble mechanisms, which involve means by which 
the primary production is increased, is presented 
by Jones (1962) in connection with the dis- 
covery of larger standing crops of zooplankton 
as the Marquesas Islands are approached. 
The pattern of fish concentrations as observed 
off western Oahu seem to accord best with a 
food resource which may be provided by the 
trapping of deep scattering layer organisms as 
suggested by Isaacs and Schwartzlose (1965). 
However, if this is the correct hypothesis, the 
observed relation among topographical features, 
fish concentrations, and deep scatters is a com- 
plex one, and is affected by elements that are 
not obvious. 
PACIFIC SCIENCE, Vol. XXII, July 1968 
CONCLUSIONS 
1. At least three well-defined terraces were 
discernible from the "Asherah” (Fig. 1): (a) 
the Lualualei Terrace deeper than 180 meters, 
(b) the Mamala Terrace at depths of 70 to 120 
meters, and (c) the Penguin Banks Terrace 
shoaler than 70 meters. 
2. Vertical and near-vertical rock escarp- 
ments separate the Mamala Terrace from the 
Lualualei Terrace. In many places these escarp- 
ments were over 35 meters high and north of 
Makua Valley there were areas of rounded 
boulders at their bases; in some areas caves were 
present (Fig. 9). Between the Penguin Banks 
Terrace and the Mamala Terrace a broken line 
of reef rock outcrops extended up above the 
level of the terraces. These outcrops were from 
5 to 10 meters in height and generally aligned 
parallel to the shore (Fig. 8). 
3. Associated with these bottom structures 
were communities of the benthic biota. There 
appeared to be two major biotopes, the terraces 
and the rocky outcrops and escarpments. Each of 
these biotopes was separable into two portions 
based on the presence or absence of dominant 
species. 
The terraces were largely sand covered, rather 
barren of fishes or obvious benthic fauna except 
for extensive beds of a hatchet clam Pinna 
muricata, which were both extensive enough and 
dense enough to constitute a biocoenosis. 
The escarpments lying between the Lualualei 
Terrace and the Mamala Terrace had at irregular 
intervals large concentrations of fish associated 
with features of the crest. Concentrations of fish 
were also observed with the reef rock outcrops 
between the Mamala Terrace and the Penguin 
Banks Terrace. These concentrations appeared 
to differ significantly in both dominant species 
and the proportions of other species. 
4. The majority of the species of fish and 
those most abundant within the range of depths 
observed from the "Asherah” were species com- 
mon or abundant in shallow water. 
5. Offshore transport of calcareous sand 
was evident to the diving limit of the "Asherah” 
(180 meters). On the Mamala and shoaler 
shelves, large sand "channels” and intercon- 
nected sand patches were present (Figs. 1 and 
