472 
PACIFIC SCIENCE, VoL XXII, October 1968 
18 
Fig. 5. Graph of numbers of stomatopod larvae taken in Kaneohe Bay during years 1964-1965. Each group 
of three bars represents one day’s haul. The bars represent runs A, B , and C in that order. 
viduals fought among themselves less frequently, 
even when the aquarium was fairly crowded. 
When put together in an aquarium, the Gono- 
dactylus falcatus killed the Pseudosquilla cili- 
ata even when the P. ciliata were considerably 
larger. This suggested the hypothesis that the 
greater aggressiveness of Gonodactylus falcatus 
might have been an important factor in its re- 
placement of Pseudosquilla ciliata in the coral 
head habitat. 
Most of the Gonodactylus falcatus collected 
for these studies were taken from dead heads of 
the coral Porites compressa. However, this spe- 
cies is also found in the dead bases of living 
heads of Pocillopora meandrina and from clumps 
of the coralline alga Porolithon , and occasional 
individuals have also been seen in bare rock or 
moving about on the surfaces of coral heads. 
When specimens of Gonodactylus falcatus were 
taken from coral heads they were consistently 
wedged between the branches of the coral near 
the base. They were usually found with the 
body flexed, the telson covering the anteroventral 
part of the body. In this position the sides of 
the animal are protected by the branches of the 
coral, the dorsal surface is covered by thick 
exoskeleton, and the ventral surface is within 
the curve formed by the animal doubling upon 
itself. 
Pseudosquilla ciliata was taken from two types 
of habitat: from burrows which the animals 
had dug in fine muddy sand and silt, and from 
areas with a similar substratum but with a great 
deal of algal growth (Fig. 2, 3), chiefly A can - 
tbophora , where the animals were seen moving 
about on the bottom in shallow water. When 
