Behaviour of the Soldier Crab — CAMERON 
slower than that of normal feeding (phase 5). 
This phase occupies 10-15 minutes. Those crabs 
which are first to emerge do not immediately 
feed in this slow fashion, but run about, stop, 
remain motionless, run further, and so only 
gradually begin to feed. The slightest disturb- 
ance at this stage, such as the flight of a bird 
overhead, results in their immediate retreat be- 
low the surface, from where they emerge again 
after about 3 minutes. 
4. Trekking Phase 
Within 15 minutes of emergence the crabs 
simultaneously begin walking toward the wa- 
ter. This spectacular migration is termed the 
"trek.” Usually the movement is down the 
beach, but if surface water is present in chan- 
nels or run-offs the direction of the trek is across 
the beach toward these channels. Walking of 
individuals is interrupted by frequent short 
stops to feed, but the movement of the popula- 
tion as a whole does not stop until the water is 
reached. Trekking occupies 15 minutes at the 
most, and is nearly always performed in a vir- 
tually straight line from the area of emergence. 
There is no cohesion among the individuals 
during the trek. Each crab makes its own way 
down the beach and the population appears dis- 
persed at random. Generally, a distance of about 
18 inches separates the individuals, which re- 
flects the fact that they have emerged in prox- 
imity, and have proceeded at the same rate. 
Juveniles do not take part in the trek proper. 
They usually move about 50 yards in the general 
direction taken by the adults, and then wander 
about feeding at this level of the beach. 
5. Feeding Phase 
When they have trekked to areas which ap- 
parently are suitably moist, the crabs increase 
their rate of feeding until it is extremely rapid. 
They progress as they feed, the direction of 
movement being more or less along the beach 
parallel to the water. Frequently, they transverse 
the Zoster a - covered areas on beaches where the 
grass occurs. 
McNeill (1926) and Lazarus (1945) have 
described the mechanics of soldier crab feeding. 
Portions of the substratum are lifted to the buc- 
cal cavity and are sampled by the mouthparts. I 
observed that material unsuitable for ingestion 
227 
accumulates at the bases of the third maxilli- 
peds, and is discarded in the form of pellets 
(Fig. 6) which either drop off, or are wiped 
away from the ventral surface of the body by a 
lateral movement of one of the chelae. A similar 
deposition of discarded material is performed 
by Uca spp. (Miller, 1961) and Dotilla mic- 
tyroides (Tweedie, 1950). Feeding pellets de- 
posited by crabs in the van are generally left 
undisturbed by following crabs. 
As the feeding phase progresses, the initial 
rapid rates of walking and of deposition of 
pellets slow down. Aggregation into groups of 
about 100 crabs takes place, and the individuals 
continue to feed, although the groups dissolve 
and reform continuously. After about V 2 hr, the 
formation of larger aggregations, called armies, 
takes place. Such groups are distinguished from 
the former feeding groups in that the indi- 
viduals of armies tend to remain in company, 
whereas previously they lacked this cohesion. 
Feeding continues. The armies increase in size 
by the addition of recruits either singly or in 
groups, until aggregations of many hundreds 
are formed. The duration of the feeding phase 
is variable. Usually it lasts for about 1 hr, but 
often can be for as long as 2 l /i hr. With dimi- 
nution of the feeding rate, the armies begin to 
move over the shore, frequently remaining in 
puddles and run-offs and performing no visible 
activity. 
Examination of the gut contents of soldier 
crabs was made by dissecting the alimentary 
canal under the microscope, and staining the or- 
ganic matter present with Toluidine Blue and 
Methylene Blue. Almost all the material in the 
gut was organic debris. Some very fine silt was 
present, but there were no sand grains in the 
cardia as reported by Lazarus ( 1945 ) . Diatoms 
occurred in all, though not as a major consti- 
tuent of the contents, and one specimen con- 
tained gastropod eggs and a nematode. These 
data agree with those of Lazarus (1945) who 
stated that the food of Mictyris is "organic frag- 
ments and minute organisms” extracted from 
the sand. However, she remarked that, although 
"the mouthparts must discard much of this use- 
less sand the gut is usually full of it either in 
its original state in the cardia or ground into a 
finer consistency in the intestine.” Since soldier 
