2 5^ 



>-20- 

 U 



O 



Li. 



5- 



1 2 3 4 5 6 7 8 



BOUT LENGTH 



Figure 6. — Frequency of bout length ( number of aggressive acts 

 per bout per animal) for the simultaneous introduction of 14 

 pairs of male Macrobrachium rosenbergii. 



gressive acts/animal per 15-min period). Simul- 

 taneous introduction animals exhibited a total of 

 68 aggressive acts occurring in 43 bouts in =28 

 animals), while the prawns from the prior resident 

 experiment exhibited only five aggressive acts in 

 five bouts in = 32 animals). 



Discussion 



The results indicate that when M. rosenbergii 

 compete for shelter at least three factors, i-elative 

 size, prior residence, and length of time contes- 

 tants are paired, play important roles in determin- 

 ing who occupies a shelter. It has long been recog- 

 nized that in crustaceans relative size plays a 

 large role in determining dominance (Allee and 

 Douglas 1945; Bovbjerg 1953, 1956, 1960; Lowe 

 1956). More recent observations have confirmed 

 the size dominance relationship (Hughes 1966; 

 Crane 1967; Griffin 1968; Hazlett 1968; Dingle 

 and Caldwell 1969; Warner 1970; Rubenstein and 

 Hazlett 1974; Jachowski 1974; Molenock 1976; 

 Sinclair 1977). However, relative size does not ap- 

 pear equally important in all species (Hazlett and 

 Estabrook 1974). 



In prawns, relative size strongly influences the 

 outcome of competition. When two prawns en- 

 counter one another in an area new to both, the 

 larger animal usually has the advantage. Often 

 these encounters are characterized by a limited 

 series of pushes with one or the other chela. The 

 function of the pushing might be threefold: 1) to 

 test their opponent's weight (rest inertia), 2) to 

 determine the opponent's molt state, and 3) to see 



if the opponent is capable of pushing back (has 

 chelae). Other crustaceans appear to measure 

 their opponent's physical strength by means of 

 physical interactions involving the chelae (Griffin 

 1968; Schone 1968). In Cambarellus shufeldtii, 

 claw removal causes dominant animals to drop in 

 rank (Lowe 1956). InM. rosenbergii deaths related 

 to agonistic behavior usually occurred near ec- 

 dysis and often the first appendages lost during an 

 agonistic encounter were the chelae (Peebles 

 1977). 



Smaller animals have been observed success- 

 fully defending shelters from attempted occupa- 

 tion by larger congeners (Bovbjerg 1953; Griffin 

 1968; Sinclair 1977). This is related to the prior 

 resident phenomenon and it is central to Nobel's 

 (1939) definition of territory. Resident M. rosen- 

 bergii, regardless of their relative size, success- 

 fully retained their shelters. The mechanism the 

 residents employed apparently was not limited to 

 direct physical interaction. Immigrants and resi- 

 dents seldom fought. Generally immigrants were 

 inactive upon placement into a tank housing a 

 resident. The immigrant's aggressive behavior 

 was well below its counterpart in the simultane- 

 ous introduction group. Only occasionally (Figure 

 1) did the immigrant seek out the shelter. This 

 latter behavior is in direct contrast to the control 

 gi'oup. A control group animal was usually back in 

 its shelter within 1 min after reintroduction. Pos- 

 sibly an exocrine was an agent of communication 

 between resident and immigrant prawns, since a 

 novel environment did not inhibit exploration in 

 animals of the simultaneous introduction experi- 

 ment; and animals from the control experiment 

 reintroduced into tanks contaminated with their 

 own exocrines, rapidly entered their shelter. 



The advantage conferred upon resident M. 

 rosenbergii appears to disappear within a short 

 period of time. The smaller resident can defend its 

 shelter against intrusion for no longer than a few 

 days (Figure 1). Apparently relative size can over- 

 come the prior resident effect if resident and im- 

 migrant continue to encounter one another. Simi- 

 lar observations were reported by Lowe ( 1956). In 

 the case of the C. shufeldtii, a dominance hierar- 

 chy was established before shelters were intro- 

 duced. Dominant C. shufeldtii displaced subordi- 

 nates from occupied shelters. In my experiments, 

 M. rosenbergii first exhibited territoriality as de- 

 termined by the presence of the prior resident ef- 

 fect. Territoriality then broke down, due to con- 

 tinued encounters, into simple dominance. 



909 



