ter at the end of a 7-day period following immigra- 

 tion. 



Push: an aggressive act where one animal 

 pushes one of its chelae against the body of 

 another animal. 



Nip: an aggressive act where one animal closes 

 down the tips of its chela on the body part of 

 another animal. 



Tete-a-tete: a type of aggressive act charac- 

 terized by a head to head confrontation with at 

 least one nip or one push. The tete-a-tete appeared 

 to be difficult enough in orientation from the push 

 and the nip to be placed in a separate category. 

 Further observation and analysis might not sup- 

 port this separation. 



Shove: an aggressive act where one animal 

 holds both chela forward and parallel while charg- 

 ing into the flanks of another animal. 



Bout: an agonistic exchange between two ani- 

 mals where at least one aggressive act occurred. A 

 bout was considered terminated when aggressive 

 acts stopped or one animal moved away and was 

 not chased. Bouts were measured in units of ag- 

 gressive acts. 



Bout length: the number of aggressive acts that 

 occurred during a bout. 



Body characteristics: standard length (cen- 

 timeters*, right and left chelae length (centime- 

 ters), weight (grams), and carapace length (cen- 

 timeters). 



Body size index: the number of body characteris- 

 tics in which an animal was larger. It was derived 

 as follows: animal A larger than animal B in stan- 

 dard length and right chela length, then A's body 

 size index is two. In the Prior Resident Experi- 

 ment three body characteristics were measured, 

 thus the maximum body size index in this experi- 

 ment was three. In the Simultaneous Introduction 

 Experiment five body characteristics were mea- 

 sured, thus in this experiment the maximum body 

 damage index was five. 



Results 



Control Experiment 



Ten out of 11 animals were in the shelter on 

 every observation period before mock immigra- 

 tion. The remaining animal was in the shelter on 

 one of the three observation periods. The same 10 

 were in the shelters on all observations following 

 mock immigration, while the same remaining one 

 was never observed in a shelter after immigration. 



It was concluded that the act of netting had no 

 effect on shelter use. 



Prior Resident Experiment 



Shelters were occupied on every observation by 

 every animal during the preimmigration week. 

 Following immigration all shelters were occupied 

 on every observation period. On several occasions 

 more than one animal was in a shelter during the 

 first two observation periods following immigra- 

 tion. However, by the end of the week, observation 

 period 4, one animal was in a shelter while the 

 other was usually at the opposite end of the tank. 

 When the data were examined by immigrant ver- 

 sus resident for shelter use over the 7-day period, 

 an interesting change became apparent (Figure 

 1). On the day of immigration, residents were oc- 

 cupying shelter significantly more often than im- 

 migrants (Binomial Test,P = 0.044, Siegel 1956). 

 By the second observation period and for the re- 

 maining two observations there were no sig- 

 nificant differences between residents and immi- 

 gi-ants in frequency of shelter use (Binomial Test: 

 day 2 after immigration,P = 1.0, day 3,P = 0.814; 

 day 7,P = 0.814). 



Examining the data for the effect of size (Figure 

 2) revealed that successful animals were sig- 

 nificantly larger than their unsuccessful paired 



12 3 4 



OBSERVATION PERIOD 



Figure l. — Shelter usage by observation period for 18 pairs of 

 Macrobrachium rosenhergii. The data from prior resident exper- 

 iments are summed for the 18 pairs. During observation period 1, 

 18 residents (circles) and 7 immigrants (dots) were inside shel- 

 ters. On observation period 1 there were seven cases of double 

 occupancy; observation period 2, two cases; and observation 

 period 3, none. 



907 



