KREKORIAN, SOMMERVILLE, and FORD: LOBSTER BEHAVIORAL INTERACTIONS 



whether one or both animals were 

 moving, and in some cases direct 

 physical contact was made. 

 2) Two or more heterospecifics were 

 touching or within 30 cm of one 

 another, were not moving, and 

 exhibited no aggressive interaction. 

 Social contact during movement 

 was distinguished from approach 

 by the absence of direct movement 

 of heterospecifics toward one 

 another. That is, the animals were 

 not on a collision course. 



II. The following categories were employed for 

 the reactor: A) No response, B) Walk away, 

 C) Abdomen flex, D) Threat, and E) Attack. 



A) No response — No change in the overt 

 behavior of the reactor. 



B) Walk away — Movement away from an 

 approaching, threatening, or attacking 

 heterospecific, using the walking legs. 



C) Abdomen flex — Movement away from 

 an approaching, threatening, or attack- 

 ing heterospecific, using rapid flexion 

 of the abdomen. 



D) Threat— The description(s) of threat 

 used for H. americanus, P. interrup- 

 tus, and C. antennarius provided above 

 for actors (I-B) was also used for the 

 reactor. 



E) Attack — The description(s) of attack 

 used for//, americanus, P. interruptus, 

 and C. antennarius provided above for 

 actors (I-C) was also used here for the 

 reactor. 



III. Additional categories used on the data sheet 

 and their definitions were as follows: 



A) Roaming — Slow or moderate walking 

 about the tank. During roaming, direct 

 interaction between heterospecifics or 

 conspecifics does not occur. For P. 

 interruptus, roaming by an individual 

 was recorded only if it occurred for 31s 

 or more during a 1-min interval. For 

 H. americanus and C antennarius, 

 the amount of time they were observed 

 roaming during each 1-min interval 

 was recorded. 



B) Wall — Like many decapods, P. inter- 

 ruptus exhibits thigmotactic behavior. 

 When shelter is absent individuals are 



found with their bodies in contact with 

 a solid object. In our no shelter con- 

 dition, the tank wall was the only solid 

 object present. A lobster that had some 

 part of its body, excluding the antennae, 

 within 30 cm of the side of the tank for 

 31 s or more of each 1-min interval 

 was recorded as wall. The location of a 

 lobster within the tank that was greater 

 than 30 cm from the side of the tank 

 for 31 s or more of each 1-min interval 

 was recorded as no wall. 

 C) Group — A congregation of two or more 

 subjects, each within 30 cm of one 

 another. Thus, in a group of five 

 P. interruptus, a maximum distance of 

 120 cm would separate lobsters at 

 opposite ends of a group. 



RESULTS AND DISCUSSION 

 Agonistic Behavior 



The frequency and outcome of agonistic behav- 

 ioral interactions between //. americanus and 

 P. interruptus for conditions involving shelter 

 and no shelter are shown in Table 2. Similar 

 data also are presented in this table for interac- 

 tions between P. interruptus and C. antennarius. 



Homarus vs. Panulirus with No Shelter 



There were a total of 2,515 //omarws-initiated 

 behavioral interactions by actors for the five repli- 

 cate H. americanus-P. interruptus no shelter 

 experiments. Twenty percent of the behavioral 

 interactions were classed as social contact, 40% 

 as Homarus approach, 24% as Homarus threat, 

 and 15% as Homarus attack (Table 2). Sixty- 

 eight percent (1,700) of the //omarus-initiated 

 behavioral interactions occurred during the 1510- 

 1700 h observation period (lights off). 



There were a total of 1,683 responses by 

 reactors to //omarus-initiated interactions. 

 Thirty seven percent of these were classed as no 

 response, 49% Panulirus walk away, and 14% 

 Panulirus abdomen flex (Table 2). Thus, P. 

 interruptus was displaced a total of 63% of the 

 time when //. americanus initiated a behavioral 

 interaction. Panulirus interruptus was never 

 observed to threaten or attack H. americanus in 

 //omarws-initiated behavioral interactions. 



A total of 227 Panulirus-ivaiiaied behavioral 

 interactions were recorded in the five replicate 



1151 



