FISHERY BULLETIN: VOL. 75. NO. 3 



RESULTS 



Interactions Prior to Spawning Season 



Prior to the onset of spawning in each study, 

 there had developed a clear dominance hierarchy 

 based on size, with the largest fish of each group, a 

 male, being dominant over a smaller male and 

 still smaller female. In turn, the smaller male was 

 dominant over the female. Prior to the reproduc- 

 tive season, the majority of interactions among the 

 three fish consisted of aggressive behavior. During 

 various hours of the day the aggression, initiated 

 particularly by the dominant male, served in part 

 to limit the access of the subordinate male and 

 female to different areas of the tank, such as the 

 feeding area and shelter site (Olla in prep.). 



Aggression was manifested at varying levels of 

 intensity with the more intense involving the pur- 

 suit of a fleeing subordinate by a dominant, which 

 we termed a chase. Prior to such an encounter a 

 dominant often rapidly approached (swam to- 

 wards) a subordinate. The subsequent chase could 

 last as long as 30 to 45 s, with the fish swimming 

 the length of the tank and at speeds reaching 100 

 to 150 cm/s. The most intense but rarest encounter 

 involved a chase accompanied by the dominant 

 biting a subordinate on any area of its body, which 

 we termed nipping. 



Aggressive encounters could also be quite sub- 

 tle, with a subordinate exhibiting a change in its 

 location, either vertically or horizontally, to a new 

 position 0.5 to 1.0 m away, which we termed dis- 

 placement. The behavior of a dominant causing 

 this response often did not appear to differ from its 

 forward swimming motion. Displacement of a 

 subordinate occurred either as a dominant ap- 

 proached or simply turned towards it, as much as a 

 full tank length away (10.6 m). Then there were 

 instances in which a similar action of a dominant 

 did not elicit any response by a subordinate. This 

 variation in response by a subordinate was due to 

 our not being able to assign an observable cause 

 with regard to the actions of the dominant. We 

 could only infer, through a subordinate's behavior, 

 the generation of an aggressive intention signal 

 by the dominant male. 



Aggression by the dominant also caused a sub- 

 ordinate to assume a posture which we interpreted 

 to be submissive, which involved the subordinate 

 tilting its dorsal surface towards the dominant at 

 an angle ranging from 5° to 90°. Frequently, when 

 a subordinate was swimming about the tank and 



588 



approaching an area in which the dominant was 

 present, it would show the submissive posture as it 

 bypassed and clearly avoided the dominant. The 

 distance at which this would occur varied, ranging 

 from 1.0 to 3.0 m. 



Onset of Reproductive Period and 

 Courtship Behavior 



The most obvious manifestation of the approach 

 of reproduction was the change in aggression di- 

 rected toward the female by the dominant male. 

 Beginning in early April 1975 (Study 1 ) and in late 

 January 1976 (Study 2), a rapid approach of the 

 male, which had previously represented the initi- 

 ation of a chase, became functionally transformed 

 into a component of the courtship repertoire. Now 

 when the male approached, when within 5 to 10 

 cm, he veered off to one side or the other. The 

 female was neither displaced nor showed any 

 change in posture. We defined these acts of the 

 male as rushes to distinguish them from ap- 

 proaches which formerly caused displacements 

 and were aggressive. Rushes were directed at the 

 female whether she was active or resting. At times 

 as the male veered off, the magnitude of the water 

 displacement from the force of the caudal thrusts 

 was great enough to stir the adjacent sand and 

 cause the female to be moved several centimeters. 

 Rushes were observed approximately 2 wk (Study 

 1) and 7 wk (Study 2) prior to the first spawning. 



The female, previously limited in her access to 

 different areas of the tank, now was more mobile 

 and concurrently began to show changes in her 

 behavior towards the dominant. Sometimes im- 

 mediately after the male's rush, the female fol- 

 lowed him at a distance of approximately 0.5 to 

 1.0 m. The duration of the following behavior was 

 usually short, lasting no more than 2 to 5 s. If the 

 male did not initiate another rush, one of the pair 

 simply swam away. 



Another change in the female's behavior to- 

 wards the dominant male was her resting in areas 

 in which the dominant was resting. While in 

 Study 2 this generally occurred along the walls of 

 the tank or in the feeding area, in Study 1 it often 

 focused around the shelter. On occasion when the 

 dominant male was resting inside the shelter, the 

 female often settled at the base of the structure, or 

 sometimes actually entered and came to rest 

 alongside the male within the same tube or in a 

 different one. 



While the female of Study 1 appeared to play a 



