Male Betta fighting

Top view of the experimental setup designed to investigate the effect of an audience on male-male (Ma, Mb) interactions. F, female audience; M, male audience.

Figure 1

Top view of the experimental setup designed to investigate the effect of an audience on male-male (Ma, Mb) interactions. F, female audience; M, male audience.

There were 20 subjects, each of which interacted twice with the same male for 10 min on each occasion. One interaction was in the presence of a female (audience), the other was in the absence of a female (no audience). Two hours separated the two interactions. The order of these interactions was balanced over the experiment (i.e., each type of interaction was the first to be presented on an equal number of occasions). Each male acted as a subject in one trial (i.e., two interactions) and also as the other male in a trial with a different male.

For the trial with an audience, males were allowed to see the female for 3 min before the male-male interaction began. Males immediately began displaying to one another through the transparent partition after the opaque partition was removed. We measured the following displays of the subject from video recordings: time with erect gill cover, number of bites, number of tail beats, and time at less than one fish length to the partition separating the two males. We analyzed these behavioral measures with Wilcoxon matched-pair signed-ranks tests.

Experiment 2: male audience

We used the same experimental design as for experiment 1 with a male audience. The same behavioral measures were collected. Twelve of the 20 subjects had been subjects in the first experiment, but they interacted with a different male in this experiment. The aquarium holding the male audience differed from that holding the female audience in that the male was placed in the outer part to prevent the male subjects from interacting with the audience rather interacting with one another (Figure 1b). Because of this difference in design, we did not statistically compare the results obtained for male and female audiences.

RESULTS

In the absence of a female audience, males behaved significantly differently from the way they behaved when a female audience was present for all the four measures collected (Table 1). Significantly more males performed more tail beats, spent more time with gill cover erect, interacted farther away from the other male, and performed fewer bites with a female audience than without.

Table 1

Male behavior during male—male interaction in presence or absence of an audience