Heather Schofield, Althia Rickard, Malisa Rai

April 20th, 2018

Bio 345 – Animal Behavior


 Betta splendens are model organisms for understanding aggressive behavior in aquatic species. Aggressive displays have been well documented and include the opercula display and hitting with the tail. These behaviors have evolutionary contexts that have created differences in the frequencies that these behaviors are displayed in the presence of various audience types. Male betta fish were observed without bystanders and with different gendered bystanders to determine differences between these displays. There were significantly less gill flares in the presence of a female bystander. The presence of a male bystander saw significantly less tail flicking. The data suggested that gill flaring is typically more common in the presence of other males whereas tail beating is more closely associated with the presence of a female bystander. It is believed that tail beating evolved from an aggressive display to a courtship display due to chase away selection.


Communication between animals occurs in a variety of forms, such as auditory, electrical, or visual signals. The use of signals allows animals to send information to one another and modify behaviors based upon the understanding of those signals[7]. A few common reasons that animals interact are for mate attraction, territory or predator defense, and social integration. Aggressive behavior in animals is common for defensive mechanisms and acquiring resources [6]. Sometimes these signals accurately depict an animal’s true size and ability which is called honest signaling. When the signals are misleading this is described as dishonest signaling, however, evolution tends to favor honest signaling [8]. Gill flaring, also called the opercula display, is an honest signal that male Betta splendens use to intimidate an opponent and suggest fighting ability and strength. This behavior of gill flaring is energetically costly because the display prevents the fish from using the gills to obtain oxygen which requires stamina [10]. Traits that are used for communication are evolutionarily and historically limited based upon the phylogeny of the species. Only a pre-existing trait such as physiological abilities and behaviors can become an evolutionary adaptation. These become incorporated into the larger population if they increase the overall fitness of the individuals with that trait [8].

Another aspect of communication is eavesdropping where an individual receives information about another from a signal that was not intended for it. This is seen in a variety of animals and often the eavesdropper behavior toward the signaler is influenced by the signal that was intercepted. An example of eavesdropping is when snakes use the mating calls of frogs to find their prey [9]. Other studies on eavesdropping as a communication mechanism have suggested that male Betta splendens alter their aggression displays depending on the context of the situation. It was found that the sex of a bystander, among other contextual elements, has significant effects on the display behavior of Betta splendens. The males showed different levels of aggression depending on the presence or absence of an audience, the gender of the audience, their own reproductive state, and the amount of resources held [2].

Siamese fighting fish, Betta splendens, are a model organism for observing fish aggression. Typical aggressive behaviors that have been previously documented on Betta splendens include, but are not limited to, gill flaring, fin spreading, tail beating, and biting. Physical displays such as tail beating and biting only occur when the benefits of territoriality outweigh the costs of potential harm of fighting [4].Tail beating has been more closely associated with courtship rituals and therefore are thought to occur more often in the presence of female Betta Splendens. The gill flaring behavior was shown to be greatest in the presence Betta splendens who were males [3]. This suggests that there may in fact be a gender-based bystander effect on the aggression displays in male Betta splendens.

This study aimed to investigate aggressive behaviors in male Betta Splendens to determine the response to the presence and absence of different gendered bystanders. This was done to shed light on the evolutionary and historical outcomes of specific aggressive behaviors in Betta splendens. We hypothesize that Betta splendens will have differential aggressive displays that are dependent on the gender of the bystander. We predicted that if a male Betta Splendens is placed in a tank with a mirror, it would display more gill flaring than tail flicking when there is a male Betta Splendens bystander. We also predicted that if a male Betta Splendens is placed in a tank with a mirror, it would display more tail beating than gill flaring when there is a female Betta Splendens bystander.


Study Species:

Five mature male crowntail Betta splendens and one mature elephant ear female from the local Petco were purchased. The males varied in size, shape, and color to be representative of the variation observed in male betta fish. They are naturally found in freshwater ponds of Southeast Asia. Captive-bred and wild males both exhibit strong and stereotyped aggression in defending their territories against intruding male conspecifics.  Each male fish was assigned a number for later identification.

Housing and Care:

 All fish were kept in a small isolated glass fish bowl with about 700mL of lukewarm tap water and black pebbles for resting on. These fishbowls were initially housed on the fourth floor of the KSC Putnam Science Center in the greenhouse and then later moved to the greenhouse office for a cooler environment. The water in all fish bowls and the experimental tank was treated with  API  Splendid Betta Complete Water Conditioner from Petco to prepare the  tanks for suitable living conditions [4].  The conditioner is used to remove chemicals from the treated tap water because these chemicals are toxic to fish [5]. Fish bowls remained segregated to prevent interactions. Fish were fed generic betta fish food from Petco daily. Fish bowls were cleaned every other day or more often if needed, all water was treated with the API conditioner before fish were placed into the water. The water was maintained at about neutral pH and the water temperature was held at room temperature.

The female betta perished the day after purchase and was replaced with a similar elephant ear female at Petco. Male #1 perished after the self-trials and was removed from the experimental data set. He was not replaced and the trials continued with only Male #2, #3, #4, #5, and the female. It was believed that the heat of the greenhouse caused these fatalities and the fish were moved inside to the greenhouse office to remain cooler.

Experimental Tank Set-up:

A half-gallon tank with a divider was filled with tap water and treated with conditioner. One male betta was placed into a plastic bag and allowed to acclimate to side A of the experimental tank. An opaque cover was placed over the divider to prevent interactions prior to the trials (Figure 1). A male or female betta acclimated to side B, the bystander side. After each trial the tank was emptied and cleaned and the water was treated with API conditioner.

Figure 1: Experimental half-gallon tank with a divider. Side A will contain the male Betta that will be exposed to the mirror. Location of the mirror is shown on the back wall of side A.  Side B will be used to house the bystander that will be male or female.

Experimental Design:

A small mirror was attached to the wall of the experimental tank on side A. Male betta #2 was allowed to acclimate to the experimental tank on side A. Then for five minutes, male aggressive displays of gill flaring and tail flicking were counted and recorded. Three trials of solitary displaying, trials without bystanders, were done for each fish. These trials served as a control to compare solitary versus bystander behaviors.

Male #3 was placed on side B of the experimental tank and was allowed to adjust to the new environment. The opaque cover for the divider prevented interactions. Once both fish in side A and B, separated by the opaque divider, were acclimated to the water, the trial began. The opaque divider was removed so that the bystander was visible. For five minutes, the number of tail flicking and gill flaring that betta #2 displayed were recorded.

After the five minutes the fish were placed in their respective fish bowls for acclimation while the experimental tank was cleaned. Male #2 was put back into side A to be tested with male #4 and this was repeated for male #5 and the female betta. All five males were tested in side A for solitary mirror displays and bystander effects. Each male was tested with the female three times. Each set of male trials (#2, 3, 4, 5) was a replicate experiment.    

Focal Betta Behaviors:

This study aimed to investigate the differences in gill flaring and tail beating behaviors of male betta fish when exposed to either a male or female bystander betta fish. The behaviors are described in Table 1. These behaviors were chosen because they are easily spotted and have been described in published papers on this species.

Table 1: Descriptions of aggressive betta fish behaviors derived from primary literature search [3] and observations.

Behavior: Description:
Gill Flaring Opercula display, extension of gills
Tail Beating Using the tail to hit or attempt to hit an opponent or object


Statistical Analysis 

            The averages of each trial condition (self, male bystander, female bystander) were calculated with standard deviations and standard errors using Microsoft Excel. A One-Way ANOVA with post hoc comparisons were done for gill flares and for tail beats using an online calculator. The post hoc comparisons were used to determine the significance of the data between the different trial conditions. The average number of gill flares for each trial condition (self, male bystander, female bystander) were calculated and compared. The average tail beats for each trial condition were calculated and compared. The significance of each type of trial condition was calculated for each behavior (gill flaring or tail beating).


            Gill flaring behavior appeared to decrease in frequency in the presence of a female bystander. The statistical analysis indicated that there were significant differences between the average number of the two aggressive displays with a male versus a female bystander for the three different trial conditions. The results showed that there were significantly less gill flares in the trials that had the female bystander (P= 0.0023, F=6.762, Figure 2) as compared to a male bystander and the self-trials. Gill flares for the self-trials and male bystander trials were not significantly different suggesting that it was the presence of the female creating the differences. The different letters above the data bars indicate if the data sets are significantly different.


Figure 2: Average number of Gill Flares exhibited by the focal male Betta during the five- minute trials. For the three trial conditions the F value = 6.7622 and P-value = 0.0023. A male Betta splendens was in the experimental tank with a mirror on side A and observed for five minutes displaying to a mirror with either nothing, a male bystander, or a female bystander in side B of the experimental tank. The number of gill flares in the female bystander trial condition was significantly less than the other two trial conditions.

The self-trials and female bystander trials appeared to have similar means for tail beating when displayed graphically. There was a significant difference between the number of tail beats in the presence of a male versus a female bystander (P=0.0015, F=7.3241, Figure 3). There were greater tail beating events in the presence of the female bystander and the self-trials when compared to the male bystander trials. This data showed that in the presence of male bystanders there were a greater number of gill flares and that there were more tail beatings in the presence of the female bystander.

Figure 3: Average number of Tail Beats exhibited by the focal male Betta during the five-minute trials. F-value = 7.3241. P-value = 0.0015. A male Betta splendens was in the experimental tank with a mirror on side A and observed for five minutes displaying to a mirror with either nothing, a male bystander, or a female bystander in side B of the experimental tank. Male bystander trial conditions had significantly less tail beating events than the other two trial conditions.


Aggressive behaviors are typically seen in response to territory defense and mate acquisition. These behaviors have evolutionary contexts that has been selected for over thousands of years. The Betta splendens is a classic fish pet that has been studied extensively for their aggressive behaviors. It has been proposed that there is gender based bystander effects that determine the frequency that specific aggressive behaviors are displayed due to the audience effect [1]. This study aimed to investigate whether there was a gender bias for gill flaring and tail beating in male Betta splendens. Various colored and sized males were observed displaying to a mirror without a bystander, with a male bystander, and with a female bystander.

It had been hypothesized that there would be greater gill flare displays in the presence of male bystander whereas a female bystander would induce more tail beatings. The experimental trials supported this hypothesis. The male bystander trials displayed significantly more gill flares compared to the female bystander suggesting that gill flaring is typically used as an aggressive defense against invaders. The female bystander trials showed significantly more tail beatings suggesting that this behavior has both aggressive and mating evolutionary contexts. This suggests that there are differences in aggressive displays due to the presence of different gendered bystanders. Future experiments or observations could look at how or if closely related species exhibit the same aggressive behaviors based upon the sex of bystanders or if this is unique to Betta splendens.


[1] Claire Doutrelant, Peter K. McGregor, Rui F. Oliveira; The effect of an audience on intrasexual communication in male Siamese fighting fish, Betta splendens, Behavioral Ecology, Volume 12, Issue 3, 1 May 2001, Pages 283–286,

[2] Dzieweczynski, Teresa L., et al. “Audience Effect Is Context Dependent in Siamese Fighting Fish, Betta Splendens .” OUP Academic, Oxford University Press, 29 Sept. 2005,

[3] Dzieweczynski, Teresa, et al. “Opponent Familiarity Influences the Audience Effect in Male–Male Interactions in Siamese Fighting Fish.” Animal Behaviour, Academic Press, 15 Mar. 2012,

[4] Egar M. J, Lynn E. S, Ramenofsky. M, Sperry S. T, Walker G. B. (2007). “Fish on Prozac: a simple, noninvasive physiology laboratory investigating the mechanisms of aggressive behavior in Betta Splendens.” American physiological society. Retrieved from

[5] Mohrman, Eric. “What Does Conditioner Do for an Aquarium?” Pets, The Nest, 2007,

[6] Romano, Donato, et al. “Multiple Cues Produced by a Robotic Fish Modulate Aggressive Behaviour in Siamese Fighting Fishes.” Scientific Reports, Nature Publishing Group UK, 2017,

[7] Rosenthal, Gil G. “Spatiotemporal Dimensions of Visual Signals in Animal Communication.”Annual Review of Ecology, Evolution, and Systematics, vol. 38, 2007, pp. 155–178. Illiad, doi:10.1146/annurev.ecolsys.38.091206.095.

[8] Schofield , Heather. “Communication Between Animals.” Schofield Investigations, KSC Open , 1 Mar. 2018.

[9] Strauss, Amy. “Eavesdropping in the Animal Kingdom: Sneaky Creatures Just Trying to Get Ahead.” Thats Life Science, 5 Sept. 2017,

[10] Verbeek, Peter, et al. “Differences in Aggression between Wild-Type and Domesticated Fighting Fish Are Context Dependent.” Animal Behaviour, vol. 73, no. 1, 2007, pp. 75–83., doi:10.1016/j.anbehav.2006.03.012.










Leave a Reply

Your email address will not be published. Required fields are marked *