Identifying Aggressive Behavior in Zebrafish in Relation to Stimuli

By:  Deanna Soper


            The zebrafish is a great model organism that has been, in recent years, studied for many purposes including developmental biology, neurology, and genetics.  It is a good model organism due to the fact that it has a relatively small genome which has been sequenced, the embryo develops in a sac that has high clarity which makes it easier to study, the short reproductive cycle, and it belongs to a class of bony fishes called teleost.  Teleost make up half of all vertebrate species on Earth, currently known.  Hence, the study of zebrafish biology makes for a good link to other teleost species by which methodologies for other studies could be modeled after (1). 

Many studies are currently underway into the workings of zebrafish molecular and cellular biology and during background research it was discovered that not much in the way of zebrafish ethology has been evaluated.  It is imperative to understand not only the organism on the genetic level but also on a behavioral level so the link between genes and gene expression can be made.  During observation of these fish we noted that while they have schooling behavior, they do exhibit aggressive acts towards one another.  It was noted that the variance of aggressive acts was notably varying among different genetic lines.  We questioned what factors affect aggression such as the presence of a predator or the presence of another group of their species.  It was at the time of this discussion that our main focus question arose.  “Are zebrafish of the F2 hybrid (SH/Nadia) genetic line more aggressive in the presence of a predator or in the presence of a school of their own species?” 



Experimental Conditions


This experiment was carried out between the dates of June 28, 2005 and June 30, 2005.  We used a total of 40 different fish making up 10 different shoals. 

The room is environmentally controlled and kept at a temperature of 26 C and has a day:night length of 14:10 h.  The water in this lab is filtered by mainly biological filtration but also carbon filtration.  A drip system is used to maintain water levels in home tanks which are 18.9 L and hold anywhere from seven to ten fish.  The tanks do not have any substrate.  Aeration of the water is maintained by a continuous system of multiple air tubes with pipettes at the end to narrow the bubbles.  Water quality parameters are measured and maintained.  Each tank is numbered to identify groups of fish and also their genetic lines. 

In this lab there are three genetic lines that are being researched.  TM1, this is a lab-created genetic line, Nadia, which is the wild-type and was originally collected from an area in India about 40 miles East of Calcutta and SH, another lab generated genetic line (2).  The fish that were chosen to work with were of a hybrid variety of SH and Nadia cross. 

  All experimental fish were approximately 28 to 32 weeks old and range in size from approximately 30 mm to 40 mm.  The experimental fish were obtained as eggs from a supplier and hatched out in this lab.               

The experimental tank is 115 liters, has white aquaria rock substrate, and is partitioned into three sections.  The first section is 24 M3 and holds a total of 22.8 Liters.  The middle section is 73.3 M3 and holds a total of 69.7 Liters.  The last section is equal in volume to the first section.  The first section has a shoal of four stimulus zebrafish of the same approximate size and age of the experimental fish.  The middle section of the tank is the area where the experimental shoal is placed for evaluation.  The last section of the tank contains a model of a predator which was made from a cast of a deceased Etroplus canarensis.  It is hung from a piece of fishing line that is anchored into the dorsal side of the fish where the dorsal fin meets the anterior portion of the fish.


Experimental Design


            Our experiment was based on a comparison of two environmental conditions:  exposure to the view of a predator and the view of another shoal.  The shoal and predator make ups are explained in the environmental conditions section.  The test ran for a total of three days.  A sum of forty fish were observed and broken up into shoals of 4 fish.  A total of 22 repeated trials for each experimental condition were observed and data collected.  Black Plexiglas blinds were used to hide both the stimulus shoal and the predator from view.  Fish were netted from the home tank into the experimental tank, utilizing a 300 ml beaker with 100 ml of water.  Once placed into the experimental tank, an acclimation period of 10 minutes was used so that fish may become accustomed to their new habitat.  The experimenter stood 1.83 M from the experimental tank, hidden behind another row of aquaria so as to not distract the fish.  The control condition was that in which data was collected while both blinds were down.  By use of the scan method the experimenter scored the number of bites and chases, as described in table 1, for a time period of five minutes.  The experimenter held one counter in each hand keeping track of bites and chases.  The data was then recorded into a data table.  The black Plexiglas blind between the stimulus shoal and experimental shoal was raised and fish were allowed a 1 minute time period to rest after distraction by the lifted blind.  Data was collected in the same manner as for the control conditions, again for a time period of five minutes.  The blind was lowered and a two minute rest period was allowed before the predator blind was removed.  Once the predator blind was removed another 1 minute rest period was allowed before data collection started.  Again, data was collected in the same manner and length of time.  The predator blind was lowered and another two minute rest period commenced.  During the next trial with the same group of fish, another sequence of presented conditions occurred.  First, the stimulus shoal was exposed, then the control condition was presented, and finally the predator was shown.  After two trials with the same group of fish, they were removed and another group was netted and experimented on in the same manner. 






Data Analysis



Data Analysis

            We took the mean between each trial of bites and chases then summed the two values to get our mean number of aggressive acts for each trial.  We then averaged the means for each trial to get the total mean for each experimental condition.  Standard deviation, standard error, confidence intervals, and a two-sample t-test were conducted.  The t-test gave us the P values which were looked at for significance.   


Data Results

            Our average number of aggressive acts for the control condition was 10.43 with a confidence interval of 15.1 + 5.8.  For the shoal condition our mean of aggressive acts was 5.5 with a confidence interval of 8.1 + 2.9.  Within the experimental condition, predator, the mean number of aggressive acts was 11.5 with a confidence interval of 19.7 + 3.3.  While the P value when comparing all three experimental conditions was the lowest between the predator and shoal its significance did not fall below .05.  It calculated to be 0.185. 








            While our results did show a noticeable difference between the conditions of shoal and predator, when analyzed statistically there was no significance.  We felt we could not ignore the difference in value as well as the fact that the standard error was less for the shoal condition.  This data tells us that the fish acted much more consistently during the shoal condition. During observation, it was noted that the fish behavior did change during the shoal condition.  Their behavior tended to be more sporadic during the predator and control portions of the experiment.  We concluded that it is not that the predator increased aggressive acts, it is that the shoal decreased aggression to some degree.



1.  Teh, C., Serguei P., and Vladimir K.  2005.  New ways to admire zebrafish:  progress in functional genomics research methodology.  BioTechniques 38:897-906


2. Westerfield, M., and Sprague, J.  Accessed June 30, 2005.  The Zebrafish Model Organism Database.  URL: