Using the FishSim Animation Toolchain to Investigate Fish Behavior: A Case Study on Mate-Choice Copying In Sailfin Mollies

This method can help to answer key questions in studying fish behavior especially in the field of mate choice and here to study the fascinating strategy of mate-choice copying. The main advantage of our technique is that we have total control of the appearance and the behavior of the fish stimuli that we use to present live fish during experiments. Visual demonstration of this method is critical as the different steps necessary to create, animate, and present computer-animated fish stimuli using our tool chain need to be precise to ensure a proper workflow.

Begin this procedure with preparation of female body textures with and without gravid spot as detailed in the text protocol. To adjust the viewpoint and set the scene for animation, start FishSim by selecting the FishSim icon in the launcher on the left side of the desktop. Configure the resolution for the presentation monitors and click Launch.

Press F1 on the keyboard to change from viewing mode to editing mode. Adjust the viewpoint to match the dimensions in the presentation monitors by adjusting the camera angle. Rotate the camera by holding the left mouse button and moving the cursor.

To pan the camera, hold the right mouse button and move the cursor. Zoom in and out by holding the middle mouse button and moving the cursor. Click Camera Settings in the editing toolbar and click copy to static cam to set the viewpoint.

Click File, Save Scene, and save the adjusted scene as the new default scene. For design of a virtual male stimuli for presentation during male-choice tests, click File, Load fish model from the dropdown menu, and load the default male sailfin molly template by selecting it from the folder models. Left double click on the loaded fish to select it.

It will be highlighted in the mesh. Click the gear icon in the toolbar to open the fish tool set. A box will pop up with the editing options used to customize the virtual male.

Untick show mesh for a better view of the fish. Change the name to male. Alter the scale of the male by changing the values for X, Y, and Z if needed and click Apply.

Edit the male's texture by clicking Textures in the edit toolbox. Click on a feature of the fish to change it. Click on a texture displayed in the list and it will directly appear and replace the prior texture on the fish.

When the desired male is created, click Apply under config in the edit toolbox and click save fish to disk. Save the new male in the models folder. For virtual male animations for presentation during mate-choice tests, plug in the gaming controller into the USB port of the operating computer and open FishSim.

Click start placing and use the controller to place the fish at any starting position in the virtual tank. Click stop placing. Start recording the fish's swimming path by clicking start recording.

Use the controller to move the fish around the scene. After stopping recording and saving the swimming path, edit the recording to add movement of the male's dorsal fin to mimic male courtship behavior during mate-choice tests. To do so, select the dorsal fin in the dropdown menu in the edit feature.

Select start editing and the complete swimming path will be replayed. Press the L1 button on the controller to raise the dorsal fin at specific points in time. Click Save to save the edited version of the swimming path as a new bag file.

Repeat these steps except to select the gonopodium to add its movement. For virtual male and model female animation for presentation during the observation period, open FishSim. Press F1 to enter the editing mode and click File, Load Scene to load a scene with male and female.

After starting fish steering, select male and female alternately to place them by clicking start/stop placing in the virtual tank. For the recording, select the female fish first from the dropdown menu of the steering panel and create a swimming path with a duration of 10 minutes as before. Click stop recording and save the swimming path on the drive.

Then successively edit the male's swimming path, dorsal fin movement and gonopodium movement as before. Save the final version for later use in FishPlayer. Prepare the MCC experiment as described in the text protocol.

Open FishPlayer on the operating computer and arrange the playlists for the first mate-choice test. Place a live focal female inside the test tank. Let her swim freely, acclimating to the tank and to the presentation of the empty tanks on both monitors for a period of 20 minutes.

After acclimatization, place the focal female in a clear acrylic cylinder in the middle of the test tank to ensure an equal distance to both monitors. Run the playlists on both monitors simultaneously by clicking Play, starting with the first mate-choice test. Before the timer reaches two minutes and 30 seconds, go slowly to the experimental tank and release the focal female from the cylinder by gently lifting it up straight out of the water.

Return to the operating computer. Observe the focal female and have two stopwatches at hand to measure association time with each virtual stimulus male. Stop measuring association time when the timer reaches seven minutes and 30 seconds.

The pause entry will then run for one minute and 30 seconds. Use the pause as handling time to again place the focal female inside the cylinder and write down the time for each virtual male on a datasheet. Before the timer reaches 11 minutes and 30 seconds, release the focal female from the cylinder.

Measure the association time for the next five minutes. Stop measuring association time when the timer reaches 16 minutes and 30 seconds. The pause entry will run for one minute.

Use this handling time to place the focal female inside the cylinder. Write down association times for the second measurement. For each male, sum up association times of both halves of the first mate-choice test.

Calculate whether the focal female had a side bias and which mate was preferred by the focal female as described in the text protocol. Rearrange the playlists so that the prior preferred male will be animated alone during the observation period and the prior non-preferred male will be animated together with the virtual model female. Click Play to continue the second part of the experiment and the entries will be replayed from top to bottom starting with the 10-minute observation period.

After the observation period, a pause of 30 seconds begins. During every pause, an empty virtual tank is visible to focal females. Before the timer reaches 10 minutes and 30 seconds, release the focal female from the cylinder and start the second mate-choice test with five-minute recording for each male.

Measure association times for the next five minutes. Stop measuring association time when the timer reaches 15 minutes and 30 seconds. The pause entry will then run for one minute and 30 seconds.

Place the focal female inside the cylinder and write down the measured time for each virtual male. Before the timer reaches 19 minutes and 30 seconds, release the focal female from the cylinder and measure association time for the last five minutes. Stop measuring association time when the timer reaches 24 minutes and 30 seconds and terminate the experiment.

As predicted for mate-choice copying, preference scores for the prior non-preferred virtual male significantly increased from the first to the second mate-choice test after focal females had been presented with the simulated mate choice of a virtual model female. This effect was found in both treatments, but not in the control where no virtual model female was present during the observation period. The copying score for the prior non-preferred male was significantly higher in treatments with a virtual model female compared to the control.

However, there was no difference between treatments whether the model female had a gravid spot or not. Compared to the control, significantly more focal females reversed their initial mate choice in favor of the prior non-preferred male after having him observed with a virtual model female. This result was again irrespective of whether she had a gravid spot or not.

While attempting this procedure, it's important to know that animating the virtual fish in a natural way using a gaming controller takes practice. Therefore, this step might actually be the most time consuming. Although our protocol is very specific for the study of mate-choice copying, our tool chain offers solutions to study various other questions on fish behavior and even cognition, for example shoreline decisions or predator-prey detection.

The use of computer animation in animal behavior research is a promising approach to manipulate the expression of visual traits and even behavioral patterns in a noninvasive way. It provides a high degree of control and standardization in comparison to use live animals as stimuli.