Thermal Imaging to Study Stress Non-invasively in Unrestrained Birds

The overall goal of the following method is to assess whether a free living animal responds to an immediate acute stressor using thermal imaging without having to handle the animal. This is achieved by luring a wild bird in front of a thermal camera. As a second step, the bird enters a box feeder of its own accord, where upon the baseline temperature of the skin around the bird’s eye is recorded by a thermal video.

Next, the box feeder is closed behind the bird acting as a stressor, and the skin temperature around the bird bird’s eye is recorded again. Ultimately, the changes in skin temperature between the baseline and the box feed are closing can be compared to determine if the animal exhibits a stress response to the external stressor. The main advantages of this technique over existing methods like the measurement of stress hormones, are that it is non-invasive, can be carried out in unrestrained animals and can even demonstrate the dynamics of the physiological response to stress.

This method can help answer key questions in the fields of ecology, conservation, biology, and animal welfare, such as do animals experience stress. Begin by creating a filming setup where the birds are encouraged to position themselves in front of the camera. The bird enters the setup through a hole in one of the end walls and has access to food near the opposite end wall.

To habituate the birds to the setup, provide a suitable food to the feeder for several weeks prior to the recording. During this period, position a dummy tripod in front of the feeder to allow the birds to become used to the camera as well. On the day of the recording, fill a transparent container bearing a small hole in the center of the top with the same food that was in the feeder.

Then place the container in the box feeder and attach the temperature probe with a small square of black insulation tape of known emissivity to one of the corners of the box so that it will appear in all of the recorded images. Next, attach a thermocouple connected to a temperature logger for recording the temperature of the black tape with a resolution of 0.1 degrees Celsius every second to the laptop. Then place the thermal imaging camera 50 centimeters from the center of the feeding box and connect the camera to the laptop.

Set the camera to record the images for an average of 7.5 frames per second with a timestamp, and to save the images to the computer’s hard drive. Then attach a fishing line to the rotating door of the setup and roll the setup about 20 meters from where the experimenter will be hidden from the view of the birds, but with the setup still in sight. Once a bird enters the feeding box, leave the animal undisturbed in the box for approximately five seconds.

Then pull the fishing line to close the feeding box, taking care that the bird is still at the far end of the box to minimize the risk of injury to the bird as soon as the feeding box is closed. Approach the setup and stand motionless behind the camera for approximately three minutes. Then retrieve the bird from the box and let it go.

To extract the maximum eye region temperature from each frame, open the thermal imaging analysis software and double click on the video of interest. Next, add a new plot for the image maximum, and right click on the plot that exports the I region temperature and time of the frame to A CSV file. Delete all of the lines from the frames where the I region of the bird was not visible from the CSV file.

Repeat for the points with low values outside of the normal range of the body temperature for the bird. These data will allow the manual identification of the temperature spikes greater than 0.2 degrees Celsius between two successive readings during which the tating membrane was pulled over the eye. To measure the ambient temperature.

Download the time and logger temperature data from the temperature logger and export it into a spreadsheet. Then in the thermal image analysis software, draw a square over the black insulation tape covering the temperature logger probe Right click on the square and select Add new plot to obtain the average temperature of the square. Right click the plot to export the data as a CSV file.

Load the three CSV files. Maximum I region temperature from the thermal video, black insulation, tape temperature from the thermal video and air temperature from the temperature logger into R.Then merge all three files. Calculate the difference between the black cancellation tape temperature of the thermal video from the air temperature of the temperature logger.

Then make the correction to the uncalibrated eye temperature data generating a calibrated maximum eye region temperature for each frame. Next, add a value of 0.2 to each eye region temperature to correct for the effect of taking the images through a mesh. Then using the peak search algorithm to extract the highest and most accurate points in the data, automatically carry out automated filtering to remove any of the less accurate low eye region temperature values.

Standardize the timing of the eye temperature response between individuals by subtracting the length of the baseline period in seconds, making zero seconds, the point at which the trap closed. Then standardized the length of each individual time series from the maximum eye region temperatures recorded from before the box was closed. Select the highest value which will constitute the baseline eye region temperature of the undisturbed bird.

Then standardized temperature data between individuals by subtracting the baseline temperature, the highest temperature measured during the period before the trap door shut from the absolute temperatures using linear interpolation. Close the gaps left by the sequences where no peak was extracted and where the eye region wasn’t visible, resulting in a single temperature value per second for each individual. Finally, use the existing R commands to plot the deviations in the maximum eye region temperature from the baseline eye region temperature of all of the individuals from after the trap was closed, along with bootstrapped 95%confidence intervals.

The exposed skin around a enes eye exhibits a yellow color in the thermal image indicating the highest body area temperature. The orange, red, purple, and blue colors observed over the rest of the body area signify progressive degrees of cooler temperatures in tests conducted on 22nd thermal video clips of undisturbed free living blue tits arriving and feeding within the feeder setup as just demonstrated correlations of R greater than 0.7 were found to exist between the maximum eye region temperature recorded during the first five seconds. In this experiment, upon closure of the feeder, the eye region temperature dropped within the first 10 seconds reaching a minimum I region temperature of approximately 1.3 degrees Celsius below the baseline temperature.

The mean temperature drop calculated for each individual separately was 2.0 plus or minus 0.2 degrees Celsius with the lowest point being reached after 9.4 plus or minus 2.8 seconds. Over the next two to three minutes, the eye reach and temperature gradually increased back toward the baseline eye region and temperature without having reached the baseline value by the end of the trial. While attempting this procedure, it’s important to remember this.

It can only be used to determine whether the animal is under stress or not. Further work will be needed to assess whether this method can also measure how stressed the animal is. Our method can demonstrate the response to acute stress, so to immediate stresses, the future work of our lab is going to focus on the response to persistent and chronic stresses.