*Study Limitations*

Being able to identify a reliable body feature to measure the body temperature of an animal with a thermal camera introduces a non-invasive method for measuring animal welfare. Focusing on one body part reduces outliers and abnormally high readings (Birri's max temp = 46.5 ◦C) that are also picked up by the thermal camera. The camera reads reflections and shadows that are present in the images therefore the camera is not reading the true temperature of the koala. This limitation has been found in similar studies, such as [10], where the solar radiation produced erratic temperature readings when reflected o ff other objects in the enclosure. This can been seen in the readings of the back and feet as there is such a high variation between the temperatures during the day for some koalas compared to another koalas in the same area. This variance is reflected in the CV%. The back and feet have the least consistent readings as koalas tend to sit in positions on the side of the trees or in branches which allows the sun to reflect o ff their backs. The nose of the koala was di fficult to image because of the variability in positioning of the koala on the perches, hence it was not used in the image analysis. The camera was also picking up reflections o ff the artificial tree in the enclosure. Other limitations that are expressed in this study are distance from the thermal camera to the subject, as well as plants and other objects obstructing the analysed object.

**Author Contributions:** Conceptualization, E.N., W.M., A.P.; methodology, E.N., W.M., A.P.; formal analysis, E.N., A.P.; investigation, E.N., A.P., W.M.; resources, E.N., W.M.; writing—original draft preparation, A.P.; writing—review and editing, E.N.; project administration, E.N.

**Funding:** This research received no external funding.

**Acknowledgments:** Thank you to James Vandersteen who carried out the data analysis under Edward Narayan's supervision. We thank the sta ff at the Wildlife Sydney Zoo for their valuable support and providing access to the study site.

**Conflicts of Interest:** The authors declare no conflict of interest.
