The ability to distinguish (in the field) the sex of wildlife that show limited sexual dimorphism is, at best, challenging but is nonetheless essential to ensure appropriate conservation strategies are implemented. We developed a technique using highly varied images, similar to those taken by amateur photographers, to accurately sex individual powerful owls (Ninox strenua) through the measurement of facial features. Powerful owls show extremely limited sexual dimorphism and have been almost impossible to sex without genetic analysis, however, we report a process that results in the unambiguous assignment of sex. Image selection was extremely important; images where the owl’s head and body faced towards the camera, in an alert posture, were selected. We derived five facial features (inter-nostril distance, inter-eye distance, head width, forehead height and facemask), all measured in pixels to account for variation in the resolution and the size of images. Ratios were then produced from these facial features enabling standardised metrics, which we tested with regard to the assignment of sex. Logistic regression analysis indicated that the ratio of head width to the inter-nostril distance was the key sexually diagnostic relationship, giving 100% accuracy in sex determination (assessed against known-sex birds). We also identified a sex-specific plumage character; where facemask feathers protrude beyond the profile of the contour feathers of the head, the bird was a female, whereas when these feathers did not protrude beyond the profile, the bird was a male. For powerful owls, simple image analysis reliably determines the sex of individuals and has potential as an extremely cost efficient approach, that can provide essential information on species dynamics such as sex ratios, social organization and behaviour. The use of photographic images is especially beneficial for species that are cryptic or are extremely difficult to capture. The quantification and analysis of images captured by amateurs enables a greater contribution of citizen scientists to conservation research.
Read full abstract