Mission Pinocchio: scientifically tracking chameleons... too small to be fitted with tracking devices

Tracking wildlife is always a scientific challenge, particularly when it comes to reliably identifying individuals. For many species, tracking relies on the use of tags, marks or transponders. But what can be done when animals are too small to be fitted with such devices without risk?

This is precisely one of the questions that the Cameleon Centre Conservation is attempting to answer with its new research project, Mission Pinocchio, dedicated to the Calumma pinocchio chameleon, a tiny endangered species endemic to the Vohimana reserve region in Madagascar, managed by our partner L'Homme et l'Environnement.

A scientific challenge: individually identifying very small chameleons

When it comes to reptiles, and chameleons in particular, knowledge of actual population numbers often remains very limited.

Counting wildlife is not just a matter of counting sightings: it involves estimating population sizes, understanding their dynamics and assessing the threats they face.

To do this, it is essential to be able to recognise individuals over time.

However, the smallest species of chameleons cannot be fitted with radio tags, and conventional marking techniques can be too invasive or inaccurate.

A completely non-invasive method of individual identification using image analysis.

Artificial intelligence for non-invasive monitoring

We recently collaborated with Alexandre Langlais, an ecological engineer, biostatistician and AI specialist, to test the feasibility of automatically identifying individuals from standardised photographs.

Initial research shows that, as with certain other species, chameleon scale patterns appear to be unique to each individual and probably relatively stable over time.

Using image classification models, it is possible to compare photographs and determine whether two observations correspond to the same animal.

The results obtained in initial tests are very encouraging and have already made it possible to identify the shooting criteria necessary for reliable recognition, optimise the field photography protocol and define the basis for robust demographic monitoring.

Beyond a certain threshold of similarity between patterns, identification becomes statistically reliable, paving the way for scientific monitoring without capture or tagging.

A promising tool for the conservation of small species

The main advantage of this approach is that it allows for minimal population estimates, tracking individuals over time, minimising animal handling, and easy deployment of the protocol in the field.

In a context where many species remain difficult to study using traditional methods, these tools open up new prospects for conservation, particularly for small reptiles and rare species.

Start of the field programme in Vohimana

With the methodological foundations now in place, Mission Pinocchio will enter its operational phase in the field in the coming weeks, in the Vohimana reserve in Madagascar.

The aim is to set up demographic monitoring over several years in order to better understand the actual size of populations, their stability over time, their use of habitat and the impact of environmental disturbances.

A 5-year programme that still requires funding

To ensure the full development of the Pinocchio Mission over the next five years, we are still looking for financial partners.

The support of donors and institutions is essential to enable fieldwork in Madagascar, the development of image analysis tools, the training of local teams and the publication of scientific results.

Help us

Photos: Lucas Orsini