Can individual small mammals be identified from their footprints?

Footprint tracking
We’ve always used footprints to track animals, from hunting and now for conservation. Tracks can reveal which animals are in the landscape, and how they use their habitats. Footprints are an excellent tool for conservation as they are non-invasive, so don’t disturb animals as much as other methods like radio-tracking.


There’s been a renewed interest in footprint tracking recently as computer technology can now help process and identify large volumes of footprint photos. Footprint identification technology (FIT) has been developed by WildTrack to identify species and even individuals from footprint tracks. The software uses markers on the footprints, placed by researchers at key points like the tip of a finger pad, to calculate many metrics used in the identification process. The precision and speed of computer measurements, coupled with computer learning, makes this a powerful tool for identifying footprints to species level, greatly exceeding even specialist tracking knowledge.
So far, FIT has been successfully applied to many large mammals, such as cheetah and rhinos. Here the species can be easily recognised by researchers by their footprints, but the FIT software also enables identification of individual animals, greatly increasing the conservation and research that can be completed. The hope is that the same techniques can be applied to small mammals, whose footprints are much less easily distinguished between species and individual.
Small mammals
Footprint tracking is already used for a few small mammal species in the UK to detect their presence. Hazel dormice and hedgehogs can both be readily identified from their footprints. Unlike mice and voles, hazel dormice have triangular shaped foot pads, whilst hedgehog footprints look a little like small handprints. Footprints are collected using paper and ink, in the form of charcoal and vegetable oil, which are placed in tunnels to protect them from the elements. Both species are under threat in the UK, so it’s vital there are as many research techniques as possible.


However, at the moment it’s only possible to identify a species by eye, not individuals, and sometimes even that is difficult for smaller hedgehogs when their footprints are similar to rats. Being able to identify individuals would mean being able to estimate population counts – a vital conservation tool.
We also collect a lot of other footprints when placing tunnels for hedgehog and dormice. All kinds of small mammals, like wood mice, bank voles and shrews, will readily run through tunnels and leave their footprints. However, most researchers are unable to identify other species by eye – the footprints are just too similar. So there are no easy non-invasive ways of investigating even species presence for most UK small mammal species. This is where FIT software comes in.
The project
PTES’s new intern, Ellie Scopes, is working with the team at WildTrack to provide more footprint data from UK small mammals for the FIT software. These footprints will help train the algorithms to identify more small mammal species, and test whether it’s possible to identify hedgehog and dormouse individuals from their footprints. These are the smallest species for which individual identification has been tried, so the concern is their prints are just too small.
Ellie is going to work with captive animals, such as those in zoos and rescue centres, to collect footprints from known species and individuals. She will feed these images into the computer to train the software and see if it can correctly identify the species and individuals. If successful, the software can be applied to a variety of research and conservation projects where the species and individuals are not known. She hopes that eventually this data can help generate software that could identify footprints from something like an app, so more people can see which small mammals are in their local area.
This project is only possible thanks to our generous donors. Can you help by donating today?
Header image credit Matt Parkins