Identifying individuals within a group is vital for wildlife ecologists. Following a specific animal can help scientists understand the social interactions, migratory routes, preferred habitats, behaviors within certain settings, life spans and reproduction rates of various species.
But distinguishing one animal from another in a group is often difficult. Many methods have been devised, such as tagging, radio-collaring, banding and fin-notching. Unfortunately, these practices can cause injury to the animal and alter its behavior; and they are labor intensive for biologists, who need to capture and sedate unwilling subjects. In addition, artificial markers can malfunction, fade, be shed, rubbed or pulled off, rendering them useless.
That’s why a software program, called StripeSpotter is gaining a lot of attention. As use of this innovation becomes more prevalent in wildlife studies and the results become increasingly available to the public, will interest in conservation and empathy for wildlife grow?
Animals of a different stripe
With StripeSpotter, a computer program that is able to identify individual zebras from a single photograph, field ecologists simply upload a digital photo of a particular animal’s flank. StripeSpotter then analyzes the pixels and assigns a “stripecode.” When a future photo of a zebra is uploaded, it’s run against the stripecode. This highly accurate ID program is currently being used to build a database of plains and endangered Grevy’s zebras in Kenya.
StripeSpotter’s abilities, however, may soon be employed to track other species in popular African safari destinations. In the future, the program may help distinguish leopards, striped hyenas or even elephants by their wrinkles.
Part of StripeSpotter’s appeal is that photo identification is usually unobtrusive and generally results in little behavioral disturbance. But there are challenges. Obtaining useful photos for identification purposes isn’t always easy. Even after long hours of searching for animals, once they’re sighted, the right weather and lighting conditions must prevail in order to take workable photos of the subjects. Even if nature cooperates, circumstances may make it difficult to maneuver into position in order to obtain an angle for a photo that will enable easy identification and comparison. For example, getting the correct orientation of a zebra in the camera frame—both vertically and horizontally—can determine a photo’s serviceability. And deciding on the one individual to focus on and follow from among a large group of animals is often random.
It’s the individual that matters
StripeSpotter’s true value might lie not only in what it can do for science but in what it can do for conservation efforts. Psychologists have known for a long time that a personal, one-on-one connection triggers empathy; the feeling rarely flows to a group. In one well-known experiment, researchers studying generosity gave participants the opportunity to donate to a global, poverty-fighting organization. The first group of participants was told that food shortages in Malawi are affecting more than three million children and given some additional information about how the need for donations was very strong.
A second group of participants was shown a photo of a seven-year-old, Malawian girl named Rokia, who is desperately poor. These participants were told that Rokia’s life would be changed for the better by their gifts. A third group was given a combination of the information that the two previous groups received. A fourth group was shown the photo of Rokia, informed about her situation and then given facts about another child, identified by name, and told that their donations would help this child, too. Interestingly, the group that was told only about Rokia (the second group) gave the most money.
Time and again, additional studies have offered similar results. In another generosity experiment, one group of people was told that a single child needed a lifesaving medical treatment costing $300,000 and was given the opportunity to contribute toward this fund. A second group was told that eight children needed a lifesaving treatment, and all of them would die unless $300,000 could be provided. More people opted to donate to the single child. This is the basis for why we’re so willing to help an individual, named victim, no matter the monetary cost, but turn a blind eye to the unknown, starving masses.
Regarding wildlife, a recent example of this can be found among the wolves of Yellowstone. When a hunter killed Wolf No. 754 last fall, it was a shot heard ‘round the world. Many said it was like hearing about the death of a friend. It’s estimated that a half-million people were familiar with Wolf No. 754, since most of this radio-collared animal’s life played out in front of binoculars and spotting scopes inside a park visited by more than three million people annually. While the single animal’s death caused outrage, appeals to protect all of Yellowstone’s wolves struggle to be heard.
Could software programs such as StripeSpotter be a boon for conservationists? If we start to see wild animals as individuals, will we spend more for their conservation and protection?
Here’s to finding your true places and natural habitats,
Candy
good approach to safeguard the wild animals coz they remain endangered and threatened by humans
Candice:
Your article on the animal ID software is most informative. The segment on identifying a single animal and getting the public’s emotional response to its needs hit home to our fawn rescue organization Kindred Spirits Fawn Rescue. We continue to educate the public on how to live in harmony with the local deer population.
Adding a individual fawn rescue story might go a long way in touching the hearts of the community.
Thank you for the creative idea!
YES, as a conservation biologist trying to get the public to embrace their local wildlife, I would have to whole-heartedly agree that these technological advancements ARE the future of wildlife conservation. Putting the power in the palm of people’s hands is crucial to taking conservation to the next level.
I would like to contact Jason and the organizations he mentioned to help me with my Box Turtle Tracking Project in an effort to save the last of the box turtle population here on Long Island, NY. Such an app would markedly help me out, and I believe would help to save this species from extirpation.
Eric
Cool!
I think as humans we see ourselves as separate to the ‘other species’ (not just another animal in the web of nature), so a program that can ‘improve’ the value we place on wildlife – separating an individual from the masses, in turn will increase our connection with that species as a whole… I hope. Jason’s response highlights the complexity of the programme – if this programme can help with identification of individual rhinos and in the fight against poaching – that would be awesome.
Yes ! If there is no need to catch wild animals, and no harm, this will be very useful for wildlife conservation.
Anyway, wildlife just need their natural habitats if available and no real need to know them that well. Leave them alone to enjoy their wilderness.
As a developer of wildlife pattern recognition software myself (see whaleshark.org) and an implementer of third party algorithms (www.mantamatcher,org), I find that the challenges in wildlife ID include:
1. Applying an algorithm to other species. It is easy to say it can be done, but rarely do algorithms get applied across species or applied correctly (i.e., users rarely understand the limitations of the algorithms).
2. Managing the large volume of mark-recapture data that the ID function is used to filter and match within.
I am working to separate ID algorithms from mark-recapture data management in the Shepherd Project:
https://www.ecoceanusa.org/shepherd
The algorithms should be pluggable and swappable, allowing for broad access and use in the research community. Good management of mark-recapture data is the more important function (e.g., peer review and collaboration, approval of algorithm match suggestions, etc.), and searchability is essential to help researchers begin analyzing he data set.
As more of these algorithms for stripes, spots, blotches, and other patterning characteristics become available, it will be essential to have good data management, especially as the size of the data set and number of marked individuals grows. If the algorithm software focuses on matching and not on data management, they could be more broadly used (in the Shepherd Project for example), allowing the algorithm to fit into multiple studies without forcing the study to match the limited data storage capabilities of the algorithms’ software.
Please check out the Shepherd Project when you have a moment.
Thanks Candice, Good information.