Scientists collar marauding baboons
Population pressure is exacerbating conflict between humans and wild animals in South Africa. Can data science bring about peace? Andrew Masterson reports.
For residents in some parts of Cape Town urban violence is a real and worsening trend. The threat, however, comes not from criminal gangs, but from baboons.
For many years, people in the South African city lived in relative harmony with an estimated 500 wild chacma baboons (Papio ursinus), congregated into roughly 11 troupes, each led by an alpha male.
The monkeys lived on a diet of plants and insects, and some even foraging along the city’s beaches for mussels and other molluscs. However in recent years a combination of urban growth and tourism operators keen to attract their attention by offering food rewards have altered the baboons’ behavior, with unpleasant results.
The monkeys now frequently raid rubbish bins, shopping trolleys, cars and houses for food, even brazenly ripping things out of people’s arms. Needless to say, this has caused alarm and concern, and largely destroyed the concordat that existed between Cape Town’s humans and their distant primate cousins.
In 2010 a team of scientists and national park officials, collectively known as the Baboon Research Unit, attempted to solve the problem by catching and killing the alpha males. The strategy remained in place until last year and the results, observers agree, have been disastrous.
Without alpha male leadership, the troupes have fragmented and – often dominated by remaining juvenile males – become smaller, less predictable, and more reckless. There are serious concerns about whether Cape Town’s famous wild baboons can survive much longer, as well as equally serious worries regarding how much damage they can do before they die out.
Now, however, a team of researchers has made an important advance in better understanding the behaviour of the remaining colonies – a result that might soon provide clues to a sustainable and humane management strategy.
Scientists led by Gaelle Fehlman of Swansea University in the UK designed and produced special collars that could be attached to the Cape Town primates. The collars were fitted with accelerometers, which broadcast a complex set of data describing the animals’ movements day and night.
Using this type of bio-logging device with wild primates is generally considered to be challenging. In part, this is because the target animals can be difficult to catch, may succeed in removing collars, and may change their behaviours in response to wearing them. In part, too, the complex systems of movement and behavior exhibited by non-human primates mean that the amounts of data generated are enormous, and require much time and massive computing power to interpret.
Fehlman’s team applied machine learning to the data collected, which reduced the number crunching considerably.
The resulting study, published in the journal Animal Telemetry, reports that 93% of the baboons’ “time budget” was taken up by just six broad state behaviours, of which foraging, running, walking and resting were the most common.
The study is the first time such extensive data on wild primate behavior has been gathered. Most wild non-human primate studies rely on a human observer – whose presence may well result in the primates changing the way they act.
Having demonstrated that custom collars and machine learning can together provide the data needed, Fehlman’s team is now hoping to discover more about the stresses present between Cape Town’s human and baboon populations.
“We are particularly interested in using acceleration data to document baboons’ behaviour in this environment to understand baboons’ behavioural responses to anthropogenic change,” they write.