Researchers analysed 3,116 paired location records collected over 112 days in an agroforestry landscape, tracking mantled howler monkey groups and the nearest worker at 15-minute intervals. The groups moved during only 14% of intervals, and their likelihood of moving was not related to worker numbers, sound levels or proximity to workers.

When the monkeys did move, their movements were not directed away from workers. On days with more workers, groups were farther from people without travelling more themselves, and much of that separation was linked to where and how the work occurred rather than clear relocation by the monkeys.