I have also read a study (can find a link if somebody wants it) that observed bushmen in the kalahari (or somewhere like that - mostly untouched by modern civilisation) and found that the distribution of sleeping patterns (i.e. night owls vs early birds) was pretty uniform. They also found that because of this, there was always somebody in the tribe who was awake at any given point. They theorised an evolutionary advantage to this in that there is always somebody awake to keep watch.
Maybe you are referring to David Samson's study of the Hadza tribe in Northern Tanzania? "Chronotype Variation Drives Nighttime Sentinel-Like Behaviour in Hunter-Gatherers", 2017. [0]
Or in some more entertaining format at Duke University's website [1].
A similar evolutionary argument was put forward in "Why We Sleep"[0] by Matthew Walker. Its mention in the book was less specific, so I'm not sure if it's referring to the same study.
