Daytime activity probably started after the dinosaurs were wiped out.
Ars Technica said:Most mammals are nocturnal, but even those mammals that are not nocturnal have large eyes that are better suited to dim light. Their eyes are more similar to the eyes of nocturnal animals in other clades than to the eyes of animals that spend their time in the daylight. After studying this peculiarity, in 1942 the optometrist Gordon Lynn Walls posited what is called the "nocturnal bottleneck" hypothesis: the ancestors of all modern mammals were once nocturnal.
Walls posited that this bottleneck occurred during the Mesozoic Era, when dinosaurs ruled the day. Mammals couldn’t venture out into the sunlight until that mass extinction event rid the planet of those pesky dinosaurs 65 million years ago. It’s definitely a nice idea, but it has been difficult to prove anatomically or morphologically—precisely because most mammals, regardless of their behavior patterns, have eyes that still look nocturnal. This makes it hard to pinpoint when exactly diurnality arose.
A survey of ancient mammalian behavior could help, but it's obviously difficult to check the behavior of an animal that's extinct. Fossil evidence is scarce, but ancestral features and behaviors can be extrapolated from a broad enough analysis of current species.
Unfortunately, most studies of mammalian activity undertaken thus far have narrowly focused on only us and our favorite lab animals: primates and rodents. Zoologists and geneticists in Israel and England have thus collated the behavior of 2,415 different mammalian species, spanning more than 90 percent of mammalian families and all 29 mammalian orders, to try to reconstruct the ancestral activity patterns of mammals and see when diurnality first showed up.
These researchers combed through the observations noted in journals, field guides, and encyclopedias, classifying each species into one of five activity groups: nocturnal; diurnal; cathemeral (active during both day and night); crepuscular (active only at dusk and dawn); and ultradian (active in cycles of a few hours). They kind of ignored the last two categories and mapped the remaining three onto phylogenic trees that indicate when the species diverged.
They found evidence supporting the main predictions made by the nocturnal bottleneck hypothesis: first, the most recent common ancestor was, in fact, nocturnal. And second, the first diurnal mammals showed up in the Cenozoic Era, around 60 million years ago. It seems that mammals went through a cathemeral phase first. And depending on your view of mammalian evolution, the first taxon to develop diurnality was either the most common recent ancestor of all apes and monkeys or the most common recent ancestor of all… elephant shrews.
...
