We mammals have some nice features (temperature regulation, big brains that seem too seldom used, etc), but we’re unremarkable when it comes to vision. I assumed this was a standard evolutionary trade-off but the story is weirder than that, as I learned this week when reading a new paper about mammalian retinas.
Recall that rods are basically night-vision photoreceptors, able to detect small amounts of light, and cones are the cells that are less sensitive but respond to color. A diurnal animal (active during the day) is expected to emphasize cones, whereas a nocturnal animal will emphasize rods. But we mammals, oddly, are different (quoting the authors):
Mammals represent a notable exception to the general trend in which diurnal vertebrate species possess cone-dominant retinas. This deviation reflects the long-term effects of a selective nocturnal bottleneck that occurred before the Cretaceous-Paleogene mass extinction event, preceding the mammalian radiation.
In other words, mammals diverge from the sensible arrangement, because we so recently emerged from a nocturnal ancestor. Apparently we humans are comfortable just getting by, but some lineages have evolved retinas that (if I may) make more sense. Specifically, diurnal ground squirrels have retinas that look, well, diurnal, with more cones than their nocturnal relatives (e.g. mice) and than their big-brained clumsy more distant relatives (you and me). How did they do this? And how did they do it so fast?
The new paper harnesses gene expression analysis in a perfect diurnal model system (the 13-lined ground squirrel which, yes, sports 13 stripes and is more beautiful than the American flag which also sports 13 stripes but let’s move on), that the authors could compare to mice (nocturnal). They find specific changes in the tempo and timing of generation of rods and cones, and show that the changes are explained by altered timing of the expression of particular transcription factors. This in turn is explained by species-specific regulatory sequences, in a new and exciting example of gene regulatory evolution beloved of evo-devo fans (like me).
The paper is very nicely written with a lot of clear and interesting discussion. Check it out! (Best to read during the day, with your lame rod-heavy retinas.)
Heterochronic transcription factor expression drives cone-dominant retina development in 13-lined ground squirrels
In eLife, 5 January 2026
From the group of Seth Blackshaw at Johns Hopkins School of Medicine
Snippet by Stephen Matheson
Image credits: Top: Figure 1A from Weir et al. linked above (CC BY). Bottom: a thirteen-lined ground squirrel from Wikipedia (public domain).
