We previously showed vertebral addition drives body elongation in African cichlids. But intraspecific variation in vertebral count doesn’t predict body shape, having more vertebrae doesn’t mean you’re more elongate. So, intraspecific variation is decoupled from macroevolutionary body shape patterns.

Despite high evolvability, vertebral counts in African cichlids show low intraspecific variation. Correcting for phylogeny: (1) variation doesn't scale with count (no sign of selection), and (2) doesn't differ between lakes & rivers, even if each system is subject to its own rate of count evolution.

Using vertebral count data from >4,500 African cichlids (~500 species), we show that axial regionalisation can shift via changes to AP patterning. However, most variation reflects differences in somite number, with homeotic transformations emerging mainly as a consequence of these somitic changes.

However, despite the focus on the lacustrine radiations, riverine taxa occupy a much wider axial morphospace than the lacustrine species. Which is partly being driven by a stochastic rate of total vertebral count evolution twice that of the highest lacustrine rate.

Elongation of the body is important for ecological adaptation. Lacustrine cichlids (those living in lakes) have repeatedly (and independently) evolved elongate, fusiform bodies supported by higher total vertebral counts, linked to demersal, pelagic, and piscivorous lifestyles.

Consistent with other teleosts, cichlid body elongation often involves adding vertebrae—but it's not the only route. Cranial and post-cranial elongation have co-evolved, revealing multiple axes of morphological change.

Despite its critical role in locomotion, the evolution of the vertebral column in cichlids has rarely been studied. We set out to change that—with the first macroevolutionary analysis of axial morphology across 4861 individuals from 583 species.