For over a century, scientists have puzzled over a fundamental mystery in our evolutionary history: how did mammals go from sprawling like lizards to striding like cats and dogs? This transition -- from a sprawled stance (like a lizard) to an upright (parasagittal) posture -- marked a pivotal moment in mammal evolution. While the earliest non-mammalian synapsids, the ancestors of living mammals, had a sprawling posture, researchers debated when and how the upright postures of modern mammals evolved.
Now, a groundbreaking study in PLOS Biology led by Dr. Robert Brocklehurst, a former postdoctoral fellow in the Department of Organismic and Evolutionary Biology (OEB) at Harvard University, offers a surprising answer: the path to upright posture wasn't linear, but full of unexpected detours, evolutionary experimentation, and dramatic anatomical upheaval.
"The evolution of mammals has previously been characterized as a series of steps from sprawling, to semi-sprawling, to upright," said Brocklehurst. "However, what we discovered was a more nonlinear evolutionary progression throughout mammalian history."
All mammals -- from bats and whales to moles and humans -- share a distinctive way of moving: they hold their limbs underneath their bodies, unlike the sprawling posture with limbs out to the side. This posture enables more efficient movement and is tied to mammals' ability to adapt to diverse lifestyles, from digging to flying. This dramatic transition was also accompanied by big changes in limb bone shape and mechanics. To assess these changes, the team analyzed the humerus (upper arm bone) of over 60 non-mammalian synapsid fossils and 140 living animals, including mammals, reptiles and amphibians.
Using a novel analytical technique, pioneered in senior author Professor Stephanie Pierce's lab (also in OEB), researchers mapped each bone's surface to measure traits like length, mass distribution, muscle leverage, and torsion (the degree to which the bone twists along its length). These traits correlate with specific modes of locomotion and allowed the researchers to reconstruct posture and locomotion in the fossils.
"By correlating bone shape and limb biomechanics with posture, we could test how well the fossil bones were optimized for specific functional tasks, like upright walking versus sprawled walking," Brocklehurst said. The researchers achieved this by mapping the fossil non-mammalian synapsids onto a functional adaptive landscape, similar to a topographic map, with peaks and valleys that relate to high and low performance of different locomotor postures.
"We expected to see a neat progression -- from sprawling pelycosaurs to a bit more upright therapsids, then cynodonts, then fully upright mammals," said Brocklehurst. "Instead, we found bursts of innovation."
The findings suggest that mammal evolution involved a series of adaptive radiations, with each major ancestral groups exploring a range of forelimb functions and postures -- some of which were closer to modern mammals, others not.
"The path to upright posture wasn't a straight line," says Pierce, "the ancestors of mammals weren't steps on a ladder with modern mammals at the top. Mammals have been evolving and radiating into many different niches and habitats throughout their history, and their postures reflect that variation."
One fossil, a close relative of today's marsupials and placentals, showed bone features consistent with a modern upright gait, suggesting that fully parasagittal postures evolved relatively late in mammalian history as opposed to previously held hypotheses. This result supports recent work from the same lab on the backbone and hindlimb.
"Our work challenges the idea that posture changed gradually and early on," said Pierce, "instead, it shows that upright posture and locomotion were a late evolutionary innovation, not an early defining trait of the mammalian lineage."
The researchers also challenge the long-held idea that the earliest non-mammalian synapsids sprawled in a similar way to living lizards or crocodiles. "Our study showed that most synapsid limbs functioned differently than those of modern reptiles. They're not just copies of reptiles, but distinctive animals in their own right that are a little different from anything that's alive today," says co-author Kenneth Angielczyk of Chicago's Field Museum.
To compare such a wide range of bones -- spanning hundreds of species, including those hundreds of millions of years apart in age and wildly different in shape -- the team had to overcome major technical hurdles. Traditional methods that describe shape in similar structures didn't work. So, the team re-engineered an existing R software package designed for a different task, transforming it into a novel "slice-based" landmarking tool tailored for this study. Co-author Magdalen Mercado, former undergraduate student in the Integrative Biology program at Harvard, helped gather the extensive dataset as part of her senior thesis and research in the Pierce lab.
The study builds on a rich scientific legacy -- both at Harvard and in paleontology. Pierce, who is also Curator of Vertebrate Paleontology in the Museum of Comparative Zoology (MCZ), noted that, "Researchers and former MCZ curators, like Alfred Sherwood Romer and Farish Jenkins, Jr., were grappling with these same questions a century ago. Now, with new tools and data, we can revisit those ideas and see the story more clearly."
This study marks the first large-scale evolutionary analysis of mammalian posture using quantitative biomechanics. But that's just the beginning. The team is now building detailed models of forelimbs in select fossil species to understand how joints and muscles functioned in ancient animals, offering even deeper insights into the evolution of mammalian motion.
As Brocklehurst put it: "Understanding how mammals came to walk upright isn't just about bones, it's about uncovering the dynamic history of life on Earth."
Funding was provided by the US National Science Foundation (DEB1754459 and DEB1754502), by the Harvard Museum of Comparative Zoology and the Wetmore Colles Fund.
Robert Brocklehurst is currently a postdoctoral research associate in the Department of Biological Sciences at the University of Massachusetts, Lowell. Magdalen Mercado is currently a graduate student in the Committee on Evolutionary Biology at the University of Chicago.
