The Pliohippus Horse and Its Place in Horse Evolution

Pliohippus was a genus of horse that roamed North America from roughly 12 to 6 million years ago, during the late Miocene. For much of the twentieth century, it was considered the direct ancestor of modern horses, but that interpretation has been revised. More recent anatomical and taxonomic work has shown that a separate lineage, Dinohippus, sits closer to the branch leading to today’s Equus. Pliohippus remains a pivotal figure in the story of equid evolution, though, because it was among the earliest horses to approach a single-toed foot structure, and its history illuminates how messy and nonlinear evolutionary “progress” really is.

Why Pliohippus Was Once Considered the Ancestor of Modern Horses

The idea of horse evolution as a straight line from tiny, multi-toed forest browsers to large, single-hoofed grazers is one of the most famous (and most misleading) narratives in paleontology. For decades, textbooks drew it as a ladder: Hyracotherium at the bottom, then Mesohippus, Merychippus, Pliohippus, and finally Equus at the top. Pliohippus held the penultimate rung because it appeared to be the first horse that was essentially one-toed, making it a logical precursor to the modern horse.

That straight-line picture started to crack as researchers looked more carefully at skull anatomy, particularly the shape of the facial fossae, the depressions in the skull in front of the eye socket. Pliohippus has deep, pocket-like facial fossae that are quite different from the shallower depressions seen in Equus. Meanwhile, another genus called Dinohippus, which overlapped with Pliohippus in time, had facial features much more consistent with those of modern horses. In 1955, the paleontologist James Quinn formally erected Dinohippus as a separate genus, pulling out species that had previously been lumped under Pliohippus. By 1988, the Italian paleontologist Augusto Azzaroli had more thoroughly differentiated the two lineages, making a strong case that Dinohippus, not Pliohippus, was the more likely ancestor (or close relative of the ancestor) of Equus.¹

This reshuffling is not just a footnote of naming conventions. It changed the understanding of which features matter most in tracing horse ancestry. Pliohippus shares the single-hoof condition with Equus, but that turns out to be a case of convergent evolution, or at least an independently reached endpoint along a closely related branch, rather than a direct inheritance. The deep skull fossae of Pliohippus are considered a derived trait that separates it from the Equus lineage, while Dinohippus retains a more conservative skull structure that matches up with Equus more cleanly.

A Tangled Taxonomy

Many mid-Miocene horse species were at various times assigned to Merychippus, Pliohippus, or Protohippus. As more specimens turned up and analytical methods improved, the boundaries between these genera shifted repeatedly. Quinn’s creation of Dinohippus in 1955 was just one of several major revisions over the latter half of the twentieth century, with additional taxonomic frameworks provided by researchers including Hulbert, MacFadden, and Kelly through the 1980s and 1990s.¹

The oldest recognized species in the genus, Pliohippus mirabilis, lived during the Middle Miocene. A 2025 revision supports the idea that P. mirabilis represents the basal species of Pliohippus and may have been the ancestral form for the broader radiation of the Equini tribe in North America during the Neogene.¹ That means Pliohippus, while not the direct ancestor of Equus, may have given rise to a broader group of lineages that includes Dinohippus and, through it, modern horses. Think of it less as a dead end and more as a close cousin whose immediate family went on to produce Equus.

The subfamily Equinae as a whole was enormously diverse. One recent synthesis recognizes 114 valid species of Equinae across North America, Central and South America, Eurasia, and Africa during the Plio-Pleistocene, with the evolutionary history of this group understood through competing cladistic hypotheses.² Pliohippus was only one branch in what was once a sprawling bush of horse-like animals, many of which looked superficially similar but belonged to distinct lineages.

The Foot That Changed the Story

The most striking thing about Pliohippus, and the feature that anchored its former starring role in the horse-evolution narrative, is its foot. Most three-toed horses in the Equinae maintained side digits that were roughly 75 to 85 percent the length of the central digit, and that proportionality stayed remarkably consistent from about 20 million years ago down to around 5 million years ago. Pliohippus was the outlier. Some specimens were functionally single-toed, while others retained three toes but with side digits that were drastically reduced compared to the standard ratio.³

That variability within the genus is itself interesting. It suggests that the transition from three toes to one was not a clean, all-or-nothing switch but a process that could fluctuate even within closely related populations. Some Pliohippus individuals were essentially monodactyl; others had vestigial side toes. This kind of variation is exactly what you would expect to see in a trait that is under active evolutionary selection but has not yet been fixed in one direction across every member of the group.

Earlier three-toed horses relied on their side digits more than they might appear to. Biomechanical modeling of fossil center metapodials, the main weight-bearing foot bones, across 13 genera spanning the equid family tree from Hyracotherium to Equus has shown that without accounting for the load-bearing role of side digits, three-toed horses as late as Parahippus would have experienced bone stresses too high to be physiologically sustainable. When the side digits are modeled as weight-bearing, horses from the base of the radiation all the way through to Equus maintained a similar safety factor against fracture.⁴ In other words, the side toes were not just evolutionary leftovers hanging on uselessly. They were load-bearing structures, and their reduction had to be compensated for by changes in the central digit’s bone geometry.

For Pliohippus, this means its greatly reduced side toes were viable only because its central metapodial had thickened and strengthened enough to handle most or all of the load alone. The evolutionary “engineering” had to keep pace with the digit loss, and in Pliohippus we see an animal that was well along in that transition, though not always all the way there.

What Did the Ancestral Three-Toed Foot Actually Look Like?

To appreciate how unusual Pliohippus was in its time, it helps to look at what the earlier horses’ feet were like. Mesohippus, which lived roughly 37 to 32 million years ago, had three functional toes on each front foot, each with a complete set of bones from the wrist to the hoof tip. The central toe was dominant, being the largest and widest, but the two flanking toes were fully formed and bore weight. Interestingly, some Mesohippus specimens show a tiny remnant of a fifth metacarpal as a short tubular projection, and both the second and fourth metacarpals carried ridges on their surfaces that resemble features still visible in the modern horse.⁵

Those ridges and fusion lines are echoes of the ancestral structure, ghosts of a multi-toed past that persist even in today’s single-hoofed horse. The splint bones in a modern horse, those thin remnant bones running alongside the cannon bone, are the evolutionary descendants of what used to be full, functional metacarpals in animals like Mesohippus. Pliohippus occupied the intermediate stage where those side structures were visibly shrunken but had not yet been pared down to the near-invisible splints of Equus.

Diet and Habitat in a Changing World

Pliohippus lived during the late Miocene, a time when North American landscapes were shifting from mixed woodland and savanna toward more open grassland. The chemical signatures locked in dental enamel can reveal what an ancient animal ate, because different types of plants leave different carbon isotope fingerprints. Analysis of enamel from Pliohippus potosinus, a species from central Mexico dating to the Clarendonian through early Hemphillian land mammal ages, shows that it was a mixed feeder, eating a combination of C3 plants (trees, shrubs, and cool-season grasses) and C4 plants (warm-season grasses and some sedges), with a substantial proportion of C4 material. The oxygen isotope data from the same teeth indicate it lived in a savanna environment.⁶

This mixed-feeding pattern fits what we know about the transitional landscapes of the period. Pliohippus was not a pure grazer the way modern plains horses are. It was an opportunist, browsing on leaves and woody material while also taking advantage of expanding grasslands. That flexibility may have been one reason the genus persisted for several million years even as habitats shifted.

The broader Equinae radiation tracked environmental change closely. In the Great Plains, distributional patterns of equine species were patchy during the middle Miocene, when speciation rates were high and multiple lineages coexisted over relatively small areas. By the late Miocene, as the clade began declining, distributions became less fragmented but overall diversity dropped. Climate change has been identified as the primary driver of both the radiation and the eventual decline, with species tracking their preferred habitats as grasslands expanded and woodlands shrank.⁷

The Late Miocene Decline and What Survived

The late Miocene saw a dramatic thinning of horse diversity in North America. At its peak, the equine adaptive radiation had produced a dizzying array of forms, from small browsers to large grazers, coexisting in overlapping ranges. As global temperatures cooled and seasonal drying intensified through the late Miocene and into the Pliocene, many of these lineages disappeared. Pliohippus was among the casualties, last appearing in the fossil record around 6 million years ago.

What survived was a much narrower set of lineages, including Dinohippus, which persisted into the Pliocene and gave rise to or closely resembled the earliest members of Equus. Equus itself eventually diversified into the horses, zebras, and wild asses we know today, spreading from North America into South America and across the Bering land bridge into Eurasia and Africa. The irony is that horses ultimately went extinct in their North American birthplace during the late Pleistocene and did not return until Spanish colonizers brought domesticated descendants back in the sixteenth century.

Pliohippus was part of the wave that crested and broke before that later chapter unfolded. Its disappearance was not unique; it was one of many equine lineages lost as climate shifts reorganized ecosystems across the continent. The geodispersal that had helped equine species colonize different regions during favorable periods could not save them when habitats contracted faster than populations could follow.⁷

Why Horse Evolution Is Not a Ladder

Pliohippus is one of the best examples of why the old “ladder of horse evolution” image is misleading. In that image, each genus neatly replaces the last, growing bigger, losing toes, and eating more grass on a straight march toward the modern horse. The reality is a dense, branching bush. During the Miocene alone, dozens of horse species existed simultaneously across North America, some with three functional toes, some with reduced side toes, and some approaching a single hoof. They occupied different ecological niches, ate different foods, and often lived in overlapping geographic ranges.

Pliohippus looked like it should be the direct predecessor of Equus because it arrived at a similar foot structure. But foot shape alone does not define ancestry. Skull anatomy, tooth structure, and now molecular evidence when available all contribute to sorting out relationships. The deep skull fossae of Pliohippus turned out to be a distinguishing trait that set it apart from the Equus lineage, even as its feet converged on the same single-hoofed plan. Convergent evolution, where distantly related organisms independently arrive at similar traits because they face similar ecological pressures, is common throughout the animal kingdom. In this case, it happened within the same family.

The competing cladistic hypotheses that researchers work with today reflect this complexity. Morphological analyses and molecular analyses do not always produce the same tree, and within morphological analyses, different character sets can yield different groupings.² Pliohippus sits in a part of the tree where these ambiguities are especially thick, partly because the mid-to-late Miocene was the period of highest equine diversity and most rapid morphological change.

What Pliohippus Tells Us About How Single Hooves Evolved

One of the persistent puzzles in equid paleontology is why horses lost their side toes at all. Multiple toes offer stability on soft, uneven ground, which is why the three-toed foot plan persisted for tens of millions of years. The shift to a single hoof seems best suited for sustained running on firm, open terrain, the kind of locomotion that becomes advantageous when grasslands replace forests and speed becomes a primary defense against predators.

The biomechanical data reinforces this. As horses got bigger and moved onto harder ground, the central toe needed to handle increasing loads. The fact that the safety factor against bone fracture stayed roughly constant from early multi-toed horses through to Equus, once side-digit contributions are accounted for, suggests that natural selection was maintaining a consistent biomechanical margin throughout the transition.⁴ Digit reduction was not reckless; the central metapodial adapted in geometry and density to absorb the functions the side digits were losing.

Pliohippus, with its variably reduced side digits, captures this transition in a way that few other genera do. It shows that the single-hoof condition did not emerge once in a single clean lineage but was “attempted” multiple times in parallel branches. Some of those branches led to Equus via Dinohippus; others, like Pliohippus itself, reached a similar endpoint independently and then died out. The evolutionary pressures toward a single hoof were widespread and persistent enough that multiple lineages converged on the same solution, and only one of those lineages happened to be the one that made it through the bottleneck of late Miocene and Pliocene extinctions.

The Modern Horse’s Vestigial Toes

If you run your hand down the leg of a modern horse, you can feel the splint bones, slender remnants of metacarpals II and IV that sit alongside the large cannon bone (metacarpal III). These are the last surviving trace of the multi-toed ancestry that Pliohippus was in the process of losing. The ridges and fusion lines on the metacarpals of Mesohippus already hinted at a consolidation process underway 35 million years ago, and modern splint bones are the endpoint of that same trajectory.⁵

Splint bones are not entirely harmless relics. In domestic horses, they can fracture or develop bony growths called splints, which cause lameness and are one of the more common orthopedic complaints in equine veterinary practice. In a sense, every lame horse with a splint-bone injury is paying a small price for the evolutionary compromise that Pliohippus and its relatives were working through millions of years ago: the body plan has committed to one toe, but the skeletal leftovers of the old design can still cause trouble.

Fossil horse footprints at sites like Laetoli, more famous for their hominin tracks, also preserve evidence of how three-toed horses actually placed their feet while walking. These trackways confirm that side digits made ground contact in three-toed species, consistent with the biomechanical models showing those digits bore load.³ By the time of Pliohippus, many individuals had crossed the threshold where those prints would have shown only a single hoof mark, indistinguishable in outline from a track left by Equus millions of years later. The anatomy was converging, even if the genealogy pointed in different directions.

1
ResearchGate / Bollettino della Societá Paleontologica Italiana. One step beyond: a revision of Middle Miocene North American Pliohippus mirabilis provides new clues about equid evolutionary history
2
PubMed Central. Evolution of the Family Equidae, Subfamily Equinae, in North, Central and South America, Eurasia and Africa during the Plio-Pleistocene
3
PubMed Central. Hipparion tracks and horses’ toes: the evolution of the equid single hoof
4
PubMed Central. Mechanics of evolutionary digit reduction in fossil horses (Equidae)
5
PubMed Central. The evolution and anatomy of the horse manus with an emphasis on digit reduction
6
Paleontological Journal. The Diet of Pliohippus potosinus Equidae, Mammalia from the Late Miocene Paso del Águila Local Fauna, San Luis Potosí, México
7
OhioLINK Electronic Theses and Dissertations Center. Paleobiogeography of Miocene to Pliocene Equinae of North America: A Phylogenetic Biogeographic and Niche Modeling Approach