Fossil Cats (That Aren't Sabretooths)

Acinonyx pardinensis

The cat family is traditionally divided into two subfamilies: the "purring cats", which are mostly small, and the "roaring cats" which are all medium to large in size. When most people think of fossil species, however, the first ones to pop into their minds are almost certainly the sabretooths, such as Smilodon. These belonged to a third subfamily (and whether they could roar or not depends on fragile bones and soft tissues that haven't survived) that left no modern descendants, dying out towards the end of the Ice Ages.

The sabretooth cats represent an early branch in cat evolution, perhaps splitting off at some point during the Early Miocene, over 20 million years ago. But this means that the cats we are familiar with must have existed - in some form - for equally long, leaving their own fossil history. If you wound back the evolutionary clock on a domestic cat, or even a tiger, you wouldn't find a sabretooth or anything that looked much like one. Exactly what you would find isn't something we can know with certainty, but we do know of a number of fossil species of non-sabretooth cat that at least give us some idea.

In the case of the domestic cat specifically, the lineage can probably be traced back to Martelli's cat (Felis lunensis) which is known to have lived in Europe around 2 million years ago. This seems to have closely resembled modern wildcats, although there are a few distinctive differences. It likely evolved into the modern species and its closest relatives during the Ice Ages. For other kinds of cats, it's a little harder to tell.

The feline cats
(click to enlarge)

The scheme I have been using for the classification of the "purring cats" through this year is based on a comprehensive genetic study published in 2006, which found evidence for seven different branches within this particular subfamily. I should, however, note that a more recent study, in 2017 found a different arrangement for the same seven branches. The key difference, perhaps is that here, the ocelots and their relatives represent the first line to diverge, although they would presumably have had to do so when they still lived solely in North America, before the Panama land bridge formed and enabled them to move south.

The fossil record doesn't really help us to decide between these two possibilities, partly because the skeletons of various cat species do look rather similar, making it hard to tell different groups apart without genetic evidence, or at least something like coat patterns. The only fossil cat that we can definitely say is more related to ocelots than to non-South American species is Leopardus vorohuensis, and that lived in Argentina during the early Ice Ages, well after the two continents had become connected.

A few living species of cat, however, are fairly distinctive, at least at the genus level, and we might expect it to be easier to trace their ancestries than those of cats whose skeletons closely resemble that of the domestic animal. Lynxes are a good example here. The Issoire lynx (Lynx issiodorensis) lived in Europe and Asia around 3 million years ago, and it is widely thought to be the ancestor of at least the three species commonly referred to as lynxes today, and probably also of the bobcat.

A detailed reconstruction of the animal's bodily proportions shows that it was longer and slimmer, with shorter legs than modern lynxes and bobcats - which is to say, closer in appearance to most other cats. Unfortunately, not enough of the tail has survived to say whether that was already shortened, although it seems plausible. Later fossils show a general decrease in size and a shift in proportions towards what we see today - the "cave lynx", which lived around 0.7 million years ago, is partway through this transition, although these days it is more commonly considered as an extinct subspecies of the modern Iberian lynx. Presumably this and the other living species of lynx, diverged from one another when their ancestors became cut off in separate regions by the Pleistocene ice sheets.

The puma/cougar/mountain lion is another "purring cat" that's clearly distinctive, if only based on its enormous size compared to almost anything else in that group. The molecular evidence shows, somewhat surprisingly that, aside from the jaguraundi, the closest living relative of the puma is the cheetah. Living a couple of continents away and having numerous adaptations for speed (something not otherwise seen in living cats) this had long been thought to be an entirely separate kind of animal - a third living subfamily distinct from both the purring and roaring cats.

In retrospect, there have been a couple of lines of fossil evidence that make this seem less surprising than it did at the time. One of these is the existence of the Eurasian puma (Puma pardoides). First described by Richard Owen back in the 1840s, this was, for a long time, thought to be a relative of living leopards, but, starting in the 1960s, new fossils began to hint that it was actually a puma... just one that lived in Europe and Asia. The oldest fossils of this animal date back to the end of the Pliocene, over 2.5 million years ago, crossing over to North America just 400,000 years ago. 

Alternate scheme

A second clue comes from the fact that not only were there pumas in Europe long before they lived in the Americas, there were also American cheetahs (Miracinonyx spp.). At least two species of this animal are known, with the later of the two having a remarkable resemblance to the modern, African, cheetah. Indeed, the similarity is so strong that the animals were at one point placed in the same genus as true cheetahs, even being considered as their ancestors.

More modern evidence, including some from DNA taken from fossils right at the end of the Ice Ages, has shown that American cheetahs were more closely related to pumas and jaguarundis than they are to true cheetahs, making them an example of parallel evolution - a cougar that ran at 60 mph for short bursts. This has, however, been questioned, so that the possibility remains of a more complex evolutionary history with cheetahs crossing back and forth over the Bering land bridge.

The earliest American and African cheetahs are of roughly similar age, both dating back around 3 million years, which doesn't really help much. Cheetahs in southern Europe date back almost as far and, at some point, spread into Central Asia and eventually, China. Some clearly belong to the modern species, but the giant cheetah (Acinonyx pardinensis) is often considered distinct. It's clearly a very close relative of the living cheetah, but stood about as tall as a lion at the shoulders. (Of course, being slimmer, it would have weighed rather less). With a longer stride it could, theoretically, have run even faster than a living cheetah, although there's no way of knowing whether it actually did so, given that there wouldn't necessarily have been much point. Indeed, some recent evidence suggests that its larger size and some details of its skull shape might imply a more leopard-like hunting style.

What of the "roaring cats"? These are sufficiently similar to one another that they are currently placed in just two genera. We know very little about the fossil history of the clouded leopards (Neofelis spp.) which leaves only the big cats of the genus Panthera. I recently talked about the two (or more) fossil species that seem to be closest to lions, the cave (P. spelaea) and American (P. atrox) lions, so I won't repeat that here, but we do also have some hints as to the origins of the other living species of big cat.

The pantherine cats

Although other candidates have been suggested from time to time, the clearest example of an ancestral tiger is the Longdan tiger (P. zdasnkyi) with fossils dating back a little over 2 million years having been discovered in China. It was noticeably smaller than the living species, although undeniably still a "big cat" and still retained a few detailed skeletal features that modern tigers lack but that are found, in, say, jaguars.

Speaking of which, the jaguar is something of an oddity, living as it does in South America, far away from any of its relatives such as lions and tigers. However, we do know that the modern species once lived much further north than it does now, reaching as far as Nebraska and Washington at times during the Late Pleistocene before reaching South America perhaps as late as 0.25 million years ago. Its closest fossil relative is the European jaguar (P. gombaszoegensis) which is known to have lived across much of that continent during the Early to Middle Pleistocene, with a number of fossils having been found near Bristol in England. Slightly larger than the modern form, it is so similar in other respects that it is sometimes considered a mere subspecies of the living animal, although it must have lived in a much colder habitat since Europe was hardly a tropical jungle at the time.

The oldest known big cat that can be related to a particular modern species is currently Panthera blythae, dating back to around 5 million years ago. This lived in the Himalayas, and is thought to be related to the snow leopard. If that's true, snow leopards would have to have evolved rather earlier than some estimates have suggested, but, given the uncertainties involved, that's not inherently unreasonable.

The only fossils of Panthera palaeosinensis are much less old, this being another species known from the Late Pliocene, 3 million years or so ago. However, while it has been argued that it might be related to lions and leopards, it has enough differences from either that this remains debated. It's equally plausible, for instance, that it represents some other line that has since died out and that diverged very early on from the ancestral big cats, even before P. blythae did. This might suggest that the first big cats evolved, not in Africa as we might expect, but in Asia, since this particular species inhabited northern China.

The "roaring" and "purring" cats probably diverged from one another around 11 million years ago, and it becomes hard to distinguish the two as we approach that date. Cats from before this time are impossible to place in any living subfamily, and consist of at least two well-recognised genera. Pseudaelurus dates back around 20 million years to the Early Miocene and, since it may include the ancestors of both sabretooth and regular cats, may well represent a mix of different lineages - it is often split into two or more genera as a result. At least some species were lynx-sized, and the genus includes the first cats to have reached North America.

The earliest fossil we can definitively say is a cat, rather than something else, belongs to Proailurus lemanensis and dates back a full 30 million years, a similar date to some molecular estimates of the ultimate origin of the cat family. Slightly larger than the modern domestic animal, it has adaptations that suggest it spent a lot of time in the trees and it did not yet have the reduction in the number of teeth that distinguishes cats from related mammal families such as mongoose and civets.

Of the 41 species of cat currently living, five are formally considered to be endangered species: the flat-headed, bay, and Andean mountain cats, the Iberian lynx, and the tiger. A great many others are considered threatened, finding coexistence with humans challenging. However, none seem to have gone extinct in the last few thousand years which is more than can be said for some other groups of animals, and a few have recovered enough in recent years to move out of the true danger zone. Once such things as killing for fur are taken out of the picture, there are at least some advantages to being charismatic and beautiful animals...

[Picture by Dawid Iurino, available under CC-BY-SA 4.0. Cladograms based on Johnson et al. 2006 and Zhou et al. 2017.]