Oncillas in the Highlands

Around 5 million years ago, as North and South America finally moved close enough together for a land bridge to form, a species of cat took the opportunity to move south. It found the southern continent devoid of any similar mammalian predators and prospered, leaving descendants from the Caribbean coast to the Magellan Straits.

We do not have a name for this lost species, or know much about where it lived prior to its journey south. However, we can tell that it existed because all of the small cats of South America are missing a pair of chromosomes found in every other species across the world - including the jaguars and pumas with which they share their continent. As confirmed by more detailed genetic analysis, this means that they all shared a single common ancestor in which this oddity first arose.

Today, all the species of cat that descend from that mystery ancestor are placed in the genus Leopardus. This was first named by John Edward Gray in a letter to a journal in which he described several mammal species that he believed were new to science. In the case of the four Leopardus species he was wrong, so the names he gave them are no longer used, but he was the first to distinguish them from other cats, so the name he gave to that group as a whole does remain. 

The best known member of the genus is the ocelot (Leopardus pardalis), but there are many others across the continent and, as currently defined, it is the cat genus with the greatest number of species. Just how many species that is is, however, a matter of debate. Much of that debate focuses on the pampas cat and its relatives but there is also contention over the exact status of the oncilla (Leopardus tigrinus) - an issue I have briefly covered before.

The oncilla was given its scientific name by Johann von Schreber in 1774, for an animal he described (in German) as a "margay". The specimen he used to define the species did not, however, belong to the animal we now call a margay, so the scientific name remains attached to what he had actually described - the oncilla. While this illustrates some of the confusion around classifying these small spotted cats, that part is no longer in dispute. What's less clear is how many species or subspecies the oncilla may represent - a question that is of more than academic interest, since it's easier to get conservation money for a species than a subspecies.

For most of the 20th century, there were four recognised subspecies of oncilla. One of these, by definition, is the one to which Schreber's original specimen belonged (L. t. tigrinus). That was collected in French Guiana, so that subspecies is the one that lives in the Guyana region and northern Brazil. A second one (L. t. oncilla) lives in Costa Rica and a small region of neighbouring Panama, and a third (L. t. pardinoides) in Colombia, Venezuela, Ecuador, and northern Peru. 

In 2013, however, it was shown that the fourth "subspecies" was genetically distinct and does not interbreed with the others even when it gets the opportunity. It is now almost universally recognised as a separate species, (Leopardus guttulus), whose common name remains unclear - although "southern tigrina" or some variation thereof seems likely to win out. Look at a map showing the distribution of this new species, and you'll notice a suspiciously straight line between it in southern Brazil and the oncilla further north, demonstrating that we're still not quite sure where the real geographic boundary lies.

In 2017, not only was there a proposal to promote the western subspecies to full species status, but also a study that showed that, firstly, we shouldn't do that, and secondly, we should be splitting off the population in northern Brazil into its own species - the northern tigrina (Leopardus emiliae). That was on the basis of physical appearance, but genetic data published last year backs it up, so it's looking quite likely that it will be accepted.

Regardless of what's a species and what's a subspecies, the upshot of all this is that we have, probably as a minimum, five different populations of oncilla-like cat that have minimal, if any, contact with one another. So, classification issues aside, why is that, and just how different are they?

When populations of a given species, or set of closely related species, end up being separated, the usual culprit, at least to begin with, is simple geography. Mountain ranges and large rivers can form a barrier to the spread of an animal, or limit contact between populations after they do get a rare opportunity to cross. In the case of the oncilla, however, it may not be mountains that are the problem, but lowlands. That population in Costa Rica inhabits the cloud forests of the Talamanca Mountains and is rarely seen below 1,500 metres (4,900 feet) elevation. If that's a real preference, perhaps due to disliking the tropical heat of the lowlands, then the Panama Isthmus would easily keep that population isolated from the one in Colombia. 

Proposed new scheme
(Southern tigrina in brown)

It is, however, less clear what barriers might exist elsewhere, or to what extent they might be artefacts of recent human population density rather than ancient preferences of the animal concerned. To try and answer this, a recent study looked at records of where oncillas have been found and compared them with data on the local climate and terrain taken from an online database.

Significantly, the study found that the preferred habitats for oncillas in the Andean region are distinctly different from those living further east, supporting the idea that the western population (L. t. pardinoides) could represent a new species. The barrier between this and the regular oncilla would probably the Llanos, a broad region of low-lying grassland between the Andes and the Guyanan highlands, through which the Orinoco River runs. 

The seasonally dry climate in this region, coupled with a lack of dense woodland means that the Llanos has never been regarded as suitable habitat for oncillas. The previous assumption, however, had been that the animals could mingle further south, in the western reaches of the Amazon. The new study suggests that the habitat in that region is far less hospitable to oncillas than thought, and points out that, while traditional distribution maps say that the cats live there, there are no definitive sightings in the region. That might, of course, be because it's very remote dense jungle, and oncillas are small, nocturnal, stealthy animals that are hard to spot even without heavy undergrowth... but it could equally well be because they just aren't there.

The reason that this hadn't been caught before may be because previous models of habitat suitability had assumed that the southern tigrina was the same species as the oncilla. In fact, the two animals seem to have quite different requirements, with the former more suited to the Atlantic Forest with its less tropical climate. Considering the two together made it look as if the oncilla was more adaptable than it really is, making it likely that it also lived in areas where we hadn't had much opportunity to look for it.

Previous work on the geographic distribution of oncillas had suggested that another barrier should exist in northern Peru where a there is a slight break in the Andes, isolating the Colombian/Venezuelan population from the Perovian/Bolivian one. This, however, has never been proposed as marking a subspecies boundary and the study shows why - there are, in fact, scattered patches of suitable habitat across the gap, close enough together that animals could easily cross from one to the other.

The upshot of this is that we can identify at least three different groups of oncilla each of which has different habitat requirements, with the Llanos and the Amazon forming the barriers between them. Only the Central American and Andean subspecies come out as having the same needs, and they are kept apart by the Panamanian lowlands - which were likely more hospitable to oncillas during the Ice Ages when cool mountain forests reached over 1,000 metres (3,300 feet) further downslope than they do today.

This does not prove that these three groups should all be recognised as distinct species. But it makes it more likely that they will do, supported by the genetic evidence that we already have. Even knowing that different populations have different requirements could help with conserving what is, by some measures, the single rarest species of South American cat.

[Photo by "Abujoy" from Wikimedia Commons. Map adapted from Bonilla-Sánchez et al. 2024.]