The Earliest Landfowl

Gallinuloides

This post will be the latest as of 1st April, so if you've been following this blog for a while, you'll know what that means... 

The humble chicken (Gallus domesticus) is a member of the pheasant family. This is a moderately-sized family, with around 180 species. Alongside the chicken and its wild ancestor, this also includes, not just pheasants, but many similar ground-dwelling birds, such as grouse, partridges, true quails, turkeys, and peacocks. 

It is, in turn, a part of a larger taxonomic group technically referred to as the Galliformes, or more commonly the "landfowl". The other four living families in this order have fewer species and are generally less well-known, but they share the same features of being generally plump, often quite large by avian standards, and having short, rounded wings unsuited for long-distance flight. 

Molecular studies indicate that the landfowl have been around for well over 50 million years, but our fossil record of their earliest history is, as so often, rather patchy. Until recently, the oldest fossil definitively belonging to the group was Gallinuloides, which lived in Wyoming around 48 million years ago. This made it the only such fossil from the early part of the Eocene epoch, and we have only a handful of fossils from the mid to late parts - notably the European specimens Quercymegapodius, Paraortyx, and Pirortyx. In recent years, we have been able to add some non-European forms, such as Xorazmortyx from Uzbekistan, and Namaortyx from Namibia. 

None of these fossil birds, however, were members of any of the living families of landfowl. In fact, there's little evidence that any living bird families predate the dawn of the Oligocene 35 million years ago. Naturally, this does depend on your definition of the family in question, since such groupings are ultimately arbitrary. In this instance, it's based on the assumption that the families we currently define are at least useful in some sense and on the concept of the 'crown group'.

A crown group is the collective set of all descendants of the last common ancestor of all the living members of a given evolutionary unit. For example, the crown group of cats would be the last common ancestor of the domestic moggie and (say) the tiger. This includes, not only all the other living kinds of cat, but also such extinct animals as the "European jaguar" or the "American cheetah". What it doesn't include are the sabretooth cats because they split off from the ancestral cat lineage before the common ancestor of domestic cats and tigers evolved. Yet, mammal palaeontologists do place sabretooths in the cat family, which - by that definition - is therefore older than the crown group.

Which, again, reinforces the idea that such dividing lines are arbitrary. Still, if we look at the crown groups defined by the currently recognised families of birds, none of them appear to have existed back in the Eocene... although, certainly, this is up for debate. Assuming we stick to it, however, to trace back the origin of the pheasant family crown group, the first thing we need to do is figure what the earliest branch in their family tree is - that is, identify which two living are the least closely related without being outside the family.

A problem here is that we don't have detailed genetic information on every one of those 180+ species, and, in some cases, we don't have anything at all. In fact, prior to the introduction of molecular genetics as a means for elucidating relationships, it wasn't entirely clear where the boundaries of the family should be, let alone how the subgroups within it related to one another. Nonetheless, a study in 2021 was able to put together all the evidence we did have and construct a landfowl family tree. It probably won't be a surprise to discover that this overturned some of what we'd previously thought, although it did at least turn out that we had the basics right.

The pheasant family, as previously defined, turned out to contain four evolutionary branches. One of these, the American quails, was sufficiently distant from the other three that it ended up being split off into its own family - and thus, technically, these birds can no longer be considered true quails, since they aren't closely related. Staying within the pheasant family were two main groups, one containing the chickens, true quails, and peacocks (among others), and the other the pheasants, grouse, and true partridges. 

Significantly, however, some partridges turned out to belong to a third lineage unrelated to what we now know to be the real sort. They're mostly found in the tropical and subtropical parts of Asia, although two species are African, but the point is that this turns out to be the oldest of the three branches inside the family. So our crown group is defined as anything descended from the last common ancestor of (say) the chicken and the Malayan jungle partridge. Which, so far as we can tell, lived somewhere between 35 and 30 million years ago.

If that's right, anything from 50 million can't possibly be part of the pheasant family crown group or, on the basis of the same study, the crown group of any of the other four families of landfowl. So what were they?

Other than misidentification, there are two possibilities here. This is because there's nothing to stop us from applying the crown group concept to a group larger (or smaller) than a family. So these early landfowl could be a member of galliform crown group without belonging to any of the living families - they're descended from the last common ancestor of chickens and megapodes, but form their own separate lineages that died out early on. This is the same principle by which we consider sabretooths to be cats, or amphicyonid bear-dogs to be carnivorans, yet not belong to any living carnivoran family.

In fact, however, that's probably not the case. They were most likely outside even the landfowl crown group but still more closely related to living landfowl than to anything else around today. We could create even high-level groups to put them in, but given that there aren't very many of them and they don't form a single branch, this would result in rather a lot of new names. Instead, we extend the landfowl group outwards to include these birds. They're technically "stem-galliforms"; we place them in the order for lack of anywhere better, but they're early experiments in evolution that left no living descendants we can be sure of.

But just how far back do these stem-galliforms stretch? Our best estimate from the molecular data puts the split between the landfowl and the waterfowl as far back as 63 million years ago, almost to the time of the non-avian dinosaurs. However, with so few fossils from any earlier than 45 million years ago, it's difficult to say much about the early history of the group. Aside from Gallinuloides, the only fossils that could date from this time are so fragmentary that it's difficult to be confident they have been classified correctly.

Until that is, last year. In August, a paper was published describing two new species of stem-galliform. They were recovered from fossil deposits in southern England thought to date back 55 million years, a new record for any fossil from the group consisting of more than a single bone. (The single bone in question, the one connecting the shoulder blade to the breastbone, belongs to an unnamed species from Mongolia, and is around the same age as the new fossils). 

One of the fossils was similar enough to existing ones to be designated as a new species of the German genus Paraortygoides, thought to be a close relative of Gallinuloides. An interesting feature of this group is the unusual strength and solidity of the wishbone, which has been interpreted as meaning that the bird must have had a small crop (the first part of the stomach, in front of the gizzard). That's not unusual for birds in general, but a large crop is a distinguishing feature of all living landfowl, so its absence here may be significant. For instance, it probably means that it fed on soft, easily digested plant matter rather than on seeds as their similarly-sized modern relatives do.

The other, however, was distinctive enough for the researchers to not only give it a new genus name, Waltonortyx, but to place it in a newly created family as well. It is, of course, a family of one, as is often the case with stem taxa that consist of little side branches near the base of a tree. 

The name Waltonortyx translates as "Walton quail". While the first part of the name comes from the fossil site in which it was found, the second refers to the bird's size. Compared with many songbirds, a quail is not especially small, at about 17 cm (6½ inches) in length, but it is certainly much smaller than most other galliform birds. It's notable that all of the early landfowl so far identified are in at least roughly the same size range, with the largest perhaps reaching the size of a partridge at 30 cm (12 inches), still well short of a pheasant or a grouse. 

This makes it likely, if perhaps unsurprising, that large size was something that developed later in landfowl evolution. Perhaps it evolved both as a means of out-competing their relatives and as a way to fend off smaller predators - something that's probably handy when you nest on the ground. 

While relationships this far down the tree are difficult to untangle, the researchers interpret its closest relative as the almost equally old Mongolian Bumbanipodius. The interesting thing if this is true is that, not only is Mongolia quite a distance from Britain, especially for a bird without long-range flight, but 55 million years ago, Europe and Asia really were distinct continents. To get from one to the other - in whichever direction it might be - would involve crossing the Turgai Strait, a body of water on the eastern side of what are now the Ural Mountains.

That crossing must have happened very early on in landfowl evolution. We have nothing that old to give us a hint as to which side of the Strait the birds originally evolved on and, if we did, they would be so primitive that it might be hard to distinguish landfowl from waterfowl. One thing we do know is that the common ancestor of both is exceptionally ancient, representing the second oldest split in the living bird family tree, after the one that gave rise to the ostriches. 

The very first fowl, before we could call them "water" or "land", lived long before the likes of Tyrannosaurus rex. 

[Photo from Mayr & Weidig, 2004, published in an open access journal. Cladogram adapted from Chen et al. 2021.]