How do paleontologists determine animal types from fossils?

I’m currently reading The Eternal Frontier: An Ecological History of North America and Its Peoples by Tim Flannery. He starts with the dinosaurs and works forward from there. I’m currently up to the last ice age.

Flannery throws around all kinds of names of animals that I’m not at all familiar with, confidently stating “this is a canid, that’s a rodent,” etc. Okay, I can see why you’d be able to make those IDs based on skeletons and teeth.

What about the larger division in mammals, though, placental vs. marsupial? How can you tell what the basic reproductive category was just by skeletons and teeth? Even if there were other kinds of remains – whole body fossils – would you be able to tell from those?

The evolutionary relationships connecting taxonomic groups are dealt with in the field of cladistics. Ideally extant species are positioned on a phylogenetic tree, or evolutionary tree, according to both morphological and molecular data. Obviously within the fossil record, genetic material is restricted to a few 100,000s years, and it is frequently only the skeletal elements which survive (though occasionally longer in the case of proteins).

In view of this, palaeontological science largely relies on comparing specific morphological traits. Such features are grouped into either plesiomorphies (ancestral, or primitive, traits shared between groups), synapomorphies (derived characteristics shared by related offspring), or homoplasies (superficially similar, but genetically unrelated traits, usually due to convergent evolution). Once indexed, this information is used to construct a phylogenetic tree, in which separate branches form monophyletic groups, which stem from a single ancestor. Nowadays this process is assisted by computational phylogenetics. Sometimes, where very little evidence is available, these relationships are somewhat tentative, and groupings may be subtly altered on the basis of later findings.

In the context of your example, the placental mammals (the Placentalia, themselves a subclass of the Eutheria) and the marsupial mammals (the Marsupialia) are both evolutionary descendants of the Theria mammal subclass. As such, they all share ancestral traits of tribosphenic molars, calcaneums with enlarged sustentacular processes, talus bones with distinctive necks, and several other memorable identifiers. However a subgroup, such as the Placentalia, may be differentiated according to derived features, notably the absence of epipubic bones and pouch, narrow stylar shelves on the upper molars, absence of a shell membrane… and so on and so forth. The systematic application of this technique has given us the present outline of taxonomic relationships where molecular evidence is unavailable.

A basic mammalian phylogenetic tree can be seen here for context.

A lot can be learned from the simplest and most minimal of fossil elements. Rodents (and lagomorphs such as rabbits) have rather unique gnawing incisors with enormous root systems; carnivores, in contrast, have tiny incisors and fairly large canines. (Look inside the mouth of a cat and a guinea pig for examples of what I’m saying.) Lagomorphs have two large incisors per half jaw, rodents only one (four vs. two per jaw; ‘half-jaw’ descriptions are used because jaws are bilaterally symmetrical). True carnivores all have a premolar called the carnassial adapted as a shearing blade; the other two carnivorous extinct groups, the mesonychids and the creodonts, have other adaptations. In addition to tusks on most forms and unusual nose structure to support and anchor a proboscis (trunk), all elephantids have a very unique molar structure. All artiodactyls (even-toed hoofed animals) have a very distinct ‘double-pulley’ ankle bone – which is shared by the extinct whale ancestors that had functional legs, one reason why whales are now classified within the artiodactyls in ‘Cetioartiodactyla’. Within the whales themselves, there is an odd sort of lengtening and foreshortening of the bones comprising the skull, which is why we know those primitive legged forms were in fact whale ncestors. And so on. The feet of one branch of the carnosaurs, along with the birds, have a very unusual modification – the reason birds are now considered dinosaur descendants. There are hundreds of these ‘synapomorphies’ – shared derived characteristics, that logically ought only to have evolved once – demonstrating closeness of relationship among various groups.

Note too that occasionally diagnostic features prove invalid on further study. When I learned paleontology, the reptiles were neatly classed into four groups according to whether they had a lower temporal fenestra, an upper parietal fenestra, both, or neither. Turns out the ‘euryapsids’, with only the upper fenestra, are at least two separate groups both derived from the diapsids, which had both, by the lower one closing – and of course lizards and snakes, considered diapsids, don’t have the lower bone closing off the fenestra, to improve their gape. One arrives at these conclusions through considering other skeletal characters on more complete skeletons.

The thing is, we are not actually identifying what kind of reproductive system the animal might have had from the bones. As The Urban Spaceman indicates, modern marsupials and modern placentals have skeletons that differ from one another in characteristic ways. Only one of these differences, the presence of epipubic bones that support the female marsupial’s pouch, is related to the reproductive system. (However, such bones are also present in some other groups, and were not originally associated with that function.)

We can trace these skeletal differences back about 125 million years to when these two lineages diverged. We can guess that these ancient forms differed in their reproductive systems in the same way as their modern descendants, but we do not know for sure. So when we refer to these fossils as marsupials and placentals, we are talking about their lineages, not their reproductive systems per se. To avoid such confusion, it is probably better to refer to them by their technical names, Metatheria and Eutheria.

So, “appears to be in the lineage that in modern times includes the marsupials” would be more accurate than “was a marsupial”?

Well, that’s probably true if you insist on describing mammals by their mode of reproduction alone. But as Colibri noted, “Marsupialia” includes a wide range of animals that have a number of synapomorphies, one of which is the marsupial mode of reproduction. Others are skeletal, dental, etc. A fossil form that shows the hard-parts synapomorphies is a member of the Marsuipialia – irrespective of whether it reproduced like modern marsupials do, used another means of reproduction, or found its young under cabbage leaves. (IIRC, there is at least one modern marsupial without a pouch – how it cares for its young during the period the young of other marsupials would remain in the pouch, is something I have no idea on.)

Depends on what you mean by marsupial. If you mean an animal that gives birth to undeveloped young that undergo most of their development in a pouch, then yes, the first statement would be more accurate. For all we know, the first members of the marsupial lineage could have laid eggs like monotremes. (They would not have had the placental mode of reproduction, however, since that is quite specialized.)

Polycarp, I love you dearly, but your use of jargon here isn’t helpful to me.

Colibri – is there some other meaning of marsupial aside from means of reproduction that would be relevant to what the author is referring to when he says “animal X was a marsupial”?

Ummm… try this: “Phiomia was a Proboscidean” does not mean “Phiomia had a long, prehensile nose called a trunk,” but rather, “Phiomia was a member of the group of animals characterized by large size, thick leathery skin, specialized dentition usually including tusks and molars that come in sequentially rather than roughly simultaneously, large heads with trunklike noses, plantigrade stance, etc.” These are all characteristics describing modern elephants, subfossil mammoths, and as regards the hard parts their extinct relatives. Those unique to Proboscidea are considered synapomorphiies, which I tried to define in my first post to the thread.

Off the top of my head I don’t remember the hard-parts characteristics of Marsuipals other than the epipubic bone (and that only because Colibri’s post reminded me of it), but saying “X was a marsupial” does not, to a taxonomist, mean “X reproduced by giving birth to small undeveloped young that complete their ‘fetal’ growth in a pouch,” but rather “X was a member of that Infraclass of Mammalia which had hard parts characteristics 1, 2, and 3” (which it displays) “and the modern members of which reproduce by the ‘marsupial’ mode of reproduction.” Does that help make it clearer?

Yes, thanks for making another effort.

My question is, though, what exactly are those characteristics A, B, and C that are visible in the fossil record? How do you recognize a marsupial as a marsupial if all you’ve got is a bunch of bones? Since the epipubic bone, as I understand what Colibri said [which of course may or may not be what he said], is neither unique to marsupials nor present in all marsupials, what are taxonomists looking at when they say animal X is of the marsupial lineage?

There are a some other details that characterize marsupials. One is the dentition – IIRC marsupials have four molars, versus three in placental mammals. For the trained expert, there are numerous little details like this that distinguish marsupials from placentals. Generally speaking, many of these features aren’t necessarily adaptive traits. Certainly, the number of molars isn’t directly related to the mode of raising young. Likely it’s just some historical accident of random and neutral evolutionary processes.

Somewhere, probably buried in academic libraries, there will be textbooks and references that list these every little character. I can’t find anything like that via google, but here’s something similar: a class handout talking about the differences and similarities between marsupials and monotremes. That should give you a flavor of what paleontologists are looking for.

Of course these things are going to be highly technical, and I’m no expert in comparative osteology. Some of the defining characteristics for the Metatheria (marsupials) given in the link are:

There are some distinctions in the cusps of the teeth as well; one character given is that the entoconid and hypoconulid cusps on the molar teeth are “twinned,” that is, very close together. This article illustrates some of these tooth characters.

Aren’t you sorry you asked. :slight_smile:

Heh. No, actually, I’m not.

Thank you for laying it out. Obviously there are a slew of significant differences, most of them visible in teeth and bones, since most of which have nothing directly to do with reproduction.

That, believe it or not, was pretty much the point of my question.

lazybratsche – thanks for the handout – it also made the same point, and I actually understood a few things that it said. :smiley:

Many thanks for your patience, all. Much appreciated.