What makes a mammal?

Which shows that an evolutionary context is necessary for a rational taxonomy. If you don't have that basis, you will have to make many arbitrary decisions and be content with an artificial classification scheme. Just one more reason why evolution is the unifying principle of all modern biology.

DS said: Which shows that an evolutionary context is necessary for a rational taxonomy. If you don't have that basis, you will have to make many arbitrary decisions and be content with an artificial classification scheme. Just one more reason why evolution is the unifying principle of all modern biology.

It's one of the ways in which Stephen Meyer was (deliberately) confused in 'Darwin's Doubt'. A current phylogenetic group might have a number of common features, but there's no reason why they had to evolve simultaneously. Or that all of the features have to evolve in all members. Or that they have to be subsequently retained.

Meyer claims that arthropods and nematodes can't have a common evolutionary ancestry because a modern arthropod can't evolve into a nematode. Or vice versa. Because the 'body plans' are so radically different.. Ignoring the fact that the body plans in both have developed over hundreds of millions of years. And that the proto-arthropod would have looked almost identical to the proto-nematode.

In my opinion, all female mammals feed their babies milk and if they didn't do that, why do they cal mammals and why do they have breasts? So they feed their babies with milk.

There are some organisms that have what superficially looks like hair- tarantulas for example.
I believe there is also a species of cockroach where the female feeds the nymphs a whitish high protein fluid produced from glands. I heard about this years ago, but I forgot where. I just remember thinking "roach milk"

...all species that descended from the common ancestral mammal population will all be mammals.

I love this post and the cheery thread, but it's not what a phylogeneticist would write :-).

Chris Lawson said:

...all species that descended from the common ancestral mammal population will all be mammals.

Will this always be true, Wilson? I'm thinking of, say, another 100M years in the future: would a hypothetical far-future taxonomist still classify all descendants of mammals as mammals even if some of them have lost all of the classic features of mammals (i.e. no fur, no endothermy, no mammaries, no live birth)? We don't classify birds as theropods. I'm wondering where the point of divergence is where we decide to give the descendants of one group their very own category?

RodW,

In the original post, Wilson does say that he is not aware of any mammals that don't lactate, but his point is that there are lots of mammals missing one or two of the classic mammalian traits, so it wouldn't necessarily mean that a hypothetical non-lactating animal would *have* to be kicked out of the mammalian clade.

Imagine a descendant of, say, bears that had fur, endothermy, a mammalian neocortex, 3 middle ear bones, and a 4-chambered heart, but whose females had vestigial mammaries that never produced milk.

Nick,

That's an really interesting point. I guess the DFTD cancer cells can be considered a separate species that reproduces asexually, is an obligate parasite, and has no fur/hair, endothermy, neo-cortex, or middle ear bones. Should we still call it a mammal? I don't know what the best answer should be.

I have a few comments. First, thanks for introducing me to pangolins, which I'd never heard of before.

Second, I think you could have made more out of the fact that all female mammals nourish their young with milk. This is the defining characteristic of mammals, and it ties into your main point because it was an evolved trait that every mammal still retains.

Third, in the third to last paragraph you have "it's" when you mean "its". "It's" is a contraction, "its" is the third person neuter singular possessive pronoun.

Chris Lawson speculated: RodW, In the original post, Wilson does say that he is not aware of any mammals that don't lactate, but his point is that there are lots of mammals missing one or two of the classic mammalian traits, so it wouldn't necessarily mean that a hypothetical non-lactating animal would *have* to be kicked out of the mammalian clade. Imagine a descendant of, say, bears that had fur, endothermy, a mammalian neocortex, 3 middle ear bones, and a 4-chambered heart, but whose females had vestigial mammaries that never produced milk.

DS said: Which shows that an evolutionary context is necessary for a rational taxonomy. If you don't have that basis, you will have to make many arbitrary decisions and be content with an artificial classification scheme.

DS said: Just one more reason why evolution is the unifying principle of all modern biology.

Nick Matzke said:

Chris Lawson said:

...all species that descended from the common ancestral mammal population will all be mammals.

Will this always be true, Wilson? I'm thinking of, say, another 100M years in the future: would a hypothetical far-future taxonomist still classify all descendants of mammals as mammals even if some of them have lost all of the classic features of mammals (i.e. no fur, no endothermy, no mammaries, no live birth)? We don't classify birds as theropods. I'm wondering where the point of divergence is where we decide to give the descendants of one group their very own category?

The future is now: http://en.wikipedia.org/wiki/Devil_facial_tumour_disease

Nick Matzke said: I love this post and the cheery thread, but it's not what a phylogeneticist would write :-).

balloonguy said: I have a few comments. First, thanks for introducing me to pangolins, which I'd never heard of before. Second, I think you could have made more out of the fact that all female mammals nourish their young with milk. This is the defining characteristic of mammals, and it ties into your main point because it was an evolved trait that every mammal still retains. Third, in the third to last paragraph you have "it's" when you mean "its". "It's" is a contraction, "its" is the third person neuter singular possessive pronoun.

RodW said: There are some organisms that have what superficially looks like hair- tarantulas for example. I believe there is also a species of cockroach where the female feeds the nymphs a whitish high protein fluid produced from glands. I heard about this years ago, but I forgot where. I just remember thinking "roach milk"

apokryltaros said: Now, if we were able to find a live "mammal-reptile" like a gorgonopsid, today, would that critter be considered a mammal, or not?

Chris Lawson said:

...all species that descended from the common ancestral mammal population will all be mammals.

Will this always be true, Wilson? I'm thinking of, say, another 100M years in the future: would a hypothetical far-future taxonomist still classify all descendants of mammals as mammals even if some of them have lost all of the classic features of mammals (i.e. no fur, no endothermy, no mammaries, no live birth)? We don't classify birds as theropods. I'm wondering where the point of divergence is where we decide to give the descendants of one group their very own category?

phylogenetically, sharks are much more different from bony fish than an echidna is from you. Bony fish split off from sharks hundreds of millions of years ago.

we really shouldn't be calling sharks fish any more (or calling hagfish "fish" either for that matter), but how does one go about changing such ancient common classification?

Uh, guys, out of all of the sites on the entire internet that I'd expect to know that the defining character of the Class Mammalia is just those three inner ear bones derived from bits of the jaw, it's this one. None of that other stuff (hair, milk, heart shape, etc) fossilizes, making it useless for 99% of all of the mammalian species who have ever lived. We know that Smilodon was a mammal not because today we have furry lions, but because it has that same inner ear structure as all the rest of the clade.

Uh, guys, out of all of the sites on the entire internet that I’d expect to know that the defining character of the Class Mammalia is just those three inner ear bones derived from bits of the jaw, it’s this one. None of that other stuff (hair, milk, heart shape, etc) fossilizes, making it useless for 99% of all of the mammalian species who have ever lived. We know that Smilodon was a mammal not because today we have furry lions, but because it has that same inner ear structure as all the rest of the clade.

Andy White said: In my opinion, all female mammals feed their babies milk and if they didn't do that, why do they cal mammals and why do they have breasts? So they feed their babies with milk.

balloonguy said:

Andy White said: In my opinion, all female mammals feed their babies milk and if they didn't do that, why do they cal mammals and why do they have breasts? So they feed their babies with milk.

Not all mammals have breasts. Monotremes don't - they secret milk through the skin. Marsupials have nipples but I think they're in their pouches.

In the original post, Wilson does say that he is not aware of any mammals that don’t lactate, but his point is that there are lots of mammals missing one or two of the classic mammalian traits, so it wouldn’t necessarily mean that a hypothetical non-lactating animal would *have* to be kicked out of the mammalian clade. Imagine a descendant of, say, bears that had fur, endothermy, a mammalian neocortex, 3 middle ear bones, and a 4-chambered heart, but whose females had vestigial mammaries that never produced milk.

perfusionman said: Uh, guys, out of all of the sites on the entire internet that I'd expect to know that the defining character of the Class Mammalia is just those three inner ear bones derived from bits of the jaw, it's this one. None of that other stuff (hair, milk, heart shape, etc) fossilizes, making it useless for 99% of all of the mammalian species who have ever lived. We know that Smilodon was a mammal not because today we have furry lions, but because it has that same inner ear structure as all the rest of the clade.

RodW said: It seems to me that if some highly derived mammal that was missing most of these traits became the base of a huge radiation, its descendents would not be considered mammals in common parlance 100mya from now, for the same reason birds arent considered reptiles. I think if we discovered an underived therapsid on some tropical island it would be sufficiently different that we wouldnt call it a mammal,

RodW said: There are some organisms that have what superficially looks like hair- tarantulas for example. I believe there is also a species of cockroach where the female feeds the nymphs a whitish high protein fluid produced from glands. I heard about this years ago, but I forgot where. I just remember thinking "roach milk"

^secrete, not secret, in my last post

RodW said: Chris Lawson said:

In the original post, Wilson does say that he is not aware of any mammals that don’t lactate, but his point is that there are lots of mammals missing one or two of the classic mammalian traits, so it wouldn’t necessarily mean that a hypothetical non-lactating animal would *have* to be kicked out of the mammalian clade. Imagine a descendant of, say, bears that had fur, endothermy, a mammalian neocortex, 3 middle ear bones, and a 4-chambered heart, but whose females had vestigial mammaries that never produced milk.

Of course. But this thread started with the assumption that we pretty much know what mammals are and the question is what should we exclude. I think its interesting to consider a situation in which we dont really know what constitutes a mammal and so we consider other organisms that have some of the 'mammal traits' It seems to me that if some highly derived mammal that was missing most of these traits became the base of a huge radiation, its descendents would not be considered mammals in common parlance 100mya from now, for the same reason birds arent considered reptiles. I think if we discovered an underived therapsid on some tropical island it would be sufficiently different that we wouldnt call it a mammal,

perfusionman said: We know that Smilodon was a mammal not because today we have furry lions, but because it has that same inner ear structure as all the rest of the clade.

apokryltaros said: Now, if we were able to find a live “mammal-reptile” like a gorgonopsid, today, would that critter be considered a mammal, or not?

M. Wilson Sayres said: And now I know about roach milk. Aw, they’re so adorable.

Chris Lawson said: Will this always be true, Wilson? I’m thinking of, say, another 100M years in the future: would a hypothetical far-future taxonomist still classify all descendants of mammals as mammals even if some of them have lost all of the classic features of mammals (i.e. no fur, no endothermy, no mammaries, no live birth)? We don’t classify birds as theropods.

The three middle-ear-bones definition isn't perfect either, as there are a variety of early mammal-ish fossils that almost or barely have this character, and it might have evolved independently a few times amongst these closely related almost-mammals. This is why there was a scientific debate, sometimes dug up by creationists, about whether or not "mammals" evolved twice.

And, anyway, picking that character as the dividing line is almost as arbitrary as picking any other. It may or may not correlate with other characters, such as mammary glands and hair, which might have evolved at other times.

This sort of thing is why node-based definitions of taxa are becoming the most popular amongst taxonomists. Nodes define the groups, and characters (as many as possible, in a phylogenetic analysis) are used to diagnose whether or not a species of interest falls into a group of interest.

Melissa wasn't trying to dig into all of this, but for people who want a starter, here's an intro:

http://en.wikipedia.org/wiki/Phylogenetic_nomenclature

John,

Thanks for clearing that up. From a phylogenetic point of view, it makes sense to classify everything from the root downwards. I had never heard of birds being called theropods, but I'm not in the field. Without meaning to get into all the no doubt innumerable complicated arguments in taxonomy, I imagine this system of classification still has controversies over what to do about endosymbiosis. (I guess this makes humans eukaryotes *and* prokaryotes *and* proteobacteria.)

In the small and recent scale, things seem reasonably easy to work out:

primates > hominoids > hominids > hominines > hominins > humans

On the larger scale I notice most phylogenetic trees for humans on the web only go back as far as eukaryotes, which is I guess where things get hard. Conversely, on the very small recent scale, say the last 5M years, there also seems to be a lot of controversy about where to put various Australopitheci and Homos on the branch.

perfusionman said: Uh, guys, out of all of the sites on the entire internet that I'd expect to know that the defining character of the Class Mammalia is just those three inner ear bones derived from bits of the jaw, it's this one. None of that other stuff (hair, milk, heart shape, etc) fossilizes, making it useless for 99% of all of the mammalian species who have ever lived. We know that Smilodon was a mammal not because today we have furry lions, but because it has that same inner ear structure as all the rest of the clade.

Tom said: phylogenetically, sharks are much more different from bony fish than an echidna is from you. Bony fish split off from sharks hundreds of millions of years ago.

we really shouldn't be calling sharks fish any more (or calling hagfish "fish" either for that matter), but how does one go about changing such ancient common classification?

apokryltaros said:

Tom said: phylogenetically, sharks are much more different from bony fish than an echidna is from you. Bony fish split off from sharks hundreds of millions of years ago.

With the discovery and study of the Silurian placoderm, Entelognathus, which has jaws anatomically similar to primitive bony fish, bony fish, including the ancestors of lungfish and tetrapods, branched off from the arthrodire+ptyctodontid branch of Placodermi during the early Silurian. Whole body fossils of the early Devonian shark Ptomacanthus, which was originally described as an acanthodian, or "spiny shark," suggest that true sharks/cartilaginous fish diverged from the acanthodians during the start of the Devonian, well after bony fish diverged from the placoderms.

we really shouldn't be calling sharks fish any more (or calling hagfish "fish" either for that matter), but how does one go about changing such ancient common classification?

The term "fish" is an imprecise term that technically encompasses hagfish and all non-tetrapod craniate vertebrates (i.e., every aquatic vertebrate that breaths through gills from lampreys to sharks), and even a few amphibians and some invertebrates (i.e., "congo eels," and "silverfish"). Having said that, it's probably easier to catch the moon in a bucket than trying to convince laypeople, or even other scientists to stop calling sharks "fish." Hell, it's a painfully uphill battle to even convince people to stop using the terms "jellyfish" and "starfish." People are very resistant to changing their vocabulary over what they perceive as a minor quibble that requires a lot of hard to understand explanation.

There are a few mammal fossils that do show hair, like the Mesozoic Eomaia, Volaticotherium, Casturocauda, and Late Paleocene coprolites containing hair of Lambdopsalis.

Karen S. said:

There are a few mammal fossils that do show hair, like the Mesozoic Eomaia, Volaticotherium, Casturocauda, and Late Paleocene coprolites containing hair of Lambdopsalis.

You can see mammoth hair at the American Museum of Natural History

DS said: Agreed. But we shouldn't just let them call whales and dolphins "fish". You gotta draw the line somewhere.

Chris Lawson said: John, Thanks for clearing that up. From a phylogenetic point of view, it makes sense to classify everything from the root downwards. I had never heard of birds being called theropods, but I'm not in the field. Without meaning to get into all the no doubt innumerable complicated arguments in taxonomy, I imagine this system of classification still has controversies over what to do about endosymbiosis. (I guess this makes humans eukaryotes *and* prokaryotes *and* proteobacteria.)

I suspect that most people use the words "fish", "reptile", etc., to mean something having some set of traits (or some subset of that set), rather than the clade that includes all descendants of the first member species of that group.

Henry

Not sure about the Echidna but I have read (god knows where), that the milk of the Platypus actually oozes through large pores in the skin. This is possibly a modified sweat? Concentrate the screting glands, expand and localise the pores (nipple), hey presto, breast!

Tom said: phylogenetically, sharks are much more different from bony fish than an echidna is from you. Bony fish split off from sharks hundreds of millions of years ago. we really shouldn't be calling sharks fish any more (or calling hagfish "fish" either for that matter), but how does one go about changing such ancient common classification?

apokryltaros said: Having said that, it's probably easier to catch the moon in a bucket than trying to convince laypeople, or even other scientists to stop calling sharks "fish."

TomS said: I presume that this means that "fish" is to be identified with Osteichthyes, the "bony fishes"? I have heard it said that if there is to be a genuine clade of fishes, one which includes both sharks and tunas, for example, then it would also include giraffes and ostriches (as well as whales and penguins).

Just yesterday, Discover magazine came out wih an article on another trait all mammals evidently share. I don't vouch for the research's accuracy, but I will tentatively vouch for its amusement value...

John Harshman said:

Chris Lawson said: John, Thanks for clearing that up. From a phylogenetic point of view, it makes sense to classify everything from the root downwards. I had never heard of birds being called theropods, but I'm not in the field. Without meaning to get into all the no doubt innumerable complicated arguments in taxonomy, I imagine this system of classification still has controversies over what to do about endosymbiosis. (I guess this makes humans eukaryotes *and* prokaryotes *and* proteobacteria.)

"Prokaryote" isn't currently a taxonomic term, as it doesn't describe a clade. But yeah, cladistic classification -- any classification, really -- has a bit of a problem with reticulation, as in endosymbiosis and hybrid speciation. I solve the problem by trying not to think about it, which works fine.

What makes a mammal?

A female mammal, a male mammal, and some romantic music.* *(Or whatever it is that members of the particular species happen to regard as romantic.)

Karen S. said:

There are a few mammal fossils that do show hair, like the Mesozoic Eomaia, Volaticotherium, Casturocauda, and Late Paleocene coprolites containing hair of Lambdopsalis.

You can see mammoth hair at the American Museum of Natural History

Thnaks You Very İnteresting

Another defining characteristic of mammals is a single lower dentary bone. All mammals possess this.

I consider monotremes to be mammalian heretics.

Or evolutionary deviants, having deviated from the proper path lo these many millions of years.*

Same thing, really.

Glen Davidson

*Well, they really are the long-diverged "exceptions" that indicate how taxonomic categories are fairly arbitrary for evolving organisms, apart from extinctions.