I have had a long time fuzzy idea about multiple feedback loops occurring between external evolutionary pressures and the DNA mechanisms. Though the DNA scheme is exceptionally adaptive it seems there are chemical and mathematical limitations at some point. If trait A is possible and expressed, then it may exclude or somewhat suppress the possibility of trait B, and maybe some others. Further, if A is expressed, it may make it easier for some others to be expressed. Maybe the genome that evolves to express milk, suppresses some other possibilities in that line and enhances the possibility of others. The gene sequence has evolved down a certain path. Other paths are now behind it, some have now been made available. If a creature has gone down that path, it will now have the opportunity to test the new paths against evolutionary pressure. It seems such major forks will lead to large numbers of species having those traits. Separate taxonomies as we choose to sort them.
All those bits of so called junk DNA may be inactive old paths that are now disconnected. Though it seems that circumstances can reactivate them at some cost.
But we don’t define mammals by all of those traits. We pick just one of them, and then remarkably find that there are many other traits that could also be used to define that exact same group. Why is it that the group of vertebrates defined by “has mammary glands” is precisely the same as the group defined by “has hair”, or the group defined by “has three earbones”? And why is it that all of those groups are almost precisely the same as the groups defined by “bears live young” or “has heterochromosomal males”?
And yes, I know that this happens with other groups, too. I’m focusing on mammals because I’m more familiar with us. But the fact that this same phenomenon, of traits clustering in groups, occurs in many groups just makes it even more remarkable.
Which is why I would expect traits to not be clustered.
All the other branches which have evolved from the cynodonts have died out, except monotremes and ‘theria’, which includes placentals and marsupials. So it is not surprising that the theria all share traits, since they retain features which have evolved since they branched off from the cynodonts.
Milk and fur are closely related developmentally, since they are both secretions from glands in the skin. But I don’t think it is clear which evolved first yet; one may have led to the other.
“Only mammaliformes have hair” is true, “only mammals have hair” is not. But non-mammalian mammaliformes don’t have the 3 inner-ear bones. So it’s not true to say the group “has hair” is precisely the same as the group “has three earbones”.
I mean that, of extant organisms, the groups are identical.
I confess I don’t see the same degree of remarkableness as you do. The class mammalia is only one of over a hundred classes and contains within it an incredible diversity of creatures that have substantial differences as well as similarities. (bats v cetaceans anyone?). That there are some fundamental defining features for that class is hardly surprising seeing as that is how the class gets defined.
All modern mammals share a common ancestor and though substantial evolution has occurred in the millions of years since that point there are some features which pretty much all mammals retain, hence the use of those features to define the class.
I feel I’m just restating what has already been said but again, I can’t quite see what the problem is. Taxonomy is a human tool of classification which is imperfect but where it does work it relies upon animals sharing similar traits (or not…in which case, into another pot you go or we’ll make a new one for you)
That they all share one defining feature is hardly surprising. But it’s a lot more than one.
Amen, but it is a useful tool for thinking about how things are in the living world. About the only division that isn’t man made is speciation: Can two critters make viable, fertile offspring when they mate? The recent shocking ( 
) discovery that 2 or 3% of the European genome is made up of Neanderthal genes has led to some arguing for classifying them as Homo sapiens neanderthalensis rather than Homo neanderthalensis.
I would find it far more surprising if they all only retained one feature from that single ancestor.
Even then, it isn’t a perfect dividing line. Darwin’s various Galapagos finches are considered different species but can certainly interbreed.
I’m honestly confused on how you can be confused. That is one of the basic and predictable evidences of evolution from common ancestors itself as opposed to individual special creation. You want traits that can be mixed and matched arbitrarily, you gotta go with baraminology.
Linnaeus picked “has mammary glands” as a defining trait of a group of animals. In modern terms, he created a monophyletic clade. Any traits that evolved in the last common ancestor of that group prior to mammary glands is an ancestoral trait in relation to mammary glands so of course it will be a trait of all mammals (or be secondarily lost but should leave developmental or genetic hints behind, like hind limbs in whales.) There are traits that evolved after mammary glands that are found in many mammals but not all of them (nipples, live birth, the SRY gene, the ability to fly, etc.) If* there are traits that are found in all mammals that evolved after mammary glands, then that is just a happenstance of which lineages have survived to today and which are extinct.
*(I say “if” because without extensive googling I don’t know if there are any traits found in all extant mammal groups that postdate mammary glands. Here is one interesting link.)
Why would you restrict it to extant organisms when discussing taxonomy?
I wouldn’t expect fully free mix-and-match of traits. You wouldn’t get that from common ancestry. What you would get is the groups based on some traits to be subsets of the groups based on other traits. If it were just that, say, all animals that have fur also produce milk, then I’d conclude that the first animal to evolve fur happened to be a milk-producing animal. But when it’s also that all animals that produce milk have fur, then I would conclude that the first milk-producer happened to have fur… In other words, when the two groups are identical, I would conclude that each trait evolved before the other, an absurdity.
I wouldn’t. I’d just conclude that we’re missing either a fur-bearing non-lactator or a lactating non-fur-bearer from the fossil record. In reality given there’s some evidence for some non-mammalian cynodonts having hair, but no evidence for them lactating, it’s likely the former.
Some creatures of different species can occasionally make viable, fertile offspring if they mate. I’ll add to @Novelty_Bobble’s finches ligers, beefalo, Savannah cats. I’m sure there are others.
It looks like precursors to full mammalian lactation are very ancient and somewhat associated with fur even in early days. So it may be that mammals are the creatures that took both lactation and fur growth to extremes (possibly because the small size of early mammal ancestors gave large advantages to both developments).
“Lactation represents an important element of the life history strategies of all mammals, whether monotreme, marsupial, or eutherian. Milk originated as a glandular skin secretion in synapsids (the lineage ancestral to mammals), perhaps as early as the Pennsylvanian period, that is, approximately 310 million years ago (mya). Early synapsids laid eggs with parchment-like shells intolerant of desiccation and apparently dependent on glandular skin secretions for moisture. Mammary glands probably evolved from apocrine-like glands that combined multiple modes of secretion and developed in association with hair follicles.” (bolding mine)
To reframe the question a bit: I recall reading somewhere that if you looked at all humans living 1000 years ago, one might naively think that if you selected 1% of those people, it would turn out that they were the ancestors of 1% of the people alive today. But it turns out that is not the case – something like 80% of people at that time are ancestors of 100% of people alive today (and nearly all the rest are ancestors of lines that died out).
And you think about it for a bit and it makes some sense. If two people hook up and have children together all of the ancestors of those two people are ancestors of all of the children, no matter how unrelated those ancestors are otherwise. And eventually that process joins up everyone.
Similarly if you think of all species alive 100 million years ago, one might naively expect that 1% of all species are the ancestors of 1% of species alive today.
What the clustering a traits implies is that at the species level, the opposite thing happens with species that does for people, and that only a disproportionately tiny fraction of species alive 100 million years ago are ancestors of currently alive species. The ancestral species of all current mammals was a furry milk-bearer, which are (clustered) traits that current mammal share because of that. Contemporary to that ur-mammal species were species that either were non-furry milk-bearers, or furry non-milk bearers, and if any of those species had living descendents, we would not see the trait clustering we observe now.
In other words, the question, “why do traits cluster?” reduces to “why do so few species have descendents a significant amount of time on?”. The answer to the latter could be just Darwinian competition (furry milk-bearers survive and reproduce significantly better than species with just one of the traits), or random chance (natural disaster wiped out all of the milk-bearers except for one species that happened to also be furry), or some combination of the two.
Thanks for this, I think it helps explain why traits cluster, and I think we can all agree that it makes sense that the common ancestor of all mammals had hair, produced milk and had three earbones - but what I think Chronos might be wondering (and I at least am wondering) is why there hasn’t been some divergence in those traits since that common ancestor. Are some traits for whatever reason relatively “sticky” and slow to change? eg. why haven’t some furry milk producers evolved to have their earbones merge, or split further, something like that? Or does having three earbones actually convey some sort of evolutionary advantage that is strong enough to have it persist over other possibilities?
The total number of bones is fixed, as they derive from jaw bones. There were proto-mammals with two ear bones, and an extra pair of bones in the jaw. But what we can say is that the ear structures for hearing is reliant on those three bones being formed in a very specific manner. Mutations to change that structure would almost certainly render the animal either deaf or with greatly reduced hearing capacity. So there is significant pressure to preserve the structure. That said, there are some mammals with noticeable modifications, but these are mutations that retained hearing capacity. This is what one would expect. One can assume that a mutation to change the number of bones used for the ear structure from the current one, and still retain functional hearing would be very rare. In animals that have no need for hearing, pressure to preserve ear structure reduces. But the actual number of bones is much more likely to remain, even if hearing becomes non-functional. Eventually there may be mutations that re-purpose the bones, or the bones become so vestigial that they are essentially lost, or the bones fuse together, and we might count that as a change in number.
The thing about the number of bones in the ear is that it is directly related to the number of bones in the jaw structure. The bones are re-purposed, but the total number of bones in the jaw and ear is constant across not just mammals, but also reptiles, and possibly others. What I find fascinating is the proto-morphology of most bony animals. The number of bones, limbs, organs, layout of structure, is common. Morphology changes, sometimes bones fuse and become difficult to distinguish, extraordinary differences exist in shape and function - jaw bones in reptiles, ear bones in mammals - for the same component. Yet the same basic unit is preserved. It is clearly very very difficult to create a mutation that changes it.
A change of the number of ear bones is possible, but it is clear there is a significant hump to crest to allow it, and there may simply not have been enough time for it to occur.