Among other things, you are a mammal. This is an assertion that no rational person is likely to dispute. Such a person might reasonably ask: “What makes people mammals?” This question would likely be answered as follows: “People have hair, make milk, have a single bone in their lower jaw, and so on.” But this answer is completely wrong. It is, in fact, the correct answer to the question, “What characteristics can be used to recognize a mammal?” The correct answer to the question above is: “People are mammals because they are descended from the most recent common ancestor of all mammals.”
The correct answer may seem trivial, but in the case of assertions about how creatures are grouped, it is crucial to understand what sort of claim is being made. When we say that a penguin is a bird, we do not mean to suggest that penguins are endotherms.
Significant confusion can be avoided if, each time we inquire about the membership of a creature in a category, we are meticulous about what we mean. That ospreys are birds is a taxonomic statement. It concerns evolutionary relatedness and should not be taken to imply anything about what a creature is like. That ospreys are endotherms is a statement about their form or function. It says nothing about the evolutionary relatedness of the creature to other organisms possessing analogous features. Blurring this distinction is a persistent hazard for two reasons: related creatures tend to look and function alike, having inherited similar characteristics from shared ancestors, and creatures that look and function in a very similar manner tend to be closely related.
Consider birds. Inevitably, one thinks of flight. Most birds can fly. Were someone to suggest that, since penguins do not fly, they are not birds, you would know something was wrong. Everyone knows penguins are flightless birds. What, then, is the logical basis of what everybody apparently knows? Is it simply agreement? Some rule like, “birds always have feathers”? Modern birds all have feathers. Would that imply that if a species of penguins were to lose their feathers, they would cease to be birds? Would they still remain penguins? What about the penguin species that retained their feathers? Why are birds allowed to lose the ability to fly, but not allowed to lose their feathers?
When taxonomic membership is assigned on the basis of particular physical characteristics, the rules are unavoidably arbitrary. Adaptive evolution adds and removes characteristics from lineages. Why are feathers more essential to birds than flight? For rhetorical purposes, one could make some kind of argument in favor of feathers, but it would be no more rigorous than an argument in favor of one name as opposed to another. The Steller’s jay, Cyanocitta stelleri, could have been designated the black-crested bluebird, Azulus negracappa.
Technical terms now become invaluable. Three of the most important are closely related: clade, taxon, and monophyletic group. Ideally, these terms are synonyms and refer to any complete branch of the tree of life.1 To be a clade, a group of creatures must include an ancestor and all of its descendants. Your mother, and all of her offspring, and all of their offspring, comprise a monophyletic group. And so does your father and all of his offspring. But your mother and father, together with all of their descendants, do not constitute a monophyletic group. The most recent common ancestor of your parents is excluded, as well as that ancestor’s other progeny.
Clades almost always include other, smaller clades. They are inherently fully nested sets. New sets may be recognized as scientific knowledge grows. In biological terms, a creature is a permanent member of every set from which it has arisen. You are: a Homo sapiens, an ape, a primate, a mammal, a vertebrate, and an animal. None of this will likely come as a surprise. You may be surprised to discover that you are a monkey as well. If one clips the tree of life, pruning the branch containing the world’s monkeys, one cannot avoid having captured every ape. Apes are monkeys. If one doesn’t wish to be classified as a monkey, the only alternative is to render “monkey” an arbitrary, and therefore biologically meaningless, term. That still would not justify the assertion that apes are not monkeys.
If you are troubled to discover that logic has made a monkey out of you, then you are likely to be scandalized by the discovery that you are a fish as well. If one prunes the tree of life of its fish, the entire clade of tetrapods falls, too. If fish is to remain a taxonomically meaningful concept, then it must be as the name of the clade whose members are more formally known as the gnathostomes, which includes the tetrapods. If one isolates the branch of tetrapods, one captures snakes, bats, birds, whales and numerous other subclades that have lost their quadrupedal nature. Quadruped is a functional designation, implying four feet, whereas tetrapod is taxonomic, implying descent from a particular ancestor. Tetrapods may have any number of limbs, including zero. Confusion, though problematic, is understandable.
The tree of life is almost entirely composed of binary branchings. The occasional, strange exceptions are fascinating, but rare enough that they need not concern us here. The result of this process of endless bifurcation is a pattern of fully nested sets on the one hand, and entirely non-overlapping sets on the other. Mammals are a monophyletic group that include the primates, and the primates include the chimps. The clade of all mammals also includes a clade of giraffes, because giraffes are within the set of even-toed ungulates, which is nested within the set of mammals. The set of primates, however, does not include giraffes, since primates and even-toed ungulates are non-overlapping sets. Each set is a distinct branch on the mammalian limb. If you prune the set of mammals, you get both these subsets, but you could also prune either set from within Mammalia without disrupting the other. Giraffes are mammals, but they are not primates.
Are you a chimpanzee? Strong evidence suggests that human ancestors branched from the ancestors of modern chimpanzees about six to seven million years ago; the chimpanzee branch itself divided into two distinct species about four million years later. The more recent split led to common chimpanzees (Pan troglodytes) and bonobos (Pan paniscus). If one prunes the chimpanzee branch, the two chimpanzee species fall from the tree together. If one prunes the branch prior to divergence, chimpanzees and humans fall from the tree together.
Does that suggest a biological ambiguity? No. We have found two clades and both are real.
By scientific convention, the term chimpanzee defines the genus Pan, which is limited to common chimpanzees and bonobos. The clade united by descent from the common ancestor of Homo and Pan is Hominini. Humans, by definition, are not chimpanzees, but sister taxa. All the members of the clade we share with chimpanzees are hominins. These designations allow us to describe our cladistic relationship to chimpanzees with precision, while implying nothing whatsoever about the size or significance of the evolutionary gap that separates us.
Biologists can decide how best to label the clades, but cladistic relationships are what they are, a part of the real world. Studying the Earth’s biota, aliens should ultimately arrive at the same tree topology as we do, although they would, of course, label each clade differently.
There is great elegance and power in a system in which taxonomic labels faithfully track cladistic relationships. It is of great value to be able to say, for example, that snakes (Serpentes) are the most successful and diverse clade of legless lizards. A given sentence tells us that snakes are a true monophyletic group, that the snake clade is nested within the larger lizard clade, and that leglessness has evolved in lizards multiple times.
This reasoning, and the precise language that it engenders, does not require us to know anything about the most recent common ancestor uniting a given clade; invariably we know little, if anything, about it. The claim is only that such a creature must have existed. It is extremely unlikely that we will have a fossil, or any other physical evidence, in this respect. Patterns of shared characteristics among its descendants suggest what it was probably like. We cannot say more with any degree of confidence; fortunately, we do not need to.
Suppose that your sister is destined to become the most recent common ancestor of a long-living clade. Her skeleton is due to be reduced to dust. Having fallen into an anoxic bog, your uncle remains uncommonly well-preserved. It would be wrong to proclaim your uncle the most recent common ancestor of this clade. Nevertheless, his bones would be quite good enough for almost any analytic purpose to which fossilized bones are relevant evidence.
Even without an appropriate fossil, we can often infer a great deal about a common ancestor simply by looking at the characteristics shared by the clade’s living representatives. The most recent common ancestor of all currently living birds, for example, was almost certain to have had feathers, a four-chambered heart, hollow bones, and no teeth. And it is pretty clear that it flew.
The most recent common ancestor of all living mammals lived on land, did not fly, had fur, made milk, and had good hearing, warm blood, a four-chambered heart, and a single bone for its lower jaw. Interestingly, it probably laid eggs instead of giving birth to live young. Although almost all extant mammal species give birth to live young, there are three species that do not, and they are, it seems, derived from the earliest mammal branch.
Why not say that wolverines are mammals because they make milk? After all, it is true that every species within the monophyletic taxon Mammalia makes milk, and it is also true that no species outside that group does. Any living animal that makes milk is sure to be a member of the mammal clade. What if some future member of the clade were to lose the capacity to produce milk? Would a milkless mammal species lose its membership in the clade? And what if something incredibly milk-like showed up in a distinct clade? It is vital that we not confuse structural/functional claims with claims of relatedness. Mammals and birds, for example, are both endotherms with four-chambered hearts. But that does not persuade us that they are descended from a common ancestor with those characteristics. There is overwhelming evidence that the most recent common ancestor of birds and mammals was coldblooded with a three-chambered heart. Mammals and birds have therefore independently evolved metabolically warmed blood and efficient four-chambered hearts. So it is true that birds and mammals are endotherms. But it is not true that endotherms are a clade.
Many descriptive biological nouns—for example, evergreen, annual, predator, glider, vine, parasite, and detritivore—are functional rather than taxonomic. Even traits like male and female are functional. When it comes to sex, male plants tend to be far less picky than female plants. Even within a given flower that has both male and female parts, the male parts have lower standards for engaging in sex with strangers than the female parts. Femaleness generally, though not always, comes with higher investment in offspring, and the greater the investment, the greater the returns are to being selective. Male plants and male newts, for example, do not behave alike because they have inherited male quirks from a distant shared male ancestor. Their behavior is similar because they face the same selective forces. For males of most species, offspring are cheap. There is little point in looking for anything beyond the obvious.
Cladistic relationships and functional groupings must always be tracked independently, whether they align perfectly, as with mammals and milk production, or not, as with birds and flight. When relatedness and function describe different groupings, we must remain vigilant about the distinction.
Defining a clade is never arbitrary because the claim is inherently factual. Clams, for example, are either descended from the most recent common ancestor of all snails, or they are not.
If we allow ourselves to decide, for example, that a given taxon has been so thoroughly modified by evolution that it is no longer a member of its ancestral taxon, we allow arbitrary biases to trump analytical rigor. One could argue that birds are sufficiently different from dinosaurs that they are not dinosaurs any more. Equally, one could argue the reverse. If we legitimize the debate, there is no objective way to say who is right. The alternative is to let the facts dictate the nomenclature, and embrace the fascinating facts that this brings to light. In the case of birds, the result is particularly delightful: one clade of dinosaurs escaped extinction at the end of the cretaceous period, and at our present position in “the age of mammals,” living dinosaur species outnumber the extant mammal species by more than two to one.
A defensible system requires that every taxon remain a permanent member of the taxa from which it is descended. With such a system, we lose no information about history, and we can describe evolutionary relationships as real as that of parent to offspring. Our present taxonomy retains a lot of pre-Darwinian baggage. Getting taxonomic labels to fully track our cladistic understanding should be a top priority among evolutionary biologists. Not only is the concordance useful for all rigorous organismal research, but it allows us to teach about the power of adaptive evolution simply by describing the evolutionary relationship between familiar groups. Adaptive evolution is the force that modifies ungulates such that they take on the characteristics of orcas.
Witness the following:
- Seals, sea lions, and walruses comprise a monophyletic group, Pinnipedia. Evidence has yet to fully settle the question whether the pinniped clade is a member of the bear clade or the weasel clade. Although pinnipeds look very little like terrestrial bears, a comparison between the skull of a modern bear and that of a seal reveals remarkable morphological similarity. The only easily identifiable difference between the two is the seal’s heavily modified teeth.
- Whales are members of the even-toed ungulate clade that returned to the sea about fifty million years ago. The intervening period between the time their first tetrapod ancestors initially emerged from the sea, and their eventual return to the sea, was more than three hundred million years. Their closest living relatives are hippos.
- In the late Devonian period, members of the lobe-finned fish clade, Sarcopterygii (a clade within Osteichthyes, the bony fish) began to exploit shallow water habitat with increasing frequency. A clade of terrestrial amphibians, ancestral to the clade of all terrestrial vertebrates (Tetrapoda), resulted, ultimately giving rise to the approximately 6,000 partly terrestrial, and nearly 22,000 fully terrestrial, sarcopterygian fish species now living. Dozens of these terrestrial fish taxa have spawned partial (e.g. seals, sea turtles, penguins) or complete (e.g. whales, sea snakes, manatees) reversions to an aquatic habit, though every secondarily-aquatic tetrapod species continues to breathe air, and many continue to breed on land.
In his great dictionary, Samuel Johnson describes the lexicographer as a “writer of dictionaries; a harmless drudge, that busies himself in tracing the original, and detailing the signification of words.”
The post-Darwinian drudge, it would seem, occupies a more conspicuous position.
Letters to the Editors
- Or, as my student Peter Cardi eloquently put it: “a clade is anything that falls from the tree of life with a single cut.” ↩