Week In Science

An atavism is a very unusual feature that an individual of a species is found to have which

• Is extremely rare or even non-existent amongst other representatives of the same species
• Is found commonly in fossil or living species in a broader taxon (a higher grouping of similar organisms in the general hierarchy of living things as classified by taxonomists)
• Has a genetic basis

A classic example of an atavism is the occasional appearance of pelvic protrusions on whales containing unossified (cartilage that hasn't turned into bone) structures that resemble the femurs, tibia, fibulae and occasionally metatarsals of land dwelling mammals with legs. These protrusions are often quite small and appear to be of no use to a whale which might possess them.

Some articles suggest that whales have occasionally been found with complete digits attached to their protrusions, though I have not found such articles specifically identifying primary sources which might contain this information, so I have been unable to verify this.

Another, perhaps less well known atavism is the true human tail.

The coccyx (tail bone) of a human is in fact not an atavism. It is generally referred to as a vestigial feature rather than an atavism (though it in fact has a purpose - it holds various muscles and tissues in place), since it is found in all humans. It consists of four fused bones similar to vertebrae. Note that it is not a tail, and is only classified as a vestigial feature by some writers.

But on very rare occasions a human is born with a tail. Most of these tails are fatty or benign growths and so are not atavisms despite their resemblance to the tails of other animals. Mammalian tails need not contain bones, and in fact one primate is known with a tail not containing bone. But a fatty deposit hardly classifies as a tail.

Even so, rarely, humans are born with tails containing vertebrae, muscles and nerves and which are covered with skin, hair and sweat glands. Such babies have even been known to have control over the tail in response to their emotional state, though they are usually removed for cosmetic reasons very early on, so it is not known if the control is voluntary.

The interesting thing is that human fetuses have for some time a small tail containing vertebrae. Under normal conditions of development these tails undergo programmed cell death (apoptosis) so that the later stage fetus and newborn infant does not have a tail.

In distinguishing an atavism, it is essential to distinguish the cause of the feature, since many defects can occur and the causes can include hereditary diseases, cytogenetic diseases (cell diseases), radiation or drug induced malformations (e.g. agent orange), maternal infection, birth trauma and maternal metabolic factors.

Two major hypotheses have been put forward (that I know of) to explain the human tail atavism (and others).

The first is that humans evolved from an ancestral animal with a tail and that the genetic information for the formation of the tail is still present but usually not activated in humans. A genetic fluke results in the tail developing in rare cases.

Of course humans are classified as apes, and one of the defining characters of apes is that they do not have tails. So any ancestor from which humans might have inherited a tail must be a long way back. This is already somewhat of a difficulty with this explanation.

Tails are not usually heritable features in humans, though at least one and possibly other cases where they have been present in more than one generation of a single family have been documented.

One of the problems with the evolutionary explanation is that atavisms are often used as an evidence for evolution, i.e. evolution is a presumed explanation of atavisms, not a demonstrated one.

Some studies have been done into the genetics of tails and indeed there does appear to be evidence that common genetics are at work in mammalian tails. I have not found specific references to studies of children actually born with tails where this genetic pathway has been demonstrated to be active.

Atavisms cannot in general be used as evidence for universal common descent, for they are not universal, i.e. they do not give an appearance of being throwbacks to the earliest lifeforms known but more recent supposed ancestors.

A major problem with using atavisms as evidence for evolution is that the existence of atavisms is only consistent with macroevolution. They don't strongly imply that evolution took place.

For example, other features not diagnostic of ancestral forms are known and occur relatively regularly. Humans have been born with a single eye, with an extra head, extra limbs, extra nipples, and many others features besides, most of which are not diagnostic of supposed ancestors.

The other major hypothesis put forward regarding human atavisms was that of Louis Bolk (1866-1930) who proposed that they are expressions of usually latent mammalian developmental patterns. In other words, all mammals are able to express these traits and that in humans those traits are simply unexpressed or repressed. Bolk took the view that they were not genetic throwbacks to an earlier generation.

This explanation need not have an evolutionary interpretation. There may be a template mammalian form, which we possess, being mammals, some features of which are supressed in our development. This would certainly align with the evidence in the case of the human tail, fetuses having tails which are then actively supressed by apoptosis.

This explanation also stretches credulity less than the evolutionary throwback explanation, since the many millions of years between supposed ancestors with those features and the appearance of them as atavisms are irrelevant. In the mammalian case, all mammals possess the template, so ancestors with tails are irrelevant. Humans simply supress this particular template trait.

This explanation also tends to align better with our modern understanding of genetics, namely that gross morphology is controlled by genetic switches and that gene networks control whether various features develop or not and how long for, etc.

Bone structures for example are no longer thought of as being controlled feature-for-feature by individual genes. Rather gene networks regulate the development of various buds of cells, features develop and are aligned in response to the chemistry controlled by the timing of the expression of various genes, and programmed cell death takes care of features which make only a transitory appearance during development.

Atavisms may be nothing more than evidence for the fact that species are built with a similar template to other animals in related taxa and that a given species may occasionally express traits which it doesn't usually express.

As studies are conducted it may become clear that higher taxa (above the species level) indeed have a genetic basis, and that there are distinct boundaries to the limits of variation within taxa.

That may be very exciting news to those who are looking forward to seeing where the current biological revolution will lead us.

An even more interesting question is whether atavisms occur in the fossil record. If an atavism was found, how would it be interpreted?

We will try to keep track of the recent news article of a dolphin found off the coast of Japan with pelvic fins (see our news page). In that case we don't know enough to say whether those fins are an atavism, but watch our main page carefully and we'll try to update this story as more information becomes available after studies are conducted by the scientists in Japan.