When evolution runs backwards
Running in reverse is not something evolution is meant to do; yet the cases pointing the other way keep multiplying, and in certain species that backward step may actually be where their future lies.
To label any creature an 'evolutionary throwback' still provokes argument. Wary of a long-standing rule holding that evolution simply cannot reverse itself, most biologists have shied away from the phrase for close to a hundred years. That rule, however, is now being rewritten, as fresh cases accumulate and genetics sheds new light on them. Throwbacks, it turns out, are not merely possible; on occasion they even help drive evolution forwards.
Scientists call such a throwback an 'atavism', a word borrowed from the Latin atavus, or forefather. Much of its unpleasant flavour can be traced to Viktor Halász, a nineteenth-century Hungarian physician. Criminality, Halász maintained, was inborn rather than acquired, and a wrongdoer could be picked out by particular bodily traits — features he read as reversions to a primitive, less-than-human condition.
As Halász took his measurements, a Dutch palaeontologist, Anton Reuvers, was poring over fossils and arriving at the reverse view. Evolution, he suggested in 1893, could not be undone: no organism, he wrote, can go back — even in part — to a condition its ancestors had already passed through. Biologists of the early twentieth century reached much the same verdict, but softened it into a matter of odds, allowing that nothing forbids evolution from running backwards; it is merely most improbable. In this form the notion of irreversibility took hold, and it became known as 'Reuvers's rule'.
Were Reuvers's rule correct, atavisms ought to be vanishingly rare, if they happened at all. Almost from the moment the idea gained ground, though, awkward exceptions began to surface. A fin whale hauled up off the coast of southern Chile in 1923, for instance, carried a pair of leg-like projections more than a metre long, each with a complete set of limb bones. For the naturalist Halvard Sørensen, only one reading made sense: the animal was a throwback to an ancestor that had lived on land. 'I can see no other explanation,' he wrote in 1925.
So many further cases have turned up since then that calling evolution all but irreversible no longer holds water. A riddle follows from this: how can features lost to a lineage millions of years earlier abruptly return?
In 1996, Milena Krause and her team at Ashford University set out to attach an actual figure to the odds of evolution reversing, and they turned to genetics to do it. Some evolutionary shifts, they noted, destroy genes outright and cannot be undone; others, though, may amount to no more than genes being switched off. Should those dormant — or 'silent' — genes ever be switched on again, the team argued, traits long since lost might resurface.
Krause's group then tried to pin down how probable such an event was. Left unused, they reasoned, a gene steadily gathers random mutations until it is good for nothing; so how long can one persist in a species once it has fallen idle? Their figures suggested that silent genes stand a fair chance of lasting as long as five million years in at least a handful of a population's members, with a few possibly enduring for eight million. Throwbacks, in short, can happen — but only reaching back into the fairly recent evolutionary past.
By way of illustration, the team turned to a group of salamanders living across Guatemala and Arizona. Most amphibians open life as juvenile 'tadpoles' before transforming into adults, and these salamanders are no exception — save for a single species, known to locals as the 'lake-child', which famously stays a juvenile all its life. The obvious reading is that this one lineage, and no other, gave up the capacity to transform. Trace the group's family tree, however, and a different picture emerges: the remaining lineages sprang from an ancestor that had itself already lost that capacity. Transformation in these salamanders is, then, an atavism — one that sits comfortably within Krause's eight-million-year window.
Lately, though, reports have appeared of cases that overshoot this limit, hinting that silent genes are not the whole account. In a paper released last year, the biologist Tomasz Brandt of Valmont University described his study of the evolutionary past of the reed lizards, a group native to the grasslands of eastern Africa. Tiny limbs are common among them; a number could easily be mistaken for snakes, and in a few the toes of the hind limbs have vanished altogether, while other species carry as many as four toes behind. That the toed kinds simply never mislaid their toes is the tidy explanation — but Brandt takes a different line. His reading of the reed lizards' family tree has the toed species growing their toes afresh from toeless forebears; more striking still, toes have been lost and regained not once but repeatedly across tens of millions of years.
What, then, lies behind this? One account holds that such traits vanish and later crop up again, no differently from the way matching structures can arise on their own in wholly unrelated animals — the wings of birds and of bats being a case in point. A second, more arresting account is that the genetic instructions for building toes lingered somehow inside the lizards for tens, even hundreds, of millions of years before being switched back on. Once revived, these atavistic traits conferred an advantage and worked their way through the population, in effect turning evolution around.
Yet if silent genes fall apart within five to eight million years, how could traits lost far earlier be switched on again? Perhaps the answer lies before birth. In many species the early embryo passes through ancestral features on its way to its final form; snake embryos, for instance, put out buds for hind limbs. Ordinarily a developmental programme telling the body to 'lose the leg' erases these as growth proceeds. Where it fails to run, the ancestral feature can stay put, and an atavism results.
