Tamed Page 3

  But then, these Ice Age dogs had always been controversial. Some researchers had called into question the canine credentials of these animals – they seemed so out of step with the rest of the archaeological evidence. The physical differences between these contentious canids and wolves are, admittedly, quite subtle, and doubts were cast on the methods used to analyse and interpret the skulls. The size of the Goyet canid was considered problematic. With such a large skull, it must have had a large body, too – and domesticated animals are generally smaller than their wild counterparts. So perhaps, some researchers argued, it was just another, now-extinct, variety of wolf, rather than a dog. Or, if Goyet and Razbo really were early dogs, they were probably dead-ends – blips, failed experiments in domestication. The bulk of the archaeological evidence still pointed to the true ancestors of our modern dogs being domesticated much later, after the peak of the last Ice Age. A later date would also go some way to explaining the extinction of Ice Age megafauna, such as woolly mammoth and woolly rhinoceros – hunted to extinction, perhaps, as humans teamed up with their deadly canine companions. The objections to the dogginess of the Goyet Cave canid seemed almost too shrill, too indignant: these early ‘dogs’ just didn’t fit with the current edifice of theory; even if they were dogs, they were unlikely to represent the ancestors of our modern hounds. Research into canine domestication is fraught with controversy. If you will forgive me, canine palaeontology is a dog-eat-dog world.

  Neither the bones nor the DNA were producing a clear-cut answer, though – in early 2015 it looked as though the weight of evidence was building for a late date of domestication, after the peak of the last Ice Age. After all the excitement over Goyet and Razbo, those early ‘dog-like’ skulls could just be odd-looking wolves – or early dogs whose descendants died out.

  But the date of domestication at 11,000–16,000 years ago, inferred from the DNA of living dogs and wolves, depended on a few crucial assumptions about mutation rates and generation times. If the actual mutation rates had been slower, or generation times longer, that would push the date earlier – it would have taken longer for the DNA differences seen between modern dogs and wolves to accumulate.

  June 2015 saw the publication of a striking new piece of genetic evidence. This time, rather than sifting through the genomes of modern dogs and wolves, looking for clues to their ancestry, the geneticists had gone after ancient DNA. The transatlantic team, with members based in Harvard and Stockholm, worked on a rib discovered on an expedition to the Russian Taimyr Peninsula in 2010. The rib was clearly canid, and it dated to 35,000 years ago. Sequencing a tiny section of mitochondrial DNA, the researchers were able to identify the species of the animal to whom this bone had belonged – it was from a wolf. The next part of the investigation involved comparing the ancient genome of the Taimyr wolf with genomes of modern wolves and dogs. The degree of difference between the ancient and modern genomes just didn’t tally with previously assumed rates of mutation. Applying the standard rates to the genetic difference between modern wolves and the Taimyr wolf suggested that the common ancestor of both of them lived 10,000–14,000 years ago – but that’s less than half the actual age of the Taimyr wolf. So mutation rates must have been less than had been previously thought – 40 per cent of the assumed rate, or even slower. Using the new, low rate of mutation, the predicted date of divergence of wolves and dogs moves from 11,000–16,000 years ago to 27,000–40,000 years ago.

  The revelations didn’t stop there. The geneticists went on to scrutinise particular patterns of variation in the DNA of modern dog breeds – looking at mutations which each involved a single nucleotide ‘letter’. These genetic variants are known as single nucleotide polymorphisms, or – more snappily – SNPs (pronounced ‘snips’). These single-letter mutations are good indicators of evolutionary history in the genome because they’re common – and often inconsequential, so not weeded out by natural selection. Comparing a handful of SNPs (170,000, to be precise) between modern dog breeds and the Taimyr wolf, the geneticists found that some breeds had more wolf in them than others. This suggests that, after the origin of domestic dogs, some populations had interbred with wild wolves. Breeds that had ended up with a bit more wolf in them included the Siberian husky, Greenland sledge dog, the Chinese Shar-Pei and Finnish spitz. The geneticists also looked at the genetic diversity of modern wolves, and found that the split between North American and European grey wolves must have occurred after the Taimyr wolf lineage had peeled away – but presumably before the sea levels rose at the end of the Ice Age, submerging the Bering land bridge that had – during the glaciation when the sea level was low – provided a link between north-east Asia and North America.

  So have Goyet and Razbo been saved, then, by the latest genetic research? It seems there’s no reason to doubt the existence of domesticated dogs 33,000–36,000 years ago, nor that their descendants could still be with us today. Genetics, though, has thrown a final spanner in the works here. Goyet’s mitochondrial DNA is unusual – distinct from that of both wolves and other dogs, ancient and modern. So we’re left wondering what Goyet actually was – an early experiment in domestication that led nowhere? Or an unusual, ancient type of grey wolf that no longer exists today? A sophisticated analysis on the 3D skull shape of the Goyet Cave canid, published in 2015, suggests that it’s more wolf-like than dog-like after all. And so the argument continues. Razbo, on the other hand, appears to fit nicely into the dog side of the mitochondrial DNA family tree. So it looks as though Razbo really could have been an early dog – he certainly has plenty of close relations alive today, in the form of our current canine companions.

  It’s just incredible how heated the debate over the origins of dogs has been over the last few years. New techniques and new discoveries seem to have the potential to radically change theories. And the story keeps changing. But with all the progress – from better dating of archaeological finds, to faster DNA sequencing – the real history of the origin of our oldest and closest ally seems to be emerging from the shadows at last. And it’s bound to be complicated. Just look at how convoluted the human history that we know about is. When we approach prehistory – our own or the unwritten histories of other species – we may start off very naively, somehow expecting a simple story that neatly summarises the complexity of interactions over thousands of years. It’s no wonder the picture changes as more scientific analyses are carried out and more detail emerges. The work done on the DNA of the Taimyr wolf and its cousins, ancient and modern, shows just how tortuous tracing the roots of domestication can be.

  Having pushed the origin of dogs back into the Ice Age, the next question that emerges is – where were dogs domesticated? And was there a single, discrete area where domestication first began, then spread – or multiple times and places where wild wolves became dogs? This might be impossible to pin down – the domestication of dogs may have started 40,000 years before the present, and interbreeding with wolves continued long after that, and can still happen today. But, armed with the latest genetic techniques, which allow us to unlock secrets from genomes, ancient and modern, we can at least try.

  Finding the homeland of dogs

  The debate over the date of domestication has rumbled on, but pinpointing the area where dogs were first domesticated is no less fraught with contention. On the one hand, the genetic results are unequivocal: dogs are clearly domesticated grey wolves. But the grey wolf has a huge range – right across most of Europe, Asia and North America today, and its geographic range was even wider in the prehistoric past. So where within the huge territory of grey wolves did the alliance with humans first take off? We can quickly eliminate North America – humans arrived too late, after the last glacial maximum, for the original transformation of wolf into dog to have occurred there. Analysis of wolf and dog genomes provides further evidence that dogs must have evolved from wolves in Eurasia. The family tree of canine genomes reveals an early branching event when North American and Eurasian wolves diverged away from
each other, and a later divergence – of Eurasian wolves and dogs. Across the Eurasian range of grey wolves, it’s been very much up for grabs: Europe, the Middle East and East Asia have all been put forward as the original homeland of our canine allies.

  Geneticists – and by now, you won’t be surprised at this – have argued and argued over this particular question. Early analysis of mitochondrial DNA pointed to a possible – single – origin in East Asia. This seemed to be supported by a peculiar shape of part of the mandible that was shared by both Chinese wolves and modern dogs. Genome-wide analyses also seemed to support a single origin, but were less clear about the site of domestication for a while, as wolves from all over Eurasia seemed to be equally related to our modern dogs. Further work on mitochondrial DNA of living dogs from across the world then appeared to settle the question. It revealed what appeared to be a clear connection between all modern dogs and the ancient dogs and wolves of Europe. This seemed to match the archaeology. The bones of ancient dogs have been discovered in East Asia and the Middle East, but the earliest of them dates to just 13,000 years ago – whereas there are prehistoric dogs from Europe and Siberia ranging from 15,000 right back to over 30,000 years ago. The original ancestors of dogs were, most likely, Pleistocene – Ice Age – wolves from Europe.

  In 2016, new evidence came to light. Firstly, there was a careful analysis of the part of the mandible that had been thought to indicate a link between Tibetan wolves (Canis lupus chanco) and modern dogs, apparently supporting an Asian origin. The coronoid process, where temporalis attaches, was a similar shape in both Tibetan wolves and modern dogs: this large bony projection was unusually hook-like and leaning backward. But a wider study showed that only 80 per cent of Tibetan wolves and 20 per cent of dog mandibles showed this particular trait. It was just too variable and inconsistent to be used to infer an Asian origin of dogs. And then, just as this morphological argument for an East Asian origin of dogs fell through, in 2016, a new genetic study appeared on the scene to liven things up again.

  This time the geneticists had surpassed themselves, with a very thorough sequencing of the genome of a 5,000-year-old dog, from the famous Neolithic site of Newgrange in Ireland. They also sequenced mitochondrial DNA from fifty-nine other ancient dogs. They compared all this genetic data with existing data from modern dogs, including eighty full genomes and another 605 sets of SNPs. Firstly, the Neolithic Newgrange dog looked similar in its genes to modern free-living dogs – it hadn’t been moulded by the highly selective breeding that would eventually lead to all our modern breeds. And although its DNA suggested that it could digest starch better than wolves, it couldn’t do it as well as modern dogs.

  However, it was the patterns of variation – or rather, the breaks in that variation – that really caught the researchers’ eyes. One modern breed, the Saarloos, stood out from the rest – a little twig on its own, isolated from the rest of the canine family tree. This was not so surprising, as the breed was created in the 1930s by crossing German shepherds with wolves – it’s a true hybrid. But there was another deep split in the DNA, driving a wedge between dogs from East Asia and those from Europe and the Middle East. The genome of the Neolithic Newgrange beast clustered, or matched up best, with the western Eurasian dogs. But mitochondrial DNA revealed something else – most of the ancient European dogs possessed different genetic signatures compared with modern European dogs. The geneticists suggested that the ancient dogs of Europe must have been largely replaced by a later wave of incomers from the east.

  Hot on the heels of that study, another one reported the results of whole genome analyses on not just one, but two Neolithic dogs – this time from Germany. One dated to the beginning of the German Neolithic, 7,000 years ago (5000 BCE), and the other to the end, around 4,700 years ago (2700 BCE). The early Neolithic dog’s genome was very similar to that of the Newgrange dog from Ireland. Yet there were also clear genetic connections, stretching across the millennia, to the late Neolithic dog – and to modern European dogs. There was no sign here of a major population replacement. But there was an intriguing additional signal of ancestry in the later German dog, suggesting that some interbreeding had taken place with dogs arriving from further east. This could be a canine echo of a major human migration, westwards from the steppe country to the north of the Black Sea, which saw the Yamnaya culture spreading across Europe. The Yamnaya people were horse-riding nomads, who buried their dead along with pottery beakers and animal offerings, under great mounds of earth. It looks like they may have brought their dogs with them, too – but they blended with European dogs, rather than replacing them. The disappearance of the Newgrange dog’s mitochondrial DNA lineage – just one tiny part of its genetic make-up – doesn’t have to indicate population replacement. These disappearances, the pruning of particular genetic lineages, happen all the time.

  But going back beyond Newgrange, to the origin of domestication itself, what is the meaning of that ancient east–west split in dog ancestry? There are two possibilities. It could be that dogs originated once, then spread out and populations became effectively separated, drifting apart genetically and creating the deep rift. Or there could have been two separate origins of modern dogs, from genetically distinct populations of wolves, one somewhere in west Eurasia, and the other somewhere in east Eurasia. Answering this question turns on the timing of the split – and the date of domestication. The genome sequencing of the two Neolithic German dogs helps to pin down those crucial events in time. Added to the existing data, geneticists came up with a date of divergence between dogs and wolves at 37,000 to 42,000 years ago. The east–west divergence then occurred between 18,000 and 24,000 years ago, after domestication. This means that a single origin – followed by a split – is most likely. What’s still up for grabs at the moment, though, is precisely where domestication first took place. The only way to settle that question will be to analyse more ancient DNA – from even earlier dogs, right back into the Ice Age. At the moment, though, the jury is out. Ancient mitochondrial DNA and archaeological evidence seem to suggest that a European origin is most likely – but genome-wide data from modern and early dogs reveals a hotspot of diversity in East Asia, suggesting that dogs have existed there longer than anywhere else.

  This isn’t the last word on the origin of dogs, clearly. But it’s extraordinary to think how much we have learnt in just the last five years. The early pathfinding of genetics showed us the slender routes laid out by the unfurling, maternal lineages of mitochondrial DNA. The latest techniques, sequencing entire genomes, allow us to see the whole genetic landscape. Questions whose answers have eluded us before are now answerable. The next few years will see our view of the past expanding even more. We already know that dogs were domesticated – most likely somewhere in Europe, when our ancestors were nomadic hunter-gatherers. Soon, we may have a better idea of exactly where that alliance was first formed.

  But how did domestication of dogs take place – and just how intentional was it? We’re so used to thinking of domestication of animals and plants as an idea that occurred to our ancestors some 11,000 years ago as part of the so-called ‘Neolithic Revolution’, when our forebears gave up their primitive hunter-gatherer lifestyles and settled down to farm, taking control of themselves and their environment and laying the foundation of civilisation itself. There are lots of things wrong with this simplistic view – not least that domestication is a gradual process that is probably much less deliberate, from the human perspective, than we have tended to presume.

  First contact

  We can only imagine how Ice Age hunter-gatherers and grey wolves teamed up. It probably happened – or nearly happened – many times, in many separate places. There may have been occasions when a tenuous alliance formed, only to break up again. History doesn’t run along a railroad, heading for a destination. It meanders, branches off, and often reaches dead ends (and we can only recognise those dead ends retrospectively). But eventually – as we know with the powerful benefit of hi
ndsight fuelled by science – at least one of these alliances prospered and became cemented so that the ongoing partnership of humans and their canine companions was ensured.

  What we don’t really know is who chose whom. Our instinct may be to suppose that our human ancestors, surely supreme masters of their own destiny, chose the wolves and enslaved them, deliberately moulding them into dogs over generations. In reality, conscious intent may have had very little to do with the transformation of certain wolves into a domesticated species. It may have all started as a gentle form of symbiosis, a loose partnership based on mutual benefit, something more like the story conjured at the opening of this chapter. Perhaps it was even the wolves who drove the process. You don’t need to imagine them having some sort of cunning, canine masterplan. By hanging around humans more and more, even if just picking over middens for scraps of food, the wolves may have unconsciously trained the humans into accepting them – first as neighbours, and then as companions.