The simplified diagram (see here) only shows the major mitochondrial groups. Within each circle are a number of breeds. They are not shown here but can be inspected in the original,3 where there are many examples of exactly the same mDNA sequence being found in several different breeds. For example, a Norwegian Buhund, a Border Collie and a Chow Chow had precisely the same mitochondrial DNA sequence. Equally, the same breeds could have different mDNA sequences and appear on different branches of the tree. For instance, the eight German Shepherds had five different sequences between them. We will consider what this means a little later.
Wilson’s Human Tree (simplified). (Image courtesy of Professor Bryan Sykes)
Wayne’s Dog Tree (simplified). (Image courtesy of Professor Bryan Sykes)
Had Darwin been alive to read it, he would have been itching to know where wolves, coyotes and jackals fitted into the tree, if at all. The answer was very clear. The coyote and jackal fell out of the main wolf/dog tree, as it were, immediately. Their DNA sequences were clearly quite different to all the dogs, and none made it into any of the four major branches. When it came to placing the wolf DNA sequences, the answer was equally striking, not because they were outside the dog tree but because they were deeply embedded within it. There was no doubt, from the mitochondrial DNA analysis, that all dogs were descended from wolves and from no other species. It was the first triumph of molecular genetics as applied to dogs – and by no means the last.
Darwin wasn’t wrong about much, but, by means he could never have foretold, his statement on dogs that ‘We shall probably never be able to ascertain their origin with certainty’ would turn out to be one of those rare exceptions. I am sure he would have been utterly delighted to be proved wrong.
Turn the clock forward another ten years to the present day and the Wayne dog tree is still alive. But, like the technical improvements we have considered in the decade between the Wilson and Wayne papers, there have been great strides in DNA analysis in the last ten years, which have led to some radical pruning of the original tree, while leaving the major branches intact.
Before we turn to the effect of these improvements in filling in the blanks in our knowledge of dog evolution there is one other important genetic system to consider. This is the Y-chromosome, the mirror image of mDNA in a genealogical sense in that it traces not the maternal but the paternal genealogy through time. Again the reason is simple enough. Only males have Y-chromosomes and they pass them on exclusively to their male offspring. In many species it is a less reliable witness than the mitochondrial equivalent because of the very variable mating success of males. In most species, including our own, males have the potential to father virtually unlimited numbers of offspring, or none at all, but females are restricted to just a few. This has major implications when we come to look at pedigree dogs.
Just as any conclusion about evolution based on mitochondria should carry the caveat that it can only reveal patterns based on females, so the Y-chromosome only traces the origins of males. Of course, ultimately they both have to tell more or less the same story, but there are fascinating twists and turns along the way.
In any sort of genetic analysis it is vital to be able to detect inherited variation, which is the lifeblood of genetics. Variation comes in many different forms – blood groups, hair colour, height or DNA sequence. You simply can’t do any genetics without it. For DNA the variation in sequence can be read directly, as it is for most mDNA comparisons, or it can use what are known as genetic markers. These are places where the sequence between, in this case, different Y-chromosomes, is known to differ. You can then test for the markers directly without having to sequence the whole chromosome, which saves a lot of time and money. But before you can use them you have to find them, which used to be an enormous bore. It is much better now, as we shall see.
The tedious process of discovering dog and wolf Y-chromosome markers was slow to get going and the first studies used a panel of only four markers. Luckily Y-chromosomes alone are spared the process of shuffling with other chromosomes, something else I will explain as we go along, and so the markers can be combined as blocks. So four markers (A–D), with two versions at each one (1 or 2), can differentiate sixteen Y-chromosomes (A1, B2, C1, D2; A2, B1, C2, D1 and so on), meaning that you can do a lot with just four markers and sixteen combinations.
A group from Sweden was the first to publish any wolf and dog results from this kind of analysis, having studied both Y-chromosomes and mitochondria in 314 dogs from 109 different pedigree breeds.4 Their wolves came from six different regions in Europe and North America, a total of 112 animals. And of course, for both dogs and wolves, all the animals were males. It came as no great surprise after the Wayne mitochondrial pattern (as illustrated here) to find that the dog and wolf Y-chromosomes were similar. Also, there was no sign of any other species, as was always a formal possibility when only the mDNA results were known. Had the original dogs been hybrids between female wolves and male jackals, for example, this would have been invisible to mDNA analysis but not to that of the Y-chromosome. The confidence that wolves really were the only ancestors of all dogs increased substantially after the Swedish study.
In a similar fashion to mDNA, the same Y-chromosome, as defined by its genetic markers, was to be found in several different breeds of dog. As an example, an identical Y-chromosome was found in a Bernese mountain dog, a Border Collie, a Dalmatian, a Greyhound, a Poodle, a Shetland sheepdog and a West Highland terrier. On the other hand, different individual dogs of the same breed often had several different Y-chromosomes. Five Collies, for example, were found to have three different Y-chromosomes between them.
The comparison of the male and female genetic contributions showed quite clearly that in domestic dogs there were many more different mDNA sequences around than there were different Y-chromosomes. What that meant became clear when the wolf results were compared. In wolves the number of different mDNA and Y-chromosome sequences was about the same, not skewed as in dogs. This is very familiar scenario in many human populations where there are lots of different mDNA types but fewer Y-chromosomes than there should be if breeding success was roughly equal across the sexes.
Wolves are almost entirely monogamous, with only one breeding male and one breeding female in a pack. As a consequence males and females make an equal overall genetic contribution to successive generations and, as the Swedish team found, this balances the mDNA and Y-chromosome diversity. In pedigree dogs, the situation is more like some human populations where a few males have a disproportionate number of offspring. The ultimate human example is Genghis Khan, the thirteenth-century Mongol emperor who has an estimated 16 million male descendants living today, each of them carrying his Y-chromosome. Genghis Khan achieved this feat by slaughtering his male enemies defeated in battle and inseminating as many women as possible, often to the point of exhaustion. ‘Try spending the night alone from time to time,’ his doctors cautioned. When he died in 1127 Genghis passed on his wealth, and his habits, to his sons. Male dogs can achieve similar breeding success