Genetic history of dog domestication

[MUSIC] Welcome to Science Sessions, the podcasts of the proceedings of the National Academy of Sciences, where we connect you with Academy members, researchers, and policymakers. Join us as we explore the stories behind the science. I'm Matthew Hartcastle. Dogs were the first species domesticated by humans at least 14,000 years ago, and they have been an integral part of human society ever since. Despite two decades of genetic research since the first dog genome was sequenced in 2005, many open questions remain about the origins and history of dogs. In this special episode, we'll hear from biologists, geneticists, and archeologists about recent advances in our understanding of how domestication has shaped dog genetics and behavior. Gregor Larson, an archeologist at the University of Oxford in the United Kingdom, is the organizer of a recent Hianneas special feature on the genetics of canine domestication, migration, and behavior. Larson explains how the history of gene flow between dogs and their wild relatives differs from what we know about other domestic species. We know a lot less about dog domestication than we do about a lot of other tax on it, which is intriguing because we know that dogs are the first organism that we form this relationship with. We know explicitly that dogs come from gray wolves. We know that very clearly, but exactly where and exactly when and even questions about like how many separate populations may have independently gone through this process are still kind of up in the air a little bit. What's been super fun for me, anyway, working on a lot of different domestic animals is this really divergent pattern that we see within domestic animals. When farmers first come into Europe, at about 8,000 years ago, they bring with them these domestic pigs that were derived from Near Eastern Wild Boar. Once those pigs come in, they start mating free and fast with the European Wild Boar, such that after only 3,000 years, the entire genome, all of the Near Eastern ancestry that those pigs arrived with, gets completely erased by the European Wild Boar dogs, just resistant to that gene flow like nobody's business. They are the antithesis of the pig, almost suggesting that you have these two peaks on a selection landscape where you have dog on one side and wolf on another side, and something in between, an F1 or a hybrid between a dog and a wolf makes for a very bad dog and a very bad wolf. Despite dogs and wolves being able to produce fertile offspring, crossbreeding between the two species is seemingly rare. Automatable exceptions include a dog gene for a black coat that is now widely found in North American wolves. However, an article in the special feature finds that low levels of wolf ancestry are found in almost two-thirds of dog breeds, accounting for around 0.14% of their genomes on average. Audrey Lynn, a biologist at the American Museum of Natural History in New York, explains her and her colleagues findings. We analyzed 2700 published ancient and modern dog and wolf genomes. Gene flow from wolves to dogs is more ancient than gene flow from dogs to wolves. 64% of modern breed dogs that have wolf ancestry, we estimated to have occurred about 1,000 generations ago on average. When it comes to breeds like arctic sled dogs or pariah breeds and hunting dogs, they have the greatest wolf ancestry and the terriers and gun dogs and sentounds. They have the least amount of wolf ancestry on average. We found that post-domestication wolf ancestry. It correlates with size, breed category, personality characteristics, and the shows that perhaps a genetic source from wolves even in small amounts does in small ways contribute to the specific traits that are valued by humans in different environments. Post-domestication gene flow from wolves may have subtly shaped dog evolution, but how were dogs domesticated in the first place? The traditional view is that humans directly and deliberately managed wolves. Using cubs that became increasingly cooperative with humans. Linus, Gergland Flink, a biomolecular archaeologist at the University of Aberdeen in Scotland is the author of an article on this virtual feature, the documents evidence of two genetic wolves on a remote Scandinavian island that may have been under human control. Humans have been present on these islands despite them not having been connected to any kind of mainland for a long time. The Canada remains that we analyzed were excavated from a cave called Stura Fervar Cave. It was excavated in 1888 to 1893, and several tons of animal bones were excavated from this cave. The vast majority of which were seals, which of course reflects large-scale seal hunting. The genetic data reveals that they were 100% of wolf genetic ancestry. We do not find any evidence of domestic dog ancestry in these individuals. They date to the middle neolithic, as well as the Bronze Age, so about 4,800 to 3,000 calibrated BP across the two different specimens. We also got carbon and nitrogen-stable isotope data from which we can infer a likely diet. They had a very high proportion of marine diet for a fairly long period prior to that. All the data together points towards human control of these wolves. Potentially over multiple generations. We detected a reduced heterocygosity, which is consistent with the reduced population size. Human control is precursor to domestication in many ways. Another line of thinking suggests that dog domestication may have been a less deliberate process. Less aggressive, more trainable wolves may have naturally been able to better take advantage of food scraps and other resources found around humans, in a sense, domesticating themselves. In another article in this special feature, Kalemond Carr, an evolutionary biologist at the University of Gdansk in Poland and his colleagues explore how the mating systems of free-ranging village dogs, which differ from an agame typically seen in wolves, could provide insight in a canine domestication. We have three different populations with very typical characteristics of free-ranging dogs. All these populations live in human-modified landscapes. They are at the same time not constrained by humans for a production. We confirmed in the three populations that free-ranging dogs are polygamous, both for males and females. Even if there is polygamy, it doesn't imply there is promiscuous, but height and the number of social interactions seem to impact the nature. It may indicate a preference for familiar partners in this population. Instead of the traditional view of the domestication process that gives an active role to humans in shaping the characteristics of domesticating species by breeding practices, here we propose instead that this transition happened naturally in the early stages of the domestication process in dogs. Due to the transition to life in human-modified landscape and the presence of more stable food, for example. Analyzing the genomes of ancient dogs and wolves can provide further insights into the domestication process. However, genetic information extracted from ancient remains is often damaged or contaminated. Katya Buyori, an ecologist at the University of Copenhagen in Denmark, is the author of another article in the Special Picture. Buyori explains how she and her colleagues use this statistical method to improve ancient genomes and what a result reveal about the history of inbreeding in dogs. One of the main challenges is the fact that ancient genomes are low coverage and also have a lot of contamination. This leads to samples having quite a lot of missing information. Genome type imputation is a statistical method used to infer missing sites in genomes and it does it by using upload type information from high quality reference panels. The reference panel we used contains around 1500 modern kind of samples. After we heard out the benchmarking, we subsequently carried out a mutation on ancient candy datasets consisting of 50 ancient dogs and 40 ancient wolves. The ancient dogs come from the past 10,000 years maximum, focusing on the different ancestries of the dogs. In most cases, we would get quite good results for coverages as low as 0.5x. We estimated the proportion of our ancient individuals, which were within runs of homozygosity and compared it to present day breeds. The main result as we would expect was that the ancient dogs showed much slower and breeding levels compared to present dog breeds. The genetic structure of modern dogs is complicated by the relatively recent establishment of dog breeds. By limiting the available breeding pool, restrictive breed standards run the risk of increasing inbreeding and associated health concerns. Lockhees Garzbrook, a geneticist at the University of Oxford in the United Kingdom, and the Ludwig Maximilian University of Munich in Germany is the author of another article in the special feature. Garzbrook explains how he and his colleagues use museum specimens to reconstruct the history of inbreeding in German shepherd dogs. Dog breeds as we know them today only really came about in the 1850s. That strong selection for a certain aesthetic criteria. We're expecting that to have a consequence at the genomic level as well. We wanted to track that in real time. So that's why we targeted the museum dogs and our study. The earliest dog we had was from 1906 and that was Lord Quirin. Because these dogs have names, they have pedigrees, we know exactly who they're related to and how they descended. Before World War II, there was a little bit of diversity that had been lost due to the process of breed formation, which is what you'd expect. But this massive popularity reduction during the Second World War caused the diversity to plummet, post-war period. A lot of these genetic health disorders would have been starting to manifest in the dogs themselves. In the American German Shepherd line, the last popular sire, so this individual that all modern German shepherds were related to, was associated with a bottleneck around the birth of this guy called Lance of Fran Joe. And he was responsible for introducing that sloping back end that you see in German Shepherd dogs into the population in America from Germany. This process of everybody breeding to a specific male dog to try and boost these traits through the population also had these genomic effects. The big finding from this is that it's not breed formation, it's driving the majority of the reduction in the diversity of dogs. It's all the things that have happened since. Most dog owners will agree, but dogs, whatever breed they may be, have their own personalities. Eleanor Rathin, a geneticist at the University of Cambridge and the United Kingdom is the author of yet another article in the special feature that analyzes genetic data and owner submitted behavioral questionnaires from 1,343 golden retrievers. Rathin describes how dog genes associated with the behavioral traits that she and her colleagues identified may relate to human temperament, mental health and cognition. We mapped a whole bunch of different traits and found multiple regions of the genome associated with some of them. Then having spotted the doggy regions, we were able to look in the same area of the human genome to see if those regions were also associated with what we've lumped together as temperamentals, psychiatric and cognitive traits in humans. When we did that, it was really striking. Of the 18 associated regions that we looked at in dogs, 12 of them were also associated in humans with these kind of traits. What we see there is that there's perhaps something shared in terms of the underlying brain biology. We're perhaps in the dogs mapping genes, not that cause a specific thing like an ability to follow instructions, but perhaps affect the dog's broader emotional state and make them more amenable to being trainable, for instance. It seems likely that those dogs displaying undesirable behaviors are actually having it driven by a degree of emotional upset that means they can't cope and act out. While genetics may play a role in canine behavior, a dog's environment and upbringing is likely to shape their personality even more. Catherine Lorde, an evolutionary biologist at the University of Massachusetts, Chan Medical School, is another organizer of the special feature. And the author of an article in the special feature, which finds that genetic testing cannot accurately predict canine behaviors. There are some now available gene tests that pet owners can do and they will get back results that have predictions about their dog's behavior because behavior is a very complicated trait. All behaviors probably have a lot of genes that contribute to them. They are also very impacted by the environment, so they are not usually terribly charitable. We have a project called Darwin's ARP, it's a community science program where people can sign up their dogs and answer questionnaires about their behavior and then also send in saliva swabs with which we get their genetic data. So we have a very large group of dogs over 3,000 that have both genetic and behavioral data now. We looked to see if any of the candidate genes that were previously found in these other studies were associated with anything. We found out quite a few of them were associated with aesthetic traits the way that the dog looked, but none of them were associated with behavioral traits in our data set. While we may not yet know the full picture of how dog domestication came about, advances in genetic analyses and techniques are starting to fill in the blanks on our 14,000 year history of coexistence with canines. Whatever new insights into dog genetics and behavior we may find in the future one conclusion is clear. They're all good dogs. Thanks for tuning in to science sessions. You can subscribe to science sessions on iTunes, Spotify, Google Play, or wherever you get your podcasts. If you like this episode, please consider leaving a review and helping us spread the word.