UNIVERSITY OF COPENHAGEN STOCKHOLM UNIVERSITY
PhD Thesis
Tatiana Richtman Feuerborn
Genomic Insights into the Population
History of Circumpolar Arctic Dogs
UNIVERSITY OF COPENHAGEN STOCKHOLM UNIVERSITY
PhD Thesis
Tatiana Richtman Feuerborn
Genomic insights into the population history of circumpolar Arctic dogs
Supervisors:
Anders J. Hansen Love Dalén
Mikkel-Holger Strander Sinding Kerstin Lidén
Submitted: 29th February 2020
Cover Image: Tatiana Richtman Feuerborn
“In point of fact they [Canadian Inuit Dogs] are probably the purest bred dogs in the world, being so securely segregated from the rest of the canine world.” Lindsay 1935
Table of Contents
List of papers 5
Candidate’s Contributions 6
English Summary 7
Dansk abstract 8
Svensk sammanfattning 9
Introduction 10
Domestication of the Dog 10
The Role of Dogs in the Arctic 10
Human History in Siberia 11
Human History in the North American Arctic 12
Dogs in the Siberian/Eurasian Arctic 14
Dogs in the North American Arctic 15
Implications of Recent Contact Between the West and the Arctic on Arctic Dogs 16
Epidemics in Arctic Dogs 18
Relationship Between Wolves and Sled Dogs 19
Objectives 21
Methods and Materials 22
Materials 22
Palaeogenetics 22
DNA Extraction 25
Next Generation Sequencing (NGS) 25
Computational Methods 26
Results & Discussion 29
Mitochondrial Haplotype Frequency Shifts Associated Human Cultures 29 Identification and Dietary Stable Isotope Analysis of Mammal Fur in Arctic Clothing 30
Population Structure in Arctic Dogs 31
Gene Flow From Non-Arctic Dogs into Arctic Dogs 33
Introgression From Wolves into Arctic Dogs 35
Mitigation Against Human Contamination in Sequenced Faunal Libraries 36
Future Directions 3 8
References 39
Acknowledgements 48
List of papers
I. Ameen, C.*, Feuerborn, T. R .*, Brown, S.K.*, Linderholm, A.*, et al. (2019) ‘Specialised sledge dogs accompanied the Inuit dispersal across the North American Arctic’, Proceeding of the Royal Society B., 286, https://doi.org/10.1098/rspb.2019.1929
II. Harris, A.* Feuerborn, T. R.* , Sinding, M.-H. S., Nottingham, J., Knudsen, R., Rey-Iglesia, A., Schmidt, A.-L., Appelt, M., Gronnøw, B., Alexander, M., Eriksson, G., Dalén, L., Hansen, A.J., and Lidén, K. (Submitted) ‘Archives of human-dog relationships: Genetic and stable isotope analysis of Arctic fur clothing’
III. Feuerborn, T. R ., Carmagnini, A . , Gopalakrishnan, S., Losey, R., Appelt, M., Grønnow, B., Schmidt, A.-L., Gilbert, M.T.P., Meldgaard, M., Larson, G., Dalen, L., Hansen, A.J.*, Sinding, M.-H.S* , Frantz, L.A* (Manuscript) ‘Gen omic insight into the population history of Siberian dogs’
IV. Feuerborn, T. R ., Gopalakrishnan, S., Fernández Díaz-Maroto, P., Appelt, M., Grønnow, B., Schmidt, A.-L., Rankin, L., Gilbert, M.T.P. , Dalen, L., Meldgaard, M., Sinding, M.-H.S.*, Hansen, A.J.* ( Manuscript ) ‘Pre-contact Inuit dog genomes show a lost wealth of dog diversity in the North American Arctic’
V. Feuerborn, T. R , Pečnerová, P., Ersmark, E., Dehasque, M., Krzewinska, M., Lagerholm, V.K., Munters, A., Rodriguez, R., Ureña, I., von Seth, J., van der Valk, T., Götherström, A., Dalen, L., Díez-del-Molino, D. (Manuscript) ‘Competitive mapping allows to identify and exclude human DNA contamination in ancient faunal genomic datasets’
* These authors contributed equally to this work.
The articles are reprinted with permission from the respective publishers.
The following paper which I contributed to is included as an appendix:
Sinding, M-H.S., Gopalakrishnan, S., Ramos-Madrigal, J., de Manuel Monter, M., Pitulko, V.V., Kuderna, L., Feuerborn, T.R. , Frantz, L.A.F., Vieira, F.G., Niemann, J., Samaniego Castruita, J.A., Carøe, C., Andersen-Ranberg, E.U., Skoglund, P., Jordan, P.D., Pavlova, E.Y., Nikolskiy, P.A., Kasparov, A.K., Ivanova, V.V., Willerslev, E., Fredholm, M., Wennerberg, S.E., Heide-Jørgensen, M.P., Dietz, R., Sonne, C., Meldgaard, M., Dalén, L., Larson, G., Petersen, B., Sicheritz-Pontén, T., Bachmann, L., Wiig, Ø., Marques-Bonet, T., Hansen, A.J., and Gilbert, M.T.P. (Submitted) ‘Emergence of Arctic-Adapted Dogs at Pleistocene-Holocene Transition’
I am an author on the following papers that were published during my PhD but are not included in this thesis:
Pečnerová P., Díez-del-Molino D., Dussex N., Feuerborn T. , von Seth J., van der Plicht J., Nikolskiy P., Tikhonov A., Vartanyan S., Dalén L. (2017) Genome-based sexing provides clues about behavior and social structure in the woolly mammoth. Current Biology 27(22): 3505-3510.
Rowan, B.A., Heavens, D., Feuerborn, T.R. , et al. (2019) ‘An Ultra High-Density Arabidopsis thaliana Crossover Map That Refines the Influences of Structural Variation and Epigenetic Features’, GENETICS;
https://doi.org/10.1534/genetics.119.302406
Candidate’s Contributions
Candidate contributions to thesis articles*
I II III IV V
Conceived
the study Significant Significant Substantial Substantial Significant Designed
the study Significant Substantial Substantial Substantial Significant Collected
the data Significant Substantial Substantial Substantial Significant Analysed
the data Substantial Substantial Substantial Substantial Significant Manuscript
preparation Significant Significant Substantial Substantial Significant
*Contribution Explanation:
Minor: contributed in some way, but contribution was limited.
Significant: provided a significant contribution to the work.
Substantial: took the lead role and performed the majority of the work.
English Summary
The Siberian and North American Arctic have both borne witness to numerous migrations of humans and with them their dogs. This PhD thesis is based on whole genome data from 22 Siberian dogs and 72 North American Arctic dogs, in addition to 186 mitochondrial genomes Siberian and North American Arctic dogs.
Mitochondrial genome data allowed for the identification of migration events that introduced distinct dog populations to North America, associated with different cultural complexes arriving to the region. A novel mitochondrial clade was also identified in dogs from eastern Siberia and Alaska. Genetic analysis was performed to confirm the macroscopic identification of fur used to make clothing in the Arctic in conjunction with stable isotope analyses to explore dietary differences of dog populations across the circumpolar region.
The whole genome data generated for this PhD also detected and explored evidence for several gene flow events from West Eurasian dogs into the dogs of Siberia starting 10,900 BP. There was an additional gene flow event that introduced Near East related ancestry to the dogs of the Siberian Steppe before the Late Bronze Age. Dogs carrying this West Eurasian ancestry spread throughout Siberia, reaching northwestern Siberia by the Iron Age, by 2,000 BP. Further gene flow was detected later in Siberia from West Eurasia a thousand years later. North American Arctic dogs universally carry the Near East related ancestry that is seen in Siberian dogs starting in the Bronze Age, showing it had reached the Bering Strait before the ancestors of the Inuit departed Siberia for Alaska. Once in North America Inuit dogs experienced several other gene flow events from pre-contact subarctic dogs, modern European dogs, and wolves. The population structure seen in North American Arctic dogs reflects geography and the subsequent isolation as well as population turnover events associated with catastrophic epidemics in the dog populations. Finally, a simple method was developed to evaluate and remove human contamination from ancient DNA datasets originating from faunal taxa. All together this thesis has compiled genomic information from 94 Arctic dogs to shed light upon the genetic history of these dogs from the early Holocene through to the present day. This dataset has been able to provide insight not only into past dynamics of Arctic dogs but also a much needed resource for understanding and preserving the indigenous dog populations still present in the Arctic that face continued challenges of globalisation and climate change.
Dansk abstract
Det Sibiriske og Nordamerikanske Arktis har være vidne til talrige folkevandringer af historiske kulture og med disse vandringer fulgte folks hunde. Denne afhandling bygger på fuld-genom-data fra 22 Arktisk Sibiriske og 72 Arktisk Nordamerikanske hunde, samt 186 mitokondrielle genomer fra Arktiske hunde fra Sibirien og Nordamerika. Det mitokondrielle genom data tydeliggøre de individuelle folkevandringer, hvormed ny hunde populationer indførtes til Nordamerika i forbindelse med indvandring af nye kulture.
Yderligere viser data en hidtil ukendt mitokondriel klade i hunde fra Østsibirien og Alaska. Genetiske undersøgelser blev brugt til at bekræfte makroskopiske artsbestemmelser af pels, brugt i klædedragter af Arktiske folk. Pels bestemt som hund, blev brugt til stabilisotopiske undersøgelser, der viser forskellige i diæt hos forskelig hunde populationer cirkumpolært i Arktis. Data fra fulde genomer, tydeliggøre adskillige bølger af genetisk diversitet fra Vesteurasiske hunde der bliver indblandet i Sibiriske hunde, allerede fra for 10,900 år siden. Der ses opblanding med Mellemøstlig hundediversitet i hunde fra den Sibiriske steppe før sen bronzealder. Denne opblandede diversitet spredes op igennem Sibirien og når Nordvestsibirien i jernalderen, hvorefter yderligere Vesteurasiske hundediversitet optræder i Sibirien efter jernalderen. Alle hunde i det Nordamerikanske Arktis, nedstammer delvist fra de Sibiriske bronzealderhunde med Mellemøstlig opblanding, hvilket bevidner at denne genetiske diversitet nåede til Beringstrædet før Inuitterns forfædre spredtes til Alaska fra Sibirien. Efter ankomsten til Nordamerika, bliver Inuithundene påvirket af andre kilder af genetisk diversitet, fra Indianerhunde, ulve og tilslut moderne Europæiske hunde. Den populationsstruktur der opstår blandt de Nordamerikanske Arktiske hunde, afspejler isolation i specifikke geografiske områder, samt kollaps af lokale hunde populationer, som følge af katastrofale epidemier. Sidst i afhandlingen præsenteres en metode til at opfange og fjerne genetisk forurening af moderne menneske DNA fra nedbrudt forhistorisk-DNA fra dyr. Som helhed samler denne afhandling fuld-genom-data fra 94 Arktiske hunde, for at belyse deres genetiske historie fra tidlig Holocæn frem til i dag. Denne data giver ikke bare indsigt i Arktiske hundes historie og dynamik gennem tid, men er også en tiltrængt resurse til at forstå og bevare de oprindelige hundepopulationer, der stadig findes i Arktis men som er pressede af globalisering og klimaændringer.
Svensk sammanfattning
Där har varit många migrationer av människor, och med dem deras hundar, i de sibiriska och nordamerikanska delarna av Arktis. Denna doktorsavhandling är baserad på hela genomdata från 22 sibiriska hundar och 72 nordamerikanska arktiska hundar, samt 186 mitokondriella genom från arktiska hundar från Sibirien och Nordamerika. De mitokondriella genomen möjliggjorde identifiering av migrationer av hundar som introducerade distinkta populationer till Nordamerika, associerade med olika kulturella komplex som anlände till regionen. En ny mitokondriell klad identifierades också hos hundar från östra Sibirien och Alaska. Genetiska analyser utfördes för att bekräfta den makroskopiska identifieringen av päls som användes för att göra kläder i Arktis tillsammans med stabila isotopanalyser för att undersöka dietskillnaderna hos hundpopulationer i den cirkumpolära regionen. Hela genomdata som genererats i denna avhandling upptäckte och undersökte bevis för flera genflöden från västra -eurasiska hundar till sibiriska hundar med början för 10 900 B.P. Det fanns ytterligare ett genflöde före den sena bronsåldern som introducerade börd från Främre Orienten till hundar från den sibiriska stäppen. Hundar som bär denna väst-eurasiska härkomst spred sig över Sibirien och hade nått de nordvästra delarna vid tiden för järnålderns början. Ytterligare genflöde upptäcktes senare i Sibirien från Västeurasien med början för 1 000 år sedan. De arktiska hundarna från Nordamerika har samma börd från Främre Orienten som ses i de sibiriska hundarna i början av bronsåldern, vilket visar att detta genflöde hade nått Berings sund innan förfäderna till inuiterna lämnade Sibirien för Alaska. När de hade nått Nordamerika upplevde de inuitiska hundarna flera andra genflöden från för-kontakt subarktiska hundar, moderna europeiska hundar och vargar. Befolkningsstrukturen i de arktiska hundarna från Nordamerika återspeglar geografi och den efterföljande isoleringen, samt befolkningsomsättningar förknippade med katastrofala epidemier i hundpopulationerna. Slutligen utvecklades en enkel metod för att utvärdera och ta bort mänskliga kontamineringar från gammalt DNA som härrör från fauna. Sammantaget har denna avhandling sammanställt genomisk information från 94 arktiska hundar för att belysa den genetiska historien för dessa hundar från tidigt Holocen till nutid. Dessa data har kunnat ge insikter i den historiska dynamiken hos arktiska hundar, samt också tjäna som en välbehövlig resurs för att förstå och bevara de inhemska hundpopulationer som fortfarande finns i Arktis och som står inför fortsatta utmaningar av globalisering och klimatförändringar.
Introduction
Domestication of the Dog
For millennia, dogs ( Canis lupus familiaris ) have played a central role in human societies following their domestication in the Upper Palaeolithic, sometime between 15,000 and 30,000 years ago (1–5) . Dogs were the first animal species to be transformed from a wild animal to a tame working animal or pet. However, it is not clear where, when, or how and potentially how many times dogs have been domesticated, although Eurasia is the most agreed upon region for the initial domestication (3,5–7) . It is often hypothesised that the process started as a mutually beneficial collaboration between humans and a grey wolf-like canid ( Canis lupus ). The most widely accepted theory about the nature of the earliest close association between dog ancestors and humans involves a process called ‘self-domestication’. This theory suggests that wolves scavenging refuse at human camps and settlements, slowly began showing ‘tame’ behaviour towards humans, eventually manifesting in full domestication. Subsequent to their domestication, the role of the early dogs in human society is unknown. During this period, humans were hunter-gatherers relying on the exploitation of natural resources and living in a world with diverse megafauna - including several carnivores that could be a threat or competition to humans. In this context, it is tempting to imagine the benefits of having dog-like canids in the vicinity of humans to provide protection, assistance in hunting, and perhaps a source of food. Utilising dog-like canids while hunting could have improved the success rates and thereby increased the stability of the communities hosting the dogs. Since then, the collaboration between humans and dogs has expanded and evolved. Human selection for specific traits, behaviour, and skills has resulted in the wide array of phenotypic diversity and specialised dog types that can be seen today.
Current research on dog domestication is confounded by the fact that the earliest domestic dogs and their wolf ancestors are, as of yet, unidentified. The absence of these pieces presents a challenge in reconstruction of the evolutionary history of dogs. However, indisputable domesticated dog remains from as early as 14,500 years ago have been discovered in Germany at the Bonn Oberkassel site (8–10) . Additional ancient dog remains found in the Siberian High Arctic have been dated to 9,500 years ago, which shows that dogs had already spread into regions with extreme environments (11) .
The first ancient dog genome to be published came from a 4,800 year old Irish Neolithic dog which indicated that dogs may have been domesticated in multiple events (5) . Additional Neolithic dog genomes from Germany dated to 7,000 and 4,700 years ago disputed a dual origin of dogs and instead supported a single origin of dogs in Eurasia between 20,000 and 40,000 years ago with continuity of dogs during the European Neolithic (4) . Genomic sequencing of an ancient wolf from the Pleistocene Siberia, dated to 35,000 years ago, showed that Arctic dogs derive part of their ancestry from Pleistocene wolves (12) . While these ancient remains are unable to tell the complete story of dogs from the point of domestication to today, they have provided a great deal of insight into the origin of dogs.
The Role of Dogs in the Arctic
Following their domestication, dogs settled into human society and played a variety of roles, these roles varied culturally, geographically, and temporarily, becoming invaluable in their ability to herd, guard, assist with hunting, and provide companionship among others (9) . Historically, dogs have been used in a variety of
ways to aid human transportation, including acting as pack animals for carrying materials and draft animals for pulling sleds and travois (13–15) . The use of dogs as draft animals is seen as a defining feature for life in the Arctic. The combined power of a team, whether only two dogs or twenty, confers a much greater strength and endurance than a person alone, enabling humans to travel further, faster, and the potential to pull many more kilograms than a human alone could (16) . They also possess the potential to find their way home when lost, for example during blizzards and whiteouts (15,17) . This use of dogs is particularly remarkable evidenced by findings of remains and sled materials dated back to at least 9,500 years old on the New Siberian Islands (11,18) . This in turn indicates that dog sledding has occurred in Siberia for millennia.
Simi lar to Siberia, dog sledding has been an important technology for transportation in the North American Arctic where dog sledding is a hallmark of Inuit culture (19) .
In the Arctic, dogs have frequently spent two lifetimes with their human companions. In the first one, their living-lifetime, they work as hunters, herders as well as pack and draft animals. In their second, their post-living-lifetime, they serve new roles such as food or protection from the elements as clothing. Across the Arctic, dog fur has been utilised as a material for making garments. Dogs are equipped for survival in the Arctic in part by their insulating fur which can also be utilised by humans post mortem. Most mammals have tripled layered fur, comprised of inner down hair, intermediate awn hair, and the outer guard hair, which together create a thick, well-insulated coat especially during winter. The guard hairs work to repel water while the inner layers primarily function for thermoregulation. The properties of dog fur when included in clothing provide much needed protection from the elements. Historically dog fur has predominantly been used as trim for the sleeves and hoods to insulate and prevent frost buildup. Dog fur is particularly valuable as a result of its resistance to repeated freeze-thaw cycles.
Archaeological sites in both Siberia and North America contain the telltale signs of butchery on the remains of dogs (13,20–22) . At sites like Ust-Polui, signs of butchery can be seen on many of the dog bones, indicating that Arctic dogs served numerous purposes in the Iron Age society including sledding, consumption, and likely ritual purposes (20,21) .
Human History in Siberia
Arctic Siberia was occupied by humans potentially as early as 45,000 cal. BP (23–25) , however the occupation was discontinuous due to the extreme climatic conditions (24,26) . At least three major prehistoric influxes of people have led to the settlement of Northeastern Siberia since the Late Pleistocene (25) . The earliest group, Ancient North Siberians, was a population of distant relation to the early hunter-gatherers in West Eurasia (25) . The second major wave brought people who were related to populations in East Asia which was ancestral to ‘Ancient Palaeo-Siberians’ and related to present-day groups in Northeastern Siberia and Native Americans (25) . The final prehistoric wave of human migration into Northeastern Siberia occurred during the Holocene bringing a population known as ‘Neo-Siberians’ who were related to the people of East Asia (25) . In succession, each of these major waves of people entering Northern Siberia resulted in the near replacement of the previous population.
The Bronze Age in Siberia and the Eurasian Steppe, 6,000 to 3,000 BP, marked a period that witnessed an era of striking cultural changes. After the Early Bronze Age, the Ancient Palaeo-Siberian ancestry which had previously been widespread across Siberia became confined to Northeastern Siberia (25) . The remarkable cultural changes and the disappearance of ancestry which was once widespread can be linked to the arrival of
numerous groups to the region (27,28) . These large-scale migrations have been detected through material culture changes in the archaeological record in addition to signals of admixture between distinct human groups and population replacements (27,29) . The Bronze Age migrations in the Steppe and Siberia brought groups from Western Eurasia into these eastern regions of Eurasia (30) . The Andronovo cultural complex, a collection of closely related cultural groups, arrived to the West Siberian Plain from the region around present-day central Kazakhstan approximately 4,000 BP (28) . The largest migration in the Late Bronze Age also brought humans from present day Kazakhstan to the West Siberian Steppe (28) .
Similar to the Bronze Age, several waves of human migration occurred during the Iron Age, 3,000-1,500 BP, in Siberia and the Steppe (31) . Starting around 3,000 BP, groups related to the Scythian cultural complex spread across huge stretches of the Eurasian Steppe (30,31) . Migrations from northern Siberia to the south, also around 3,000 BP, has been postulated to be connected to a period of cooling and in turn making the northern regions less habitable (32) .
Currently, there are over 40 groups with distinct languages and cultures in Northwestern Russia and Siberia (29) . Genetic studies of modern groups in Siberia have revealed different levels of ancestry from non-Siberian populations regionally (29) . Over the last few centuries there have been new migrations into Siberia from outside of the region. According to oral tradition in the Yamal region, the Nenets arrived at the Lower Ob and Yamal Peninsula after the 17th century AD when they brought their nomadic reindeer herding with them (33) . Archaeological investigations and oral traditions both show that the earlier inhabitants of the region lived in houses underground and used dogs instead of reindeer as transportation (33) . Reindeer herding in Kamchatka, on the other hand, was introduced as late as the early 20th century AD (34) . The onset of the Russian exploration and occupation of Kamchatka in the 17th and 18th centuries AD brought the presence of people from the west along with their diseases to the region and a gradual change in lifeways was initiated (34,35) .
Human History in the North American Arctic
The general consensus in archaeological research supports a scenario where the first humans to settle North America arrived from Siberia at the Pleistocene-Holocene transition or at the end of the Late Pleistocene (36–39) . This initial migration of people from Siberia to North America brought the ancestor to the native groups of both North and South America (40–42) .
Fig. 1. Distinct migrations of people into the Americas.
A second migration arriving from Siberia to North America during the middle of the Holocene, around 5,000 BP, stayed concentrated in the Arctic region. This second wave brought cultures related to the Arctic Small Tool Tradition of the western Bering Strait in a cultural complex known as the Pre-Inuit, Paleo-Eskimo or Paleo-Inuit (43) . Paleo-Inuit cultures are thought to have arrived in the North American Arctic around 5,000 years ago and include cultures such as the Ipiutak, Norton, Kachemak, Pre-Dorset, Dorset, Saqqaq, and Independence (44,45) . Humans first entered Greenland around 4,500 BP when the Saqqaq, a subset of the Paleo-Inuit cultural complex, arrived. The Saqqaq occupation lasted until approximately 2,800 BP, making it the longest-lasting culture in Greenland to date (46) . Coinciding with the Saqqaq occupation in West Greenland a culture known as Independence I occupied North Greenland between 4,500 and 3,800 BC (46,47) . Two more Paleo-Inuit cultures occupied Greenland Following the Saqqaq and Independence I eras of Greenland’s history, these are known as the Greenlandic Dorset (2,800-2,000 BP) in South Greenland (48) and the Late Dorset (1,250 - 700 BP) in Northwestern Greenland (47,49) .
The ‘Thule culture’ was first recognised by Therkel Mathiassen in Greenland, who named the archaeological culture after the district where the first remains were found and the expedition which found the remains in 1925, the Thule District and the Second Thule Expedition respectively (50) . The district and expeditions in Greenland were named after the most northerly trading station in the world, which was called such by Knud Rasmussen as inspired by the “Ultima Thule” in classics (51,52) . In keeping with the resolution of the Inuit Circumpolar Council (53) , academics transitioned away from the use of the term Eskimo to Inuit which they themselves have adopted (43) . A further shift has also been recommended to transition to the use of ‘Inuit culture’ rather than ‘Thule culture’ for the same rationale (51) . The Inuit culture, which spread from the western shores of Alaska through to the eastern shores of Greenland, originated in the Bering Sea region, not from any of the Paleo-Inuit cultures which already occupied the North American Arctic (54) . The Old Bering Sea, Birnirk, and Punuk cultures of eastern Siberia and the Bering Strait preceded the Thule culture, from one of these cultures it is thought the Inuit culture arose by 1,000 BP, although it remains debated as to where in Beringia the culture first arose (45,54) .
The Inuit culture rapidly spread across the North American Arctic (45,51) . The swift migration from Alaska to Greenland was enabled by the universal use of dog sleds, in addition to their specialised maritime technology - umiaks and kayaks (45,50) . Eastern migration of the Inuit culture began from Alaska travelling via the Canadian Arctic. The Inuit arrived in the Eastern Arctic, eastern Canada and Greenland, around 800 years ago, with many suggesting that the migration occurred within a single generation (45,51,55,56) . In Greenland, archaeological and genetic evidence suggests the Inuit colonisation of the country consisted of multidirectional waves of migration originating from Ellesmere Island, Canada, and arriving in the Thule District in Northwest Greenland (45,51,56) . From the Thule District, the Inuit travelled in two directions east and south. The eastern migration travelled across northern Greenland, reaching Northeast Greenland shortly after arriving in the Thule District. The southern migration travelled south from the Thule District down the west coast, reaching the Sermilik and Ammassalik Fjords around 600 years ago (57,58) . In the past it has also been suggested that East Greenland was populated at least in part from Northeast Greenland, although the general accepted theory is that populations from West Greenland settled East Greenland (58) . FIGURE Studies of the genetic ancestry of people in North America and Chukotka determined that Na-Dene and Eskimo-Aleut speaking people, such as Athabaskans and Eskimo-Aleut speakers, universally possess ancestry attributed to a Paleo-Inuit related ancestor (59) . Furthermore, current day Inuit, Yup’ik, and Aleutian
Island populations in the North American Arctic, in addition to Na-Dene speakers, share Siberian Paleo-Inuit related ancestry (59) .
A few centuries before the arrival of the Inuit, the Norse settled Greenland. In the 10th century CE the Norse arrived in Greenland, initiating the first contact between Europe and Greenland (60) . Archaeological evidence from Norse and Thule archaeological sites suggests that there was contact between the two cultures, although the extent and nature of this contact is unknown (60) . While the Inuit spread across much of Greenland’s coast the Norse occupied only three settlements, referred to as the East, middle and West Settlements. On their settlements the Norse are thought to have maintained their agricultural lifestyle, they had even brought with them their livestock (60,61) . The Norse occupation of Greenland lasted for approximately 450 years, ending around 500 years ago (60,61) . The disappearance of the Norse from Greenland has been attributed to many causes, including climate change, decline in contact with Europe, failure to adapt their lifestyle to the environment, and conflict with local Thule populations (60) .
Several centuries after the disappearance of the Norse from Greenland, European contact resumed with the arrival of Dutch, English, Norwegian and Danish whalers visiting Greenland during the seventeenth and eighteenth centuries CE (55,62) . Later, Denmark and Norway renewed their connection with Greenland during their eighteenth century AD colonisation of the country (55) . The arrival of Hans Egede at the behest of King Frederik IV of Denmark-Norway marked the beginning of the later only Danish colonisation of Greenland in 1721 (55) . The founding of permanent settlements or villages in west Greenland by missionaries led to the transition from seasonal occupations to sedentary settlement establishment on the West coast during the eighteenth century. Contact between Europeans and the Inuit on the East coast and Thule District occurred during the nineteenth century (55) .
As a result of the interconnected migration histories of humans and dogs, patterns of ancestry connecting the dogs of North America to dogs of Siberian origin will likely be revealed through genomic study of ancient Arctic dogs similar to those seen in humans.
Dogs in the Siberian/Eurasian Arctic
Wolves have occupied Siberia for longer than humans, dogs on the other hand have only been in Siberia since the early Holocene. Debate about the dog/wolf status of some of the earliest ‘dogs’ in Siberia remains a contentious issue (63–65) . Dogs arrived early to the High Arctic alongside humans, evidence of this can be seen at the Zhokhov site on the New Siberian Islands when the islands were still attached to the Siberian mainland (11) . Direct radiocarbon dating of a dog from Zhokhov revealed that these dogs represent the earliest known dogs in the High Arctic starting by 9,500 cal. BP (7) . In addition to dog remains, excavations at the Zhokhov site also recovered organic materials associated with sleds such as sled runners (11,18) . The combination of both dogs and sleds being present on the site contemporaneously has been suggested to represent the earliest evidence of not only dogs in the High Arctic but also dogs being used for the pulling of sleds (11) . Morphological studies of the shape and size of skulls from the dogs at Zhokhov showed that these skulls belonged to fully domesticated dogs and not dogs at a midway point during the domestication process, which has been corroborated by genetic studies (7,11,66) . Excavations at other Siberian sites, such as Afontova Gora, Cape Vhodnoi, and Ust-Polui materials related to dog sledding have also been found, including toggles for harnessing dogs to sleds (11,67) .
In the Trans-Baikal region of Siberia, various sites dated to the Siberian Early Neolithic around 8,000 years ago, have been excavated that contained the remains of dogs (13) . The dogs in this region appear to have been used for hunting and potentially burden carrying. During this time dogs have also been found in articulated burials (13) . Later during the Iron Age, a shift appears to have occurred when dogs began to be consumed (13) . In the neighbouring Cis-Baikal, dog remains have been recovered from around 7,300 years ago in an articulated burial which is the oldest known dog burial in Cis-Baikal (68) . Articulated burials of animals are often associated with ritual and personal or societal significance as a result of the nature of the burials with ‘grave goods’ and the deliberate exertion needed for interments (9,68) .
On the Yamal Peninsula and around the River Ob, there has been widespread and varied usage of dogs (20,33,69) . The Iron Age site, Ust-Polui, was occupied from 2,200 to 1,800 BP, during this period of settlement numerous dog remains were deposited on the site (20) . Sled materials were also recovered from the site indicating that the dogs at Ust-Polui were likely being used as draft animals for pulling the sleds (20,21) . Cutmarks on numerous dog bones suggests that pulling sleds was not their only use at Ust-Polui during the Iron Age (20) . Later in the same region arose the settlement of Tiutei-Sale-I around 1,000 BP (33) . The limited number of reindeer bones and equipment associated with reindeer herding has been interpreted as the occupants of the site not being reindeer herders like the later groups in the area, however similar to the earlier settlements in the area dogs were present on the site (33) .
To the easternmost extreme of Siberia, the Bering Strait the oldest dog and dog sled materials are found associated with the Old Bering Sea and Punuk Cultures from the Bering Sea regions starting around 2,000 years ago (70) . Humans along with their dogs departed Siberia from the western coast of the Bering Strait and eventually arrived in Alaska.
Dogs in the North American Arctic
The first port of entry for dogs entering and staying in the North American Arctic from Siberia was Alaska with the Paleo-Inuit. From Alaska, they spread with people throughout the North American Arctic. Alaska’s shores and islands are scattered with archaeological sites left by these early arrivals. Some of these sites contain an abundance of dog remains like the Uyak site on Kodiak Island where hundreds of remains were found (71,72) . Cape Krusenstern, also in Alaska, has an archaeological record stretching over thousands of years, from early Paleo-Inuit occupations to the present day, containing a rich assortment of archaeological materials including dogs (73) . Other sites have more limited preservation or evidence for dogs.
The earliest culture to arrive in Greenland was the Saqqaq who arrived together with th eir dogs to northwestern Greenland from Ellesmere Island in eastern Canada. Dog remains have been recovered during archaeological excavations of Saqqaq sites in West Greenland - such as Qeqertasussuk, Qajaa, and Nipisat I (16) . Many of the dog bones from Qeqertasussuk and Qajaa were discovered with cut marks consistent with butchery, suggesting that dogs were used as a food source during times of need (16) . Based on the low number of individuals (two or three dogs) per Saqqaq site, it has been concluded that dogs smaller populations during the Saqqaq era (16) and thus have played a different role than the sled dog of the later Inuit occupation (74) .
Throughout the geographical range of the Dorset culture in the Eastern Arctic, there is a remarkable near absence of dog remains (16) . Dorset sites across Canada, such as Tikilik, Lagoon, and Mill Island, have all
yielded dog bones, showing that dogs were not completely absent during the Dorset era (75–77) . Archaeological excavations of Dorset sites in Greenland have not revealed the presence of dogs through skeletal material. However, DNA extracted from sediments in Greenland have contained dog DNA, suggesting that, while the skeletal material has not been preserved, dogs were present in Greenland on Dorset sites (78) . Unfortunately, due to the absence of physical remains and the limited amount of dog DNA in the sediment, little is known about the dogs of the Dorset people. A Dorset age site in the Disko Bay region yielded organic materials that have been interpreted as being the remains of a dog sled, currently the only known Dorset dog sled materials (79) . The limited representation of dogs in Dorset and Independence sites is likely due to a combination of poor preservation, small populations of dogs, and possible confusion between wolf and dog remains (16) . For the first three thousand years of human history in Greenland, there appears to have been sporadic use of dogs in small numbers possibly for hunting, individual pack animals, and occasional food sources (16) .
While Paleo-Inuit dogs may have been used as pack animals, the absence of sled materials and small population sizes indicates that their use with sleds was not likely to have been prevalent (16) . This contrasts with the Inuit culture, which is the first North American Arctic culture to have possessed large groups of dogs along with the first known appearances of sleds in North America (9) . In Greenland, the Inuit culture spread from the Thule District east to the north of Greenland and south to West Greenland and further on to East Greenland. The success of the Thule settlement of the North American Arctic and Greenland has been connected to the technology and the specialised dogs that came with them (16,74,80) .
Implications of Recent Contact Between the West and the Arctic on Arctic Dogs
Indigenous dogs cross the Arctic share a common origin in Siberia, supported by archaeological and modern genetic data. From there, they travelled across the Bering Strait into Alaska. There has been varied studies of the mixture of Arctic dogs in Siberia and North America with non-Arctic dogs as a result of Industrialisation and Globalisation in the last three centuries. The same stands true for the genetic impact of industrialisation and colonisation on Arctic dogs and the changes they have undergone over the last three hundred years.
Genetic studies of Arctic dogs have generally focused on Arctic breeds, dogs officially recognised by kennel clubs, such as Siberian Huskies, Samoyeds, Siberian Laikas, Kamchatskaya ezdovaya, Alaskan Malamute, Canadian Inuit/Eskimo Dog, and Greenland Sled Dog. Despite their official recognition, not all of these breeds have been studied and there have been few studies specifically centred around the genetic study of Arctic dogs (81–83) . Recent genetic studies on these breeds have revealed a mosaic of European ancestry.
Nevertheless, not much in known about current or past populations of these indigenous dogs as finding have focused on populations of the Arctic breeds living outside of the Arctic. Some of this European ancestry likely occurred during the Gold Rush in Alaska and Canada when many European dogs were brought to the North American Arctic for the first time (82) . The Alaskan Husky, a racing sled dog, can also be found in Alaska and beyond, although they are not a cohesive population representative of a ‘breed’. Alaskan huskies have a more diverse ancestry compared to other Arctic dogs. During the Gold Rush, local dogs from Alaska were mixed with European breeds and Siberian Huskies to create a new variety of dogs for pulling sleds and racing, which is reflected in their genetic composition. Genetic studies of partial mitochondrial and nuclear genomes have confirmed the shared heritage of modern Arctic Dogs rooted in Siberia, while simultaneously highlighting some differences between populations as is the case with the Alaskan huskies (3,7,82,84) .
Conversely, partial mitochondrial genomes have been used to suggest that Greenland Sled Dogs and the Canadian Inuit Dog are a single breed, reflecting their shared origin and ancestry (81,85) .
In North America, Greenland was the earliest to witness contact between Europeans and the local Inuit populations. The first dogs of European origin to arrive in Greenland were brought by the Norse. Although the Norse disappeared from Greenland, it remains unknown if there has been a lasting legacy of the Norse occupation left through the contact between the Norse and Inuit dogs. Later, when the whalers began visiting Greenland, starting around 300 years ago, ships brought their dogs ashore to Greenland where they reportedly mixed with the local dogs (86) . In the late 19th century, the explorer Robert Peary attempted to crossbreed Greenland Sled Dogs with Newfoundland Dogs in an attempt to produce ‘a better breed’ (86) . While these dogs are said to have been large and strong, they lacked the resistance to cold and starvation necessary for survival in winter and lasted only three to four generations (86) . In Greenland, the more recent contact between Arctic dogs and non-Arctic dogs has been more limited compared to other regions of the North American Arctic in part as a result of legislation (87) .
Admixture between Arctic dogs and non-Arctic dogs has not been the only consequence of the contact between the Arctic and the West. Colonisation and industrialisation have resulted in changes to settlement patterns and by default dog movements, pressures on population size, and the redistribution of dogs on top of health implications of the comingling of Arctic dogs and non-Arctic dogs.
Early 20th century Arctic explorations in the East Arctic (encompassing the Arctic Archipelago, the neighbouring Canadian mainland, and Greenland) embarked upon archaeological and ethnographic studies of the communities of the past and present (88) . These exploration expeditions utilised local dogs to pull sleds with the teams and their equipment (86) . During the Danish Literary Expedition, Knud Rasmussen brought dogs from West Greenland to Cape York to ‘introduce new blood’ to the dog population in Northwestern Greenland (86) . Later the Second Thule Expedition brought dogs from Upernavik north to Cape York during their journey, facilitating long distance movement of dogs in Greenland (86) . Degerbøl reported that in the Cape York region, it was not customary for dog owners to trade or sell dogs with each other until the
‘baptised’ Greenlanders in West Greenland began communicating with people in Cape York (86) . Knud Rasmussen’s expedition by dog sled across the breadth of the North American Arctic as part of the Fifth Thule Expedition was powered by dogs from the Eastern Arctic, making it the first European expedition to traverse the Northwest Passage in its entirety by dog (89) .
Before the introduction of reindeer (Rangifer tarandus) to Kamchatka, and later mechanical transportation methods, the Kamchatskaya ezdovaya dog breed, registered in 1991, were the main means of transportation in the region by the indigenous Kamchadal, Itelmen, and Koryaks (34,35) . In the eighteenth century, Steller noted the already apparent changes to the mobility patterns in Kamchatka resulting from the Russian occupation and the demanding fishing quotas they placed on the local populations (35) . The dog population in Kamchatka dramatically increased at the turn of the 20th century to keep up with the fish quotas (34) . Conversely, during and following World War II, the dog population declined in Kamchatka (34) . The Beringia Dog Race was founded in 1990 which later contributed to the dog sledding tourism industry. As a result of the race new dogs, Siberian Huskies, were imported in 2014 to “revitalise” the local populations (34) . Records also show the large scale movements of dogs according to the regional demand for them in Canada for use by the Canadian Royal Mounted Police (RCMP). For example, in 1899 during the Yukon Gold Rush,
140 dogs were brought from the coast of Labrador (90) . Although, the RMP are reported to have regionally favoured the use of Siberian Huskies over the local Canadian Inuit Dogs (17,91) . Community or family relocations were also responsible for the movement of large numbers of dogs. For example, approximately one hundred dogs were moved hundreds of kilometres from Mittimatalik (Pond Inlet) and Qausuittuq (Resolute Bay) when seven families relocated in 1953 (15) . There was also said to be ‘frequent’ exchange of dogs between Canada and Greenland by Arctic expedition teams and by the local populations during at least the nineteenth and twentieth centuries, although this was most likely on a small scale.
European/colonial contact with dogs in the Canadian Arctic is frequently discussed in connection to the police slaughter of dogs in the 1950s and 1960s in the East Canadian Arctic (17,90–92) . The killing of dogs was in relation to the Ordinance Respecting Dogs 1929, officially to stop the spread of disease. However, it is widely held to have been “a direct attempt to force the Inuit population of Eastern Canada to assimilate by destroying Inuit dog sociality and thus removing their mode of transportation” (90) . Until the early 2000s the slaughter was widely denied, a changing tide arrived in the form of research and independent commission in the last decade (93) . Regardless of the rational thousands dogs are thought to have been killed during the mid twentieth century (91,92) . Between the nineteenth and twenty-first centuries the number of Canadian Inuit Dogs are thought to have dropped from around twenty thousand to only a few hundred individuals (91) . Over the last hundred years, there has been a transition away from having free roaming dogs to picketting them, tethering them to a set location with a chain. Picketing eliminates the dogs’ ability to move freely through the community. The motivation for this transition has been in large part to improve safety by reducing conflict between dogs and people, principally children and the elderly (91) . The health implications of this transition have been both advantageous and adverse. Dogs that are picketted cannot run throughout the settlement spreading contagion to the rest of the dog population. Furthermore, picketting dogs has led to a shift away from a free breeding dog population to more explicit mate selection for the dogs by people.
However, picketing can result in unhygienic settings for the dogs to live in, repeatedly exposing dogs to the same pathogen as can be the case with the canine distemper virus.
Following outbreaks of disease in Greenland, initially thought to have been brought by European dogs, the official veterinarian of Greenland at the time, S.R Hjortlund, advised restricting the import of non-indigenous dogs to the sledding regions of Greenland in 1904 (87) . Decades later, in 1998, legislation was introduced officially prohibiting the import of dogs to the sled dog district of Greenland in order to preserve the health and genetic purity of the dogs (Act. No 18).
Epidemics in Arctic Dogs
Danish colonists in Greenland brought their European dogs with them. European dogs were thought to have carried Eurasian pathogens after the occurrence of disease outbreaks amongst local Greenland Sled Dog populations (91) . The apparent increased virulence of these diseases in Greenland was linked to the isolation of dogs in the Arctic, thereby reducing their resistance to diseases to which they had not previously been exposed (87) . Rabies-like epidemics in Greenland have been recorded for over 150 years, the earliest recorded occurred in 1859 (94) . Particularly in Cape York, Thule District, there were reports of epidemics that had a heavy toll on the dog population. An epidemic in 1898, thought to be rabies, resulted in the survival of only three dogs (86) . Later investigation indicated that local Arctic fox populations were the cause of the rabies epidemics in northern Greenland, rather than European dogs (94) . A distemper virus
outbreak spread across the North American Arctic in the 1980s, reaching Greenland in 1988 with devastating results. In Qaanaaq alone, 80% of the dogs died as a result of the disease (85,95) . Epidemics such as these resulted in large population turnover and have undoubtedly left a lasting mark on the present-day populations of dogs in these locations, thereby making the modern populations distinct from the populations in the same location in earlier decades and centuries. The Qaanaaq area was repopulated with dogs from various settlements in West Greenland and even imported Canadian Inuit Dogs from North Baffin.
Parallelling many of the outbreaks of disease in Greenland were contemporaneous epidemics in Canada. As previously discussed, the health status of the dogs in Eastern Canada, particularly Nunavut, has been at the centre of their dwindled population size during the twentieth century. The distemper epidemic in the 1980s was first reported in Canada before making its way to Greenland. The first reported connection between arctic foxes and sled dogs in Canada occurred in 1916 (94) . The hamlet of Naujaat in Nunavut, also known as Repulse Bay, lost up to 90% of their dogs in 1930 to the distemper virus (85) .
Similarly, in Kamchatka, an epidemic in 1889 reduced the population of dogs on the peninsula by nearly half, from approximately 11,000 to 6,500 (34) . The spread of disease during the 1889 epidemic in Kamchatka has been attributed to the Russian colonisation (34) . The picketing of dogs has also been suggested to be a contributing factor to the decline in population size and health of the dogs due to an increased exposure to mosquitoes in the summer (34) .
In the Arctic, where dogs live in comparatively high density and undertake long journeys, the spread of disease can take hold and spread rapidly through a population and into others. Given the record of periodic contagions that have decimated the dog populations, strong genetic effects have undoubtedly been observed in these populations. For example, significant decrease of genetic diversity. In cases where population replacements took place, a new genetic signature should be present in the locality where the replacement occurred. On the other hand, in cases where a near replacement took place, the population while carrying this new genetic signature may inadvertently benefit from the influx of ‘new blood’ and thus increased diversity.
Despite the potential for heavy inbreeding, the general health and low levels of genetic diseases in Arctic dogs have been attributed to the intense survival and performance-based pressure placed on the dogs (85) . Relationship Between Wolves and Sled Dogs
“For times immemorial the Eskimos have wanted to get wolf blood into their dogs, the idea being to impart the ferocity of the wolf to the bear-hound.” - Peter Freuchen (96)
Arctic explorers and anthropologists have reported that Arctic groups encouraged the hybridisation of their sled dogs with local wolf populations to strengthen their dog teams (96–100) . On the other hand, some reports state that hybrids often make poor sled dogs and as such wolf traits are selected against (96,101) . Hybridisations of dogs and wolves are genetically and physically possible as the result of their close genetic relationship. Despite the widespread reports of hybridisation between dogs and wolves, there is very limited evidence in the offspring of these events outside of the context of breeds such as the Saarloos Wolfhound and Czechoslovakian Wolfdog. Deliberate hybridisation of dogs and wolves typically involved picketing, or tying-up, a female dog in oestrous in an area visited by wolves in to attract male wolves (100,102) .
A bias has been observed in the wild showing evidence for female wolf-male dog hybridisation being more dominant than male wolf-female dog hybrids (102,103) . The appearance of hybrids with a wolf father and dog mother has rarely been observed in populations studied in Europe (102) . The reason for the limited
evidence for wolf-dog hybridisation remains unclear, although this has been suggested to be linked to biological and behavioural constraints, such as male wolf aggression towards dogs, social compatibility, and differences in fertility cycles (102,103) .
Fig. 2. Global distribution grey wolves (Canis lupus). The present-day distribution (as of 2003) is depicted in the darkest shade of grey and historical distribution is indicated in a lighter grey. The map was downloaded on 26 February 2020 from Wikimedia commons where it is freely available. The original colour scheme was converted into greyscale.
Gene flow between wolves and dogs over the thousands of years since domestication has been seen very infrequently (104,105) . In the context of dog sledding, there are conflicting reports and opinions of deposition of hybrids for sledding. During the Fifth Thule Expedition, a female Greenland Sled Dog was bred with a male wolf. They reported that while the offspring were “large, powerful, and ferocious” their long legs were a hindrance when pulling sleds over uneven ice and problematic on bear hunts (96) . Similarly, Knud Rasmussen embarked upon a similar attempt by bringing a female Greenland Sled Dog back to Denmark with him to the Zoological Gardens in Copenhagen to breed with a captive wolf (96) . The attempt was unsuccessful, but he was able to acquire two hybrids from the Stockholm Zoological Gardens, one (a female) survived the journey to Cape York, Greenland in 1916 (96) . She was uninterested in pulling sleds despite enticement or punishment. Furthermore, she was unsuccessful in her attempts to attract the male dogs (96) . These ethnographic accounts demonstrate the feasibility (or infeasibility) of crossing Greenland Sled Dogs and wolves. However, they also highlight the frequent inaptitude of these hybrids with regard to sledding. Nevertheless, anecdotal reports of wolf-dog hybrids continue to be widespread in Greenland today.
Due to the bias in wolf-dog hybridisation towards female wolf-male dog events these offspring are most likely to be raised by the mother in the wild and the mitochondrial genome of the wolf mother will be passed to the offspring. These events are difficult to detect genetically in past populations as a result of the reduced likelihood of finding one of these individuals in a domestic context. If the reports of Knud Rasmussen and the Fifth Thule Expedition represent the general outcome of dog-wolf hybrids then these admixture events do not often become widespread within the population. Whole genome studies of Greenland Sled Dogs, specifically from regions of Greenland where overlap can be seen between wolf and dog/human territories are needed to address the question of whether introgression from local wolf populations can be detected in the Greenland Sled Dog after their arrival to Greenland from Canada.
