RESEARCH
ANCIENT DNA the nearby Semiahmoo Bay region to low cov-
erage (0.05×; “SB dog” hereafter, USNM 3512;
The history of Coast Salish “woolly dogs” revealed by collected 1858). For additional genomic con-
text, we increased the coverage of an ancient
ancient genomics and Indigenous Knowledge dog fromPort auChoix,Newfoundland [AL3194;
4020 calibrated years B.P. (cal B.P.)] (3), from
Audrey T. Lin1,2*, Liz Hammond-Kaarremaa1,3*, Hsiao-Lei Liu1, Chris Stantis1,4, Iain McKechnie5, 1.9× to 11.9×, and sequenced the genome of
Michael Pavel6, Susan sa'hLa mitSa Pavel6,7,8, Senaqwila Sen’ ákw Wyss9, Debra qwasen Sparrow10, an ancient dog from Teshekpuk Lake, Alaska
Karen Carr11, Sabhrina Gita Aninta12, Angela Perri13,14, Jonathan Hartt15, Anders Bergström16,17, (ALAS_015; 3763 BP; 1.23×); three modern
Alberto Carmagnini18, Sophy Charlton19,20, Love Dalén21,22,23, Tatiana R. Feuerborn24,25, coyotes; and 59 modern dogs representing
Christine A. M. France26, Shyam Gopalakrishnan24, Vaughan Grimes27, Alex Harris25, 21 breeds (data S1). We also undertook d
13C and
15
Gwénaëlle Kavich26, Benjamin N. Sacks28,29, Mikkel-Holger S. Sinding30, Pontus Skoglund16, d N stable-isotope analysis of Mutton and the
David W. G. Stanton18,31, Elaine A. Ostrander25, Greger Larson19, Chelsey G. Armstrong15, SBdog to test for substantial differences in their
Laurent A. F. Frantz12,18, Melissa T. R. Hawkins32, Logan Kistler1* dietary life histories. Lastly, we interviewed
seven Coast Salish Elders, KnowledgeKeepers,
Ancestral Coast Salish societies in the Pacific Northwest kept long-haired “woolly dogs” that were bred and wool weavers about family histories and
and cared for over millennia. However, the dog wool–weaving tradition declined during the 19th century, traditional knowledge surrounding woolly dogs
and the population was lost. In this study, we analyzed genomic and isotopic data from a preserved to provide a cultural framework for interpret-
woolly dog pelt from “Mutton,” collected in 1859. Mutton is the only known example of an Indigenous ing the genomic analyses (9). The interviewees
North American dog with dominant precolonial ancestry postdating the onset of settler colonialism. We spanseveralCoast Salish communities, including
identified candidate genetic variants potentially linked with their distinct woolly phenotype. We Stó:lō, Squamish, Snuneymuxw, andMusqueam
integrated these data with interviews from Coast Salish Elders, Knowledge Keepers, and weavers about Nations in British Columbia (BC) and Suquamish
shared traditional knowledge and memories surrounding woolly dogs, their importance within Coast and Skokomish/Twana in Washington.
Salish societies, and how colonial policies led directly to their disappearance.
Woolly dog origins
Throughout northwestern North America there
ogs were introduced to the Americas machine-made blankets by British and Amer- are numerous oral histories and origin stories
D from Eurasia via northwestern North ican trading companies in the early 19th cen-America ~15,000 years ago and have tury (11, 13). However, this explanation ignores 1Department of Anthropology, National Museum of Naturalbeenubiquitous in Indigenous societies of the cultural importance of woolly dogs, as re- History, Smithsonian Institution, Washington, DC, USA.2Richard Gilder Graduate School, American Museum ofthe Pacific Northwest (PNW) for millen- flected through their enduring use by weavers, Natural History, New York, NY, USA. 3Vancouver Island
nia (1–4). Coast Salish peoples in the Salish particularly for high-status items such as re- University, Nanaimo, BC, Canada. 4Department of Geology
Sea region (Fig. 1A) kept multiple different galia (7, 14). Given their role in Coast Salish and Geophysics, University of Utah, Salt Lake City, UT, USA.
5Department of Anthropology, University of Victoria, Victoria,
types of dogs: hunting dogs, village dogs, and societies, it is unlikely that the entire dog-wool BC, Canada. 6Twana/Skokomish Indian Tribe, Skokomish
“woolly dogs” with a thick woolen undercoat tradition would have been abandoned simply Nation, WA, USA. 7Coast Salish Wool Weaving Center,
8
that was shorn for weaving (4, 5). Dog-wool because of the ready availability of imported Skokomish Nation, WA, USA. The Evergreen State College,
9
blankets, often blended with mountain goat textiles. Furthermore, this explanation ignores Olympia, WA, USA. Skwxwú7mesh Úxwumixw (SquamishNation), North Vancouver, BC, Canada. 10Musqueam First
wool, waterfowl down, and plant fibers such weavers’ efforts tomaintain culturally relevant Nation, Vancouver, BC, Canada. 11Karen Carr Studio, Silver
as fireweed and cattail fluff, were prestigious practices in the face of settler colonialism. The City, NM, USA. 12School of Biological and Behavioural
cultural belongings (6–8). Woolly dogs, known use of blankets and robes served not only a func- Sciences, Queen Mary University of London, London, UK.13
w w Department of Anthropology, Texas A&M University,as “sqwemá:y,” “ske'-ha,” “sq əméy,” “sq baý,” tional purpose but also a spiritually protective College Station, TX, USA. 14Chronicle Heritage, Phoenix, AZ,
and “q’əbəl̃” in some Coast Salish languages role in Coast Salish cultures. Wearing a cere- USA. 15Department of Indigenous Studies, Simon Fraser
16
(9), were emblems of some communities, as monial blanket was spiritually transformative University, Burnaby, BC, Canada. Ancient Genomics
Laboratory, The Francis Crick Institute, London, UK. 17School of
depicted in a 19th-century Skokomish/Twana because it intertwined the creator of the blan- Biological Sciences, University of East Anglia, Norwich, UK.
basket (Fig. 1B) (10). ket, the wearer, and the community (13–15). 18Palaeogenomics Group, Institute of Palaeoanatomy,
The first comprehensive book on Salish weav- The only knownpelt of an extinct Coast Salish Domestication Research and the History of Veterinary Medicine,
Ludwig-Maximilians-Universität, Munich, Germany. 19PalaeoBARN,
ing (11) scrutinized most Coast Salish woven woolly dog is of “Mutton,” a dog cared for by nat- School of Archaeology, University of Oxford, Oxford, UK.
blankets in museums around the world, ques- uralist and ethnographer George Gibbs during 20BioArCh, Department of Archaeology, University of York, York,
21
tioning if any contained primarily dog wool and theNorthwest Boundary Survey (1857–1862). Ac- UK. Centre for Palaeogenetics, Stockholm, Sweden.
22Department of Zoology, Stockholm University, Stockholm,
disputing the fiber’s spinnability. More-recent cording to Gibbs’s field journal and Smithsonian Sweden. 23Department of Bioinformatics and Genetics, Swedish
proteomic analysis of 19th-century blankets con- ledgers [National Museum of Natural History Museum of Natural History, Stockholm, Sweden. 24Center for
firmed the use of dogwool in Coast Salish weav- (USNM) A4401 to A4425], Mutton became ill Evolutionary Hologenomics, The Globe Institute, University of
ing (12). In addition, zooarchaeological remains and died in late 1859 (9, 15). His pelt and lower Copenhagen, Copenhagen, Denmark.
25National Human Genome
Research Institute, National Institutes of Health, Bethesda, MD,
thought to be fromwoolly dogs have been found leg bones are housed at the Smithsonian Insti- USA. 26Museum Conservation Institute, Smithsonian Institution,
in dozens of archaeological sites in Coast Salish tution (USNM 4762) (figs. S2 and S4). Suitland, MD, USA. 27Department of Archaeology, Memorial
territories beginning~5,000 years before present In this study, we combined genomic analy- University of Newfoundland, St. Johns, NL, Canada.
28Mammalian
Ecology and Conservation Unit, Veterinary Genetics Laboratory,
(B.P.) (2, 4) (Fig. 1A). The last Coast Salishwoolly sis, ethnographic research, stable isotope and School of Veterinary Medicine, University of California, Davis,
dogs likely lived in the late 19th and early 20th zooarchaeological analysis, and archival re- Davis, CA, USA. 29Department of Population Health and
centuries (5, 13). Later photographs and records cords to investigate this iconic dog’s history, Reproduction, School of Veterinary Medicine, University of
California, Davis, Davis, CA, USA. 30Department of Biology,
referring to woolly dogs extend into the 20th including ancestry, the genetic underpinnings University of Copenhagen, Copenhagen, Denmark. 31Cardiff
century, but these examples likely reflect mixed of woolliness, and their ultimate decline. We School of Biosciences, Cardiff University, Cardiff, UK.
32
ancestry or non-Indigenous breeds (9). sequenced Mutton’s nuclear genome to a mean Department of Vertebrate Zoology, National Museum of Natural
History, Smithsonian Institution, Washington, DC, USA.
The decline in dog-wool weaving has previ- 3.4× depth of coverage and, for comparison, a *Corresponding author. Email: linat@si.edu (A.T.L.);
ously been attributed to the proliferation of nonwoolly village dog (figs. S3 and S5) from liz.hammond-kaarremaa@viu.ca (L.H.-K.); kistlerl@si.edu (L.K.)
Lin et al., Science 382, 1303–1308 (2023) 15 December 2023 1 of 6
Downloaded from https://www.science.org on December 14, 2023
RESEARCH | RESEARCH ARTICLE
involving the woolly dog. Skokomish/Twana
Elder Michael Pavel reports that in a former
time, when all beings including woolly dogs
were recognized as relatives, all were “people”
and were as family. High-status Qw’ó:ntl’an
women are an example of those who trace their
lineages from the woolly dog at a time when all
beings were one family (16). According to Pavel:
“And out of [the origin story], [woolly dogs]
were given the gift of the wool, and they were
able to teach the women how to gather the
wool, how to process the wool, how to spin the
wool, and how to weave with the wool” (9).
Early colonial explorers and scholars specu-
lated that woolly dogs originated in Japan
(17) or were recently introduced to the Coast
Salish by the Dene people from their home-
lands in northern boreal Canada (18). However,
zooarchaeological remains of morphologically
distinct dogs in Coast Salish territories sug-
gest that woolly dog husbandry was present
for ~5000 years before European colonization
(2, 4). Furthermore, longstanding oral histories
and traditional knowledge hold that woolly
dogs have been part of Coast Salish society for
millennia (9).
To test whether Mutton has precolonial or
settler dog ancestry, we first compared hismito-
chondrial genome with 207 ancient and mod-
ern dogs from a global sampling.Mutton carries Fig. 1. Domestic dogs in the culture and society of Indigenous Coast Salish peoples. (A) Coast Salish
the A2b mitochondrial DNA (mtDNA) haplo- ancestral lands include the inner coastal waterways of the Salish Sea in southwest British Columbia and
type, which emerged after dogs initially arrived Washington State. Archaeological woolly dog data are from (2). Distribution of the Coast Salish languages in the
fromEurasia (3). Most of this mtDNA lineage of 19th century are as indicated by colored areas. [The map is modified from https://commons.wikimedia.org/wiki/
so-called precolonial dogs (PCDs) disappeared File:Coast_Salish_language_map.svg and licensed under CC BY-SA 4.0.] (B) Woven Skokomish/Twana basket
after European colonization (3, 19, 20). Mutton’s with woolly dog iconography, depicted with upturned tails. Woolly dog puppies are inside pens represented
nearest mtDNA neighbor is an ancient dog by diamond shapes (10) [courtesy of Burke Museum, catalog no. 1-507]. (C) Forensic reconstruction of a woolly
(PRD10, ~1,500B.P.) fromPrinceRupertHarbour, dog based on Mutton’s pelt measurements and archaeological remains (9). Sketches of Arctic and spitz dog
BC (Fig. 2A and fig. S16). PRD10 is the only ar- breeds are shown for scale and comparison of appearance and do not imply a genetic relationship.
chaeological dog from the PNW in the mtDNA
dataset, and this similarity reflects the deep roots To assess Mutton’s nuclear ancestry, we an- European ancestry if the true contributor of
ofMutton’smaternal ancestry in the region.Apair alyzed 217 globally distributed ancient and this ancestry was equally related (an outgroup)
of modern and ancient (~620 B.P.) dogs from modern dogs. Outgroup-f3 statistics reveal that to the two European breeds in the tests. How-
Alaska form a sister clade of the Mutton-PRD10 Mutton carries substantially greater shared ever, estimates across all permutations are
grouping, further underscoring the long-term genetic driftwithPCDs thanwith anyotherdogs, broadly consistent (Fig. 2D and fig. S18), sug-
maternal population structure in northwestern specifically, with the archaeological remains gesting European ancestry roughly on the order
NorthAmerica. By contrast, the SBdog carries an of a dog from Port au Choix, Newfoundland of one great-grandparent in Mutton’s back-
A1a haplotype, which is similar to that of most (4020 cal B.P.), and from Weyanoke Old Town, ground. By contrast, outgroup-f3 statistics indi-
modern Europeandogs and is themost common Virginia (~1,000 B.P.) (Fig. 2B and fig. S17). cate that the contemporaneous SB dog appears
present-day haplotype worldwide (found in 64 Because Mutton lived after European colo- highly admixed, showing the greatest similar-
out of 207 dogs in our analysis) (21). nization and waves of precolonial dog intro- ity to ancient dogs from Siberia and Alaska
To place a timeframe on the divergence of ductions (3, 21), we tested for gene flow from (fig. S17). The distribution of PCD versus Euro-
Mutton’s maternal lineage, we performed a introduced lineages usingD-statistics.We found pean ancestry tracts in Mutton can provide
molecular-clock analysis on the mitochon- that European breeds yielded strongly positive some additional insight into the timing of ad-
drial phylogeny (data S1). The results suggest D-statistics, indicating that Mutton’s non-PCD mixture. Although this method is imprecise
a mitochondrial common ancestor estimated ancestry most likely stemmed from introduced because of recent admixture and the scarcity
between 4776 and 1853 years B.P. for the sub- European dogs (Fig. 2C). of PCD source–population data, we estimate
clade containing Mutton, PRD10, and the two To refine these results, we used f4-ratio tests that Mutton’s European admixture occurred
Alaskan dogs (95% highest posterior density; with six modern European breeds (Chinese 10.8 ± 4.9 generations before (1 SE). Assuming
Fig. 2A and fig. S16). Although we are limited Crested dog, English Cocker Spaniel, Dalmatian, a 3-year generation time, this analysis suggests
by the analysis of a single individual, this tim- German Shepherd, Lagotto Romagnolo, and admixture ~32 years before Mutton’s birth,
ing is generally consistent with the increasing Portuguese Water Dog), estimating that Mutton consistent with postcolonial admixture (9).
occurrence of small-sized woolly dog zoo- had 84% PCD ancestry and 16% European an- To test for dietary differences betweenMut-
archaeological remains in the regions sur- cestry (11.9 to 19.9%2 SE range; Fig. 2D). The f4- ton and the SBdog,we performed stable isotope
rounding the Salish Sea (2). ratio test may slightly overestimate Mutton’s analysis of d13C and d15N on bone collagen and
Lin et al., Science 382, 1303–1308 (2023) 15 December 2023 2 of 6
Downloaded from https://www.science.org on December 14, 2023
RESEARCH | RESEARCH ARTICLE
Fig. 2. Genetic ancestry of
woolly dogs. (A) mtDNA tree
of 207 dogs with A2b (Mutton)
and A1a (SB Dog) haplotypes
expanded. The map points
correspond to colored tree tips
for the most similar archaeo-
logical and historic dog mtDNAs,
highlighting the subclades of
interest and the broader haplo-
types. Samples used are listed
in data S1. (B) Outgroup-f3
statistics (f3(Gray Fox; Mutton,
B)) or estimation of shared
drift between Mutton and 229
other dogs revealed that Mutton
has the highest similarity to PCDs.
Black-point estimates indicate
ancient genomes. (C) D-statistics
(((PCD, Mutton), Test Dog), Gray
Fox) consistent with gene flow
into Mutton’s background,
with European breeds appearing
the most likely contributors
to Mutton’s non-PCD ancestry.
(D) f4-ratio tests (f4(A, Out;
Mutton, AL3194-Port au Choix):
f4(A, Out; B, AL3194-Port au
Choix)) to estimate the proportion
of European settler-dog ancestry
in Mutton’s background, performed
by using six modern European
breeds as proxies for Mutton’s
European ancestry component.
hair keratin. The SB dog has high d13C and Mutton lived and illustrates how interbreed- maintaining their distinct hair characteristics:
d15N values similar to those of archaeological ing with settler-introduced dogs could have soft guard hairs with an unusually long, crimpy
dogs from the PNW(22), indicating a traditional threatened the survival of woolly dogs. undercoat (fig. S2), which was highly spinna-
marine-based diet (figs. S13 and S14). Mutton’s ble and could be made into warm blanket yarn.
isotope values reveal amore terrestrial and com- The influence of people on the woolly These management practices likely contrib-
plementC3 component–richdiet, likely reflecting dog genome uted to Mutton’s PCD ancestry long after the
Mutton’s life and travels with Gibbs from an Woolly dogs were treated as beloved extended onset of settler colonialism.
early age (fig. S14, B and C, and fig. S15) (9). family members. According to Debra qwasen Long-term husbandry for woolly hair likely
The persistence of a high proportion of post- Sparrow, aMusqueamMasterWeaver, her grand- limited woolly dogs’ effective population size,
colonial PCD ancestry may reflect concerted father [Ed Sparrow (1898–1998)] told her that which would be reflected in nucleotide diversity
efforts by Coast Salish peoples to maintain “every village had [woolly dogs], that they were and thus in Mutton’s heterozygosity. We found
the breed against the pressure of gene flow likegoldbecause theyweremixedwith themoun- that Mutton’s heterozygosity is in the lowest
from nonnative dogs. Mutton lived near the tain goat and then rove and spun” (9). Dogs range of living breeds (n = 51) and village dogs
end of traditional woolly dog husbandry also comprised a form of wealth and status for (n = 42) downsampled to the same coverage
(5, 9, 13). Although he had mixed ancestry, Coast Salishwomen, who carefullymanaged the (Fig. 3A). Additionally, runs of homozygosity
Mutton’s background is dominated by PCD dogs to maintain their woolly coats, isolating (ROH) better reflect recent demography than
ancestors, as compared with that of the con- them on islands or in pens to strictly manage global heterozygosity. Using an ROH method
temporaneous SB dog. This finding may in- their breeding (9, 17, 23). Island names often optimized for low coverage (9, 24), we estimate
dicate careful reproductive management to reflect their connection with dogs, such as that 15.7% of Mutton’s genome is in ROH of
maintain woolly dogs’ distinct genetic make- “sqwiqwmi'” (“Little Dog”) village on Cameron 2.5 mega–base pairs (Mbp) or greater, again
up and phenotype until their decline.Mutton’s Island in Nanaimo, Snuneymuxw territory, in the range of modern breeds. The ancient
fraction of European ancestry also highlights BC. The prevention of interbreeding wool dogs Port au Choix dog also has low genomic het-
the turbulent cultural moment at the time with hunting or village dogs was critical for erozygosity and 11.3% ROH, so Mutton’s low
Lin et al., Science 382, 1303–1308 (2023) 15 December 2023 3 of 6
Downloaded from https://www.science.org on December 14, 2023
RESEARCH | RESEARCH ARTICLE
Fig. 3. Genomic outcomes
of management and selection.
(A) Global heterozygosity and
long runs of homozygosity over
transversions in Mutton com-
pared with modern dogs and the
ancient Port au Choix dog. All
dogs have been downsampled to
Mutton’s coverage level for
analysis. (B) Tree schematic
used in dN/dS analysis to
identify genes under selection
in Mutton compared with other
canids. The branching order
is based on (50). dN/dSgenome
estimates were done separately
including one of the four dogs
plus all other canids. Genes
with elevated dN/dSgenome
values in multiple dogs could
reflect more ancient shared
selection before the separation
of the woolly dog lineage.
Therefore, likely candidates for
selection in woolly dogs were
conservatively assessed where
dN/dSgenome > 1.5 in Mutton
(9) but dN = 0 in the other
three dogs, including one PCD.
(C) Genes with an excess of
nonsynonymous mutations in
Mutton. Black points are the
125 selection candidates identified
on the basis of dN/dSgenome ≥
1.5 in Mutton but dN = 0 in three
other dogs, including one PCD
(9). Several genes with high
dN/dSgenome in Mutton (shown in
gray) are excluded as selection
candidates because they carry
at least one nonsynonymous
mutation in other dogs. This
approach is designed to conservatively highlight genes in which selection is more likely specific to Mutton’s lineage rather than during dog domestication or in the
common ancestors of PCDs. Candidate genes discussed in text are indicated.
heterozygosity may partly reflect shared de- mentally limited with only a single genome, member of the keratin gene family responsible
mographic history from a small PCD founding we identified a candidate set of genes with for the structural integrity of cells in the epi-
population (Fig. 3A). Because of recent Euro- high lineage-specific dN/dS values. We iden- thelium and hair follicles. Mutations in kera-
pean admixture, Mutton’s genome is inevita- tified 125 genes as candidates for positive se- tin genes are linked to curly-hair phenotype in
bly more heterozygous than that of his recent lection in woolly dogs (data S2). Among these, other dogs, rats, and mice (31), and to woolly
woolly dog ancestors. 28 have plausible links to hair growth and hair and hereditary hair loss in humans (26, 30);
To search for evidence of geneticmechanisms follicle regeneration according to a model of and multiple KRT genes underwent selection
for woolliness, we used maximum likelihood– the hair-growth cycle (fig. S12) and are asso- in woollymammoths (25). CERS3, PRDM5, and
based estimation of the enrichment of nonsyn- ciated with cell replication, proliferation, the HAPLN1 are associated with maintaining the
onymousmutations (dN/dS) (ratio of nonsynon- formation of extracellular matrix components, integrity of the skin or connective tissue in hu-
ymous to synonymous mutations) observed vascularization, and related processes (25–31) mans (27, 28). GPNMB is involved in multiple
within Mutton’s coding regions (9). We eval- (Fig. 3C and data S3). cellular functions in the epidermis, potentially
uated 11,112 genes with sufficient sequence Candidate selection genes in Mutton include mediating pigmentation (29). We also manu-
coverage for all dogs and outgroups (data S1) KANK2, a steroid-signaling regulator responsi- ally evaluated 15 specific variants fromprevious
and restricted selection-candidate identifi- ble for hereditary diseases of the hair shaft in literature that are linked with hair character-
cation to geneswith elevated dN/dS inMutton humans (32). A distinct nonsynonymous muta- istics in living dog breeds (data S4). Apart from
but lacking any nonsynonymous mutations in tion in Mutton lies in the amino acid adjacent a widespread FGF5mutation conferring long
three other dogs, including one PCD (Fig. 3B). to the KANK2 mutation, causing a “woolly” hair (33, 34), Mutton showed the ancestral
Although power to detect selection is funda- hair phenotype in humans (32). KRT77 is a allele in all cases with data present (data S4),
Lin et al., Science 382, 1303–1308 (2023) 15 December 2023 4 of 6
Downloaded from https://www.science.org on December 14, 2023
RESEARCH | RESEARCH ARTICLE
illustrating the independent origins of woolly were not completely lost, because many cul- 10. “Burke Museum Record,” Burke Museum basketry exhibition,
dogs’ distinct phenotype. tural teachings and types of expertise were https://www.burkemuseum.org/static/baskets/idgame/
dreport.html.
carried on in secret. Bolton said: “Our people
The impact of colonialism on the iconic 11. P. Gustafson, Salish Weaving (Douglas & McIntyre, 1980).were not allowed to spin on shxwqáqelets [tra- 12. C. Solazzo et al., Antiquity 85, 1418–1432 (2011).
breed’s disappearance ditional spindle whorls]. They could spin on a 13. R. L. Barsh, J. M. Jones, W. Suttles, in Proceedings of the 9th
Woolly dogs’ decline throughout the 19th cen- European one but not on the shxwqáqelets. Conference of the International Council of Archaeozoology,Durham, August 2002, L. M. Snyder, E. A. Moore, Eds. (Oxbow
tury is not fully understood. The narrative that They couldn’t use their looms, and they would Books, 2006), pp. 2–11.
the influx of trade blankets into the region led take them out and burn them or they would 14. L. H. Tepper, J. George, W. Joseph, Salish Blankets (Univ. of
to the abandonment of woolly dog husbandry give them to museums or collectors … The Nebraska Press, 2017).
15. G. Gibbs, Journal, Northwest Boundary Survey, 1857–1862,
oversimplifies a complex scenario. By 1857 (a generation that was there when the Europeans 1859; https://doi.org/10.5962/bhl.title.97030.
year before Mutton’s birth) in Stó:lō territory, came and colonized us, that’s where it ended, 16. K. T. Carlson, in A Stó:lō-Coast Salish Historical Atlas,
where Mutton was most likely acquired, the and there [were] just a few people who went K. Carlson, A. J. McHalsie, Eds. (Douglas & McIntyre, 2001), p. 25.
17. J. K. Lord, The Naturalist in Vancouver Island and British
settler population consisted of only a few dozen underground. And my grandmother and my Columbia (R. Bentley, 1866).
permanent settlers at Fort Langley (35, 36). The mother were two of them” (9). 18. F. W. Howay, The Washington Historical Quarterly 9, 83–92 (1918).
following year, more than 33,000 miners ar- A growing body of research demonstrates 19. A. Bergström et al., Science 370, 557–564 (2020).
20. S. Castroviejo-Fisher, P. Skoglund, R. Valadez, C. Vilà,
rived at present-day BC during the 1858 Fraser how peoples of the PNW cared for and man- J. A. Leonard, BMC Evol. Biol. 11, 73 (2011).
River Gold Rush. This large-scale migration set aged their ancestral lands, cultivating diverse 21. C. Ameen et al., Proc. Biol. Sci. 286, 20191929 (2019).
off conflicts between miners, colonial govern- and highly localized plants and marine foods 22. D. Hillis, I. McKechnie, E. Guiry, D. E. St Claire, C. T. Darimont,
Sci. Rep. 10, 15630 (2020).
ments, and Indigenous peoples. Indigenous (43–45). Woolly dogs may have also been sim- 23. M. Eells, The Indians of Puget Sound: The Notebooks of Myron
populations declined by an estimated two- ilarly localized and diverse.We focused on Coast Eells, G. P. Castile, Ed. (Whitman College, 1985).
thirds between 1830 and 1882 (37). Smallpox Salish dogs, but non-Salish peoples in the PNW 24. K. G. Daly et al., Proc. Natl. Acad. Sci. U.S.A. 118, e2100901118
epidemics—almost one every generation from also kept woolly dogs. For example, Nuu-chah- (2021).25. D. Díez-del-Molino et al., Curr. Biol. 33, 1753–1764.e4 (2023).
the 1700s to 1862 (38)—are estimated to have nulth peoples of western Vancouver Island kept 26. Y. Shimomura, M. Wajid, L. Petukhova, M. Kurban,
killed more than 90% of Indigenous people in a different wool dog that was reportedly bigger A. M. Christiano, Am. J. Hum. Genet. 86, 632–638 (2010).
some villages across BC (38), along with steady andhadcoatsofdifferent colors, includingbrown, 27. F. P. W. Radner et al., PLOS Genet. 9, e1003536 (2013).
28. E. M. M. Burkitt Wright et al., Am. J. Hum. Genet. 88, 767–777
depopulation due to other introduced diseases spotted, black, gray, or white (46–48). These dif- (2011).
such asmumps, tuberculosis, and influenza (37). ferences could be population-specific, or they 29. K. B. Biswas et al., Sci. Rep. 10, 4930 (2020).
Survival of woolly dogs depended upon the could be a result of widespread phenotypic di- 30. N. Wasif et al., Hum. Genet. 129, 419–424 (2011).
31. S. Harel, A. M. Christiano, J. Invest. Dermatol. 132, 2315–2317 (2012).
survival of their caretakers. In addition to dis- versity, as noted by explorers in the 18th and 32. Y. Ramot et al., J. Med. Genet. 51, 388–394 (2014).
ease, expanding colonialism increased cultural 19th centuries (17), reflecting trade among the 33. C. Dierks, S. Mömke, U. Philipp, O. Distl, Anim. Genet. 44,
upheaval, displacement of Indigenous peo- different Indigenous communities. 425–431 (2013).
34. E. Cadieu et al., Science 326, 150–153 (2009).
ples, and a diminished capacity to manage the Weaving andwoolly dogs are intertwined in 35. J. R. Gibson, Farming the Frontier: The Agricultural Opening of
breed. Policies targeted Indigenous governance Coast Salish culture and society, which cannot the Oregon Country, 1786–1846 (Univ. of British Columbia
and inherent rights, resulting in the deliberate be separated from the long-time management Press, 1985).
36. K. Carlson, in A Stó:lō-Coast Salish Historical Atlas, K. Carlson,
disenfranchisement and criminalization of of their ancestral homelands. Weavers, artists, A. J. McHalsie, Eds. (Douglas & McIntyre, 2001), pp. 76–83.
Indigenous cultural practices (39). Indigenous and Elders continue to promote the renewal of 37. K. T. Carlson, in A Stó:lō-Coast Salish Historical Atlas, K. Carlson,
women, the caretakers of woolly dogs and traditional or customary weaving knowledge A. J. McHalsie, Eds. (Douglas & McIntyre, 2001), pp. 92–93.
38. R. Boyd, BC Stud. 101, 5–40 (1994).
weaving knowledge, were specifically targeted. and practices. Artist Eliot Kwulasultun White- 39. B. Lawrence, Indians and Others: Mixed-Blood Urban Native Peoples
Missionization efforts reduced women’s roles Hill (Snuneymuxw) said (9): “It starts to un- and Indigenous Nationhood (Univ. of Nebraska Press, 2004).
in society, and legislation such as the Indian ravel, in a way, people’s understanding of us as 40. E. Hanson, D. P. Gamez, A. Manuel, The Residential School System
(Indigenous Foundations, 2020); https://indigenousfoundations.
Act (1876) explicitly prohibited women from a hunter gatherer society … Our relationship arts.ubc.ca/the_residential_school_system/.
participating in local governance, deniedwomen with thewoolly dogs, our relationshipwith the 41. R. Fisher, Contact and Conflict: Indian-European Relations in
basic property rights, and restricted their move- camas patches and the clam beds, theway that British Columbia, 1774–1890 (UBC Press, ed. 2, 1992).
42. J. S. Lutz, Makúk: A New History of Aboriginal-White Relations
ment (39). In the 20th century, transference of we tended the land and tended the forests … (UBC Press, Vancouver, 2008).
cultural knowledge was further disrupted by these all show the systems in place that are 43. C. G. Armstrong, J. Earnshaw, A. C. McAlvay, J. Archaeol. Sci.
mandatory residential schooling designed to far more complex than what people take for 143, 105611 (2022).
44. D. Lepofsky et al., Ecosystems 24, 248–260 (2021).
remove children from their families and sup- granted about Coast Salish culture.” 45. N. J. Turner, Ancient Pathways, Ancestral Knowledge:
press culture (40). Ethnobotany and Ecological Wisdom of Indigenous Peoples of
Through these compounding waves of co- Northwestern North America, (2 vols.), no. 74 of Mcgill-Queen’s
REFERENCES AND NOTES Indigenous and Northern Studies, J. Borrows, S. Carter,
lonialism, the transmission of important knowl- 1. D. Fedje, Q. Mackie, D. McLaren, B. Wigen, J. Southon, Quat. A. J. Ray, Eds. (McGill-Queen’s Univ. Press, 2014).
edge relating to woolly dog husbandry and Sci. Rev. 272, 107221 (2021). 46. J. T. Forrest, P. Kane, J. R. Harper, West. Hist. Q. 3, 79–81 (1972).
hair processing, spinning, and weaving was 2. I. McKechnie, M. L. Moss, S. J. Crockford, J. Anthropol. 47. J. G. Swan, The Indians of Cape Flattery: At the Entrance to the Strait
interrupted. Stó:lō Elder Rena Point Bolton, Archaeol. 60, 101209 (2020).
of Fuca, Washington Territory (Smithsonian Institution, 1868).
3. M. Ní Leathlobhair et al., Science 361, 81–85 (2018). 48. C. H. Smith, The Natural History of Dogs: Canidae or Genus
95 years old in 2022, recalls how Th’etsimiya, her 4. S. J. Crockford, Osteometry of Makah and Coast Salish Dogs Canis of Authors: Including Also the Genera Hyaena and
great-grandmother, had kept woolly dogs, but (Archaeology Press, Simon Fraser University, 1997). Proteles (W.H. Lizars, 1839).
was forced to give them up: “They were told 5. R. Schulting, Can. J. Archaeol. 18, 57–76 (1994).
49. C. Stantis, github/stantis/PNW-dogs-isotopes, Version v1.0,
6. W. H. Dall, G. Gibbs, J. W. Powell, Tribes of the Extreme Zenodo (2023); https://10.5281/zenodo.10247167.
they couldn’t do their cultural things. There Northwest, and Tribes of Western Washington and
was the police, the Indian Agent and the priests. Northwestern Oregon, vol. I of Contributions to North American ACKNOWLEDGMENTS
The dogs were not allowed. She had to get Ethnology series (Cosimo Classics, 1877). We wish to express our deep gratitude to the Honorable S. Point,
7. W. Suttles, in Indian Art Traditions of the Northwest Coast, Grand Chief, and to G. Point of the Stó:lō Nation for giving us
rid of the dogs” (9). The dogs represented high R. L. Carlson, Ed. (Archaeology Press, Simon Fraser University, permission and encouragement for this research. Thanks to
status and traditional practices that threatened 1982), p. 70. C. Wellman for her role in rediscovering Mutton, assistance with
British and later Canadian dominion and 8. H. G. Barnett, The Coast Salish of British Columbia (University history of the area, and photographs. We raise our hands in thanks to
of Oregon, 1955), vol. 4 of University of OregonMonographs: all people within the Coast Salish communities who have graciously
as such were removed through policies of as- Studies in Anthropology. shared their time and knowledge to realize this project, specifically
similation (40–42). The weaving traditions 9. Materials and methods are available as supplementary materials. Xweliqwiya. R. Point-Bolton (Stó:lō Nation); D. Morsette (Suquamish/
Lin et al., Science 382, 1303–1308 (2023) 15 December 2023 5 of 6
Downloaded from https://www.science.org on December 14, 2023
RESEARCH | RESEARCH ARTICLE
Shxwhá:y Village); E. Kwulasultun White-Hill (Snuneymuxw First provided by Smithsonian Museum Conservation Institute federal coyote from Wyoming is PRJNA734649. Stable isotope data are
Nation); Sulqwan P. Williams (Cowichan); V. Snu’Meethia Elliott and trust funds. P.S. was supported by EMBO, the Vallee available (49). All other public genomic data sources are provided in
(Snuneymuxw); T. Sesemiya Williams (Skwxwú7mesh Úxwumixw/ Foundation, the European Research Council (grant no. 852558), the data S1. License information: Copyright © 2023 the authors, some
Squamish Nation); A. Fritz, Norris family (Lyacksun); T. Jones Wellcome Trust (217223/Z/19/Z), and Francis Crick Institute core rights reserved; exclusive licensee American Association for the
(Tulalip); T. Hohn (Puyallup); and q́wat́ələmu N. Bob (Lummi). funding (FC001595) from Cancer Research UK, the Medical Advancement of Science. No claim to original US government works.
Interviews were carried out under Institutional Review Board and Research Council, and the Wellcome Trust. V.G. was supported by https://www.science.org/about/science-licenses-journal-article-
Research Ethics Board approvals from the Smithsonian Institution an SSHRC-IG. Author contributions: Conceptualization: A.T.L., reuse. This research was funded in whole or in part by The Wellcome
(Human Subjects Protocol no. HS220007) and Vancouver Island L.H.-K., and L.K. Methodology: A.T.L., L.K., H.-L.L., L.H.-K., S.G.A., Trust (217223/Z/19/Z), a cOAlition S organization. The author will
University (no. 101410), with informed consent including explicit opt-in C.S., C.A.M.F., and K.C. Investigation: A.T.L., L.K., C.S., S.G.A., H.-L.L., make the Author Accepted Manuscript (AAM) version available under
permissions to reprint quotations with personal attribution. M.T.R.H., L.H.-K., J.H., I.M., G.K., T.R.F., M.-H.S.S., S.G., L.F., A.B., a CC BY public copyright license.
Computations performed for this paper were conducted on the A.C., A.H., and S.C. Formal analysis: A.T.L., L.K., C.S., C.A.M.F., S.G.A.,
Smithsonian High Performance Computing Cluster, Smithsonian D.W.G.S., and A.H. Visualization: A.T.L., L.K., C.S., K.C., M.H., G.K., and SUPPLEMENTARY MATERIALS
Institution (https://doi.org/10.25572/SIHPC), and the Leibniz I.M. Resources: L.K., M.T.R.H., V.G., B.N.S., I.M., and E.A.O. Funding
science.org/doi/10.1126/science.adi6549
Supercomputing Centre (LRZ). Portions of the laboratory work were acquisition: L.K., P.S., and L.D. Supervision: L.K. and L.H.-.K.
Materials and Methods
conducted in and with the support of the Laboratories of Analytical Writing – original draft: A.T.L., L.K., and L.H.-.K. Writing – review
Figs. S1 to S19
Biology (LAB) facilities of the National Museum of Natural History. and editing: all authors. Competing interests: The authors declare
Tables S1 and S2
Thanks to T. Gilbert for funding the processing/sequencing of that they have no competing interests. Data and materials
References (50–161)
AL3194, J. Ososky for specimen-handling assistance, and L. Orlando availability: Genomic sequencing data for Mutton, SB dog, the
MDAR Reproducibility Checklist
and S. Harding for providing helpful comments on the manuscript. Port au Choix dog (AL3194), and ALAS_015 are available for
Data S1 to S5
Funding: Research was supported by Smithsonian Institution funds to noncommercial use through NCBI SRA Project accession no.
L.K., A.T.L., H.-L.L., and C.S. were supported by Smithsonian PRJNA1005336 and BioSample accession nos. SAMN36985984 to Submitted 12 May 2023; accepted 25 October 2023
postdoctoral fellowships. Funding for stable isotope analysis was SAMN36985987. The SRA Project accession no. for the modern 10.1126/science.adi6549
Lin et al., Science 382, 1303–1308 (2023) 15 December 2023 6 of 6
Downloaded from https://www.science.org on December 14, 2023
The history of Coast Salish “woolly dogs” revealed by ancient genomics and
Indigenous Knowledge
Audrey T. Lin, Liz Hammond-Kaarremaa, Hsiao-Lei Liu, Chris Stantis, Iain McKechnie, Michael Pavel, Susan sa'hLa mitSa
Pavel, Senaqwila Sen#á#w Wyss, Debra qwasen Sparrow, Karen Carr, Sabhrina Gita Aninta, Angela Perri, Jonathan
Hartt, Anders Bergström, Alberto Carmagnini, Sophy Charlton, Love Dalén, Tatiana R. Feuerborn, Christine A. M. France,
Shyam Gopalakrishnan, Vaughan Grimes, Alex Harris, Gwénaëlle Kavich, Benjamin N. Sacks, Mikkel-Holger S. Sinding,
Pontus Skoglund, David W. G. Stanton, Elaine A. Ostrander, Greger Larson, Chelsey G. Armstrong, Laurent A. F. Frantz,
Melissa T. R. Hawkins, and Logan Kistler
Science 382 (6676), .  DOI: 10.1126/science.adi6549
Editor’s summary
Multiple Indigenous groups, such as the Coast Salish in the Pacific Northwest, bred and maintained dogs with a
distinctive woolly undercoat, and hair from these dogs was used in weaving. Lin et al. sequenced the only known
specimen of a Coast Salish woolly dog named Mutton from 1859 (see the Perspective by Orlando). Mutton was found
to have limited introgression from European colonial dogs and otherwise lacked much genetic diversity, suggesting
that these dogs were carefully reproductively managed to maintain their coats. By incorporating traditional knowledge,
historical records, and genetic data, the cultural importance of these dogs and the role of colonialism in their loss are
illuminated. This study joins the growing number attempting to incorporate Indigenous groups in historical studies. —
Corinne Simonti
View the article online
https://www.science.org/doi/10.1126/science.adi6549
Permissions
https://www.science.org/help/reprints-and-permissions
Use of this article is subject to the Terms of service
Science (ISSN 1095-9203) is published by the American Association for the Advancement of Science. 1200 New York Avenue NW,
Washington, DC 20005. The title Science is a registered trademark of AAAS. 
Copyright © 2023 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of Science. No claim
to original U.S. Government Works
Downloaded from https://www.science.org on December 14, 2023