Access to this full-text is provided by Wiley.
Content available from Animal Genetics
This content is subject to copyright. Terms and conditions apply.
Genetic structure and admixture in sheep from terminal breeds in the
United States
K. M. Davenport* , C. Hiemke
†
, S. D. McKay
‡
, J. W. Thorne*
,§
, R. M. Lewis
¶
, T. Taylor** and
B. M. Murdoch*
*Department of Animal and Veterinary Science, University of Idaho, Moscow, ID 83844, USA.
†
Niman Ranch and Mapleton Mynd
Shropshires, Stoughton, MA 53589, USA.
‡
Department of Animal and Veterinary Sciences, University of Vermont, Burlington, VT 05405,
USA.
§
Texas A&M AgriLife Extension, San Angelo, TX 76901, USA.
¶
Department of Animal Science, University of Nebraska–Lincoln, Lincoln,
NE 68583, USA. **Department of Animal Science, Arlington Research Station, University of Wisconsin–Madison, Arlington, WI 53911,
USA.
Summary Selection for performance in diverse production settings has resulted in variation across
sheep breeds worldwide. Although sheep are an important species to the United States, the
current genetic relationship among many terminal sire breeds is not well characterized.
Suffolk, Hampshire, Shropshire and Oxford (terminal) and Rambouillet (dual purpose) sheep
(n=248) sampled from different flocks were genotyped using the Applied Biosystems Axiom
Ovine Genotyping Array (50K), and additional Shropshire sheep (n=26) using the Illumina
Ovine SNP50 BeadChip. Relationships were investigated by calculating observed heterozy-
gosity, inbreeding coefficients, eigenvalues, pairwise Wright’s F
ST
estimates and an identity
by state matrix. The mean observed heterozygosity for each breed ranged from 0.30 to 0.35
and was consistent with data reported in other US and Australian sheep. Suffolk from two
different regions of the United States (Midwest and West) clustered separately in eigenvalue
plots and the rectangular cladogram. Further, divergence was detected between Suffolk
from different regions with Wright’s F
ST
estimate. Shropshire animals showed the greatest
divergence from other terminal breeds in this study. Admixture between breeds was
examined using ADMIXTURE, and based on cross-validation estimates, the best fit number of
populations (clusters) was K=6. The greatest admixture was observed within Hampshire,
Suffolk, and Shropshire breeds. When plotting eigenvalues, US terminal breeds clustered
separately in comparison with sheep from other locations of the world. Understanding the
genetic relationships between terminal sire breeds in sheep will inform us about the
potential applicability of markers derived in one breed to other breeds based on relatedness.
Keywords genetic admixture, genetic relationships, sheep, terminal sheep breeds
Introduction
The production of lamb and wool is an important agricultural
industry in the United States, with approximately 5 million
sheep and 80 000 operations (USDA ERS 2019). According to
the American Sheep Industry National Animal Health Mon-
itoring System’s most recent study, 81.6% of operations raise
sheep for meat purposes (American Sheep Industry 2011).
The most popular breeds used formeat production include the
Suffolk, Hampshire, Shropshire, Oxford, and Southdown
(American Sheep Industry 2011). To make progress in their
own flocks, some US lamb and wool producers have imple-
mented quantitative genetic selection strategies using esti-
mated breeding values through the National Sheep
Improvement Program (NSIP) to identify and select animals
with desirable traits (Wilson & Morrical 1991; Notter 1998;
Lupton 2008). As this program is more widely utilized, the
improvement of product quality and yield of lamb and wool
products in the United States is anticipated to accelerate.
Previous research indicates that selection for various
traits such as wool or growth within breeds of sheep has led
to greater breed specialization across the world (Kijas et al.
2012; Zhang et al. 2013). However, many breeds of sheep
have retained greater heterozygosity in comparison with
other species, including cattle (Bovine HapMap Consortium
Address for correspondence
B. M. Murdoch, Department of Animal and Veterinary Science,
University of Idaho, Moscow, ID 83844, USA.
E-mail: bmurdoch@uidaho.edu
Accepted for publication 12 December 2019
doi: 10.1111/age.12905
284 ©2020 The Authors. Animal Genetics published by
John Wiley & Sons Ltd on behalf of Stichting International Foundation for Animal Genetics, 51, 284–291
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and
distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
et al. 2009; Kijas et al. 2012). Furthermore, sheep from
similar locations have been reported to have high levels of
admixture (Blackburn et al. 2011; Kijas et al. 2012).
The current genetic structure and level of admixture
among terminal sire breeds in the United States have not
been well characterized (Zhang et al. 2013). The objective of
this study was to examine population structure and
admixture in sheep from terminal breeds from US sheep
operations in collaboration with producers engaged with
NSIP. Understanding the genetic relationships between
terminal sire breeds in the United States will allow us to
better understand the genetic relatedness of these breeds of
sheep and assess the potential applicability of information
based on breed relatedness. Further, this study can help
elucidate how biological differences segregate in different
breeds, as well as between breeds of sheep.
Materials and methods
Sample collection and DNA isolation
A total of 248 sheep from terminal breeds of sheep including
Hampshire (n=45 from six flocks), Suffolk (n=68 from nine
flocks in the Midwest and n=37 from one flock, the
University of Idaho Suffolk flock, in the West), Oxford
(n=11 from two flocks) and Shropshire (n=44 from five
flocks), as well as wool/dual-purpose Rambouillet (n=43
from one flock), were genotyped for this study. Blood, semen
or tissue samples were collected by individual producers and
shipped to the University of Idaho and DNA was isolated using
the phenol chloroform method previously described (Sam-
brook et al. 1989).
Genotyping and quality control
Samples were genotyped using the Applied Biosystems
TM
Axiom
TM
Ovine Genotyping Array (50K) consisting of 51 572
SNPs (Thermo Fisher Scientific, catalog number 550898). A
subset of Shropshire samples (n=26) previously genotyped
on the Ovine Illumina SNP50 Bead Chip consisting of 54 241
SNPs (Illumina catalog number WG-420-1001) was also
included in this dataset. The genotypic data for these samples,
from each platform, were merged by SNP name and location
in PLINK version 1.90, with a total of 47 485 SNPs overlapping
between the two panels. Quality control of genotype data was
performed using PLINK version 1.90 specifically excluding
SNPs with a call rate of less than 0.90 and MAF less than
0.01, resulting in 45 864 SNPs remaining in the analyses
(Purcell et al. 2007; Chang et al. 2015).
Observed heterozygosity, inbreeding coefficients, and
F
ST
calculations
The observed heterozygosity was estimated for each animal
using PLINK version 1.90 and averaged by breed (Purcell et al.
2007; Chang et al. 2015). Inbreeding coefficients were
calculated for each animal based on the observed and expected
homozygosity in PLINK version 1.90, and the mean and 95%
confidence intervals were calculated with the Rpackage
‘rcompanion’ in Rversion 3.6.1. To remove redundancy and
provide a more accurate representation of variation, LD
pruning was performed using the --indep-pairwise function in
PLINK version 1.90 with an r
2
=0.5, a sliding window size of 50
SNPs and shifts of five SNPs (Visser et al. 2016; Gilbert et al.
2017). After LD pruning, 40 121 SNPs remained for further
analyses. Pairwise F
ST
was estimated in PLINK version 1.90
between breeds of sheep using the LD pruned dataset (Purcell
et al. 2007; Chang et al. 2015).
Eigenvalue analyses
Eigenvalues were calculated using the filtered SNP dataset
for terminal breeds only and then with Rambouillet in SNP
and Variation Suite version 8.7.2 (Golden Helix, Inc.,
www.goldenhelix.com). The top two eigenvalues were
plotted against each other in SNP and Variation Suite.
Hierarchical clustering
An identity by state matrix was calculatedfrom the LD pruned
dataset pairwise between all sheepusing the PLINK version 1.90
--distance flag (Purcell et al. 2007; Chang et al. 2015). The
matrix was read into Rversion 3.6.1 and hierarchical
clustering based on the identity by state matrix of Hamming
distances between each animal using the ‘hclust’ function.
The Bioconductor package ‘ctc’ was used in Rversion 3.6.1 to
write a Newick file to import into DENDROSCOPE 3 software
(Huson & Scornavacca 2012). A rectangular cladogram was
drawn from the Newick file in DENDROSCOPE version 3.5.9
(Huson & Scornavacca 2012). Individual branch labels were
colored according to producer-reported breed of sheep.
Admixture analysis
The program ADMIXTURE version 1.3.0 was implemented to
examine admixture between all samples using the LD pruned
genotypes in BED format (Alexander et al. 2009; Decker et al.
2014). The most probable number of Kgiven populations was
estimated using the lowest cross-validation error (Alexander
et al. 2009; Akanno et al. 2018). Euclidean distances were
calculated in Rversion 3.6.1 with the adegenet package and an
analysis of molecular variance (AMOVA) was performed with
the pegas package with 1000 permutations to statistically
examine differences between populations (McKay et al. 2008;
Paradis 2010; Jombart & Ahmed 2011).
International breed comparisons
Genotypes from 2819 sheep from 74 breeds across the
world were retrieved from the International Sheep Genome
©2020 The Authors. Animal Genetics published by
John Wiley & Sons Ltd on behalf of Stichting International Foundation for Animal Genetics, 51, 284–291
Genetic structure and admixture in sheep 285
Consortium Sheep HapMap Database and used in compar-
ison with US terminal breeds including the addition of n=5
Dorset and n=7 Southdown sheep from the United States.
The same set of 45 864 SNPs used with the US terminal
breeds was then merged with the same SNPs from the Sheep
HapMap dataset. Eigenvalues were calculated between US
terminal breeds and the same breeds from other locations in
the HapMap dataset, all US breeds in this study and the
same breeds present from other locations in the HapMap
dataset, and all US breeds in this study and the Sheep
HapMap dataset.
Results
Observed heterozygosity and inbreeding coefficient
To examine the relatedness of animals within each of the
breeds, observed heterozygosity and average inbreeding
coefficient were calculated. These statistics were calculated
based on observed and expected homozygosity, estimated
for each individual, and averaged for each breed (Table 1).
The Oxford animals exhibited the greatest (0.35) observed
heterozygosity and lowest inbreeding coefficients. Similar
observed heterozygosity was exhibited by Shropshire (0.34),
Western Suffolk (0.34), Suffolk (0.33) and Hampshire
(0.33). Shropshire had the lowest inbreeding coefficient
(0.09) in comparison with the Suffolk (0.13), Western
Suffolk (0.14) and Hampshire (0.14). The group with the
lowest observed heterozygosity (0.30) and highest inbreed-
ing coefficient (0.16) was Rambouillet.
Wright’s F
ST
Wright’s F
ST
was calculated pairwise between each group of
animals to examine differentiation between breeds (Table 2;
Wright 1965; Weir & Cockerham 1984; Lenstra et al.
2012). In general, values between 0 and 0.05 are catego-
rized as ‘little to no differentiation,’ values between 0.05
and 0.15 as ‘moderate differentiation’, values between 0.15
and 0.25 as ‘great differentiation’, and values above 0.25 as
‘very great differentiation’ between populations tested (Weir
& Cockerham 1984; Frankham et al. 2002). Rambouillet is
considered greatly differentiated from all terminal breeds.
Interestingly, Western Suffolk are considered moderately
differentiated from other terminal breeds. Little to no
difference was detected between Hampshire and Suffolk or
Hampshire and Shropshire. Furthermore, although Western
Suffolk and other Suffolk are not reported as different
breeds, they too exhibit moderate differentiation.
Eigenvalue analyses
To investigate how individuals from reported terminal
breeds the US group or cluster, eigenvalues were calculated
and plotted for all samples (Fig. 1). An eigenvalue plot for
only terminal breeds of sheep (Fig. 1a) as well as terminal
breeds and Rambouillet sheep (Fig. 1b) is displayed. In
Fig. 1a, the largest difference of eigenvalues is between
Western Suffolk and Shropshire and can be observed on the
x-axis of the plot shown. Further, the animals sampled for
the Shropshire breed exhibited the largest spread of eigen-
value points. Interestingly, all Suffolk did not group
together. Most of the Suffolk animals sampled cluster closely
with Hampshire animals; however, the Western Suffolk
flock clustered separately from Hampshire and other Suffolk
animals.
In Fig. 1b, Rambouillet animals cluster together, and the
entire breed clusters distinctly and away from the terminal
sheep breeds on the largest eigenvalue axis. Similar to
Fig. 1a, sheep cluster primarily by breed with the exception
of four Shropshire animals. The Suffolk samples do not all
group together, with Western Suffolk clustering separately
from other Suffolk animals. With these notable exceptions,
animals within a breed cluster together.
Hierarchical clustering based on identity by state
To examine how animals from breeds of sheep in the United
States are related to those from other breeds, hierarchical
Table 2 Pairwise F
ST1
between breeds of sheep.
Hampshire Suffolk
Western
Suffolk Oxford Shropshire
Hampshire 0
Suffolk 0.03 0
Western
Suffolk
0.09 0.07 0
Oxford 0.06 0.06 0.13 0
Shropshire 0.05 0.06 0.11 0.06 0
Rambouillet 0.17 0.17 0.23 0.18 0.16
1
Wright’s F
ST
values between 0 and 0.05 are categorized as no
differentiation, 0.06–0.15 as moderate differentiation, 0.16–0.25 as
great differentiation, and >0.26 as very great differentiation.
Table 1 The mean observed heterozygosity and average estimated
inbreeding coefficient including the 95% confidence interval for each
group.
Breed
Observed
heterozygosity
Inbreeding
coefficient
1
95% Confidence
interval for inbreeding
coefficient
Hampshire 0.33 0.14 0.12–0.15
Suffolk 0.33 0.13 0.12–0.15
Western
Suffolk
0.34 0.14 0.13–0.15
Oxford 0.35 0.05 0.01–0.09
Shropshire 0.34 0.09 0.04–0.11
Rambouillet 0.30 0.16 0.15–0.17
1
Inbreeding coefficients are reported as Fhat2 and calculated by:
(observed heterozygosity expected)/(total expected).
©2020 The Authors. Animal Genetics published by
John Wiley & Sons Ltd on behalf of Stichting International Foundation for Animal Genetics, 51, 284–291
Davenport et al.286
clustering was performed using an identity by state matrix.
A rectangular cladogram was constructed to visualize the
hierarchical clustering (Fig. 2). All Western Suffolk, Oxford
and Rambouillet animals clustered together by breed.
Rambouillet animals clustered in a distinct, separate branch
from all other breeds, which was consistent with the
eigenvalue plot. In general, most sheep were more identical
by state to other animals within the same breed with a few
notable exceptions.
Several reported Shropshire animals clustered with the
Hampshire branches; these were the same animals that
clustered with the Hampshire breed in the eigenvalue plots.
A branch of Shropshire animals also clustered closely with a
larger branch of Hampshire sheep. Additionally, Suffolk and
Hampshire animals overlapped and appeared to cluster
closely within the branches of the cladogram. Still, overall
most breeds clustered independently with the few excep-
tions mentioned before.
Admixture analysis
An admixture analysis was performed using the program
ADMIXTURE to investigate the extent of admixture between
different breeds of sheep in this study (Alexander et al. 2009;
Decker et al. 2014; Getachew et al. 2017). The analysis was
conducted using two to 10 given populations. The best fit of K
given populations was determined as K=6 based on the cross-
validation values calculated in ADMIXTURE (Fig. S1; Akanno et al.
2018). Further, the AMOVA analyses showed significant
(P<0.01) differences betweenthe K=6 assigned populations.
In the best fit K=6 plot, admixture was detected within
terminal breeds (Fig. 3). Admixture between terminal
breeds was observed in Hampshire, Oxford, Suffolk and
Shropshire, but the Western Suffolk population showed
little admixture with other terminal breeds except Suffolk.
Not surprisingly, the dual-purpose Rambouillet sheep were
different from the US terminal breeds examined.
Figure 1 Plot of calculated eigenvalues for
breeds of US sheep. (a) Eigenvalues plotted for
US terminal breeds of sheep. (b) Eigenvalues
plotted for US terminal breeds and Rambouil-
let sheep. Each point represents an individual
animal and points are colored by reported
breed.
©2020 The Authors. Animal Genetics published by
John Wiley & Sons Ltd on behalf of Stichting International Foundation for Animal Genetics, 51, 284–291
Genetic structure and admixture in sheep 287
Eigenvalue plots of US and international comparisons
To examine how US sheep compare with other sheep
across the world, genotyping data from this study were
merged with data from the Sheep HapMap (Kijas et al.
2012; Kijas 2013). Eigenvalues were calculated and
plotted with US terminal breeds including additional
Dorset and Southdown sheep from the United States, and
animals of the same breeds from the Sheep HapMap
dataset (Fig. 4a). Interestingly, the US terminal breeds
clustered closer to other breeds from the United States
than the same reported breed, including Suffolk and
Dorset, from other locations. When the genetic informa-
tion for wool breeds of sheep was included, they clustered
apart from the terminal breeds (Fig. 4b). Figure 4b also
shows the Irish Suffolk clustering closely with Suffolk from
the United States. Finally, when all samples were consid-
ered, the US terminal breeds clustered with similar breeds
from Australia and the UK (Fig. 4c). In summary, animals
clustered closest with those of similar geographic location
in the eigenvalue plots.
Discussion
The observed heterozygosity results from this study were
consistent with data reported in other breeds of sheep across
the world (Kijas et al. 2012; Ciani et al. 2014; Gaouar et al.
2017). More specifically, the observed heterozygosity in
most breeds was close to what was reported in Australian
sheep (Kijas et al. 2012; Al-Mamun et al. 2015). In addition,
the observed heterozygosity was consistent with other US
sheep including Suffolk, Rambouillet, Columbia, Polypay
and Targhee (Zhang et al. 2013). However, the breeds in
this study had lower observed heterozygosity when com-
pared with Boutsko, Karagouniko and Chios breeds from
Greece (Michailidou et al. 2018).
In our study, Oxford sheep exhibited the lowest average
inbreeding coefficient and highest observed heterozygosity,
similar to Finnsheep (Li et al. 2011). This is probably
because these sheep were selected based on pedigree
diversity from NSIP, whereas Western Suffolk had one of
the highest inbreeding coefficients and was only represented
by one flock. However, to our surprise, the inbreeding
coefficient for Western Suffolk was similar to that of Suffolk,
which included animals from 10 separate flocks. Perhaps
this is because these animals are the result of and
representative of the breeding strategies of purebred flocks.
Other work in 97 sheep breeds across the world includ-
ing Ethiopian sheep reported inbreeding coefficients
between 0.07 and 0.16 and observed heterozygosity
between 0.061 and 0.343, which are similar to our results
(Edea et al. 2017; Zhang et al. 2018).
Figure 2 Rectangular cladogram of individuals clustered based on identity by state and colored by reported breed.
Figure 3 ADMIXTURE model clustering output with K= 6 populations. Each bar represents an individual animal for each terminal breed and Rambouillet,
and the six colors represent each Kpopulation cluster.
©2020 The Authors. Animal Genetics published by
John Wiley & Sons Ltd on behalf of Stichting International Foundation for Animal Genetics, 51, 284–291
Davenport et al.288
Despite similarity in inbreeding coefficient and heterozy-
gosity estimates, Western Suffolk shows moderate differen-
tiation from Suffolk whereas Hampshire, Oxford, Shropshire
and Suffolk show little to moderate differentiation from each
other. The Western Suffolk consists of representatives from
a ‘closed flock’, which may explain the divergence from the
more broadly sampled Suffolk. The lack of differentiation
observed between the Suffolk, Hampshire and Shropshire is
not surprising considering the prevalence of crossbreeding
in many US terminal breed flocks. It is worth noting that the
Southdown is thought to be a common ancestor for
Hampshire, Shropshire and Oxford breeds (Ryder 1964).
These points are strongly supported by the results of the
ADMIXTURE analysis. Furthermore, these results concur with
previous research that reported a Wright’s F
ST
=0.1621
between Suffolk and Rambouillet; these breeds differ in
origin as the Rambouillet breed was derived from Merino
bloodlines (Dickinson & Lush 1933; Zhang et al. 2013).
Differences between breed groups can be visualized in the
eigenvalue plots, where sheep cluster primarily by reported
breed with the exception of a few animals. The separation of
Suffolk from Western Suffolk is apparent, which is consis-
tent with previous work that identified regional differences
in Suffolk from the United States (Kuehn et al. 2008). The
Figure 4 Eigenvalue plots of US sheep in this
study compared with other breeds across the
world as part of the Sheep HapMap study. (a)
Eigenvalue plot of US terminal breeds and
Dorset and Suffolk HapMap breeds. (b)
Eigenvalue plot of all US sheep in this study
compared with HapMap terminal and wool
sheep. (c) Eigenvalue plot of US sheep in this
study compared with all breeds present in the
Sheep HapMap study.
©2020 The Authors. Animal Genetics published by
John Wiley & Sons Ltd on behalf of Stichting International Foundation for Animal Genetics, 51, 284–291
Genetic structure and admixture in sheep 289
Shropshire breed has a large spread of eigenvalues and a
few animals cluster with Oxford and Hampshire, suggesting
the occurrence of crossbreeding. The distinct clustering of
the Rambouillet away from other breeds clearly displays the
genetic difference between terminal and wool/dual-purpose
breeds in the United States.
The K=6 plot, supported by the AMOVA analysis, shows
that sheep cluster primarily by breed with some level of
admixture between all terminal breeds, with the exception
of Western Suffolk, which exhibits little admixture except
with other Suffolk. The observed admixture within Hamp-
shire, Suffolk, Oxford and Shropshire is potentially due to
the use of sires with composite influence from other breeds
in US commercial operations (Ercanbrack & Knight ;
Norberg & Sørensen 2007). Rambouillet sheep showed
little to no admixture with the US terminal breeds examined
in this study.
When US sheep were compared with other populations
across the world, sheep primarily clustered closest to other
animals in similar geographic locations rather than to the
same reported breeds in other parts of the world (Kijas et al.
2012). More specifically, Suffolk and Dorset animals clus-
tered closer to other US groups than to Suffolk from
Australia and Ireland, or Dorset from Australia or the UK.
This observation may be partially attributed to the differ-
ences in selection and breeding strategies and in production
systems across the world (Andersson 2012;
Curkovi
cet al.
2016; Wang et al. 2015). In addition, the difference
between terminal breeds and wool breeds is clear, suggest-
ing that there are genetic differences between breeds that
have been selected for alternative production objectives and
purposes (Blackburn et al. 2011; Zhang et al. 2013; Fariello
et al. 2014).
In summary, we characterized relationships between
sheep from terminal sire breed populations in the United
States. Internationally, there has been an increased empha-
sis on genetic selection of sheep for a variety of traits and
purposes. Marker-assisted selection is growing in popularity
as new technology is being rapidly developed, along with an
increase in the use of quantitative genetic programs that
calculate estimated breeding values. By better understand-
ing the population structure and admixture between
terminal breeds in the United States compared with breeds
across the world, we can improve the effectiveness of this
developing technology. Our research provides insight into
the current relatedness of the popular terminal breeds in the
United States and the framework for future analyses on a
larger scale.
Acknowledgements
This project was supported by Agriculture and Food
Research Initiative Hatch grant no. IDA01566 from the
USDA National Institute of Food and Agriculture. We would
like to thank Thermo Fisher Scientific for genotyping, and
Curt Stanley, Shane Kirtchen, Virginia Tech/Scott Greiner,
Hayden’s Hamps, North Dakota State University, Drewry,
Reombke, Adams, University of Wisconsin, Kindred Cross-
ing, Bingen, Mapleton Mynd, Knepp Shropshires, All Forage
Farm, Dr. Fred Groverman, Formo, Bishop, Bar-Zel Suffolks,
JMG Suffolks, Virginia Tech, Culham & Stevens, Reau
Suffolks, Double L Livestock and Dry Sandy for sample
contributions.
Data availability
Data (50K SNP) have been deposited in Open Science
Framework (https://osf.io/d7s59/?view_only=9c85566d
0ac542d89a62150524eaad0e).
References
Akanno E.C., Chen L., Abo-Ismail M.K., Crowley J.J., Wang Z., Li C.,
Basarab J.A., MacNeil M.D., Plastow G.S. (2018) Genome-wide
association scan for heterotic quantitative trait loci in multi-breed
and crossbred beef cattle. Genetics, Selection, and Evolution: GSE
50, 48.
Alexander D.H., Novembre J., Lange K. (2009) Fast model-based
estimation of ancestry in unrelated individuals. Genome Research
19, 1655–64.
Al-Mamun H.A., Clark S.A., Kwan P., Gondro C. (2015) Genome-
wide linkage disequilibrium and genetic diversity in five popu-
lations of Australian domestic sheep. Genetics, Selection, Evolution:
GSE 47, 90.
American Sheep Industry (2011) National Animal Health Moni-
toring System (NAHMS) Sheep 2011 Study. https://sheepusa.
org/researcheducation-animalhealth-nahms.
Andersson L. (2012) How selective sweeps in domesticated animals
provide new insight into biological mechanisms. Journal of
Internal Medicine 271,1–14.
Blackburn H.D., Paiva S.R., Wildeus S. et al. (2011) Genetic
structure and diversity among sheep breeds in the United States:
Identification of the major gene pools. Journal of Animal Science
89, 2336–48.
Bovine HapMap Consortium, Gibbs R.A., Taylor J.F. et al. (2009)
Genome-wide survey of SNP variation uncovers the genetic
structure of cattle breeds. Science 324, 528–32.
Chang C.C., Chow C.C., Tellier L.C., Vattikuti S., Purcell S.M., Lee
J.J. (2015) Second-generation PLINK: rising to the challenge of
larger and richer datasets. Gigascience 4,7.
Ciani E., Crepaldi P., Nicoloso L. et al. (2014) Genome-wide
analysis of Italian sheep diversity reveals a strong geographic
pattern and cryptic relationship between breeds. Animal Genet-
ics 45, 256–66.
Curkovi
c M., Ramljak J., Ivankovi
c S., Mio
c B., Iankovi
c A., Pavi
c
V., Brka M., Veit-Kensch C., Medugorac I. (2016) The genetic
diversity and structure of 18 sheep breeds exposed to isolation
and selection. Journal of Animal Breeding and Genetics 133,71–80.
Decker J.E., McKay S.D., Rolf M.M. et al. (2014) Worldwide patterns
of ancestry, divergence, and admixture in domesticated cattle.
PLoS Genetics 10, e1004254.
Dickinson W.F., Lush J.L. (1933) Inbreeding and genetic diversity of
Rambouillet sheep in America. Journal of Heredity 24,19–33.
©2020 The Authors. Animal Genetics published by
John Wiley & Sons Ltd on behalf of Stichting International Foundation for Animal Genetics, 51, 284–291
Davenport et al.290
Edea Z., Dessie T., Dadi H., Do K.T., Kim K.S. (2017) Genetic
diversity and population structure of Ethiopian sheep populations
revealed by high-density SNP markers. Frontiers in Genetics 8,
218.
Ercanbrack S. K., Knight A. D. Effects of inbreeding on reproduction
and wool production of Rambouillet, Targhee, and Columbia
ewes1. Journal of Animal Science.69, 4734–4744.
Fariello M.I., Bertrand S., Tosser-Klopp G., Rupp R., Moreno C.,
International Sheep Genomics Consortium, San Cristobal M.,
Boitard S. (2014) Selection signatures in worldwide sheep
popualations. PLoS ONE 9, e103813.
Frankham R., Ballou J.D., Briscou D.A. (2002) Introduction to
Conservation Genetics. Cambridge University Press, Cambridge.
Gaouar S.B.S., Lafri M., Djaout A., El-Bouyahiaoui R., Bouri A.,
Bouchatal A., Maftah A., Ciani E., Da Silva A.B. (2017) Genome-
wide analysis highlights genetic dilution in Algerian sheep.
Heredity 118, 293–301.
Getachew T., Huson H.J., Wurzinger M. et al. (2017) Identifying
highly informative genetic markers for quantification of ancestry
proportions in crossbred sheep populations: implications for
choosing optimum levels of admixture. BMC Genetics 18, 80.
Gilbert E., Carmi S., Ennis S., Wilson J.F., Cavaller G.L. (2017)
Genomic insights into the population structure and history of the
Irish Travellers. Scientific Reports 7, 42187.
Huson D.H., Scornavacca C. (2012) Dendroscope 3: an interactive
tool for rooted phylogenetic trees and networks. Systematic
Biology 61, 1061–7.
Jombart T., Ahmed I. (2011) adegenet 1.2-1: new tools for the
analysis of genome-wide SNP data. Bioinformatics 27, 3070–1.
Kijas J. (2013) ISGC SNP50 HapMap and Sheep Breed Diversity
Genotypes v1. CSIRO, Canberra, Australia. Data collection.
https://doi.org/10.4225/08/51870B1E8EE56. CSIRO, Australia.
Kijas J.W., Lenstra J.A., Hayes B. et al. (2012) Genome-wide
analysis of the world’s sheep breeds reveals high levels of historic
mixture and strong recent selection. PLoS Biology 10, e1001258.
Kuehn L.A., Lewis R.M., Notter D.R. (2008) Connectedness in
Targhee and Suffolk flocks participating in the United States
National Sheep Improvement Program. Journal of Animal Science
87, 507–15.
Lenstra J.A., Groeneveld L.F., Eding H. et al. (2012) Molecular tools
and analytical approaches for the characterization of farm
animal genetic diversity. Animal Genetics 43, 483–502.
Li M.H., Strand
en I., Tiirikka T., Sev
on-Aimonen M.L., Kantanen J.
(2011) A comparison of approaches to estimate the inbreeding
coefficient and pairwise relatedness using genomic and pedigree
data in a sheep population. PLoS ONE 6, e26256.
Lupton C.J. (2008) Impacts of animal science research on United
States sheep production and predictions for the future. Journal of
Animal Science 86, 3252–74.
McKay S.D., Schnabel R.D., Murdoch B.M. et al. (2008) An
assessment of population structure in eight breeds of cattle using
a whole genome SNP panel. BMC Genetics 9, 37.
Michailidou S., Tsangaris G., Fthenakis G.C., Tzora A., Skoufos I.,
Karkabounas S.C., Banos G., Argiriou A., Arsenos G. (2018)
Genomic diversity and population structure of three
autochthonous Greek sheep breeds assessed with genome-wide
DNA arrays. Molecular Genetics Genomics 293, 753–68.
Norberg E., Sørensen A.C. (2007) Inbreeding trend and
inbreeding depression in the Danish populations of Texel,
Shropshire, and Oxford Down. Journal of Animal Science 85,
299–304.
Notter D.R. (1998) The U.S. National Sheep Improvement Program:
across-flock genetic evaluations and new trait development.
Journal of Animal Science 76, 2324–30.
Paradis E. (2010) pegas: an R package for population genetics with
an integrated-modular approach. Bioinformatics 26, 419–20.
Purcell S., Neale B., Todd-Brown K. et al. (2007) PLINK: a toolset
for whole-genome association and population-based linkage
analysis. American Journal of Human Genetics 81, 559–75.
Ryder M.L. (1964) The history of sheep breeds in Britain. The
Agricultural History Review 12,1–12.
Sambrook J., Fritsch E.F., Maniatis T. (1989) Molecular Cloning: A
Laboratory Manual (No. Ed. 2). Cold Spring Harbor Laboratory
Press, New York, NY.
United States Department of Agriculture Economic Research
Service (USDA ERS) (2019) Sector at a glance. https://www.e
rs.usda.gov/topics/animal-products/sheep-lamb-mutton/sector-
at-a-glance/.
Visser C., Lashmar S.F., Marle-Koster E.V., Poll M.A., Allain D.
(2016) Genetic diversity and population structure in South
African, French and Argentinian Angora goats from genome-
wide SNP data. PLoS ONE 11, e0154353.
Wang H., Zhang L., Cao J., Wu M., Ma Z., Liu Z., Liu R., Zhao F, Wei
C., Du L. (2015) Genome-wide specific selection in three domestic
sheep breeds. PLoS ONE 10, e0128688.
Weir B.S., Cockerham C.C. (1984) Estimating F-statistics for the
analysis of population structure. Evolution 38, 1358–70.
Wilson D.E., Morrical D.G. (1991) The national sheep improvement
program: a review. Journal of Animal Science 69, 3872–81.
Wright S. (1965) The interpretation of population structure by F-
statistics with special regard to systems of mating. Evolution 19,
395–420.
Zhang L., Mousel M.R., Wu X., Michal J.J., Zhou X., Ding B., Dodson
M.V., El-Halawany N.K., Lewis G.S., Jiang Z. (2013) Genome-
wide genetic diversity and differentially selected regions among
Suffolk, Rambouillet, Columbia, Polypay, and Targhee sheep.
PLoS ONE 8, e65942.
Zhang M., Peng W.F., Hu X.J., Zhao Y.X., Lv F.H., Yang J. (2018)
Global genomic diversity and conservation priorities for domestic
animals are associated with the economies of their regions of
origin. Scientific Reports 8, 11677.
Supporting information
Additional supporting information may be found online in
the Supporting Information section at the end of the article.
Figure S1 ADMIXTURE cross-validation output plotted across
Kpopulations.
©2020 The Authors. Animal Genetics published by
John Wiley & Sons Ltd on behalf of Stichting International Foundation for Animal Genetics, 51, 284–291
Genetic structure and admixture in sheep 291
Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.