e offprint.pdfthere were two primary populations (k = 2). surprisingly, the us breeds were near...
TRANSCRIPT
1 23
GeneticaAn International Journal of Genetics andEvolution ISSN 0016-6707 GeneticaDOI 10.1007/s10709-011-9619-4
Genetic diversity of Ovis aries populationsnear domestication centers and in the NewWorld
H. D. Blackburn, Y. Toishibekov,M. Toishibekov, C. S. Welsh,S. F. Spiller, M. Brown & S. R. Paiva
1 23
Your article is protected by copyright and
all rights are held exclusively by Springer
Science+Business Media B.V. (outside the
USA) . This e-offprint is for personal use only
and shall not be self-archived in electronic
repositories. If you wish to self-archive your
work, please use the accepted author’s
version for posting to your own website or
your institution’s repository. You may further
deposit the accepted author’s version on a
funder’s repository at a funder’s request,
provided it is not made publicly available until
12 months after publication.
Genetic diversity of Ovis aries populations near domesticationcenters and in the New World
H. D. Blackburn • Y. Toishibekov • M. Toishibekov •
C. S. Welsh • S. F. Spiller • M. Brown •
S. R. Paiva
Received: 21 March 2011 / Accepted: 12 November 2011
� Springer Science+Business Media B.V. (outside the USA) 2011
Abstract Domestic sheep in Kazakhstan may provide an
interesting source of genetic variability due to their prox-
imity to the center of domestication and the Silk Route.
Additionally, those breeds have never been compared to
New World sheep populations. This report compares
genetic diversity among five Kazakhstan (KZ) and 13
United States (US) sheep breeds (N = 442) using 25
microsatellite markers from the FAO panel. The KZ breeds
had observed and expected measures of heterozygosity
greater than 0.60 and an average number of alleles per
locus of 7.8. In contrast, US sheep breeds had observed
heterozygosity ranged from 0.37 to 0.62 and had an aver-
age number of alleles of 5.7. A Bayesian analysis indicated
there were two primary populations (K = 2). Surprisingly,
the US breeds were near evenly split between the two
clusters, while all of the KZ breeds were placed in one of
the two clusters. Pooling breeds within country of sample
origin showed KZ and US populations to have similar
levels of expected heterozygosity and the average number
of alleles per locus. The results of breeds pooled within
country suggest that there was no difference between
countries for these diversity measures using this set of
neutral markers. This finding suggests that populations’
geographically isolated from centers of domestication can
be more diverse than previously thought, and as a result,
conservation strategies can be adjusted accordingly. Fur-
thermore, these results suggest there may be limited need
for countries to alter the protocols for trade and exchange
of animal genetic resources that are in place today, since no
one population has a unique set of private alleles.
Keywords Genetic diversity � Sheep � Conservation of
genetic resources � Kazakhstan
Introduction
A premise of livestock conservation genetics is that breeds
at or close to the center of domestication are more
genetically diverse than geographically distant breeds
(Tapio et al. 2010; Peter et al. 2007; Loftus et al. 1999).
For sheep, it is generally considered that northern Zagros
to southeastern Anatolia was the center of domestication
(Zeder 2008; Rezaei et al. 2010). However, there is some
evidence that China was the origin of a third maternal
lineage (Guo et al. 2005; Sulaiman et al. 2011; Wang et al.
2007). Kazakhstan’s proximity to centers of domestication
and gene flow between those centers, via the Silk Route,
could suggest that sheep from this country could have
Mention of a trade name or proprietary product does not constitute a
guaranty or warranty by the USDA and does not imply approval to the
exclusion of other products that may be suitable. USDA, Agricultural
Research Service, Northern Plains Area, is an equal opportunity/
affirmative action employer. All agency services are available without
discrimination.
H. D. Blackburn (&) � C. S. Welsh � S. F. Spiller
National Center for Genetic Resources Preservation, National
Animal Germplasm Program, USDA-ARS, Ft. Collins, CO, USA
e-mail: [email protected]
Y. Toishibekov � M. Toishibekov
Institute of Experimental Biology, Almaty, KZ, USA
M. Brown
Grazing Lands Research Laboratory, USDA-ARS, El Reno, OK,
USA
S. R. Paiva
EMBRAPA Recursos Geneticos e Biotecnologia (EMBRAPA
Genetic Resources and Biotechnology - CENARGEN, Brasillia,
DF, USA
123
Genetica
DOI 10.1007/s10709-011-9619-4
Author's personal copy
substantial genetic variability. But this hypothesis has not
been previously addressed, and from 1992 to 2009, the
number of sheep in Kazakhstan decreased by 19.7 million
(-58%) due to the dissolution of the former Soviet Union
(FAO 2011), which underscores the need to better quantify
these populations. As a result of this perceived importance
of breeds/populations close to centers of domestication,
there have been suggestions that such populations should
be high on conservation priority lists (Tapio et al. 2010).
In contrast to Kazakhstan, the United States had no
indigenous domestic sheep and a relatively short history of
sheep production. However, throughout the country’s his-
tory, a wide range of sheep breeds that originated from
Europe, Africa, and Western Asia have been brought to the
United States. A recent study confined to US sheep breeds
determined that many of the breeds were genetically distinct,
and there was a range of genetic variability (Blackburn
et al. 2011). By comparing genetic diversity between sheep
breeds from these two countries should provide further
insights into the genetic distinctness of sheep populations
within and between countries and have ramifications upon
the formation of national livestock conservation policy.
The results can also provide a basis for the growing dis-
cussions on access and benefit sharing of farm animal
genetic resources in various forums (e.g., Convention of
Biological Diversity).
Materials and methods
A total of eighteen sheep breeds were evaluated in this
study (Table 1); five breeds (117 head) from Kazakhstan
(Fig. 1) and 13 from the United States with a total of 442
animals. A brief description of Kazakh breeds follows
based upon FAO (1989) and Medeubekov et al. (2008).
The Arkhara-Merino is a composite breed developed by
harvesting semen from slaughtered Arkhar rams (Ovis
ammon, 2n = 56) and inseminating Novocaucasian Merino
ewes (FAO 1989). The F1 rams were mated to Precoce or
Rambouillet ewes and then backcrossed again resulting in
rams that were 12.5% Arkhara-Merino and 87.5% Merino/
Rambouillet type. These rams were then crossed to ewes
from the preceding generation (0.25:0.75) to form the
Arkhara-Merino breed, which would have an expected
composition of 18.75% Arkhar (Ovis ammon) and 81.25%
Merino/Rambouillet. The Degresskaya breed was formed
by crossbreeding Shropshire rams to fat-rumped ewes
(circa 1931–1936, FAO 1989), and it is suggested during
the 1940s that some Precoce (Merino type) was introduced.
The Sary-Arka breed has been considered one of the best
fat-rumped breeds that originated in southeastern Turk-
menistan. Purportedly, since the 1950s, breeders routinely
crossbred Sary-Arka with Dedresskaya. Edil’baevskaya
sheep were developed in the nineteenth century by mating
Table 1 Breeds and sample sizes (N) for US and Kazakhstan sheep analyzed with 25 microsatellites
Population Abbrev. Country Country of origin N Na Nar (20)** PAR (20)** Fis He Ho HWE*
Arkhara-Merino AM KAZ Kazakhstan 25 7.88 6.20 0.17 0.040 0.75 0.72 1
Blackbelly Barbados BB USA Caribbean 18 4.80 4.34 0.10 0.180* 0.65 0.53 0
Chyisskaya CHY KAZ Kazakhstan 17 7.68 6.45 0.20 0.053 0.74 0.70 0
Cotswold COT USA UK 09 3.92 – – 0.061 0.59 0.56 0
Degeresskaya DEG KAZ Kazakhstan 26 7.88 6.00 0.18 0.003 0.71 0.72 0
Dorper DORP USA South Africa 44 7.60 5.30 0.08 0.096* 0.69 0.62 2
Edil’baevskaya EDI KAZ Kazakhstan 27 9.12 6.67 0.26 0.051 0.76 0.72 1
Hampshire HAMP USA UK 29 8.08 5.46 0.13 0.055 0.68 0.64 2
Hog Island HOG USA UK 24 4.04 3.16 0.13 0.070 0.42 0.39 3
Karakul KAR USA Central Asia 19 5.72 4.77 0.14 0.232* 0.63 0.49 2
Leicester Longwool LELW USA UK 29 5.24 4.02 0.11 0.109* 0.55 0.49 6
Lincoln LINC USA UK 22 5.16 4.36 0.04 0.099 0.60 0.55 2
Rambouillet RAMB USA France 47 8.24 5.60 0.09 0.108* 0.71 0.64 4
Romanov ROMA USA Russia 24 5.08 – – 0.057 0.62 0.58 2
Sary-arkinsskaya SA KAZ Kazakhstan 22 7.52 6.05 0.09 0.037 0.74 0.71 0
St. Croix STC USA Caribbean 26 6.08 4.93 0.06 0.097* 0.67 0.60 1
Texel TEX USA the Netherlands 20 6.28 5.18 0.23 0.049 0.66 0.63 0
Tunis TUN USA Tunisia 14 4.96 4.58 0.25 0.140* 0.64 0.55 1
Mean number of alleles per locus (Na); allelic richness obtained with rarefaction method (Nar); privative allelic richness (PAR); inbreeding
coefficient (Fis); expected heterozygosity (He); observed heterozygosity (Ho); and number of loci significantly deviating from Hardy–Weinberg
equilibrium (HWE)
* P \ 0.05 after Bonferroni correction; ** Number of observations in each breed
Genetica
123
Author's personal copy
Kazakh fat-rumped ewes to coarse-wooled rams from
Astrakhan, Russia. Chuisk sheep were developed from
indigenous fat-rumped, meat, and coarse-wooled breeds in
addition to Stavropol and Precoce admixed at various
points in time.
In Kazakhstan, the animals were randomly selected in
various parts of the country and transported to an institute
farm near Almaty, KZ. At the farm, semen and embryos were
collected and cryopreserved as a conservation measure. For
this study, blood samples were collected on Whatman FTA
paper and transported to the United States for analysis.
The US animals were part of a larger project (Blackburn
et al. 2011), where samples were collected from 222 pro-
ducers in 38 states. Criteria for within-flock animal sam-
pling included the acquisition of tissue samples from both
sexes and no known genetic relationship. In addition,
efforts were made to sample from flocks that had been
relatively independent in their breeding programs. Blood
samples were collected by the animal’s owner or a col-
laborator and shipped to this laboratory for processing.
Upon receipt, blood samples were cryopreserved until
ready for DNA extraction and analysis. The semen utilized
in the study was acquired while developing breed collec-
tions for the gene bank. The blood and semen cryopreser-
vation protocols used are found at the website: http://www.
ars.usda.gov/Main/docs.htm?docid=16979 (Accessed March
2011). DNA was isolated from cryopreserved blood and
semen samples using the BloodPrepTM
chemistry protocol
and the NucPrepTM
chemistry protocols (Applied Biosytems
2004), respectively, in conjunction with the ABI PrismTM
6100 Nucleic Acid PrepStation.
Analysis of microsatellite data
For this analysis, the Food and Agriculture Organization of
the United Nations/International Society for Animal
Genetics (FAO/ISAG) panel of 31 microsatellite markers
were used (FAO 2004) to maintain the option for combining
this data set with other studies using the same complement of
markers. The consensus panel markers located across 22
chromosomes were unlinked (FAO 2004). A commercial
company (GeneSeek) constructed the multiplex system,
amplified the DNA samples by PCR, and made the allele
calls. ‘‘Appendix 1’’ lists the markers (chromosome num-
ber), number of alleles, and, respectively, the percent of
missing data from each microsatellite. Markers OarFCB20
and BM1329 did not amplify well and were deleted from the
analysis. In addition, another four markers (BM1824, ILS-
TS5, OarFCB304, and SRCRSP5) were not used as a result
of more than 50% of the populations evaluated not being
within Hardy–Weinberg Equilibrium.
The GENALEX 6 program (Peakall and Smouse 2006)
was used to compute the average and effective number of
alleles, allele frequency per locus, observed and expected
heterozygousity (Ho and He, respectively), a breed’s private
alleles (Slatkin 1985), and principal coordinate analysis. The
analysis of molecular variance (AMOVA) was performed by
the Arlequin software (Excoffier et al. 2005) using the
codominant allelic distance matrix with 1000 permutations.
Due to unequal sample sizes for breeds and countries, rar-
efaction methods were used (Szpiech et al. 2008) to adjust
allelic richness and private alleles per population. During
these computations, it was apparent that the Cotswold (likely
due to sample size) and Romanov (due to missing data) were
deleted from this portion of the analysis. Inbreeding (Fis)
was calculated using FSTAT (Goudet 2002). The Dtl genetic
distance was used because it simultaneously accounts for the
impact of mutation and genetic drift on gene frequencies
(Tomiuk and Loeschcke 1995). In addition, Dtl has been
shown to be more robust to deviations in the assumptions
concerning mutation and genetic drift (e.g., Tapio et al.
2003). This distance was estimated with the software Molkin
Fig. 1 Geographic locations for
the five Kazakhstan sheep
breeds: 1 Arkhara-Merino, 2Sary-Arkinsskaya, 3Edil’baevskaya, 4 Chyisskaya, 5Degeresskaya
Genetica
123
Author's personal copy
(Gutierrez et al. 2005), and a phylogenetic tree was com-
puted with the Neighbor-Net method implemented in
SplitsTree4 package (Huson and Bryant 2006).
STRUCTURE (Pritchard et al. 2000) was run using an
admixture model with correlations between loci and a
burn-in of 100,000 iterations followed by an additional
500,000 iterations. Within each specified cluster (K) rang-
ing from 1 to 18, three replicates were run and the averaged
likelihood at each K was used to calculate DK (Evanno
et al. 2005; McKay et al. 2008), which was used as an ad
hoc indicator of cluster number. The program DISTRUCT
(Rosenberg 2004) was used to plot STRUCTURE results.
Upon completion of the previously mentioned analysis,
breeds were pooled into one population per country and
selected analyses were performed again comparing genetic
differences between the two countries.
Results
Analysis of within-breed genetic diversity showed that all
Kazakh breeds had a relatively large number of alleles per
breed (7.52–9.12) compared to US breeds (3.92–8.08)
(Table 1), but the ranges for both countries are similar to
previously reported values (Peter et al. 2007). Within the
US breeds, Barbados Blackbelly, Cotswold, and Hog Island
(all rare breeds) had fewer alleles per loci present for the
microsatellites evaluated. Using the rarefaction method
(and dropping the Cotswold and Romanov breeds), an
across-breed reduction in allelic richness (Hedrick 2011)
occurred for breeds from both countries. Most notable were
the decreases in allelic richness of the Dorper, Hampshire,
and Rambouillet. The rarefaction approach yielded private
allele estimates that were less than one per breed (Table 1).
Average observed heterozygosity levels for most breeds
tended to range above 0.50, but the Hog Island, Karakul,
and Leciester Longwool were below this level (0.39, 0.49,
and 0.49, respectively). In addition, the Cotswold and Hog
Island had loci that were monomorphic. Most loci within
breed tended to be within HWE (Table 1). Inbreeding (Fis)
was computed, and it was found that among US breeds,
Karakul and Barbados Blackbelly had the highest
inbreeding levels that were of a magnitude that maybe
biologically important for production characteristics.
Among the Kazakhstan breeds, Fis was slight and ranged
from 0.003 to 0.053 (Table 1).
Breeds were pooled within country and then analyzed
across countries for various genetic diversity measures
(Table 2). The countries were similar for expected heter-
ozygosity, average number of alleles per locus, average
number of effective alleles per locus, allelic richness using
rarefaction, and private alleles; Ho was notably different
between the two countries.
Variation among the 18 breeds was measured at 12.8% via
AMOVA procedure (P \ 0.000001). But when breeds were
pooled within country, the AMOVA between the two
countries decreased to 0.7% (P \ 0.01662). Further evalu-
ation was performed by placing the Arkhara-Merino with US
breeds, but this did not change the partitioning of the vari-
ance. Genetic distances (Dtl) among all breeds (Appendix 2)
were calculated using the Tomiuk and Loeschcke (1995)
approach. Among the Chuisk, Edil’baevskaya, Degresskaya,
and Sary-Arkinsskaya breeds genetic distances were less
than 0.12. Among the Kazakh breeds, the Arkhara-Merino
tended to be the most distant population, and it had the
shortest genetic distance to the Rambouillet (0.057). This
result was similar to genetic distance values between the
Arkhara-Merino and other fine-wool breeds found in the
former Soviet Union (Ozerov et al. 2008). The remaining
four Kazakh breeds were not particularly distant from the US
breeds. Rather they tended to be similar in magnitude to the
distances for breeds like the Rambouillet versus Lincoln. The
most extreme genetic distances were associated with the US
minor breeds. These large differences may be a result of the
relatively small number of alleles the minor breeds exhibited
for this neutral set of markers.
Neighbor-Net tree, using Dtl (Fig. 2), placed the four fat-
rumped Kazakh breeds in close proximity to one another, and
it would appear the four breeds (Chuisk—Edil’baevskaya
and Degresskaya—Sary—Arkinsskaya) could be further
partitioned into two groups, confirming the FAO (1989)
report. Intermediate to these four breeds was the US Karakul.
As anticipated, the Arkhara-Merino was placed close to the
Rambouillet. Interestingly, the two US hair breeds where
also placed in the Kazakh half of the net. At the opposite end
of the net were the English longwools, while Texel, Tunis,
Table 2 Genetic diversity measures for pooled Kazakhstan and United States sheep breeds (standard errors)
Country He Ho Na Nae Nar* (85) PAR* (85)
Kazakhstan 0.76 (0.02) 0.72 (0.01) 12 (3.63) 4.71 (0.34) 11.69 (0.71) 2.03 (0.23)
United States 0.75 (0.02) 0.58 (0.01) 12.8 (4.31) 4.83 (0.49) 11.01 (0.75) 1.48 (0.25)
* Number of observations in each breed
Expected heterozygosity (He), observed heterozygosity (Ho), average number of alleles per locus (Na), average number of effective alleles per
locus (Nae), allelic richness obtained with rarefaction method (Nar) and privative alleles (PAR)
Genetica
123
Author's personal copy
Hampshire, and Hog Island were intermediate and matches
their known history (Porter and Mason 2002).
Three principal coordinate analyses (Fig. 3) explained
over 60% of the genetic variation and generally supported
the Neighbor-Net results. As Fig. 3 demonstrates, there are
groupings that contain multiple breeds; for example, the
English longwools, breeds from Central Asia and Russia,
and the hair breeds. To further evaluate these breed
groupings, an analysis of private allelic richness was per-
formed (Table 3). Across the seven groups, differences in
allelic richness were uniformly small and less than one
allele. This result suggests the differentiation noted in
Fig. 3, which is due to differences in allele frequencies and
that no one group or combination of groups possesses a
high degree of genetic uniqueness.
Using the Bayesian approach (Pritchard et al. 2000) and
the method of Evanno et al. (2005) indicated that the number
of clusters for this set of breeds was two (Appendix 3). In this
analysis, for a breed to be an exclusive member of a cluster, it
needed to have greater than 50% of its composition in one
cluster. The cluster size of two and the combination of breeds
within each cluster were unexpected. In one cluster, all Ka-
zakh and approximately one-half the US breeds, including
the hair breeds (Fig. 4), were found. The second cluster was
dominated by breeds with British origins or foundation. The
low proportion of admixture for the Degresskaya with the
second cluster suggests that the Shropshire (a down breed
similar to the Hampshire), which was used in the Degress-
kaya’s formation, has been reduced in a manner similar to the
situation in the Arkhara-Merino and the Arkhara. A minor
peak was observed when K = 7, which reflected mainly a
multi-breed series of clusters formed with four Kazakh
breeds, fine-wooled, long-wooled, and hair breeds as
observed in Figs. 2 and 3.
Discussion
Kazakhstan holds a unique geographic position in terms of
the flow and use of sheep genetic resources via pastoral
nomadism since approximately 2500 B. C. (Wagner et al.
2011) and proximity to the center of sheep domestication.
Conversely, the United States has a relatively recent col-
lection of genetic resources from a wide range of geographic
regions, and these have been assessed on a within-country
basis and in comparison to Brazilian breeds (Blackburn et al.
2011; Paiva et al. 2010, 2011). But, comparisons between US
breeds and sheep populations close to the center of domes-
tication have not been made until this study.
Among the US breeds, there was a wide range of genetic
diversity measures (heterozygosity, number of alleles per
locus, genetic distance, etc.) as a result of diverse countries of
origin, selection pressure, genetic drift, and population size.
For the rare breeds, the results suggest a general lack of
genetic diversity, while breeds like Rambouillet, Hampshire,
and Dorper do not appear hampered by genetic diversity
issues. The STRUCTURE analysis indicated that all US
breeds shared a proportion of their genotypes with the Ka-
zakh breeds and vice versa. As expected, the Karakul and
Romanov had higher proportions of their cluster assignments
in common with the Kazakh cluster, and they were placed in
close proximity to the coarser-wooled Kazakh breeds in the
principal coordinate analysis (Fig. 3). These results confirm
those of Tapio et al. (2010), who also showed the Romanov
to be grouped with the fat-tailed Kazakh breeds.
Fig. 2 Neighbor-Net tree of United States and Kazakhstan sheep
breeds using Dtl genetic distance (abbreviations as is Table 1)
-0.10
-0.05
0.00
0.05
0.10
0.15
0.20
-0.20 -0.15 -0.10 -0.05 0.00 0.05 0.-0.10-0.050.000.050.100.150.20
PC
A3
PCA1
PC
A2
COT
LINC
LELW TEX
TUNHOG
HAMPDORP
RAMB
AM
BB
STC
ROMAEDI
CHY DEG
SA KAR
Fig. 3 Breed relationships among United States and Kazakhstan
breeds based upon three principal coordinates that explained:
PCA1 = 28.8%, PCA2 = 21.6%, and PCA3 = 15.5% of the varia-
tion, abbreviations as in Table 1
Genetica
123
Author's personal copy
Among the Kazakh breeds, it was anticipated that a
portion of the Arkhara-Merino would be placed in a unique
cluster (Fig. 4) due to the use of the Arkhara during its
formation (Rezaei et al. 2010). Given the description of
how the breed was formed, the theoretical expectation
would have been approximately 18% Arkhara (FAO 1989).
However, there was no such assignment. To explore this
aspect, additional STRUCTURE analysis was performed
using only the five Kazakh breeds plus the Rambouillet,
Romanov, and Karakul. In this analysis, no unique cluster
within the Arkhara-Merino emerged from the analysis.
Two potential explanations for the lack of an Arkhara
cluster include: during the inter-se mating of the final
composite population, animals with the expected compo-
sition of breed types were culled; or there may have been a
substantial contribution of the Arkhara to the other Kazakh
breeds (Sulaiman et al. 2011). Evaluating sheep breeds in
northwestern China, Sulaiman et al. (2011) used Ovis
ammon sheep in their analysis, which was distinct when
compared to domesticated Ovis aries breeds using a
mitochondrial DNA control region. However, the mito-
chondrial DNA of the Arkhara-Merino was derived from
Merino-type ewes used to construct the F1 population.
Further analysis using markers specific to the Y chromo-
some would add clarity to this issue.
The fat-rumped Kazakh breeds tended to have moderate
genetic distances from US breeds (Appendix 2). This result
was also reflected in the principal coordinate analysis
(Fig. 3), where the five Kazakh breeds were placed among
the various US breeds. At the extremes of Fig. 3 are the
hair, English long-wooled, and Hog Island breeds similar to
Blackburn et al. (2011) suggesting that these rare breeds
are isolated populations that through genetic drift have
become distinct in their microsatellite profile. Peter et al.
(2007) used the same FAO panel on 57 sheep breeds and
had similar values of observed and expected heterozygosity
with the present study, but they observed a lower genetic
distance from European and Middle-Eastern breeds. Gen-
erally, other studies with similar markers and Asian
breed(s) have similar within-breed genetic diversity values
(e.g., Tapio et al. 2010; Lawson-Handley et al. 2007).
Previous analyses have shown how breeds from close to
centers of domestication tend to be more variable (Bruford
et al. 2003; Sulaiman et al. 2011; Tapio et al. 2010). This
study found similar results at the breed level. However, by
pooling breeds within countries and comparing the resulting
diversity measures, this study showed each country had
similar genetic diversity measurements (Table 3). That said,
the observed level of heterozygosity was lower for the US
population. The lower observed heterozygosity was
Table 3 Pairwise values for private alleles (upper diagonal) and standard error (lower diagonal) for major breed groups (excluding Romanov) at
a standardized sample size of 34 and using rarefaction method
KAZ ? KAR Hair Long wool Dorper Hampshire Hog Island Merino type
Kazakh ? Karakul 0.1744 0.1422 0.1656 0.1680 0.0573 0.2710
Hair 0.0499 0.0713 0.0762 0.0291 0.0150 0.0424
Long wool 0.0246 0.0211 0.0833 0.0699 0.0274 0.0983
Dorper 0.0448 0.0233 0.0390 0.0285 0.0082 0.0675
Hampshire 0.0566 0.0092 0.0268 0.0109 0.0398 0.1539
Hog Island 0.0386 0.0128 0.0108 0.0036 0.0178 0.0064
Merino type 0.0741 0.0132 0.0238 0.0204 0.0467 0.0036
Fig. 4 Breed assignment when
cluster size (K) equaled 2 based
on a Bayesian cluster
(STRUCTURE) analysis. Plots
were constructed using the
program DISTRUCT, and the
width of each segment is
relative to breed sample size.
Labels at the top of the figure
are country where animals
within breed were sampled
followed by the country of
origin (US breeds only),
abbreviations as in Table 1
Genetica
123
Author's personal copy
potentially due to genetic drift and/or bottlenecks established
during breed formation. While we realize the limitation of
neutral markers, these results suggest large portions of genetic
variability found in sampling Kazakh breeds had flowed to the
United States. While this result was unexpected, there are
similar reports where considerable genetic variation in non-
Central Asian sheep breeds has been found (Pereira et al.
2006). In general, these data illustrate the well-known concept
that while genetic variation can be reduced in subpopulations,
genetic variability of the entire population does not change
(Falconer and Mackay 1996). While theory predicts such an
occurrence, we found the results interesting, given the lack of
direct and indirect gene flow between Central Asia and the
United States in a more recent time frame. Furthermore, they
support Groeneveld et al. (2010) assertion that sheep tend to
have little phylogeographic structure. In addition, the results
suggest that breeder development and use of genetic resources
tend to have transient and non-permanent effects. In other
words, breeds/subpopulations are developed, recombined,
and discarded routinely as conditions dictate (Wood and Orel
2001) across time.
The lack of large differences in genetic diversity mea-
sures between the pooled US and Kazakhstan populations
implies that countries may wish to explore utilization of
indigenous and locally adapted genetic resources before
importing new and untested genetic resources from other
countries. Both Young (1992) and Blackburn and Gollin
(2009) suggest this to be a more effective strategy for
developed countries, while FAO (2007) promotes such a
concept on an international basis.
The results derived from this study have implications for
national and international policies. Our findings suggest
limited existence of private alleles when breeds were pooled
at the country level. Furthermore, the STRUCTURE analysis
did not show any clear demarcation between sheep breeds
based upon country of origin (Kazakhstan vs. United States).
These findings suggest there may be limited need for coun-
tries to be concerned with the current trade and exchange
policies for animal genetic resources since no one population
has a unique set of private alleles. Therefore, any new
mechanism to monitor germplasm exchange via the Nagoya
Protocol (Convention on Biological Diversity 2011) should
be minimal at best and the opportunities for altering benefit
sharing strategies appear limited. The reported results also
have potential implications for the utilization of genetic
diversity and its conservation. They suggest that strong
national conservation programs, focused upon the conserva-
tion of their indigenous genetic resources, would be an
effective approach to insure availability of sufficient genetic
variation for future utilization. The suggestion that breeds
from particular geographic regions (e.g., centers of domesti-
cation) have global priority over other breeds (Tapio et al.
2010) in different regions or countries appear unjustified,
given this set of neutral markers. The findings presented do
not nullify the need for further and more detailed exploration
of specific genes or gene complexes that may confer adapt-
ability to environmental vagaries, the likes of which are found
in Kazakhstan and various regions of the United States.
Appendix 1
See Table 4.
Table 4 Genotyped loci, their chromosome, number of alleles per locus and percent missing data per locus
Locus (chromosome #) # alleles % missing data Locus (chromosome #) # alleles % missing data
BM1824 (1) 10 4.8 OarAE129 (5) 10 2.8
BM8125 (17) 9 3.5 OarCP34 (3) 9 18.7
DYMS1 (20) 20 3.7 OarCP38 (10) 10 13.9
HUJ616 (13) 24 5.0 OarFCB128 (2) 14 2.7
ILSTS11 (9) 9 15.5 OarFCB193 (11) 21 2.6
ILSTS28 (3) 14 9.2 OarFCB304 (19) 22 6.4
ILSTS5 (7) 13 15.0 OarFCB226 (2) 15 2.4
INRA063 (14) 19 13.9 OarHH47 (18) 16 4.8
MAF209 (17) 14 3.5 OarJMP29 (24) 18 4.8
MAF214 (16) 11 6.6 OarJMP58 (26) 18 1.7
MAF33 (9) 13 2.3 OarVH72 (25) 9 10.6
MAF65 (15) 11 3.1 SRCRSP1 (CHI13) 10 1.6
MAF70 (4) 20 3.7 SRCRSP5 (18) 7 3.7
MCM527 (5) 13 4.3 SRCRSP9 (12) 13 4.0
MCM140 (6) 15 24.1 – –
Markers not used OarFCB20 (2), BM1329 (6)
Genetica
123
Author's personal copy
Ap
pen
dix
2
See
Tab
le5
.
Ta
ble
5G
enet
icd
ista
nce
sfo
r1
8sh
eep
bre
eds
(ab
bre
via
tio
ns
asin
Tab
le1
)w
ith
dis
tan
ces\
0.1
0in
ital
ican
d[
0.3
5in
bo
ldu
sin
gD
tl(T
om
iuk
and
Lo
esch
cke
19
95
)
AM
BB
CH
YC
OT
DE
GD
OR
PE
DI
HA
MP
HO
GK
AR
LE
LW
LIN
CR
AM
BR
OM
AS
AS
TC
TE
XT
UN
AM
0
BB
0.2
20
CH
Y0
.11
90
.25
0
CO
T0
.21
0.4
24
0.2
68
0
DE
G0
.10
60
.22
70
.10
10
.27
40
DO
RP
0.1
02
0.1
90
.11
70
.20
80
.11
60
ED
I0
.09
50
.18
60
.07
10
.26
30
.07
70
.10
HA
MP
0.0
85
0.2
12
0.1
15
0.2
03
0.1
29
0.0
71
0.1
03
0
HO
G0
.34
0.4
18
0.2
78
0.3
95
0.3
34
0.2
42
0.3
04
0.1
91
0
KA
R0
.17
90
.29
10
.16
60
.35
40
.17
20
.15
90
.13
80
.18
60
.38
10
LE
LW
0.1
71
0.3
02
0.2
30
.15
40
.21
10
.21
10
.20
50
.14
0.3
29
0.3
04
0
LIN
C0
.16
50
.30
80
.21
60
.11
70
.23
40
.16
0.2
09
0.1
66
0.3
87
0.2
54
0.0
99
0
RA
MB
0.0
57
0.2
09
0.1
18
0.2
04
0.0
87
0.0
79
0.1
09
0.0
63
0.3
20
.18
10
.15
90
.19
90
RO
MA
0.2
08
0.2
96
0.2
08
0.3
55
0.2
12
0.1
98
0.2
14
0.1
69
0.3
39
0.2
81
0.3
04
0.3
44
0.1
80
SA
0.1
04
0.2
36
0.1
04
0.2
49
0.0
71
0.1
25
0.0
72
0.1
24
0.3
28
0.1
63
0.2
24
0.2
21
0.1
06
0.1
91
0
ST
C0
.16
40
.14
60
.17
10
.32
40
.16
80
.11
70
.13
0.1
32
0.3
20
.24
60
.27
0.2
83
0.1
33
0.2
89
0.1
59
0
TE
X0
.14
20
.27
90
.19
80
.17
60
.20
80
.11
0.1
61
0.1
15
0.2
57
0.2
64
0.1
51
0.1
85
0.1
29
0.2
28
0.1
99
0.2
01
0
TU
N0
.21
80
.31
20
.26
60
.31
90
.22
90
.24
80
.28
10
.18
0.4
57
0.3
0.2
16
0.2
34
0.1
88
0.3
45
0.2
61
0.3
15
0.2
69
0
Genetica
123
Author's personal copy
Appendix 3
See Fig. 5.
References
Blackburn HD, Gollin D (2009) Animal genetic resource trade flows:
the utilization of newly imported breeds and gene flow of
imported animals in the United States of America. Livestock Sci
120:240–247
Blackburn HD, Paiva SR, Wildeus S, Getz W, Waldron D, Stobart R,
Bixby D, Purdy PH, Welsh C, Spiller S, Brown M (2011)
Genetic structure and diversity among US sheep breeds:
identification of the major gene pools. J Anim Sci 89:2336–2348
Bruford M, Bradley D, Luikart G (2003) DNA markers reveal the
complexity of livestock domestication. Nat Rev Genet
4:900–910
Convention on Biological Diversity (2011) The nagoya protocol on
access and benefit-sharing. http://www.cbd.int/abs/. Accessed
July 2011
Evanno G, Regnaut S, Goudet J (2005) Detecting the number of
clusters of individuals using the software STRUCTURE: a
simulation study. Mol Ecol 4(8):2611–2620
Excoffier L, Laval G, Schneider S (2005) Arlequin ver. 3.0: an
integrated software package for population genetics data anal-
ysis. Evol Bioinform Online 1:47–50
Falconer DS, Mackay TFC (1996) Introduction to quantitative
genetics, 4th edn. Prentice Hall, England 464
FAO (1989) Animal genetic resources of the USSR. FAO Animal
Production and Health Paper 65. FAO, Rome, p 516
FAO (2004) Secondary guidelines for development of national farm
animal genetic resource management plans: measurement of
domestic animal diversity (MoDAD): recommended microsatel-
lite markers. FAO, Rome
FAO (2007) Global plan of action for animal genetics resources and
the interlaken declaration. FAO, Rome
FAO (2011) http://faostat.fao.org. Accessed Feb 2011
Goudet J (2002) FSTAT, a program to estimate and test gene
diversities and fixation indices (version 2.9. 3.2). University of
Lausanne, Department of Ecology and Evolution, Lausanne,
Switzerland
Groeneveld L, Lenstra J, Eding H, Toro M, Scherf B, Pilling D,
Negrini R, Finlay E, Jainlin H, Roeneveld E, Weigend S, the
GLOBALDIV Consortium (2010) Genetic diversity in farm
animals–a review. Anim Genet 41(Suppl. 1):6–31
Guo J, Du L, Ma Y, Guan W, Li H, Zhao Q, Li X, Rao S (2005) A
novel maternal lineage revealed in sheep (Ovis aries). Anim
Genet 36:331–336
Gutierrez JP, Royo LJ, Alvarez I, Goyache F (2005) MolKin v2.0: a
computer program for genetic analysis of populations using
molecular coancestry information. J Hered 96:718–721
Hedrick PW (2011) Genetics of populations, 4th edn. Jones and
Bartlett, Sudbury, p 675
Huson DH, Bryant D (2006) Application of phylogenetic networks in
evolutionary studies. Mol Biol Evol 23(2):254–267
Lawson-Handley LJ, Byrne K, Santucci F, Townsend S, Taylor M,
Bruford M, Hewitt G (2007) Genetic structure of European
sheep breeds. Heredity 99:620–631
Loftus RT, Ertugrul O, Harba AH, El-Barody MAA, MacHugh DE,
Bradley DG (1999) A microsatellite survey of cattle from a
centre of origin: the Near East. Mol Ecol 8:2015–2022
McKay SD, Schnabel R, Murdoch B, Matukumalli L, Aerts J,
Coppieters W, Crews D, Dias Neto E, Gill C, Gao C, Mannen H,
Wang Z, Van Tassell C, Williams J, Taylor J, Moore S (2008)
An assessment of population structure in eight cattle breeds
using a whole genome SNP panel. BMC Genet 9:37. doi:
10.1186/1471-2156-9-37
Medeubekov KU, Nartbayev A, Abishev B, Berus V, Kanapin K,
Terentyev V, Ombaev A, Aryngaziev S, Netyaga MO, Trayisov
B, Esentaev E, Kineev M (2008) Small ruminant breeds of
Kazakhstan. In: Iniguez L, Mueller J (eds) Characterization of
small ruminant breeds in central Asia, the caucasus. International
Center for Agricultural Research in the Dry Areas (ICARDA),
Aleppo, p 416
Ozerov M, Marzanov N, Tapio M, Feizullaev F, Burabaev A,
Amerkhanov Kh, Petrov S, Marzanova L, Gostishchev S,
Kantanen J (2008) Microsatellite characterization of closely
related fine-wool sheep breeds. Russian Agricultural Science
34:343–347
Paiva SR, Blackburn H, da Mariante AS (2010) Meta-analysis of
microsatellite data from US and Brazil sheep breeds. 9th world
congress genetics applied livestock production, Leipzig, vol 660,
pp 1–4
Paiva SR, da Mariante AS, Blackburn H (2011) Combining US and
Brazilian microsatellite data for a meta-analysis of sheep (Ovisaries) breed diversity: facilitating the FAO global plan of action
for conserving animal genetic resources. J Heredity (in press)
Peakall R, Smouse PE (2006) GENALEX 6: genetic analysis in
Excel. Population genetic software for teaching and research.
Mol Ecol Notes 6:288–295
Pereira F, Davis S, Pereira L, McEvoy B, Bradley D, Amorim A
(2006) Genetic signatures of a Mediterranean influence in
-50000
-45000
-40000
-35000
Log-
Like
lihoo
d
Cluster Number
a
0
50
100
150
200
250
300
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Cluster
Δ K
b
Fig. 5 Mean Log-likelihood (a) and DK (b) based on 3 replicated
STRUCTURE runs for Kazakh and US breeds
Genetica
123
Author's personal copy
Iberian Peninsula sheep husbandry. Mol Biol Evol
23:1470–1476
Peter C, Buford M, Perez T, Dalamitra S, Hewitt G, Erhart G, the
ECONOGENE Consortium (2007) Genetic diversity and subdi-
vision of 57 European and Middle-Eastern sheep breeds. Anim
Genet 38:37–44
Porter V, Mason I (2002) Mason’s world dictionary of livestock
breeds, types, and varieties. CABI, New York
Pritchard JK, Stephans M, Donnelly P (2000) Inference of population
structure using multilocus genotype data. Genetics 155:945–959
Rezaei HR, Naderi S, Chintauan-Marquier I, Taberlet P, Virk A,
Naghash H, Rioux D, Kaboli M, Pompanon F (2010) Evolution
and taxonomy of the wild species of the genus Ovis (Mammalia,
Artiodactyla, Bovidae). Mol Phylogenet Evol 54:315–326
Rosenberg NA (2004) DISTRUCT: a program for the graphical
display of population structure. Mol Ecol Notes 4:137–138
Slatkin M (1985) Rare alleles as indicators of gene flow. Evolution
39:53–65
Sulaiman Y, Wu C, Zhao C (2011) Phylogeny of 19 indigenous sheep
populations in Northwestern China inferred from mitochondrial
DNA control region. Asian J Anim Vet Adv 6:71–79
Szpiech ZA, Jakobsson M, Rosenberg NA (2008) ADZE: a rarefac-
tion approach for counting alleles private to combinations of
populations. Bioinformatics 24:2498–2504
Tapio M, Miceikiene I, Vilkki J, Kantanen J (2003) Comparison of
microsatellite and blood protein diversity in sheep: inconsisten-
cies in fragmented breeds. Mol Ecol 12:2045–2056
Tapio M, Ozerov M, Tapio I, Toro M, Marzanov N, Cinkulov M,
Goncharenko G, Kiselyova T, Murawski M, Kantanen J (2010)
Microsatellite-based genetic diversity and population structure of
domestic sheep in northern Eurasia. BMC Genet 11:76
Tomiuk J, Loeschcke V (1995) Genetic identity combining mutation
and drift. Heredity 74:607–615
Wagner M, Wu X, Tarasov P, Aisha A, Ramsey C, Schultz M,
Schmidt-Schultz T, Gresky J (2011) Radiocarbon-dated archae-
ological record of early first millennium B.C. mounted pasto-
ralists in the Kunlun Mountains, China. PNAS 108:15733–15738
Wang X, Ma Y, Chen H, Guan W (2007) Genetic and phylogenetic
studies of Chinese native sheep breeds (Ovis aries) based on
mtDNA D-loop sequences. Small Rumin Res 72:232–236
Wood RJ, Orel V (2001) Genetic prehistory in selective breeding: a
prelude to mendel. Oxford University Press, New York
Young LD (1992) Effects of Duroc, Meishan, Fengjing and Minzhu
boars on productivity of mates and growth of first cross progeny.
J Anim Sci 70:2020–2029
Zeder MA (2008) Domestication and early agriculture in the
Mediterranean Basin: origins, diffusion, and impact. PNAS
105:11597–11604
Genetica
123
Author's personal copy