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: Harvey.blackburn@ars.usda.gov

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