doi:10.1111/j.1365-2052.2010.02102.x

Genetic diversity in the Maremmano horse and its relationship withother European horse breeds

M. Felicetti*,1, M. S. Lopes†1, A. Verini-Supplizi*, A. da Camara Machado†, M. Silvestrelli*,

D. Mendonca† and O. Distl‡

*Department of Pathology, Diagnostic and Veterinary Clinic, University of Perugia, Via San Costanzo 4, 06126 Perugia, Italy.†Biotechnology Centre of Azores, Department of Agriculture, University of Azores, Terra-Cha, 9701-851 Angra do Heroısmo, Portugal.‡Institute for Animal Breeding and Genetics, University of Veterinary Medicine Hannover, Bunteweg 17p, 30559 Hannover, Germany

Summary The Maremmano is an Italian warmblood horse breed from central Italy. We characterized

the genetic diversity and the degree of admixture in Maremmano in comparison to 14 other

European horse breeds using 30 microsatellites. Between-breed diversity explained about 9

per cent of the total genetic diversity. Cluster analysis, genetic distances and genetic dif-

ferentiation coefficients showed a close relationship of Maremmano with Hanoverian and

Lusitano in accordance with breed history.

Keywords clustering analysis, Equus caballus, Maremmano.

The Maremmano is an Italian warmblood horse mostly bred

in the provinces of Grosseto and Viterbo (Central Italy). It is

believed that the origin of this horse breed goes back to local

Etruscan horse populations which were crossbred with

modern breeds in the last centuries. Pedigree analysis of the

Maremmano horses revealed four male lines contributing

11.1% to the genetic diversity of all Maremmano stallions

(Silvestrelli 1991). In 1980, Maremmano breeders estab-

lished a stud book for Maremmano.

The objective of this study was to investigate the genetic

structure and the degree of admixture of Maremmano to-

gether with the Italian nucleus of Lipizzan and Lusitano as

well as twelve further horse breeds previously characterized

by Aberle et al. (2004). This data set consisted of Hanove-

rian, Arabian, Exmoor, Icelandic, Przewalski, Sorraia and

all German coldblood horse breeds. We used the same

marker set of 30 autosomal microsatellites and the same

PCR conditions and reference samples as described else-

where (Aberle et al. 2004).

A total of 146 animals that were not closely related were

genotyped in this study: Maremmano (n = 50), the Italian

nucleus of Lipizzan (n = 49) and Lusitano (n = 47) from

Portugal. DNA was extracted from EDTA-blood using

standard methods.

Genetic variability of the breeds genotyped in this study

was determined by the mean number of alleles, observed

and expected heterozygosities, breed-specific alleles and the

molecular variance (AMOVA) using GENALEX 6 (Peakall &

Smouse 2006). Excess and deficiency of heterozygotes,

which are deviations from Hardy-Weinberg equilibrium,

were estimated using GENEPOP (Raymond & Rousset

1995). Molecular genetic relationships among populations

were estimated using Wright�s FST and Nei�s standard

genetic distance (GST; Nei 1973) by bootstrapping 1000

replicates using MICROSAT (Minch 1997). Phenograms

based on Nei�s GST and genetic distances among all 549

animals were drawn using the unweighted pair group

method with arithmetic mean algorithm (UPGMA) by

PHYLIP (Felsenstein 1989) and displayed by TREEVIEW

(Page 1996). The Bayesian clustering procedure of

STRUCTURE was employed to investigate the genetic

structure and the degree of admixture of the 15 horse breeds

(Pritchard et al. 2000). A 20 000 initial burn-in was used,

followed by 100 000 MCMC iterations as recommended by

Falush et al. (2007) with 10 independent replicates each.

All runs used an admixture model with correlated fre-

quencies and the parameter of individual admixture alpha.

The mean number of alleles was 4.7 for the Lipizzan, 6.7

for the Lusitano and 7.3 for the Maremmano. For Marem-

mano and Lipizzan breeds the total observed heterozygosity

was higher than the expected, whereas for Lusitano breed

Address for correspondence

Prof. Dr O. Distl, Institute for Animal Breeding and Genetics, University

of Veterinary Medicine Hannover, Bunteweg 17p, 30559 Hannover,

Germany.

E-mail ottmar.distl@tiho-hannover.de

1These authors have contributed equally to this work.

Accepted for publication 20 September 2010

� 2010 The Authors, Journal compilation � 2010 Stichting International Foundation for Animal Genetics, 41 (Suppl. 2), 53–55 53

19 out of the 30 loci showed observed heterozygosity values

lower than the expected ones (Table S1).

The AMOVA indicated that for Maremmano, Lipizzan and

Lusitano 9% of the total genetic variability is attributed to

significant differences between the horse breeds, whereas

91% of the observed variation was from within the breeds.

The AMOVA performed for the five riding horse breeds

(Maremmano, Lipizzan, Lusitano, Arabian, Hanoverian),

considered separately from the others, showed 11% of the

genetic variability due to breed differences. When the Ger-

man coldblood breeds were included, the variation among

breeds increased only slightly to 12%; with all 15 breeds the

total genetic variability rose to 14.6%.

Maremmano showed the lowest genetic differentiation

with the Hanoverian (5.5%); in addition, the Lipizzan and

Lusitano, when matched with the Maremmano, showed the

lowest FST values (11.4% and 6%, respectively). Among the

three breeds genotyped here, the Maremmano showed

the least differentiation when compared to the coldblood

breeds (Table S2). Similar results were obtained when the

genetic differentiation based on Nei�s GST among breed pairs

was used. The only exception was for the Maremmano,

which showed lowest genetic differentiation (14.7%) with

the Lusitano.

The phylogenetic tree based on Nei�s GST (Fig. 1) and the

dendrogram based on the proportion of shared alleles

(Fig. S1) displayed three main clusters representing riding

horses, the Exmoor, Przewalski and Sorraia group and the

German draught horses. The Icelandic horses did not cluster

with any of the other breeds.

Clustering using STRUCTURE separated for K = 3 horses

into riding horse breeds, ancient and isolated breeds, and

German draught horses. For K = 9, Maremmano clustered

together with Lusitano, but Hanoverian and Arabian were

separate clusters. When K = 14, Maremmano and Lusitano

also clustered in their own pre-defined populations (Fig. S2;

Table S3).

From the three breeds genotyped in this study, both the

Maremmano and the Lusitano showed a high level of

genetic variability and similar to that observed for the same

breeds in other studies (Luıs et al. 2007; Zuccaro et al.

2008). The high level of genetic differentiation observed for

the Maremmano may partly reflect contributions from

several breeds, although a significant proportion of the

Maremmano is descending from a reduced number of male

lines. The Italian nucleus of Lipizzan horses showed low

levels of variation similar to those observed by Achmann

et al. (2004). The small number of founders resulting in a

small effective population size and the traditional pure-

breeding system within a nucleus without any crossbreed-

ing may explain the reduced variability within the Lipizzan.

In agreement with the results of Luıs et al. (2007) in the

Lusitano, levels of observed heterozygosity were lower than

their expected counterpart.

Pairwise FST values among Maremmano, Lipizzan, Lu-

sitano and Hanoverian were in a similar range as those

among closely related coldblood breeds (Aberle et al. 2004;

Druml et al. 2007). Inclusion of Hanoverian and German

coldblood increased the proportion of variance among

breeds only slightly, indicating a close relationship of these

breeds with Maremmano, Lipizzan and Lusitano in contrast

to Arabian, pony and primitive horse breeds. The two

genetic differentiation measures and the two model-based

clustering approaches also revealed a genetic proximity of

Maremmano with the Hanoverian and the Lusitano. Two

male lines founded by thoroughbred stallions (Aiace and

Ingres) and a Trakehner stallion might have created the

relationship between Maremmano and Hanoverian, as

thoroughbred and Trakehner stallions have been intensely

used in the Hanoverian warmblood (Hamann & Distl 2008).

Regarding the close genetic proximity between Maremmano

and Lusitano, it is believed that Iberian horses, the ancestors

of the Lusitano and the Andalusian, were in the Stato dei

Reali Presidi di Spagna (1557–1800), located between Stato

Pontificio and Granducato di Toscana, and therefore might

have influenced the founder lines of the Maremmano breed.

Figure 1 Dendrogram showing the genetic relationships among the 15

horse breeds inferred from microsatellite data. The tree is based on Nei�sstandard genetic distance (GST). Bootstrap values higher than 50 are

shown in the tree.

� 2010 The Authors, Journal compilation � 2010 Stichting International Foundation for Animal Genetics, 41 (Suppl. 2), 53–55

Felicetti et al.54

In conclusion, the Maremmano retained a high genetic

diversity and the results reported here can be used to pre-

vent genetic erosion of the Maremmano breed.

Acknowledgements

Thanks to Dr Luca Buttazzoni for technical support and Mr

Gianluca Alunni for expert technical assistance. This work

was supported by SELMOL-MIPAF (Silvestrelli) and by FTC

and FEDER within the projects POCTI/CVT/41890/2001

and POCI2010 Ref. GG/GGP/ME611. CBA-UAc was sup-

ported by FCT and DRCT and Lopes by DRCT M3.1.1/I/

017A/2005.

Conflict of interest

The authors have declared no potential conflicts.

References

Aberle K.S., Hamann H., Drogemuller C. & Distl O. (2004) Genetic

diversity in German draught horse breeds compared with a group

of primitive, riding wild horses by means of microsatellite DNA

markers. Animal Genetics 35, 270–7.

Achmann R., Curik I., Dovc P., Kavar T., Bodo I., Habe F., Marti E.,

Solkner J. & Brem G. (2004) Microsatellite diversity, population

subdivision and gene flow in the Lipizzan horse. Animal Genetics

35, 285–92.

Druml T., Curik I., Baumung R., Aberle K., Distl O. & Solkner J.

(2007) Individual-based assessment of population structure and

admixture in Austrian, Croatian and German draught horses.

Heredity 98, 114–22.

Falush D., Stephens M.W. & Pritchard J.K. (2007) Inference of

population structure using multilocus genotype data, dominant

markers and null alleles. Molecular Ecology Notes 7, 574–8.

Felsenstein J. (1989) PHYLIP – Phylogeny inference package. Cla-

distics 5, 164–6.

Hamann H. & Distl O. (2008) Genetic variability in Hanoverian

warmblood horses using pedigree analysis. Journal of Animal

Science 86, 1503–13.

Luıs C., Juras R., Oom M.M. & Cothran E.G. (2007) Genetic diversity

and relationships of Portuguese and other horse breeds based on

protein and microsatellite loci variation. Animal Genetics 38, 20–

7.

Minch E. (1997) MICROSAT, Version 1.5b. Stanford University

Medical Center, Stanford, California.

Nei M. (1973) Analysis of gene diversity in subdivided populations.

Proceedings of the National Academy of Sciences of the United States

of America 70, 3321–3.

Page R.D.M. (1996) TREEVIEW: an application to display phylo-

genetic trees on personal computers. Computer Applications in the

Biosciences 12, 357–8.

Peakall R. & Smouse P.E. (2006) GENALEX 6: genetic analysis in

Excel. Population genetic software for teaching and research.

Molecular Ecology Notes 6, 288–95.

Pritchard J.K., Stephens M. & Donnelly P. (2000) Inference of

population structure using multilocus genotype data. Genetics

155, 945–59.

Raymond M. & Rousset F. (1995) GENEPOP (version 1.2): popu-

lation genetics software for exact tests and ecumenicism. Journal

of Heredity 86, 145–55.

Silvestrelli M. (1991) The Maremmano horse. Animal Genetic

Resources Information (FAO/UNEP) 8, 74–83.

Zuccaro A., Bordonaro S., Criscione A., Guastella A.M., Perrotta G.,

Blasi M., D�Urso G. & Marletta D. (2008) Genetic diversity and

admixture analysis of Sanfratellano and three other Italian horse

breeds assessed by microsatellite markers. Animal 2, 991–8.

Supporting information

Additional supporting information may be found in the

online version of this article.

Figure S1 UPGMA dendrogram constructed from allele-

sharing distances among 549 animals from 15 different

horse breeds.

Figure S2 Graphical presentation of the population struc-

ture analyses for a sample of 549 horses from 15 different

horse breeds obtained by STRUCTURE.

Table S1 Number of alleles, observed and expected hetero-

zygosity for Maremmano, Lusitano and Italian Lipizzan

horses (146 horses) based on 30 microsatellite loci.

Table S2 Nei�s standard genetic distance and Wright�s dis-

tance among 15 horse breeds.

Table S3 Estimated memberships to inferred clusters

obtained by STRUCTURE and individual assignments of 549

horses according to their own predefined breed or to

another breed as obtained using GeneClass2.

As a service to our authors and readers, this journal

provides supporting information supplied by the authors.

Such materials are peer-reviewed and may be re-organized

for online delivery, but are not copy-edited or typeset.

Technical support issues arising from supporting informa-

tion (other than missing files) should be addressed to the

authors.

� 2010 The Authors, Journal compilation � 2010 Stichting International Foundation for Animal Genetics, 41 (Suppl. 2), 53–55

Genetic diversity of the Maremmano 55