Funpec-RpAbout The JournalEditorial BoardCurrent IssueAll IssuesSearchIndexersInstructions For AuthorsContactSponsorsLinks

Morphological evolution and heritability estimates for some biometric traits in the Murgese horse breed
C. Dario, D. Carnicella, M. Dario and G. Bufano
Department of Animal Health and Welfare, University of Study of Bari,
strada prov.le per Casamassima, km 3, 70010 Valenzano, Ba, Italy
Corresponding author: C. Dario
Genet. Mol. Res. 5 (2): 309-314 (2006)
Received January 18, 2006
Accepted January 20, 2006
Published May 17, 2006

ABSTRACT. A data set concerning 1,816 subjects entered in the Italian Horse Registry from 1925 to 2002 was analyzed to investigate the morphological evolution of the Murgese horse and to obtain useful elements to enhance breeding practices. Three basic body measurements (height at withers, chest girth, and cannon bone circumference) were considered for each subject. Heritabilities were calculated for each parameter to infer the growth and development traits of this breed. Over the past 20 years the Murgese horse has undergone considerable changes, passing from a typical mesomorphic structure (height at withers: 156.30 and 151.04 cm; chest girth: 185.80 and 176.11 cm; cannon bone: 21.10 and 19.82 cm for males and females, respectively) to a mesodolichomorphic structure (height at withers: 160.31 and 156.44 cm; chest girth: 187.89 and 182.48 cm; cannon bone: 21.07 and 20.37 cm, for males and females, respectively). Due to these changes and to its characteristic strength and power, the Murgese, which was once used in agriculture and for meat production (at the end of its life), is now involved in sports, mainly in trekking and equestrian tourism. The heritability estimates for the three body measurements were found to be 0.24, 0.39 and 0.44.

Key words: Horse, Murgese, Biometric traits, Heritability



Over the Murgia hillsides interspersed with trulli (typical cone-shaped houses) and bordering the Apulian provinces of Bari, Brindisi and Taranto, there is a breed of carefully selected horses, called Murgese. This horse has Spanish, Neapolitan, Berber, and Arabian ancestors and its origins date back to the 15th and 19th centuries (Basile, 1956; Pagano, 1975; Aurigi, 1986; Buonavolontà and Silvestrelli, 1986; Catalano, 1993). This horse breed was fully developed in the 1920s; it originated from local horses found in the arid and rocky Italian hills of Apulia, from which takes its name: le Murge (Mason, 1996). The Murgese horse has a blackish, iron-grey coat; it has great rusticity and a fine powerful appearance because it was grown wild on the open range.

The Murgese was a useful, easily manageable farm horse and a good worker; it was highly valued for its rustic nature, which was necessary for its survival in such a harsh environment. Its conformation and characteristics made it particularly suitable for raising cattle. The mares were also used to breed strong mules when the production of these hybrids was required (especially for the army during World War II).

The Murgese today

The modern Murgese is a light draft horse with variable types; it is an extremely docile animal and easily becomes used to saddles and harnesses. It is amenable to equestrian tourism and cross-country riding because of these qualities.

Another important characteristic, along with its obedience, is its learning ability, which qualify the Murgese for the complicated exercises of haute cole, steeple chasing and for physiotherapy. The Murgese horse is likewise a superb grazer and makes good use of poor pastures, especially in the mountain regions. The Italian Forestry Service uses the Murgese for service all over Italy. The Murgese is also highly resistant to diseases, so much so that organic disorders, such as heaves and intestinal diseases, are practically non-existent in this breed. It has a robust skeleton and muscles, as well as a tough and thick epidermis, which plays an important role in protection against insect bites and thorny vegetation.

Effective selection for the improvement and conservation of a pure breed has ensured these valuable aptitudes and extremely desirable traits in the Murgese horse, which is now receiving rapid and well-deserved notoriety. The Murgese herdbook (a breeder’s census) has only been kept since the 1920s, when the Italian government’s horse breeding association took an active interest in preserving this unique breed. There are currently 1089 herdbook entries (AIA, 2003). Since the endangered status of animal breeds is determined by the size of the breeding stock (Bodò, 1990), the Murgese horse was classified as vulnerable, and measures were taken to prevent a further decrease in its numbers.

We analyzed three basic body measurements (height at withers, chest girth and cannon bone circumference) and estimated the heritabilities for the biometric traits measured at registration in order to assess the morphological evolution and genetic characteristics of this breed.



Data collected in the period from 1925 to 2002, regarding 1,816 animals (905 mares and 911 stallions) registered in the official genealogical herdbook of the Murgese horse were analyzed. Based on the year of birth and stallion reproduction period (an average of seven years) the dataset was divided into eight classes covering seven years each (1925-49; 1950-57, 1958-65, 1966-72, 1973-79, 1980-87, 1988-95, 1996-2002), except for the first period, which was extended because the number of findings was too small to obtain reliable estimates.

The traits were the three basic biometric measures used for horses (height at withers, chest girth and cannon bone circumference). Following a standard protocol, measurements were taken as follows: height at withers (measuring tape) is the distance from the highest point of the processus spinalis of the thoracic vertebra to the floor; chest girth (measuring tape) measured at the saddle girth; circumference of the cannon bone (measuring tape): smallest circumference of the forelimb cannon bone.

Statistical analysis

Statistical analysis was performed by SAS (SAS, 1998). The dataset was analyzed using the GLM procedure, considering the fixed effect of classes, sex type and sire. Means were compared by the t-test.

In order to estimate the genetic parameters (heritabilities) of the biometric traits, a second dataset, including only contemporary animals (belonging to classes 6, 7 and 8) generated by 128 sires, was considered. The model included the fixed effect of year, sex and the random effect of sire, using the restricted maximum-likelihood method from the Mixed Procedure of the SAS software (SAS, 1998). Error was assumed to be randomly and independently distributed, with a mean of zero and a variance of s2e. Heritability estimates were based on the sire component of variance (s2s), utilizing the following expression: h2 = 4 s2s /(s2s + s2e). The standard error of heritability was approximated using the method described by Becker (1968).


Biometric traits

The mean values of the three basic measurements (average ± SE) are presented in Table 1. The male population has progressively and significantly (P < 0.01) increased in height at withers from about 156 cm for subjects included in classes 1 and 2 to 160 cm in classes 7 and 8. The same trend was observed in the females, measuring about 151 cm (class 1) and reaching 156-157 cm over the last 20 years; the differences found were significant (P < 0.01). This continuous and constant trend to become taller was more evident in the last 20 years, and it clearly identifies the historical period during which these animals better responded to the selection process, incorporating the desired trait.

Also the average chest girth increased both in males and in females from 185.80 and 176.11 cm to 198.92 and 190.78 cm in class 6, respectively. These values significantly (P < 0.05) decreased to 187.89 and 182.48 cm in the last two periods for males and females, respectively. This trend was replicated by the average values for the circumference of the cannon bone, with an initial progressive increase from 21.10 and 19.82 to 22.89 and 21.83 cm in class 6, followed by stabilization at 21.07 (P < 0.05) and 20.37 cm for males and females in class 7, respectively.


The heritability value calculated for height at withers was rather low (0.24, Table 2) and generally consistent with the results in the literature (0.26, von Butler-Wemken, 1990; 0.27, Baban et al., 1998), although Zechner et al. (2001) reported a value of 0.52 for Lipizzan horses and Miserani et al. (2002) found 0.61 for the Pantaneiro breed.

Molina et al. (1999) estimated heritabilities in Andalusian horses of 0.49 for chest girth and of 0.35 for forelimb cannon bone circumference (vs 0.39 and 0.44 reported in the present study, respectively). Baban et al. (1998) estimated heritabilities for the Lipizzan horse population of Djakovo, which were 0.25 for chest girth and 0.18 for cannon bone circumference. Pikula et al. (1998) estimated lower values for height at withers for cold-blooded horses with 0.16, 0.09 for chest girth and 0.23 for cannon bone circumference; the highest value (0.83) for thorax perimeter was found for the Pantaneiro horse breed by Miserani et al. (2002). The values in our study of the Murgese horse breed (Table 2) are based on a rather small number of observations for a conclusive analysis.


Thanks to its highly appreciated rustic nature and conformation, the Murgese horse has preserved its high value over the course of time. Today this horse is bred for special performance traits that are different from those of the past.

Current selection is now directed towards obtaining a taller, more distinct and better saddle horse; our morphometric findings at least partially reflect these new breeding goals. The heritability estimates that we found, though moderate (ranging from 0.24 to 0.44), suggest that mass selection is resulting in a positive response.


AIA - Associazione Italiana Allevatori (2003).

Aurigi M (1986). Il Morello della Murgia, un cavallo diverso. Terra Vita 36: 82-83.

Baban M, Rastija T, Caput P, Knezevic I, et al. (1998). Estimation of heritability of Lipizzaner horses for morphological traits by means of various methods. Czech. J. Anim. Sci. 43: 299-303.

Basile A (1956). La razza cavallina murgese. L’Italia agricola 8: 631-638.

Becker WA (1968). Estimation of variance components and heritability. In: Manual of procedures in quantitative genetics. 2nd edn. The program in genetics. Washington State University, Pullman, WA, USA, pp. 7-59.

Bodò I (1990). Methods and experiences with in situ preservation of farm animals. FAO, Animal Production and Health Paper, Rome, Italy, pp. 85-102.

Buonavolontà G and Silvestrelli M (1986). II Murguese. II più bel cavallo del mondo che nessuno vuole. Ed. Equestri, Milano, Italy.

Catalano AL (1993). Valutazione morfo-funzionale del cavallo. Igiene ed etnologia. Ed. SBM, Noceto, Italy.

Mason IL (1996). A world dictionary of livestock breeds, types and varieties. 4th edn. C.A.B International, Liphook, Hants, UK.

Miserani MG, McManus SA, Santos JA, Silva AS, et al. (2002). Heritability estimates for biometric measures of the Pantaneiro horse. Arch. Zootec. 51: 107-112.

Molina A, Valera M, Dos Santos, R and Rodero A (1999). Genetic parameters of morphofunctional traits in Andalusian horse. Livest. Prod. Sci. 60: 295-303.

Pagano MG (1975). La razza cavallina delle Murge. Fiera di Verona, Italy.

Pikula R, Grezesiak W and Stepien J (1998). Genetic conditioning of body conformation traits in cold-blooded horses. Zootechnica 36: 17-22.

S.A.S. (1998). Guide for personal computers. 6th edn. S.A.S. Institute Inc., Cary, NC, USA.

von Butler-Wemken I (1990). Genetiche parameter fur exterieurmerkmale bei reitpferdestuten. Landw. Jb. 67: 581-586.

Zechner P, Zohman F, Solkner J, Bodo I, et al. (2001). Morphological description of the Lipizzan horse population. Livest. Prod. Sci. 69: 163-177.

   Copyright © 2006 by FUNPEC