Biodiversity in goats in the Tunisian oasis

M Nafti, Z Khaldi*, B Rekik and A Ben Gara

Ecole Supérieure d’Agriculture de Mateur. 7030 Mateur, Tunisie
* Centre Régional de Recherche en Agriculture Oasienne. 2260 Degueche Tunisie
rekik.boulbaba@iresa.agrinet.tn

Abstract

Goats are well adapted to Tunisian climatic conditions. Goat breeding is primarily found in the mountains, natural- pasture regions, and the arid (Sahara) and oasis regions in the south of the country. Goat variability is high in Tunisia and breeds are under extensive production systems. The objective of this study was the identification and phenotypic characterisation of various goat breeds or populations in the oasis in six localities in the governorate of Tozeur. Qualitative and quantitative traits of horns (form and length), the coat, ears, hair, teeth, the cutting, etc. were used to distinguish among phenotypes representing sub- populations of goats.

There were four locally identified goat populations. The first population called “Arbi” is characterised by a long black to black and white coat with frequently shammy, generally with bears but without wattles. The second population called “Cherki” has short-hair-coat coloured with several colours often shammy. The third population is the “Damasquine”. The latter includes animals that are high on their paws showing greater features than the two first populations. Finally, the “Alpine” with her specific shammy colour and is made of animals that are too high on their paws compared to other populations. Morphological differences among various populations indicate high diversity of goats in the Tunisian oasis. This study may be complemented by molecular characterisation of various populations for conservation and genetic improvement schemes.

Key words: locality, phenotype, polymorphism, population

Introduction

The goat population in Tunisia counted 1412000 live animals in 2004, among of which 809330 were females. Female population size increased up to 820000 in 2006, with a growth rate of 1.2% (Ministry of Agriculture and Hydraulic Resources 2006). The abundance of goats mirrors the increase of aridity from the north to the south of the country. Goats in Tunisia represent a high diversity and heterogeneity, which is the result of uncontrolled brewing among various breeds populating different areas in the country. Therefore, it is difficult to distinguish among well-defined genetic entities. The study of the features and the productive behaviour of a population are of considerable utility, especially, when animal resources display some authentic details, biologic, technical or environmental (Narjisse et al 1989). The characterization of local goats, raised in the arid Tunisian regions, presents scientific and technical interests especially to the genetic improvement and also to the conservation of animal genetic resources. The aim of this study was to characterize goats in the Tozeur region, in the south of Tunisia, on their morphological appearances and to link morphological characteristics to genetic determinism and environment.

Materials and methods 

Survey

Our study took place in 6 localities (El Hamma, Tozeur, Degueche, Ibn Chabbat, Legwifla and Nefta) in the region of Tozeur. The choice of the localities within the whole region was fully developed according to the importance and sizes of herds. For instance, herds with at least ten animals were privileged especially when they included two-year-old females. This approach is justified by the fact that female goats spend more time in the herd than males. A total of 36 herds were surveyed and sampled animals were 450 females and 38 males.

Trait definition and measures

The genetic profiling was done following the standards found in Lauvergne (1986), Cogovica (1987) and Lauvergne et al (1987). Those standards recommend various scales such as form and length of horns, colour pattern, dentition, the length of the coat, length and carriage of the ear, etc. Phenotypes that are qualitative in nature were marked in terms of presence or absence while quantitative variables were measured.

Analysis

Univariate analysis of quantitative and qualitative profiles

Data were organised in a I x J contingency table, where the I (line) represents observations and J (column) represents traits. Then, Xij is the i^th observation on the j^th trait. Elementary statistics (mean, standard deviation, frequencies, etc.) and distributions of quantitative measures of various traits were determined. Firstly, we studied the representativeness index of the total value (Table 1). Secondly, attributes (qualitative measures) were analysed according to norms defined by Lauvergne (1986) and Cogovica (1987). Absolute frequencies of phenotypes were determined in each locality. These frequencies were complemented by relative gene frequencies coding for the observed phenotypes with presumably established genetic determinism.

Multivariate analysis

The methods of studying whole characters rest on the principles of multi-dimensional analysis (Jivotovski 1985). Multivariate analyses were used to determine relationships among various variables. A principal component analysis was performed on the quantitative variables: the cutting, the wither height, the turn of the cannon, the ears length, the horns length and the length of the hairs. Axes were intended to represent various breeds or populations through specific common observables characters. A general linear model (SAS 1989) was then used to explain the variation of observed characters (body length, wither height, heart girth, ears length, horns length, hair length) by explanatory variables such as locality, breed, sex, age, etc. The general form of the linear model was:

Y_ijkl = µ+ locality_i + breed_j + sex_k + age_l + breed (locality) _ij+ sex (breed)_jk+ε_ijkl

Where

y_ijkl = a quantitative variable (body length, wither height, heart girth, ears length, horns length, hair length),

µ = general mean of population,

locality_i = effect of the i^th locality,

breed _j = effect of the j^th breed,

sex_k = effect of the k^th sex,

age _l = effect of the l^th age,

breed (locality) _ij = effect of breed within locality,

sex (breed) _jk = effect of sex within breed and

ε_ijkl = residual error.
 

Results 

Descriptive statistics for quantitative characters

The largest recorded value of the range was for the length of the horns which was 58 cm (Table 1). The coefficient of variation was less than 10% for the wither height, the heart girth and the cutting and was equal to 15.6% for the ears length (Table 1).  However, for the horns length, this coefficient reached 32% and was up to 58.2% for the hairs length. The average body length of sampled animals varied from a locality to another (Table 2). The highest value was observed in Nefta (mean = 104.2, STD = 10.1) while the lowest value was observed in El Hamma (mean = 92.6 cm, STD = 9.4 cm). We could identify two classes of height, the first with a mean height around 65 cm in Degueche, El Hamma and Tozeur localities and the second around 68 cm in Ibn Chabbat, Legwifla and Nefta localities. On the other hand, in Nefta, Legwifla and Ibn Chabbat we recorded large means of heart (76 cm < mean < 80 cm) compared to that found in other localities with the lowest value found in El Hamma (72.9, STD = 5.8 cm). Important ear length differences were also observed among the different localities. The longest ears were found in Nefta (19.2 cm). As for the length of horns (for horned animals only), Legwifla and Ibn Chabbat had the largest means. The long hairs distinguish the localities of Legwifla and Ibn Chabbat localities among the others, while animals in Nefta and Degueche localities had the shortest hair lengths.

Nature of phenotypic frequencies

The visible phenotypic profile

Frequencies of observable phenotypes tended towards one. The obedient characters to this tendency were the normal length ears, the presence of the horns, the absence of the wattles and the straight chamfer. The sites of Tozeur, Ibn Chabbat and Legwifla were good examples of that tendency.

The genetic profile

The genetic profile was deduced directly from the phenotypic frequencies. Derived allelic frequencies are given in Table 3. These frequencies show the dominance of the wild character which supposes that the studied goat populations belong to the traditional type according to Lauvergne (1986). Principally normal length ears, the presence of horns, the absence of the wattles, the absence of the beard and the relatively short hairs characterize the goat populations of the Tozeur area.

The locus « length of the hairs » (HL)

A pair of autosomal alleles controls the length of the hairs with intermediate dominance of the short hair HL⁺ in relation to the HL^L gene with long hairs (Lauvergne et al 1987, Pattie and Restall 1989). In this study, we observed a greater proportion of the long-haired animals HL, with 57% of HL^L and 43% of HL⁺. Similar results were reported by Lanari et al (2003) with proportion 53% and 47% for long and short hairs, respectively, in Criollo goat population.

The locus «length of the ears» (EL)

A gene controls the length of the ears with intermediate dominance. Short ears result from the EL^r/El^r heterozygote (Cogovica 1987). Ear size appears to be conditioned by a gene series where small ears are dominant or additive to large ears (Pattie et Restall 1989). In the present study, the length of ears is dominated by the EL^r allele with a frequency equals to 1.

The locus «wattles presence» (Wa)

The presence of the wattles is controlled by a dominant autosomal gene (Wa^w) (Lauvergne 1987). The presence of wattles on the neck is dominant and is common in some strains of feral and milking goat (Pattie and Restall 1989). But in this study, the absence of wattles dominates with the wild allele Wa+ and an allelic frequency nearing 0.86.

The locus «presence of horns» (Ho)

The polled character (HO^p) is an autosomal gene bound to the sex. In the homozygous state, HO^p exercises a pleitropic action leading to partial or total sterility (French 1971, Dolling 1999, Pattie and Restall 1989; Vivicorsi 1998). About 83% do not have horns, therefore threatened by the inter-sexuality or sterility.

The locus «beard presence » (Br)

The beard is controlled by the autosomal allele Br bound to the sex, dominating in males and recessive in females (Lauvergne 1987). The greatest portion of the studied populations do not have beards (60%), similar results were presented by Vivicorsi (1998) in the French Rove goats.

Classification of local goats on quantitative characters

Large phenotypic correlations (Table 4) were found between heart girth and wither height (0.84), heart girth and length of the horns (0.73), body length and heart girth (0.69) and wither height and body length (0.64). However, ears length and hairs length were not highly correlated. The lowest correlation coefficient (0.10) was found between body length and hairs length. Similar results were reported by Pattie and Restall (1989) on Australian goat populations. The principal component analysis revealed distinct populations. The first component was represented by the wither height, heart girth and body length; the second component by the hairs length and finally the third component by the ears length. These components may be representative of four populations: the first one represents “Arbi goats”, the second “Serti goats” (a combination of “Cherki” and “Malti” breeds) and the third may represent both “Damasquine” and “alpine” breeds.

Classification of local goats on qualitative characters

Local goat populations can be divided into: The Cherki goat that results essentially from melting of the Malti and Arbi breeds. The close-cropped hairs (92.4%) with ears of all forms characterize this goat population. The intermediary form of horns was abundant (76%); polled animals represent 20% of this population. All possible hair colours exist within this population with the dominance of the shammy colour. Neither wattles nor beards generally exist and the straight head form is common in Cherki goats. The black-long-haired animals constitute 66.5% of Arbi goats. The ears form is the falling in 80% of the cases. The horns are often present with 72% of intermediate form. The wattles are absent in 87% of the population. The presence or the absence of the beard has an equal percentage. The chamfer is straight for 76% of this population and the concave form of the chamfer was rarely found. The Malti goat presents characters very close to those of the Cherki population. The coat is predominantly close-cropped (80%) and the rest are medium-length coats (20%). A dominance of the shammy colour (80%) was noticed. The Damasquine goat, that is either pure breed or crossed with the local goat, possess long falling ears (80%). Several colours exist within this breed with a dominance of the reddish colour. The wattles are quasi-absent and the beards are almost absent (in more than 70% of the cases). The chamfers concave type is absent and the straight type represents 60% in this population. All Alpine goats have close-cropped coat, 85.5% of animals with trained ears, 63% present ibex form of horns and 91% with shammy’s hairs. The wattles and the beards are absent (in more than 75% of cases). The chamfers are straight and concave.

Description of identified goat populations

Arbi population

This population is characterized by a black, black and white coat, sometimes shammy, grey or white. Their hairs are essentially long with means of 11.5 cm for males and 10.2 cm for females. The ears form is falling. The chamfer is straight and the head has horns of 18.8 cm of length for males and 18.1 cm for females. The measurements at the tourniquet are around 75 cm for males and 66.7 cm for females. Males have a heart girth of 84.2 cm and that of females is equal to 75.3 cm. The mean body length of males of this breed is 104.2 cm while that of females is around 95.8 cm.

Serti population

It contains the two Cherki and Malti breeds having similar features. The profile of the head is straight to concave. Its coat is multicoloured, frequently shammy. All forms of ears have been recorded; the head is horned, generally without neither bears nor wattles. The hair is close-cropped. The recorded measurements of body length were 96 cm and 93.7 cm for males and females, respectively. The heights at the tourniquet of males were close to 64 cm in males and 65 cm in females. Measured heart girths were of 73 cm for both sexes who also have similar ear lengths.

Damasquine population

It has features of the native breed and exotic goats. Animals of this population present different types of hair with a dominating close-cropped type. The ears are broad, long and falling. The colours of the hairs are variable. The beard and the wattles are absent in the majority of animals. Mean body length was 127cm and 106.2 cm for males and females, respectively. Heights at the tourniquet are around 87 cm in males and 68.2 cm in females, while heart girths reached 101 cm and 83cm in males and females, respectively.

Alpine population

Specific features distinguish this population. Hairs are of the close-cropped type, ears are trained and sometimes stalked, and the ibex form of horns is dominating. Wattles and beards are generally absent. The height at the tourniquet was 80 cm, the heart girth was more than 96.6 cm and body length was 117.5 cm in males; respective measures in females were 68, 76.7 and 104 cm.

Variation of quantitative variables

Identified breeds or populations, sex and age of the goat and locality (Table 5) were used to explain variation of quantitative variables (body length; wither height, thoracic perimeter, ears length, hairs length and the horns length). The coefficient of determination (R²) ranged from 28.6 % for ears length to 64% for hairs length. All variables varied with the breed (p< 0.05). The locality was also an important source of variation for most of the studied variables except for hairs and horns lengths.

Conclusion 

• Goat populations in the Tozeur region showed a high polymorphism. This polymorphism results from the introduction of exotic breeds and also mirrors breeding management in the region. Although local farmers distinguish only between two breeds Arbi and Cherki according to the unique criterion length of the hair, four distinct breeds Serti, Arbi, Damasquine and Alpine were found in the region. This work may constitute the basis for microsatellite studies to help implementing breeding strategies for the genetic improvement and conservation of Tunisian goat populations

References 

Cogovica 1987 Parts for goats allelic series visible traits other than colour. Tech. et Doc. Editions Lavoisier, Paris, pp 37-38.

Dolling C 1999 “standardized genetic nomenclature for cattle” CAB international. The genetics of cattle. Editors: R Fries and Ruwinsky. pp 657-665  

French H 1971 Observation sur la chèvre. Etude Agricoles, Ed. FAO, Rome n° 80, pp 75-91.

Jivotovski L 1985 “Génétique. Evolution et environnement” Editor MIR. Moscou. pp 122-150.

Lanari M R, Taddeo H, Domingo E, Pérez Centeno M and Gallo L 2003 Phenotypic differentiation of exterior traits in local Criollo Goat Population in Patagonia (Argentina). Arch. Tierz. 46, 347-356. http://www.inta.gov.ar/bariloche/info/documentos/animal/genetica/Cabras%20Criollas/Ct-433.pdf

Lauvergne J J 1986 Méthodologie pour l’étude des caprins Méditerranéens 1986. Les colloques de l’INRA, France, 47, pp 77-91

Lauvergne J J, Renieri C and Audiot A 1987 Estimating erosion in phenotypic variation in a French traditional goat population. Journal of Heredity 307-314

Ministry of agriculture and hydraulic resources 2006 General direction of studies and agricultural development, enquête sur les structures des exploitations agricoles, (2004-2005, Avril 2006).

Narjisse H, Nastis A, Rubino R, Santucci P and Steinbach J 1989 Méthodologie pour l’identification des systèmes d’élevage caprin: In Flamant J C et Morand-Fehr (editors) : Symposium « Philostios »: Agriculture; Programme de recherche Agrimed: L’évaluation des ovins et des caprins méditerranéens. Fonte-Boa. Portugal. 567 p.

Pattie WA and Restall BJ 1989 The inheritance of Cashmere in Australian Goats. 2. Genetic Parameters and Breeding Values. Livestock Production Science 21 : 251-261.

SAS User’s Guide 1989., Version 6.10 for Windows. SAS Inst. Inc., Cary, NC

Vivicorsi M P 1998 Contribution à l’étude de la sauvegarde des races domestiques menaces de disparition. L’exemple de la chèvre de Rove. Thèse de doctorat. Université Claude-Bernard Lyon I.  http://www3.vet-lyon.fr/bib/fondoc/th_sout/dl.php?file=1998lyon056.pdf

Received 6 May 2009; Accepted 26 August 2009; Published 1 October 2009

Go to top