ANIMAL GENETIC RESOURCES RESSOURCES GÉNÉTIQUES ANIMALES RECURSOS GENÉTICOS ANIMALES. an international journal. un journal international

Transcription

1 54 ANIMAL GENETIC RESOURCES 2014 ISSN an international journal RESSOURCES GÉNÉTIQUES ANIMALES un journal international RECURSOS GENÉTICOS ANIMALES una revista internacional

2 Animal Genetic Resources, 2014, 54, i. Food and Agriculture Organization of the United Nations, 2014 doi: /s Editorial Dear reader, The 54 th volume of our journal includes 20 papers describing different aspects of animal genetic resources and their management in Africa, Asia, Europe and Latin America. The current volume shows that our journal is truly international due to its international readership and authorship. Furthermore the papers presented cover a range of topics from characterization of breeds to the description of their production systems. Similar to the 53 rd issue, the high percentage of papers dealing with phenotypic and or genetic characterisation indicates the continuing importance of the first strategic priority area of the Global Plan of Action for Animal Genetic Resources 1. Animal genetic resources represent an important part of biodiversity. Therefore I would also like to draw your attention to the next edition of the Global Biodiversity Outlook, the flagship publication of the Convention on Biological Diversity 2. It is a periodic report that summarizes the latest data on the status and trends of biodiversity and draws conclusions relevant to the further implementation of the Convention. The reports are based on different types on information including national reports, scientific literature, information from the Biodiversity Indicators Partnership and supplementary studies. The fourth edition of the Global Biodiversity Outlook (GBO-4) 3 will be prepared to provide a midterm assessment towards the progress in the implementation of the Strategic Plan for Biodiversity and the Aichi Biodiversity Targets. 4 You might me be especially interested in Aichi Target 13 which refers directly to the genetic diversity of farmed and domesticated animals. The Commission on Genetic Resources for Food and Agriculture, at its Fourteenth Regular Session in April 2013, requested FAO to prepare The Second Report on the State of the World s Animal Genetic Resources for Food and Agriculture, as an update to the first report published in By 25 th of March 2014, 100 countries had submitted country reports to FAO and another 30 countries had indicated that they were in the final stages of preparing their reports. International organizations and regional focal points for the management of animal genetic resources 6 have also been invited to report on their activities supporting the implementation of the Global Plan of Action for Animal Genetic Resources. 7 The Second Report will be made available to the Eighth Session of the Intergovernmental Technical Working Group on Animal Genetic Resources for Food and Agriculture, which is scheduled to take place November 2014 in Rome. 8 The Working Group will also advise the Commission on whether the Global Plan of Action for Animal Genetic Resources needs to be updated. Please encourage your country s National Coordinator 9 to attend this important meeting. You might also consider providing him/her with your comments on the Second Report, which will be published online six weeks prior to the meeting at our webpage 10. Yours sincerely, Roswitha Baumung html Intergovernmental_process.html html i

3 Animal Genetic Resources, 2014, 54, iii. Food and Agriculture Organization of the United Nations, 2014 doi: /s Éditorial Cher lecteur, Le 54 ème volume de notre journal inclut 20 articles décrivant différents aspects des ressources zoogénétiques et leur gestion en Afrique, Asie, Europe et Amérique Latine. Le présent volume fait ressortir le caractère vraiment international de notre journal vue l internationalité des lecteurs et des auteurs. En outre, les articles présentés couvrent un large éventail de sujets allant de la caractérisation des races à la description de leurs systèmes de production. Comme dans le 53 ème volume, le haut pourcentage d articles traitant de la caractérisation phénotypique et génétique indique que le premier domaine prioritaire du Plan d Action Mondial pour les Ressources Zoogénétiques 1 demeure toujours important. Les ressources zoogénétiques représentent une part importante de la biodiversité. Ainsi, je voudrais aussi attirer votre attention sur la prochaine édition de Perspectives Mondiales de la Diversité Biologique, publication phare de la Convention sur la Diversité Biologique 2.Ils agit d un rapport périodique qui synthétise les plus récentes données sur l état et les tendances de la biodiversité et qui tire des conclusions pertinentes pour poursuivre la mise en œuvre de la Convention. Les rapports sont basés sur différents types d informations, y compris des rapports nationaux, de la littérature scientifique, des informations du Partenariat sur les Indicateurs de la Diversité Biologique et des études complémentaires. La quatrième édition de Perspectives Mondiales de la Diversité Biologique (GBO-4) 3 va être préparée dans le but de fournir une évaluation à moyen terme des progrès réalisés dans la mise en œuvre du Plan Stratégique pour la Biodiversité et des Objectifs d Aichi pour la Biodiversité. 4 Il se peut que le 13 ème Objectif d Aichi vous intéresse tout particulièrement vu qu il se rapporte directement à la diversité génétique des animaux domestiques. À sa Quatorzième Session Ordinaire en Avril 2013, la Commission des Ressources Génétiques pour l Alimentation et l Agriculture a demandé à la FAO d élaborer Le Deuxième Rapport sur l État des Ressources Zoogénétiques pour l Alimentation et l Agriculture dans le Monde, comme une mise à jour du premier rapport publié en Vers le 25 Mars 2014, 100 pays avaient soumis leurs rapports nationaux à la FAO et 30 autres pays avaient indiqué qu ils se trouvaient dans les dernières étapes de la préparation de leurs rapports. Des organisations internationales et les centres de coordination régionaux pour la gestion des ressources zoogénétiques 6 ont aussi été invités à informer sur leurs activités de soutien à la mise en application du Plan d Action Mondial pour les Ressources Zoogénétiques. 7 Le Deuxième Rapport sera mis à disposition à la Huitième Session du Groupe de Travail Technique Intergouvernemental sur les Ressources Zoogénétiques pour l Alimentation et l Agriculture, programmée pour le 26 au 28 Novembre 2014 à Rome. 8 Le Groupe de Travail fera aussi savoir à la Commission si, à son avis, le Plan d Action Mondial pour les Ressources Zoogénétiques doit être mis à jour. Veuillez encourager le Coordonnateur National de votre pays 9 à assister à cette importante réunion. Vous pourrez également transmettre à votre Coordonnateur vos commentaires sur le Deuxième Rapport, qui sera publié sur notre site web 10 six semaines avant la réunion. Cordialement, Roswitha Baumung html Intergovernmental_process.html 9 f227a3aeee1c3b28be6,contacts 10 html iii

4 Animal Genetic Resources, 2014, 54, iv. Food and Agriculture Organization of the United Nations, 2014 doi: /s Editorial Estimado lector, El 54 volumen de nuestra revista incluye 20 artículos que describen diferentes aspectos de los recursos zoogenéticos y su ordenación en África, Asia, Europa y América Latina. El presente volumen da muestras del carácter verdaderamente internacional de nuestra revista dada la internacionalidad del público lector y de los autores. Asimismo, los artículos presentados cubren una amplia variedad de temas, desde la caracterización de las razas a la descripción de sus sistemas de producción. Como ya ocurrió en la 53ª entrega, el elevado porcentaje de artículos que abordan la caracterización fenotípica y genética da señas de que sigue revistiendo importancia la primera área estratégica prioritaria del Plan de Acción Mundial sobre los Recursos Zoogenéticos 1. Los recursos zoogenéticos representan una parte importante de la biodiversidad. Así, me gustaría también atraer su atención hacia la próxima edición de la Perspectiva Mundial sobre la Biodiversidad, publicación insignia del Convenio sobre la Diversidad Biológica 2. Se trata de un informe periódico que resume la información más reciente sobre el estado y las tendencias de la biodiversidad, y que extrae conclusiones relevantes para seguir aplicando el Convenio. Los informes están basados en distintos tipos de información, entre los cuales informes nacionales, bibliografía científica, información de la Asociación de Indicadores de Biodiversidad y estudios complementarios. La cuarta edición de la Perspectiva Mundial sobre la Biodiversidad (GBO-4) 3 va a ser preparada con el fin de proporcionar una evaluación, en mitad de periodo, del grado de avance en la aplicación del Plan Estratégico para la Diversidad Biológica y de las Metas de Aichi para la Diversidad Biológica. 4 Tal vez le pueda resultar interesante la 13ª Meta de Aichi que alude directamente a la diversidad genética de los animales domésticos. La Comisión de Recursos Genéticos para la Alimentación y la Agricultura, en su Decimocuarta Reunión Ordinaria en abril de 2013, solicitó a la FAO la preparación de El Segundo Informe sobre la Situación de los Recursos Zoogenéticos Mundiales para la Alimentación y la Agricultura, como una actualización del primer informe publicadoen Hacia el 25 de marzo de 2014, 100 países habían enviado sus informes nacionales a la FAO y otros 30 países habían indicado que se encontraban en la fase final de preparación de sus informes. También se ha invitado a organizaciones internacionales y a los centros de coordinación regionales para la gestión de los recursos zoogenéticos 6 a informar sobre sus actividades de apoyo a la aplicación del Plan de Acción Mundial sobre los Recursos Zoogenéticos. 7 El Segundo Informe se hará disponible en la Octava Reunión del Grupo de Trabajo Técnico Intergubernamental sobre los Recursos Zoogenéticos para la Alimentación y la Agricultura, programada para los días de noviembre de 2014 en Roma. 8 El Grupo de Trabajo también notificará a la Comisión si, en su opinión, el Plan de Acción Mundial sobre los Recursos Zoogenéticos debe ser actualizado. Anime por favor al Coordinador Nacional de su país 9 a asistir a este importante encuentro. También podrá Usted hacer llegar a su Coordinador sus comentarios sobre el Segundo Informe, que será publicado en nuestra página web 10 seis semanas antes del encuentro. Atentamente, Roswitha Baumung Intergovernmental_process.html 9 b0aeecf1,contacts 10 html iv

5 Animal Genetic Resources, 2014, 54, 1 9. Food and Agriculture Organization of the United Nations, 2014 doi: /s x Phenotypic and morphological characterization and reproduction attributes of native pigs in Bangladesh C.H. Ritchil, M.M. Hossain and A.K.F.H. Bhuiyan Department of Animal Breeding and Genetics, Bangladesh Agricultural University, Mymensingh, Bangladesh Summary The study was undertaken for a period of 30 days to investigate the phenotypic, morphological and reproduction attributes of native pigs in 14 villages of Haluaghat, Durgapur and Nalitabari upazilas in Mymensingh, Netrokona and Sherpur districts, respectively. A total of 200 experimental animals were evaluated, of which 81 were male and 119 were female. Various parameters were evaluated separately at different age groups in males and females. A pre-tested questionnaire was used to record various phenotypic information. Body weight, body length, chest girth, length from ear to tail, height at withers, height at loin region, head length, width of head, ear length, hair length, tail length and number of teats were the morphometric traits recorded. The predominant coat colour of the pigs was black, with 95 percent of them having dark skin pigmentation. The majority of pigs possessed a long, straight head (75 percent) and the rest had a shape classified as a wide face. All of them had erect ears with upward orientation, whereas 98.5 percent had thin, straight tails. The pigs with stocky body shape (27.5 percent) showed larger body measurements than animals with angular body shape (72.5 percent). The mean values along with the Standard Error for traits were statistically analysed. Three main independent variables (i.e. upazila, age and sex) were considered for an univariate analysis of variance and the correlations among traits were measured. The average, minimum and maximum values for age at first farrowing, farrowing interval, litter size, gestation period and weaning period were also measured. Keywords: Dome pig, Wakmandi pig, phenotypic traits, morphometric traits, reproductive parameters Résumé L étude s est étendue sur une période de 30 jours dans le but de caractériser d un point de vue phénotypique, morphologique et reproductif les porcs autochtones de 14 villages des upazilas de Haluaghat, Durgapur et Nalitabari, dans les districts de Mymensingh, Netrokona et Sherpur, respectivement. Un total de 200 animaux expérimentaux (81 mâles et 119 femelles) a été évalué. Plusieurs paramètres ont été évalués séparément, à différents âges, aussi bien pour les mâles que pour les femelles. Un questionnaire prétesté a été utilisé pour rassembler les informations phénotypiques. Les traits morphométriques retenus ont été le poids et la longueur du corps, la circonférence thoracique, la distance des oreilles à la queue, la hauteur au garrot, la hauteur à la région lombaire, la longueur et la largeur de la tête, la longueur des oreilles, la longueur des poils, la longueur de la queue et le nombre de mamelles. La couleur de robe prédominante a été le noir, avec 95 pour cent des porcs ayant une pigmentation cutanée foncée. La plupart des porcs ont eu une tête longue à profil droit (75 pour cent) alors que pour le reste, la tête avait une forme connue sous le nom de visage large. Dans tous les cas, les oreilles étaient dressées vers le haut et pour le 98,5 pour cent des porcs, la queue était mince et droite. Les porcs au corps râblé (27,5 pour cent) ont présenté des mesures corporelles plus grandes que celles des animaux au corps anguleux (72,5 pour cent). Les valeurs moyennes, ainsi que l erreur standard, ont été analysées statistiquement. Trois variables indépendantes principales (i.e. l upazila, l âge et le sexe) ont été considérées pour l ANOVA et les corrélations entre les paramètres ont été mesurées. Les valeurs moyenne, minimale et maximale pour l âge à la première mise-bas, l intervalle entre mises-bas, la taille de la portée et la durée de la gestation et de la lactation ont aussi été déterminées. Mots-clés: porc dome, porc wakmandi, traits phénotypiques, traits morphométriques, paramètres reproductifs Resumen El estudio se extendió sobre un periodo de 30 días con el fin de investigar los atributos fenotípicos, morfológicos y reproductivos de los cerdos autóctonos de 14 pueblos de las upazilas de Haluaghat, Durgapur y Nalitabari, en los distritos de Mymensingh, Netrokona y Sherpur, respectivamente. Se evaluó un total de 200 animales experimentales, de los cuales 81 eran machos y 119 eran hembras. Se evaluaron varios parámetros de forma separada, a distintas edades, en los grupos de machos y hembras. Se utilizó un cuestionario probado para recopilar la información fenotípica. Los rasgos morfométricos registrados fueron el peso y la longitud corporal, la circunferencia torácica, la distancia de las orejas a la cola, la altura a la cruz, la altura a la región lumbar, la longitud y la anchura de la cabeza, la longitud de las orejas, la longitud del pelo, la longitud de la cola y el número de mamas. El color de capa predominante de los cerdos fue el negro; el 95 por ciento de los ejemplares tenía pigmentación cutánea oscura. La mayoría de los cerdos tenía una cabeza larga, con perfil recto (75 por ciento), y el resto presentaba una forma conocida como cara ancha. Todos los ejemplares tenían orejas erectas en posición vertical y la cola era fina y recta en el 98,5 por ciento de los casos. Los cerdos con cuerpo fornido (27,5 por ciento) presentaron medidas corporales mayores que las de los animales con cuerpo anguloso (72,5 por ciento). Se analizaron estadísticamente los valores medios, junto con el error estándar, de los distintos parámetros. Se consideraron tres variables independientes principales (i.e. upazila, edad y sexo) para el ANOVA y se midieron las correlaciones entre los parámetros. También se Correspondence to: C.H. Ritchil, Department of Animal Breeding and Genetics, Bangladesh Agricultural University, Mymensingh, Bangladesh. ritchil.himel@gmail.com 1

6 2 Ritchil et al. determinaron los valores medio, mínimo y máximo para la edad al primer parto, el intervalo entre partos, el tamaño de la camada y la duración de la gestación y la lactación. Palabras clave: cerdo dome, cerdo wakmandi, rasgos fenotípicos, rasgos morfométricos, parámetros reproductivos Submitted 25 March 2013; accepted 30 January 2014 Introduction More than 45 different ethnic groups live in Bangladesh. As a group, they are called the indigenous people and they raise pigs for a contribution to their livelihoods and are the main consumers of pork and rearers of pig (Sus crofa domesticus) in Bangladesh. In addition, the Christians and some Hindus also consume pork in large amounts. There is also a community of nomadic people called the Dome, who rear pigs and sell piglets to the indigenous people. They are the main source of pigs for those people. The pig bought from Dome is called Dome s pig. Another type of native pig is called the Wakmandi. A distinctive trait of this pig is a drooping mid-section, which differentiates it from the Dome. About 95 percent of the pigs in Bangladesh are of these native types (Dome and Wakmandi) and the remaining 5 percent are exotic or crosses between native and exotic. There has been no thorough investigation carried out to characterize these native pigs in Bangladesh. In this context, this study was formulated to evaluate the morphological, phenotypic and reproduction attributes of native pigs, specifically in the rural areas of Haluaghat, Durgapur and Nalitabari upazilas of Bangladesh. Materials and methods Study area The study was undertaken at four randomly selected villages of the Haluaghat Upazila located at N and E in Mymensingh district (Figure 1), seven villages of the Durgapur Upazila located at N and E in Netrokona district (Figure 2) and three villages of the Nalitabari Upazila located at N and E in Sherpur district (Figure 3). The villages were selected because they were known to have a reasonably high density of pig populations. Before collection of data, a questionnaire (Appendix 1) was prepared in accordance with objectives of the study. Later on, the questionnaire was validated against field conditions. Study population All the experimental animals were of the native type (i.e. Dome or Wakmandi) pigs. All pigs were reared by the Garo(s) the indigenous people in the studied areas. A total of 200 household pig owners from the aforementioned sites were selected randomly and interviewed relevant for the study. A total of 200 experimental animals were evaluated, of which 81 were male and 119 were female pigs. One pig from each household was selected randomly. Data collection The study was conducted from 1 September 2012 to 30 September The phenotypic traits such as coat colour, skin pigmentation, head shape, ear direction, ear orientation, tail shape and body shape were recorded based on subjective visual observation. Body weight of piglets was recorded by using a weighing scale, whereas body weights of boars and gilts were estimated by using the following formula:- Chest girth 2 (m) length from ear to tail (m) 69.3 = body weight of the pig (kg) where, M = meter 69.3 = constant kg = kilogram Source: The Old Farmer s Almanac Figure 1. Map of field unit at Haluaghat upazila in Mymensingh district.

7 Phenotypic and morphological characterization and reproduction attributes of native pigs in Bangladesh 3 Figure 2. Map of field unit at Durgapur upazila in Netrokona district. Figure 3. Map of field unit in Nalitabari upazila in Sherpur district. Body length was determined by measuring the distance between points of shoulder to the pin bone. Chest girth was considered as the largest circumference of the body immediately behind the shoulder. The length from ear to tail was measured from the base of the ear to the base of the tail. Height at wither was considered as vertical distance between the ground and the point of wither. Height at loin region was considered as vertical distance between the surface and the loin region. Head length was measured from the distance between snout and forehead of pig, while head width was the distance between the eyes. The distance between the base and the tip of the pinna was considered as the ear length. Shoulder hairs were used to determine the hair length. The distance between the base and the end of the tail was constituted the tail length. The number of pairs of teats was recorded by visual observation. The litter size was based on the number born alive. Gestation and weaning period were also estimated. Statistical analysis The accumulated data were arranged by using Microsoft Excel 2007 and these data were compiled and analyzed by using Statistical Package for Social Science (SPSS) v Univariate analysis of variance (ANOVA) was carried out with upazila, sex, age group and sex*age as explanatory variables. The raw and residual correlations among traits were also calculated. Results and discussion 1. Phenotypic characterization Coat characteristics The observed combinations of coat colour of the native pigs were black (95 percent), black and white (2.5 percent), greyish-black (2 percent) and brownish (0.5 percent) (Figure 4). The coat colour of native pig had been recorded as black, white, black with white legs, white with black spot, black with white belly and brown in a previous study conducted by Yaetsu et al. (1987) in Bangladesh, which is similar to the present study. Most of the native pigs in the study areas had dark skin pigmentation (95 percent). Head and ear characteristics The shape of the head was long and straight in most of the native pigs (75 percent) and the rest were classified as wide face (25 percent) (Figure 5). Yaetsu et al. (1987) conducted an experiment on native pigs in different regions of Bangladesh and reported that a straight face with a pointed snout and erect ears was common for native pig. The present study showed that the native pigs had erect ears that pointed backwards (100 percent) (Figure 5). Body and tail shapes Native pigs were found to have either angular or stocky body shapes where 72.5 percent of them belonged to

8 4 Ritchil et al. Figure 4. Coat characteristics of native pig. angular body type and rest were stocky (Figure 6). Thin, straight as well as curled tails were observed among the native pigs. It was found that 98.5 percent of the pigs had a thin, straight tail. The curled tails (1.5 percent) were generally thin and showed an upward curl forming one circle. Yaetsu et al. (1987) conducted an experiment on native pigs in different regions of Bangladesh and reported that some pigs with somewhat concave back and pendulous belly were found. They also reported that the straight tail was common for all native pigs. Figure 5. Head and ear characteristics of native pig.

9 Phenotypic and morphological characterization and reproduction attributes of native pigs in Bangladesh 5 Figure 6. Body shape of the native pig. 2. Morphological characterization For morphological characterization, body measurements of 81 male and 119 female pigs were taken. Means along with standard errors of means (SEM) of different morphometric characters for pigs are presented in Tables 1 3. Interactions between dependent and independent variables are presented in Table 4. The raw and residual correlations among the studied traits are presented at Table 5. Not surprisingly, the morphological traits of native pigs were found to increase (P < 0.001) with the advancement Table 1. Morphological characteristic of pigs affected by age (Mean ± SEM). Parameter Body weight (kg) Body length (cm) Chest girth (cm) Ear to tail length (cm) Height at wither (cm) Height at loin (cm) Head length (cm) Head width (cm) Ear length (cm) Hair length (cm) Tail length (cm) Age months months months months (n = 56) (n = 105) (n = 28) (n = 11) 16.5 a ± b ± c ± d ± a ± b ± c ± d ± a ± b ± c ± d ± a ± b ± c ± d ± a ± b ± c ± d ± a ± b ± c ± d ± a ± b ± c ± d ± a ± b ± c ± d ± a ± b ± c ± d ± a ± b ± c ± d ± a ± b ± c ± d ± 1.5 Different superscripts indicate that values are significantly different at 0.1% level of probability (P < 0.001); n = number of animals. of age (Table 1). In general, adult (18 24 months) males tended to be larger than females, whereas no clear pattern was observed at younger ages. Babu et al. (2004) conducted an experiment in the rural areas of India and reported that the average body weight of Large White pigs at puberty was 82.3 ± 1.8 kg (mean ± SE), which is similar to the result of present study. The adult body length (7 24 months) ranged from cm for males and cm for females. Subalini, Silva and Demetawewa (2010) reported that the average body length of Sri Lankan village adult male and female pig was 75.1 ± 1.9 and 69.9 ± 2.4 cm, respectively, which is 15 cm greater than in the present study. The average chest girth for the adult male and female pigs was 82.8 ± 1.7 and 81.3 ± 1.1 cm, respectively. The average length from the base of the ear to base of the tail was higher for females at 0 6 and months of age. The height at wither, height at loin, head length, head width, ear length and tail length were higher for females than males at 0 6 months of age. A total of only 20 male and 36 female piglets were measured at 0 6 months of age. Hence, the small sample size might be responsible for this result. Yaetsu et al. (1987) conducted an experiment on native pigs in different regions of Bangladesh and reported that the average heights at wither at months of age in males and females were 56.0 ± 0.0 and 52.8 ± 4.8 cm, respectively, which is lower than the present study. The average pairs of teat of the female pig at 0 6, 7 12, and months of age was found to be 4.8 ± 0.1, 5.0 ± 0.0, 5.1 ± 0.1 and 6.0 ± 0.0 pairs, respectively. Yaetsu et al. (1987) conducted an experiment on native pigs in different regions of Bangladesh and reported that the teat pattern of Bangladeshi pigs ranged from four to six pairs. A pattern of five pairs was most frequent in all the populations examined, which is similar to the result of the present study. The correlations among traits were also calculated (Table 5). The correlations were all positive and were significant at the 0.01 level (two-tailed). However, the present study failed to record a significant difference (P < 0.05) between the two sexes (Table 2) except for chest girth and ear length at and months of age group,

10 6 Ritchil et al. Table 2. Morphological characteristics of pigs as a function of sex (Mean ± SEM). Parameter Sex Male Female Significance Male Female Significance Male Female Significance Male Female Significance (n = 20) (n =36) (P-value) (n = 45) (n = 60) (P-value) (n =10) (n = 18) (P-value) (n =6) (n =5) (P-value) At 0 6 months of age At 7 12 months of age At months of age At months of age Body weight (kg) 16.1 a ± a ± a ± a ± a ± a ± a ± a ± Body length (cm) 35.2 a ± a ± a ± a ± a ± a ± a ± a ± Chest girth (cm) 61.8 a ± a ± a ± a ± a ± a ± a ± b ± Ear to tail length (cm) 56.3 a ± a ± a ± a ± a ± a ± a ± a ± Height at wither (cm) 39.4 a ± a ± a ± a ± a ± a ± a ± a ± Height at loin (cm) 40.6 a ± a ± a ± a ± a ± a ± a ± a ± Head length (cm) 19.2 a ± a ± a ± a ± a ± a ± a ± a ± Head width (cm) 11.3 a ± a ± a ± a ± a ± a ± a ± a ± Ear length (cm) 7.5 a ± a ± a ± a ± a ± b ± a ± a ± Hair length (cm) 6.3 a ± a ± a ± a ± a ± a ± a ± a ± Tail length (cm) 17.5 a ± a ± a ± a ± a ± a ± a ± a ± Pair of teat 4.8 ± ± ± ± 0.0 a,b = Means from the same age group that differ significantly between sexes have different superscripts; n = number of animals. respectively. Subalini, Silva and Demetawewa. (2010) conducted an experiment on village pigs in Sri Lanka and found no significant difference (P < 0.05) between two sexes. Sudhakar and Gaur (2006) also found that effect of sex was not significant on all body weights and measurements in Indian native pigs, except for weight at birth. The diet for male and females was almost same though. At the present study the differences between two populations (Dome and Wakmandi) was not done for the characteristics measured as the population of Wakmandi was very little. Morphological characteristics of mature pigs affected by upazila (area) are presented in Table 3. Area contributed significant differences (P < 0.05) between males and females in case of parameters such as wither height, loin height, head length, ear length and tail length. 3. Reproduction attributes In present study, the average age at first farrowing was found to be ± 0.08 months (Table 6). The study conducted by Hossain et al. (2011) using surveyed data on indigenous pigs in hilly areas of Chittagong in Bangladesh, reported that the average ages at sexual maturity in Bangladeshi boar and sow were 8 and 6 months, respectively. Naskar et al. (2007) conducted an experiment on crossbred pigs under the hilly conditions of Meghalaya, India and reported that the average age at first farrowing of pig was 15.3 ± 1.2 months. Subalini, Silva and Demetawewa (2010) conducted an experiment on village pig in Sri Lanka and reported that the average age at first farrowing was 9.50 ± 2.61 months. The average farrowing interval of native pigs was found to be 6.09 ± 0.02 months. Babu et al. (2004) conducted an experiment in the rural areas of India on Large White pig and reported that the average farrowing interval was ± 1.03 days. Subalini, Silva and Demetawewa (2010) conducted an experiment on village pig in Sri Lanka and reported that the average farrowing interval was 8.91 ± 2.49 months. The average litter size of native pigs in this study was 6.71 ± Hossain et al. (2011) reported that the average litter size of indigenous pig at hilly region of Bangladesh was 9.5 ± Prakash et al. (2008) reported that the average litter size of Indian indigenous pig was 6.78 ± Nath and Deka (2003), conducted an experiment on growth performance of large black pig in Assam, India and reported that the average litter size at birth and weaning were 9.70 ± 0.29 and 8.40 ± 0.32, respectively. The average gestation period and the average weaning period of these native pigs were ± 0.50 and ± 0.37 days, respectively. Rajiv and Pandey (2000) conducted an experiment on the economics of pig rearing in Haryana, India and reported that the average gestation period was days. Hossain et al. (2011) reported that the average weaning period of indigenous piglets in Bangladesh was 42 days.

11 Phenotypic and morphological characterization and reproduction attributes of native pigs in Bangladesh 7 Table 3. Morphological characteristics of mature pigs as a function of upazila (Mean ± SEM). Parameter Upazila Haluaghat Durgapur Nalitabari Male Female Significance Male Female Significance Male Female Significance (n = 15) (n = 26) (P-value) (n = 34) (n = 19) (P-value) (n = 12) (n = 38) (P-value) Body weight (kg) 37.1 a ± a ± a ± a ± a ± a ± Body length (cm) 48.5 a ± a ± a ± a ± a ± a ± Chest girth (cm) 82.4 a ± a ± a ± a ± a ± a ± Ear to tail length (cm) 77.0 a ± a ± a ± a ± a ± b ± Height at wither (cm) 51.6 a ± a ± a ± b ± a ± b ± Height at loin (cm) 53.8 a ± a ± a ± b ± a ± b ± Head length (cm) 22.2 a ± b ± a ± a ± a ± b ± Head width (cm) 14.5 a ± a ± a ± a ± a ± a ± Ear length (cm) 8.5 a ± a ± a ± b ± a ± a ± Hair length (cm) 7.6 a ± a ± a ± b ± a ± a ± Tail length (cm) 25.3 a ± a ± a ± b ± a ± b ± Pair of teat 5.0 ± ± ± 0.0 a,b = Means from the same age group that differ significantly between sexes have different superscripts; n = number of animals. 4. Management system of native pigs Girth tethering was found to be the most popular and widely used housing system for pigs in this study. Pigs were fed thrice a day. Pig owners used to supply boiled rice, rice polish, rice bran, etc. as conventional feeds. Some unconventional feeds were also fed to them. Tube-well or natural reservoir was the source of water for pigs. According to Ritchil et al. (2013), pigs in Bangladesh are reared mostly by poor and landless people (52 percent). The majorities of farmers are housewife (55 percent) or day labour (17 percent). The literacy level is satisfactory. Rearing systems are different and the mean figure is 97 percent for girth tethering. The sanitation condition is not satisfactory. The pig owners use to supply boiled rice, rice polish, rice bran, rice husk, fermented rice, etc. Table 4. Interactions between dependent and independent variables. Parameter Significance (P-value) Error (mean square) Area Sex Age Sex*age Body weight (kg) *(0.030) Body length (cm) NS (0.287) Chest girth (cm) NS (0.146) Ear to tail length NS (0.452) (cm) Height at wither NS (0.420) (cm) Height at loin NS (0.578) (cm) Head length (cm) NS (0.112) Head width (cm) NS (0.586) 4.33 Ear length (cm) NS (0.808) 4.21 Hair length (cm) NS (0.707) 3.74 Tail length (cm) NS (0.436) Pair of teat ***(0.000) 0.03 Non-significant (NS) = P > 0.05; * = P < 0.05; *** = P < and some unconventional feeds such as cauliflowers, boiled arum, grass or other vegetables for pigs. Pigs are fed thrice a day. Most prevalent diseases are diarrhoea, coccidiosis, pneumonia and haemorrhagic septicaemia. Easy management, adaptability, cost-effectiveness and disease tolerance are the main advantages of native pig farming while malnutrition, lack of scientific knowledge, lack of veterinary and extension services are the major constraints. Ritchil et al. (2013) also found that rearing native pig is very easy for those people who rear and consume it. They believe that it is very much cost effective, available, adaptable and merely affected by diseases. The socio-economic status of the pig rearers is appreciable. Although many constraints are taking part in the management of pig, the meat from pig is the main source of protein for those people who rear and consume it. Therefore, according to Ritchil et al. (2013) lack of mature breeding boars, technical knowhow and preventive healthcare services were noticed which need to be extended among the pig rearers for better livelihood and welfare of their keepers. Hence, the observed body weight and reproduction attributes of the native pigs may serve as a base for designing any pig improvement programme in Bangladesh. Conclusion Native pigs in Bangladesh showed diversity in their phenotype and morphology and positive reproductive performance, which indicates their potential for economic sustainability through purebred genetic improvement. However, these native pigs are usually raised in conditions with less than optimal husbandry and hence, this aspect needs immediate attention to ensure their conservation through utilization for the food and livelihood of indigenous, Christian and some Hindu people in Bangladesh. This

12 8 Ritchil et al. Table 5. Raw (on the diagonal) and residual (below the diagonal) correlations among traits. Body weight Body length Chest girth Ear to tail length Height at wither Height at loin Head length Head width Ear length Hair length Tail length Body weight Body length Chest girth Ear to tail length Height at wither Height at loin Head length Head width Ear length Hair length Tail length ** 0.96** 0.92** 0.81** 0.81** 0.65** 0.75** 0.32** 0.50** 0.57** ** 0.80** 0.69** 0.69** 0.59** 0.67** 0.22** 0.42** 0.49** ** 0.83** 0.84** 0.66** 0.73** 0.35** 0.54** 0.60** ** 0.82** 0.64** 0.71** 0.29** 0.54** 0.62** ** 0.59** 0.69** 0.37** 0.43** 0.62** ** 0.69** 0.42** 0.45** 0.63** ** 0.30** 0.49** 0.63** ** 0.40** 0.59** ** 0.38** ** **Correlation is significant at the 0.01 level (two-tailed). Residuals are computed between observed and reproduced correlations. There are 30 (45.0%) non-redundant residuals with absolute values >0.05. Table 6. Reproductive performance of native pig. Parameter Mean ± SEM can be done through a well-designed national pig improvement programme, which may include training of the pig rearers on general pig husbandry, supply of preventive vaccines and production and dissemination of quality (heavier but prolific) boars for use in the subsistence farms by the government, non-government organization or organized pig farmers of the country. References Minimum Maximum Age at first farrowing ± (months) Farrowing interval (months) 6.09 ± Litter size (Number) 6.71 ± Gestation period (months) 3.8 ± Weaning period (months) 1.5 ± Babu, G.N., Naidu, K.V., Rao, A.S. & Vijay, S Certain reproductive parameters in Large White pigs maintained with garbage feeding in rural areas. Indian J. Anim. Sci., 74(4): Hossain, M.E., Chakma, S., Khatun, M.M., Hasanuzzaman, M., Miah, M.Y. & Biswas, M.A.A Production systems of swine in the rural areas of Rangamati and Khagrachari districts of Bangladesh. Bangladesh J. Anim. Sci., 40(1 2): Naskar, S., Khan, M.H., Anubrata, D. & Bordoloi, R.K Performance of T & D (Tamworth Desi) crossbred pigs under hilly condition of Meghalaya. Environ. Ecol., 25(2): Nath, D.R. & Deka, D Litter trait and preweaning growth performance of large black pig in Assam. Indian Vet. J., 80(3): Prakash, M.G., Ravi, A., Kumari, B.P. & Srinivas Rao, D Reproductive and productive performance of crossbred Pigs. Indian J. Anim. Sci., 78: Rajiv, J. & Pandey, U.K. (2000). Economics of pig rearing in Haryana. Indian J. Anim. Sci., 70(12): Ritchil, C.H., Faruque, M.O., Tabassum, F., Hossain, M.M. & Bhuiyan, A.K.F.H The socio-economic status of pig rearers and the management system of native pigs in Bangladesh. Indian J. Anim. Sci., 83(11): , November Subalini, E., Silva, G.L.L.P. & Demetawewa, C.M.B Phenotypic characterization and production performance of village Pigs in Sri Lanka. Trop. Agric. Res., 21(2): Sudhakar, K. & Gaur, K Pre-weaning growth in indigenous pigs of eastern region. Indian J. Anim. Sci., 3(4): Website: Yaetsu, K., Takashi, A., Ikuo, O., Katuaki, O., Takao, N., Yoshizane, M., Hasnath, M.A., Mostofa, K.G., Faruque, O.M. & Majid, M.A Morphological studies of native pigs in Bangladesh. Genet. Stud. Breed Differ. Native Domest. Anim. Bangladesh, 2:47 58.

13 Phenotypic and morphological characterization and reproduction attributes of native pigs in Bangladesh 9 Appendix Appendix 1. An interview schedule for baseline survey on native pig farming in Bangladesh Sl. No Primary Information 4. Production & Reproduction Parameters Age at first farrowing Farrowing interval Litter size Gestation period Weaning period Time Name : Address : Father s Name/Husband s Name : Mother s Name : Age : Household land : Land for cultivation : Experience in pig farming : Family Member : Primary Occupation : Secondary Occupation : Level of education : Mobile No. : Monthly Income (Family) : GPS Location : Alt.: Lat.: 2. Type of pig Type Breed and no. Age Milking Pregnant Sow Gilt Boar (Castrated) Boar (Not Castrated) Piglet (male/ female) Pure breed Crossbred Dome Native Wakmandi 5. Pig morphology information Piglet Character Male Female Gilt Boar Sow Body weight (kg) Body length (cm) Chest girth (cm) Length from ear to tail (cm) Height at withers (cm) Height at loin region (cm) Head length (cm) Head width (cm) Ear length (cm) Hair length (cm) Tail length (cm) Pair of teats (number) 6. Phenotypic characters Trait Male Female Coat colour Skin pigmentation Head shape Ear direction Ear orientation Tail shape Body shape Date: Name & Signature of enumerator 3. Feed Age class Feed type Purchased feed Amount feed (kg/day) Water (litre/day) Boar Gilt Sow Piglet (male/ female)

14 Animal Genetic Resources, 2014, 54, Food and Agriculture Organization of the United Nations, 2014 doi: /s Caractéristiques phénotypiques de la chèvre du sahel au Niger par analyse des indices de primarité et des paramètres qualitatifs M. Mani 1, H. Marichatou 1, M. Issa 2, I. Chaïbou 3, A. Sow 4, M. Chaïbou 1 and J.G. Sawadogo 4 1 Faculté d Agronomie, Université Abdou Moumouni, BP 10960, Niamey, Niger; 2 Faculté des Sciences et Techniques, Université Abdou Moumouni, BP 10662, Niamey, Niger; 3 Faculté d Agronomie, Université de Maradi, Niger; 4 Ecole Inter-Etats des Sciences et Médecines Vétérinaires, Dakar, Sénégal Résumé Au Niger, le troupeau caprin est composé de trois races dont deux majoritaires: la chèvre rousse de Maradi (17,5%) qui a fait l objet de nombreuses investigations et la chèvre du sahel (80%) qui est peu ou pas connue. Ainsi, afin d établir les caractéristiques phénotypiques de la chèvre du sahel au Niger, des investigations sur les paramètres qualitatifs (port des appendices et les patrons pigmentaires) par observation directe des animaux ont été conduites dans la partie Nord-Ouest du pays (régions de Tillabéry, Tahoua et Niamey). Sur la base de l adhésion volontaire des éleveurs, 443 caprins (77% de femelles, 39% ayant 4 paires d incisives permanentes) répartis dans 145 exploitations et 60 sites (7,38 ± 6,87 têtes/site) ont été caractérisés. L analyse des indices de primarité et l analyse multivariée (ACM) font ressortir qu il n existe pas de différence significative (P < 0,05) entre sexe pour tous les paramètres qualitatifs à l exception de la présence de la barbiche (mâle 55%, femelle 41%). La distribution des paramètres est significativement différente selon les régions à l exception du type des cornes (majoritairement ibex dans toutes les régions). En effet, les oreilles sont majoritairement tombantes à Tillabéry, pédonculées à Tahoua et Niamey et les patrons pigmentaires sont dominés par eumélanique noir à Tahoua, eumélanique chocolat à Niamey et phaeomélanique à Tillabéry. En définitif, les caprins des trois régions sont significativement différents sur le plan phénotypique qualitatif. Mais, dans l ensemble, cette population caprine présente les caractéristiques d une population primaire, ce qui offre de grandes possibilités d amélioration génétique. Mots-clés: caractéristiques phénotypiques, qualitatifs, chèvre du sahel, Niger Summary In Niger, there are three goat breeds including two main ones: the Maradi red goat (17.5%) which has been the subject of numerous investigations and the Sahel goat (80%) which is little studied or unknown. Thus, to establish the phenotypic characteristics of the Sahel goat breed in Niger, investigations into qualitative parameters (appendices and pigmentation) by direct observation of animals were conducted in the northwestern part of the country (Tillabery, Tahoua and Niamey regions). Based on the voluntary participation of farmers, 443 goats (77% female, 39% with 4 pairs of permanent incisors and therefore considered adult) in 145 farms and 60 sites (7,38 ± 6,87 animals/site) were characterized. Analysis of primarity indices and multivariate analysis (ACM) emphasizes that there is no significant difference (P < 0.05) between sexes for all quality parameters except the carrying of a beard (male 55%, female 41%). The parameters distribution is significantly different between regions except the type of horn (mostly ibex in all regions). Indeed, ears are mostly drooping in Tillabéry, stalked in Tahoua and Niamey and pigmentation is dominated by black eumélanique in Tahoua, chocolate eumélanique in Niamey and phaeomélanique in Tillabéry. Goats of the three regions are significantly different based on phenotypic qualitative parameters. But, generally, the goat study population has the characteristics of a primary population which offers great opportunities for genetic improvement. Keywords: phenotypical features, qualitative features, Sahel goat, Niger Resumen En Níger, la cabaña caprina se compone de tres razas, de las cuales dos son mayoritarias: la cabra roja de Maradi (17,5 por ciento), que ha sido objeto de muchas investigaciones, y la cabra del Sahel (80 por ciento), de la cual se sabe poco o nada. Así, con el fin de caracterizar fenotípicamente la cabra del Sahel en Níger, se han investigado parámetros cualitativos (apéndices y patrones de pigmentación), por observación directa de los animales, en la parte Noroeste del país (regiones de Tillabéry, Tahoua y Niamey). Contando con la participación voluntaria de los ganaderos, se caracterizaron 443 ejemplares de ganado caprino (77 por ciento hembras, 39 por ciento con 4 pares de incisivos permanentes) repartidos entre 145 explotaciones y 60 emplazamientos (7,38 ± 6,87 cabezas/ emplazamiento). El análisis de los índices de primariedad y el análisis multivariante (ACM) muestran que no existen diferencias significativas (P < 0,05) entre los sexos para ninguno de los parámetros cualitativos, exceptuando la presencia de barba (machos 55 por ciento, hembras 41 por ciento). Los parámetros estudiados, con la excepción del tipo de cornamenta (principalmente de tipo ibex Faculté d Agronomie, Université Abdou Moumouni, BP 10960, Niamey, Niger. Tel: (00227) Fax: (00227) maricha@refer.ne / marimani_m@yahoo.fr 11

15 12 Mani et al. en todas las regiones), se distribuyen de forma estadísticamente diferente según las regiones. Así, las orejas están mayoritariamente caídas en Tillabéry, son pedunculadas en Tahoua y Niamey, y los patrones de pigmentación dominantes son de diseño eumelánico negro en Tahoua, eumelánico chocolate en Niamey y feomelánico en Tillabéry. En definitiva, existen diferencias significativas, en el plano fenotípico cualitativo, entre la ganadería caprina de las tres regiones. Sin embargo, esta población caprina presenta, en conjunto, características de una población primaria, lo cual ofrece grandes posibilidades de mejora genética. Palabras clave: Palabras clave: características fenotípicas, cualitativo, cabra del Sahel, Níger Soumis: 27 Le juin 2013; admis: 16 Le décembre 2013 Introduction Au Niger, les caprins représentent 36% du cheptel (Niger, 2007) et occupent une place de choix dans la vie socioéconomique et culturelle de la population. Ils constituent une spéculation importante en matière de commercialisation de bétail sur pied, mais aussi de production. C est également une source principale de protéines animales en milieu rural. L espèce caprine se présente sous trois races au Niger: la chèvre du sahel (79,54%) qui est présente dans toutes les régions du pays, la chèvre rousse (17,48%) prédominante dans la bande sud du pays et la chèvre naine d Afrique de l Ouest (chèvre Djallonké) dite localement chèvre «gourma» (2,98%) qui est présente dans la partie ouest de la région de Tillabéry à la frontière avec le Burkina Faso (Rhissa, 2010). Comparativement à la chèvre rousse de Maradi qui a fait l objet de nombreuses investigations sur les plans ethnologique et zootechnique (Robinet, 1967; Marichatou et al.,2002; Hamidou,1995; Saadou, 2005; Mani, 2009; etc.), il existe peu ou pas de données scientifiques sur la chèvre du sahel, malgré son importance numérique et son omniprésence dans toutes les zones climatiques du Niger. Cette étude vise à établir les caractéristiques phénotypiques qualitatives de la chèvre du sahel au Niger. En effet, si la caractérisation phénotypique vise à identifier et établir la diversité inter et intra races à travers des traits physiques, les paramètres qualitatifs, partie intégrante de cette caractérisation phénotypique, renseignent autrement sur les caractères d adaptation des races à leur milieu d élevage mais aussi, peuvent traduire les préférences des acteurs (éleveurs et consommateurs) ou des aspects culturels (FAO, 2012). Matériel et Méthodes Echantillon d étude L étude a été conduite dans la partie Nord-Ouest du Niger, notamment, les régions de Tahoua (départements de Tahoua, Abalak et Tchintabaraden), Tillabéry (département de Téra et Ouallam) et la Commune Urbaine de Niamey (figure 1) de juillet 2011 à mai Figure 1. Zone d étude

16 Caractéristiques phénotypiques de la chèvre du sahel 13 Tableau 1. Nombre de sites et de caprins par région de l échantillon d étude Régions Tahoua Tillabéry Niamey Total N sites (*) N caprins Moyennes/site 6,87 ± 2,92 13,22 ± 8,13 3,78 ± 4 7,38 ± 6,67 Maximum/site Minimum/site N = nombre (*) Site = village ou quartier Les échantillons de sites et d individus ont été calculés à partir de la formule de proportion suivante: n = (Z/e) 2 p(1 p) n est la taille de l échantillon; Z est le niveau de confiance, en général, Z = 1,96 à un intervalle de confiance de 95%; P est le niveau initial des indicateurs; e est la marge d erreur. Du fait de l importance de la chèvre du sahel dans la zone d étude, les valeurs de P et e choisis pour les calculs du nombre de sites et d individus sont respectivement de 80% et 10% et 50% et 5%. En outre, s agissant de l échantillon d individus, un forfait supplémentaire de 15% de la valeur calculée a été ajouté. En raison du manque de données sur les effectifs caprins au niveau villageois d une part et en fonction du mode d élevagedanslazoneéchantillond autre part, le nombre d individus dans les sites n a pas été pondéré en fonction de la densité des caprins. Au total 443 caprins (77% des femelles) appartenant à 145 exploitations sur 60 sites choisis au hasard en fonction de l adhésion volontaire des éleveurs ont été caractérisés. Le tableau 1 présente le nombre de sites et de caprins par région de l échantillon d étude. Selon l âge, apprécié à partir de l échelle de dentition (Hamito, 2009), l échantillon est composé de 22,1% de caprins de 0 1 an; 16,5% de 1 2 ans; 22,3% de 2 3 ans; 18% de 3 4 ans et 20,3% de plus de 4 ans. Indicateurs de caractérisation Plusieurs critères de classification des races caprines africaines ont été relatés par Bouchel et Lauvergne (1996). Les indicateurs de caractérisation considérés dans cette étude, extraits de FA0 (1986), Lauvergne et al. (1993) et FAO (2012) sont les ports des oreilles et des cornes, la présence/absence de la barbiche et de pampilles (pendeloques), la structure (longueur et type) de pelage et les patrons pigmentaires. Ces paramètres ont été appréciés par observation directe et les informations sont répertoriées sur une fiche de caractérisation élaborée à cet effet. Cependant, pour la détermination des patrons pigmentaires, l échelle visuelle présentée dans la note technique de Danchin-Burge (lien électronique: idele.fr/?eid = cmis_download&oid = workspace://spacesstore/415d632d c-9e02-87c53ab30 fcc) a été utilisée. Cette échelle distingue neufs (9) patrons pigmentaires dont: deux patrons unis: eumélanique (noir ou chocolat) et phaeomélanique; cinq patrons présentant des variations de mélanine au niveau dorsal, ventral et de la face: joue rouge, eumélanique et feu, eumélanique uni et feu à ventre clair, chamoisé (ou blaireau, ou badger face) et patron sauvage; Deux patrons mantelés: mantelé antérieur et mantelé postérieur. Sur ces patrons, il peut y avoir des dépigmentations de la phaeomélanine mais aussi, des panachures. Les dépigmentations peuvent s étendre sur le corps entier de l animal se traduisant par un mélange de poils blancs soit sur un animal de couleur entièrement noir ce qui donne le gris, soit sur un animal de couleur entièrement rouge, ce qui donne le rouan ou au niveau de la tête (à la base des oreilles, des yeux ou du nez) donnant l impression des mouchetures, on parle de «frosting». Les panachures quant à elles sont des altérations pigmentaires qui, se caractérisent par la présence des tâches blanches à contour irrégulier sur une ou plusieurs parties du corps allant de la simple tâche (panachure localisée), à la grande panachure étendue à la quasi-totalité du corps (panachure généralisée), rendant souvent même difficile la lecture du patron pigmentaire (patron illisible). A partir des données des paramètres visuels, les indices de primarité (ou de traditionalité) tels que définis par Machado, Lauvergne et Souvenir zafindrajaona (1992) et Lauvergne et al. (1993) ont été calculés. Indice de primarité «loci à effet visible en ségrégation» (IPs): c est la fréquence (ou pourcentage) du nombre des loci à effet visible en ségrégation (n s ) rapporté au nombre total de loci à effet visible (Ns) identifiés dans l espèce. IPs = ns/ns Indice de primarité allèles au locus agouti (IPa): Soit Na le nombre total d allèles en agouti considéré dans une population et na le nombre d allèles identifiés dans cette même population, l indice de primarité allèles au locus agouti se définit comme suit: IPa = (na 1)/(Na 1) Analyses statistiques Les différents indicateurs étant codifiés, une maquette élaborée au logiciel SPSS 17.0 a été utilisée pour enregistrer toutes les informations recueillies. Ces dernières

17 14 Mani et al. ont fait l objet d une analyse descriptive préliminaire et de test de comparaison (Test de Khi carré) au logiciel SPSS et une analyse en composante multiple (ACM) au logiciel XLSTAT Résultats Analyse descriptive des paramètres qualitatifs étudiés Le tableau 2 présente les fréquences observées selon le sexe des paramètres qualitatifs étudiés. Il ressort de ce tableau que le port d oreilles est tombant (60%) aussi bien chez les mâles que chez les femelles, les cornes sont de type ibex dans plus de 85% des cas (chez les deux sexes). La barbiche et les pendeloques sont observées sur les deux sexes en moyenne respectivement sur 44,2% (54,9% des mâles et 41,1% des femelles) et 52,3% (56,4% des mâles et 51% des femelles) des animaux. Le pelage chez les deux sexes est entièrement court à ras (plus de 97%) et lisse à 56,1 % (51% des mâles et 57,6% des femelles). Il ressort en outre que tous les patrons pigmentaires ont été observés avec une prédominance des patrons phaeomélanique (30% en moyenne avec 28% des mâles et 31% des femelles), eumélanique chocolat (29% en moyenne dont 28% des mâles et 29% des femelles) et eumélanique noir (27% en moyenne dont 24,4% des mâles et 28% des femelles). Il faut cependant noter que la majorité de patrons observés (68,84%) est panachée (tache blanche) dont 65% panachure généralisée (64% des mâles et 67% des femelles); le pourcentage de dépigmentation est de 11,5% dont les plus dominantes sont le rouan avec en moyenne 45,5% (11,1% des mâles et 52,2% des femelles) et le frosting qui représente 40% (66,7% des mâles et 34,8% des femelles). La figure 2 présente quelques patrons observés. En considérant les variables (tableau 2) dont les modalités ont des effectifs supérieurs à 5 (ports d oreilles et de cornes, présence/absence de barbiche et des pampilles, types de poils), la répartition des fréquences selon le sexe est statistiquement non significative (P > 0,0 5) à l exception de la présence/absence de barbiche (P = 0,014). S agissant de la répartition selon les régions, le test de Khi carré indique qu elle est statistiquement significative pour les paramètres présence/absence de barbiche et des pampilles, port d oreilles, type de poils et non significative pour port de cornes (tableau 3). Tableau 2. Répartition des fréquences des paramètres qualitatifs selon le sexe Paramètres Modalités Mâle Femelle Tout sexe N % N % N % Port d oreilles Dressé 7 6,9 a 25 7,3 a 32 7,2 Pédonculé 34 33,3 a ,1 a ,2 Tombant 61 59,8 a ,5 a ,6 Port de cornes Ibex 87 86,1 a ,7 a ,0 Autre 14 13,9 a 52 15,3 a 66 15,0 Présence/absence de barbiche Présence 56 54,9 a ,1 b ,2 Absence 46 45,1 a ,9 b ,8 Présence/absence de pendeloques Présence 57 56,4 a ,0 a ,3 Absence 44 43,6 a ,0 a ,7 Longueur de poils Court à ras 99 97, , ,0 Mi-long 3 2,9 6 1,8 9 2,0 Types de poils Lisse 50 49,0 a ,1 a ,7 Non lisse 52 51,0 a ,6 a ,1 Patrons pigmentaires Eumélanique noir 20 24, , ,1 Eumélanique chocolat 23 28, , ,5 Phaeomélanique 23 28, , ,3 Joue rouge 2 2,4 4 1,4 6 1,6 Eumélanique et feu 0 0,0 1 0,3 1 0,3 Eumélanique uni et feu ventre claire 0 0,0 4 1,4 4 1,1 Chamoisé ou blaireau 4 4,9 10 3,4 14 3,7 Sauvage 1 1,2 3 1,0 4 1,1 Mantelé antérieure 5 6,1 1 0,3 6 1,6 Mantelé postérieur 1 1,2 6 2,0 7 1,9 Invisible 3 3,7 8 2,7 11 2,9 Dépigmentation Frosting 6 66, , ,0 Rouan 1 11, , ,5 Gris 2 22,2 6 13,0 8 14,5 Panachure Localisée 26 36, , ,1 Générale 46 63, , ,9 Les fréquences selon le sexe portant des lettres différentes sur la même ligne sont statistiquement différentes.

18 Caractéristiques phénotypiques de la chèvre du sahel 15 Figure 2. Quelques patrons pigmentaires observés [eumélanique noir Téra (a), eumélanique noir Niamey (b), eumélanique chocolat Téra (c), phaeomélanique Niamey (d), patron illisible Niamey (e), patron illisible Téra (f), chamoisé Niamey (g et h), sauvage Niamey (i et j), mantelé postérieur Niamey (k), mantelé antérieur Niamey (l), grise Niamey (m), chèvre rouanée Téra (n), frosting Niamey (o), panachure Téra et Oullam ( p et q)]. Indices de primarité Les tableaux 4 et 5 présentent respectivement les différents loci à effet visible en ségrégation et les allèles au locus agouti identifiés. Dans les départements d Abalak et de Tchintabaraden (1 ers sites d enquête), les données pour les indices de primarité n ont pas été abordées. Il ressort (tableau 4) que sur la population totale, IPS = 0,92. En effet, un seul locus à effet visible en ségrégation pressenti notamment «polled» caractérisant les chèvres mottes n a pas été identifié. La valeur de l IPS varie selon les départements, elle est de 0,83 dans le département de Tahoua et à Niamey et 0,67 dans les départements de Téra et Ouallam (région de Tillabéry). Cependant, le tableau 5 indique que les allèles au locus agouti pressentis et recherchés ont tous été identifiés dans la zone d étude, IPA = 1. Selon les départements, l IPA varie de 0,57 à 0,86. Analyse multi-variée des paramètres qualitatifs Au-delà de l étude de la traditionalité, une Analyse en Composante Multiple (ACM) a été conduite afin de mieux apprécier la dispersion des paramètres phénotypiques visibles. Les deux premiers axes factoriels F1 et F2 totalisent 51,47% de l inertie cumulée (tableau 6). La proportion des contributions élevées des modalités des différentes variables est la plus importante sur ces axes, mais aussi, ces axes représentent la plus grande proportion des cosinus carrés élevés des modalités. Le premier plan factoriel (F1-F2) peut cependant être utilisé pour l interprétation. La carte factorielle des modalités (figure 3) indique que toutes les modalités sont bien représentées. Il ressort que: le port d oreilles est majoritairement pédonculé dans les régions de Tahoua et Niamey et tombant dans la région de Tillabéry; le port d oreilles dressées bien que faiblement représenté dans les trois régions est majoritaire dans la région de Tahoua; les cornes conformément au test de Khi carré, sont majoritairement de type ibex dans les trois régions avec une forte proportion dans la région de Tillabéry; la barbiche et les pampilles sont présentes sur la majorité de chèvres de la région de Tillabéry et majoritairement absentes dans les populations caprines de la région de Niamey et Tahoua; les poils, de types majoritairement lisses dans la région de Tahoua et non lisses dans les régions de Tillabéry et Niamey, sont presque exclusivement courts à ras dans les trois régions; S agissant des patrons pigmentaires, la figure 3 montre que le patron phaeomélanique est majoritairement observé à Tillabéry (majorité des chèvres),

19 16 Mani et al. Tableau 3. Paramètres qualitatifs selon les régions Variables Modalités Tahoua Tillabéry Niamey Total N % N % N % N % Port d oreilles Dressé 21 20,6 a 2,8 b 9 8,7 c 32 7,2 Pédonculé 44 43,1 a 24 10,1 b 79 76,7 c ,2 Tombant 37 36,3 a ,1 b 15 14,6 c ,6 Port des cornes Ibex 83 82,2 a ,1 a 83 80,6 a ,0 Autre 18 17,8 a 28 11,9 a 20 19,4 a 66 15,0 Barbiche Présence 46 45,1 a ,7 b 27 26,2 c ,2 Absence 56 54,9 a ,3 b 76 73,8 c ,8 Pampille Présence 32 31,7 a ,6 b 31 30,1 c ,3 Absence 69 68,3 a 70 29,4 b 72 69,9 c ,7 Type de poils Lisse 90 88,2 a 92 38,8 b 11 10,7 c ,7 Non lisse 12 11,8 a ,8 b 92 89,3 c ,1 Longueur de poils Mi-long 8 7,8 1,4 0,0 9 2,0 Cours à ras 94 92, , , ,0 Patrons pigmentaires Eumélanique noir 22 43, , , ,1 Eumélanique chocolat 3 5, , , ,5 Phaeomélanique 7 13, ,1 9 8, ,3 Joue rouge 1 2,0 0,0 5 4,9 6 1,6 Eumélanique et feu 1 2,0 0,0 0,0 1,3 Eumélanique uni et feu ventre claire 2 3,9 2,9 0,0 4 1,1 Chamoisé ou blaireau 2 3,9 8 3,6 4 3,9 14 3,7 Sauvage 0,0 2,9 2 1,9 4 1,1 Mantélé antérieure 4 7,8 0,0 2 1,9 6 1,6 Mantelé postérieure 4 7,8 1,5 2 1,9 7 1,9 Illisible 5 9,8 6 2,7 0,0 11 2,9 Dépigmentation Frosting 1 33,3 7 29, , ,0 Rouan 1 33, , , ,5 Gris 1 33,3 4 16,7 3 10,7 8 14,5 Panachure Localisée 11 32, , , ,1 Générale 23 67, , , ,9 Les fréquences sur la même ligne, portant des lettres différentes sont significativement différentes selon les régions. eumélanique noir est observé dans les trois régions avec des proportions beaucoup plus importantes à Tahoua et Niamey, eumélanique chocolat est majoritairement Tableau 4. Indice de primarité «loci à effet visible en ségrégation» (IPs) observé à l ouest (Niamey et Tillabéry), joue rouge est majoritairement identifié à Niamey et absent dans la région de Tillabéry, eumélanique et feu est exclusivement signalé dans la région de Tahoua, eumélanique et Tableau 5. Indice de primarité allèles au locus agouti (IPa) Loci à effet visible en ségrégation identifiés Départements Tahoua Téra Ouallam Niamey Total Allèles au locus agouti identifiés Départements Tahoua Téra Ouallam Niamey Total Agouti Brown Roan Frosting Beard Ear Length Ear Carriage Ear Curling NR NR NR NR NR Polled Horn Reduction Hair Length Wattles Spotting ns Ns Ips 0,83 0,67 0,67 0,83 0,92 NR = Non recherché Ns: nombre total de loci à effet visible identifiés, n s: nombre des loci à effet visible en ségrégation rapporté. Eumélanique (noir, non agouti) Chamoisé Eumélanique et feu ventre clair Eumélanique et feu NR NR NR NR NR ventre coloré Joue rouge Mantelé antérieur Mantelé postérieur Phaeomélanique (fauve) Sauvage fauve Na NA IPa = (na 1)/(Na 1) 0,71 0,57 0,57 0,86 1 NR = Non recherché; NA: nombre total d allèles en agouti considéré dans une population; na : na le nombre d allèles identifié.

20 Caractéristiques phénotypiques de la chèvre du sahel 17 Tableau 6. Valeurs propres et pourcentages d inertie selon les axes F1 F2 F3 F4 F5 F6 F7 F8 F9 Valeur propre 0,242 0,201 0,185 0,176 0,153 0,152 0,147 0,146 0,143 Inertie (%) 9,413 7,799 7,206 6,848 5,965 5,906 5,718 5,684 5,556 % cumulé 9,413 17,212 24,419 31,267 37,232 43,138 48,856 54,539 60,095 Inertie ajustée 0,013 0,005 0,002 0,002 0,000 0,000 0,000 0,000 0,000 Inertie ajustée (%) 38,459 13,006 7,042 4,319 0,433 0,317 0,068 0,042 0,000 % cumulé 38,459 51,466 58,508 62,827 63,261 63,578 63,646 63,688 63,688 feu ventre clair est observé à Tahoua et Tillabéry, chamoisé est indiqué dans les trois régions et majoritairement dans la région de Tillabéry, sauvage à Niamey et Tillabéry, mantélé antérieur et postérieur majoritairement à Tahoua et les chèvres totalement panachées (patron invisible) sont identifiées dans les trois régions. Discussion Seuls ou associés à d autres, les indicateurs choisis dans cette étude sont classiques pour la caractérisation des petits ruminants en Afrique et à travers le monde (Lauvergne et al., 1993; Lanari et al., 2003; Traoré et al., 2006; Traoré et al., 2008; Hamito, 2009; Yakubu, 2010; Halima et al., 2012). Le port d oreilles est à environ 60% tombant et 33% pédonculé. Pareils constats ont été rapportés sur la chèvre du sahel au Burkina et au Tchad (Traoré et al., 2008; Dumas, 1980). En effet, selon Dumas (1980), les oreilles de la chèvre du sahel tchadienne sont longues, larges, pendantes sur les joues, dépassant le museau lorsqu elles sont rabattues vers l avant. Tous les animaux (100%) sont cornus (aucun animal motte a été rencontré). Cela est conforme aux résultats rapportés par Yakubu, Raji et Omeje (2010) sur les chèvres locales du Nigéria, Lauvergne et al. (1993) au Figure 3. Carte factorielle des modalités des paramètres phénotypiques qualitatifs

21 18 Mani et al. Nord Cameroun. Les cornes sont l un des appendices importants en milieu tropical car intervenant dans le mécanisme de thermorégulation (Robertshow, 2006; FAO, 2012). Dans le cadre de cette étude, les dimensions des cornes varient en fonction de l âge et du sexe. Elles sont longues chez les animaux adultes, fines chez les femelles et épaisses et souvent torsadées chez les mâles. Les cornes sont de type ibex (plus de 84% chez les deux sexes), conformément aux résultats de Traoré et al. (2008) chez les chèvres locales du Burkina en général et particulièrement sur la chèvre du sahel, et de Dumas (1980) qui précise pour la chèvre du sahel tchadienne «les cornes sont assez longues chez le mâle, elles sont dirigées en arrière et en haut en divergeant». Par contre ces résultats diffèrent de ceux rapportés par Lauvergne et al. (1993) au Nord Cameroun (11,7% type ibex). La présence de la barbiche a été identifiée chez les animaux des deux sexes (mâles 55% et femelles 41%). La présence/absence de la barbiche est le seul paramètre qualitatif dont la différence entre sexe a été statistiquement significative. Ce dimorphisme sexuel quant au port de la barbe a été rapporté par plusieurs auteurs avec des proportions variables selon le sexe: Wilson (1991) sur la chèvre West African Long-Legged (chèvre du sahel), Traoré et al. (2006) sur la chèvre locale mossi du Burkina Faso, Yakubu, Raji et Omeje (2010) sur les races caprines locales du Nigéria. Traoré et al. (2008) ont constaté un taux d absence de barbe de plus de 75% sur les chèvres locales du Burkina. La présence de la barbe chez les deux sexes avec précellence chez le mâle tient d une part, du fait que la barbe constitue un caractère sexuel secondaire chez le mâle et d autre part, le gène codant pour la barbe existe à l état récessif chez les femelles et dominant chez les mâles (Yakubu, Raji et Omeje, 2010). Comme autre appendice, la présence de pendeloques a été observée sur plus de 50% chez les deux sexes. Ces résultats se rapprochent de ceux de Traoré et al. (2008) qui rapportent que bien que la fréquence du port de pendeloques dans la population caprine locale soit relativement faible (30%), la chèvre du sahel spécifiquement recouvre plus de 75%. Cependant, des fréquences inférieures ont été rapportées par Yakubu, Raji et Omeje (2010) au Nigéria sur les chèvres de races West Africain Dwarf Goat (mâle 31,64% et femelle 5,78%) et Red Sokoto Goat (mâle 1,69% et femelle 3,78%). Selon Ricordeau et Bouillon (1967), dans l espèce caprine, les pendeloques sont présentes avec une fréquence variable dans les différentes races et il est bien difficile de dire si c est la présence ou l absence de pendeloques qui constitue le caractère normal ou sauvage. Le pelage est presque entièrement court à ras, les animaux à poils mi-longs sont rarement rencontrés (2%) et cela conformément aux races caprines locales de la sous région, notamment la chèvre rousse (red sokoto goat), la chèvre Djallonké (WAD goat), la chèvre du sahel (West African Long-Legged) et la chèvre mossi (Wilson, 1991; Meyer, 2001; Traoré et al., 2006; Traoré et al., 2008; Yakubu, Raji et Omeje, 2010). La robe est multicolorée, tous les patrons pigmentaires identifiés chez l espèce caprine ont été observés avec dominance chez les deux sexes des patrons Phaeomélanique (Brown), Eumélanique chocolat et Eumélanique noire (30,3%, 28,5% et 27,1% respectivement). Cependant, dans la plupart des cas (plus de 68%) la robe est panachée et le pourcentage de dépigmentation est non négligeable (11,5%). Ces résultats sont comparables à ceux de Lauvergne et al. (1993) au Nord Cameroun (tous les patrons observés avec prédominance des patrons phaeomélanique, sauvage et eumélanique respectivement) et de Traoré et al. (2008) chez la chèvre du sahel. En outre, ces résultats se distinguent de ceux de Yakubu, Raji et Omeje (2010) sur les races WAD et Red Sokoto goats au Nigéria avec prédominance respective de eumélanique noir et phaeomélanique (red). L importance des panachures d une part et le pourcentage non négligeable de dépigmentation d autre part peuvent trouver une explication dans le cadre de la thermorégulation. En effet selon Acharya et al. (1995), les caprins à poils courts sont peu résistants aux radiations solaires et consomment moins que ceux à poils longs. De même, les caprins noirs sont moins résistants que les rouges, à leur tour moins résistants que les blancs. L émergence de la panachure et de dépigmentation sur les autres patrons, autrement dit la présence des taches blanches d une part et le mélange des poils blancs aux poils rouges et noirs d autre part, rehaussera probablement la résistance aux radiations solaires et peut donc traduire une certaine adaptation au milieu. L ACM indique que tous les paramètres qualitatifs étudiés sont régulièrement répartis dans toutes les régions de la zone d étude. Cependant, elle fait ressortir que les caprins dans les trois régions se distinguent par le port des oreilles (majoritairement tombant dans la région de Tillabéry, pédonculé à Niamey et pédonculé et dressé dans la région de Tahoua), la présence de la barbe et des pendeloques (majoritairement présents dans la région de Tillabéry et absents dans la région de Tahoua et Niamey) et le patron pigmentaire (majoritairement eumélanique noir à Tahoua, eumélanique chocolat à Niamey et phaeomélanique à Tillabéry). Cette hétérogénéité selon les régions reflète l originalité de la race caractérisée par une diversité de forme dans sa zone d expansion (Wilson, 1991; Meyer, 2001). Conclusion Ces résultats font ressortir une hétérogénéité sur le plan phénotypique qualitatif des caprins dans la zone d étude. En effet, selon les régions, la répartition des paramètres visuels étudiés a été statistiquement différente à l exception du type des cornes. Visiblement la chèvre dans la zone d étude est multicolorée. La fréquence des

22 Caractéristiques phénotypiques de la chèvre du sahel 19 panachures est élevée (plus de 68%) d où une proportion importante des robes pies (pie rouge dans la région de Tillabéry et la CUN et pie noire dans la région de Tahoua). Les animaux des deux sexes portent des cornes (majoritairement de type ibex), une proportion importante tous sexes confondus porte la barbiche (proportion statistiquement plus élevée chez le mâle) et des pendeloques. Le pelage est court à ras et les oreilles sont majoritairement tombantes. L analyse des indices de primarité indique que cette population présente globalement, les caractéristiques d une population primaire. Cependant, l analyse multivariée fait ressortir quelques spécificités selon les régions. Ces résultats doivent être complétés par des données biométriques afind affiner l image phénotypique type de la chèvre du sahel dans la zone d étude. Remerciements Les auteurs remercient le partenaire financier qui est l Agence Universitaire de la Francophonie (AUF), précisément son Bureau Afrique de l Ouest et le CNF de Niamey. Ils remercient également le Programme de Productivité Agricole en Afrique de l Ouest (PPAAO) qui a contribué matériellement et financièrement aux activités sur les caprins du Sahel. Références bibliographiques Acharya, R.M., Gupta, U.D., Sehgal, J.P. et Singh, M Coat characteristics of goats in relation to heat tolerance in the hot tropics. Small Ruminant Research. 18: Bouchel, D. et Lauvergne, J.J Le peuplement de l Afrique par la chèvre domestique. Revue d Elevage et de Médecine Vétérinaire des Pays Tropicaux 49 (1): Danchin-Burge, C. [en ligne]. La couleur chez les caprins. Institut de l Elevage Dept. Génétique. (Disponible à l adresse électronique: idele.fr/?eid = cmis_download&oid = workspace://spacesstore/ 415d632d c-9e02 87c53ab30fcc). Dumas, R Contribution à l étude des petits ruminants du Tchad. Revue d Elevage et de Médecine Vétérinaire des Pays Tropicaux 33 (2): FAO Animal genetic resources data banks. 2. Descriptor lists for cattle, buffalo, pigs, sheep and goat. FAO. Rome, Italia. Animal Production and Health 59 (2): 151. (available at org/docrep/009/ah760e/ah760e00.htm) FAO Phenotypic characterization of animal genetic resources. Food and Agriculture Organization of the United Nations. Rome, Italia. Animal Production and Health Guidelines 11: 144. (available at Halima, H., Michael, B., Barbara, R. and Markos, T Phenotypic characterization of Ethiopian indigenous goat populations. African Journal of Biotechnology 11 (73): Hamidou, I Contribution à l analyse des paramètres de reproduction de la chèvre Rousse de Maradi. Thèse de Médecine Vétérinaire Dakar 11: 83. Hamito, D Goat breeds of Ethiopia: A guide for identification and utilization. Ethiopia Sheep and Goat Productivity Improvement Program (ESGPIP).Technical Bulletin 23: 11. Lanari, M.R., Taddeo, H., Domingo, E., Perez-Centeno, M. and Gallo, L Phenotypic differentiation of exterior traits in local Criollo Goat Population in Patagonia (Argentina). Archiv fur Tierzucht Dummerstorf 46 (4): Lauvergne, J.J., Bourzat, D., Souvenir, Z.P., Zeuh, V. et Ngo Tama, A.C Indices de primarité de chèvres au Nord Cameroun et au Tchad. Revue d Elevage et de Médecine Vétérinaire des Pays Tropicaux 46 (4): Machado, T.M., Lauvergne, J.J. et Souvenir Zafindrajaona, P Peuplement caprin du Brésil: Scénario du peuplement caprin du Brésil depuis la découverte. Archivos de Zootecnia 41 (extra): Mani, M Le cycle sexuel de la chèvre rousse de Maradi: Etude descriptive et progestéronomie. Mémoire de Master II. EISMV Dakar 5: 31. Marichatou, H., Mamane, L., Banoin, M. et Baril, G Performances zootechniques des caprins au Niger: étude comparative de la chèvre rousse de Maradi et de la chèvre à robe noire dans la zone de Maradi. Revue d Elevage et de Médecine Vétérinaire des Pays Tropicaux 55 (1): Meyer, C Races d animaux d élevage en Afrique intertropicale et méditerranéenne: Les caprins. Mémento de l Agronome. CIRAD-EMVT. France, 22. NIGER. Recensement Général de l Agriculture et du Cheptel (RGAC) Résultats définitifs (Volet cheptel) (2): 170. Rhissa, Z Revue du secteur de l élevage au Niger. Ministère de l Elevage, des Pêches et des Industries Animales. Niger, 115. Ricordeau, G. et Bouillon, J Hérédité des pendeloques en race saanen: différences de fécondité entre les génotypes avec et sans pendeloques. Annales de Zootechnie 16 (3): Robertshaw, D Mechanisms for the control of respiratory evaporative heat loss in panting animals. Journal of Applied Physiology 101: Robinet, A.H La chèvre rousse de Maradi son exploitation et sa place dans l économie et l élevage de la République du Niger. Revue d Elevage et de Médecine Vétérinaire des Pays Tropicaux 20 (1): Saadou, M Performance de reproduction et de production de la chèvre rousse de Maradi en milieu rural au Niger. Thèse de Médecine Vétérinaire Dakar 16: 82. Traoré, A., Hamidou, H.T., Adama, K., Luis, J.R., Ivan, F., Isabel, A., Sangaré, M., Bouchel, D., Jean Paul, P., Dominique, F., Sawadogo, L. and Goyache, F Multivariate analyses on morphological traits of goats in Burkina Faso. Archiv fur Tierzucht Dummerstorf 51 (6): Traoré, A., Tamboura, H.H., Kaboré, A., Yaméogo, N., Bayala, B. et Zaré, I Caractérisation morphologique des petits ruminants (ovins et caprins) de race locale Mossi au Burkina Faso. GRI : Wilson, T.R Small ruminant production and the small ruminant genetic resource in tropical Africa. FAO. Rome. Animal Production and Health Paper, 194. Yakubu, A., Raji, A.O. and Omeje, J.N Genetic and phenotypic differentiation of qualitative traits in Nigerian indigenous goat and sheep populations. ARPN Journal of Agricultural and Biological Science 5 (2):

23 Animal Genetic Resources, 2014, 54, Food and Agriculture Organization of the United Nations, 2014 doi: /s Caractérisation de la chèvre du sahel au Niger par analyse des indices biométriques et des paramètres phénotypiques quantitatifs M. Mani 1, H. Marichatou 1, M.M.M. Mouiche 2, M. Issa 3, I. Chaïbou 4, A. Sow 5, M. Chaïbou 1 and J.G. Sawadogo 5 1 Faculté d Agronomie, Université Abdou Moumouni, BP 10960, Niamey, Niger; 2 ESMV de Ngaoundéré, Cameroun; 3 Faculté des Sciences et Techniques, Université Abdou Moumouni, BP 10960, Niamey, Niger; 4 Faculté d Agronomie de l Université de Maradi, Niger; 5 Ecole Inter-Etats des Sciences et Médecines Vétérinaires, Dakar, Sénégal Résumé Très répandue en Afrique sèche, la chèvre du sahel représente environ 80% du troupeau caprin Nigérien. Conformément aux orientations nationales au Niger et dans la perspective d une meilleure gestion de cette ressource génétique aux niveaux local et national des travaux de caractérisation phénotypique de la race ont été conduits au Nord-Ouest du pays de juillet 2011 à mai Sur la base de l adhésion volontaire des éleveurs, des investigations de caractérisation des paramètres quantitatifs (hauteur au garrot et à la croupe, tour de poitrine, profondeur du thorax, largeur du bassin et longueurs du corps, des cornes et des oreilles) ont concerné 443 caprins (77% de femelle, 39% ayant 4 paires d incisives permanentes) repartis dans 145 exploitations, 60 sites (7,38 ± 6,87 têtes/site), dans les régions de Tillabéri, Tahoua et Niamey. L analyse des indices biométriques couplée à l analyse des variances (ANOVA) et des analyses multivariées (ACP, AFD) ont fait ressortir que le cheptel caprin dans cette zone comporte quatre sous types morphologiques (Tillabéri, Tahoua, Niamey et un sous type intermédiaire Niamey-Tillabéri). Du fait de la coexistence à Niamey de plusieurs races (locales et exotiques) et son corolaire le brassage génétique, la chèvre du sahel correspondrait à deux sous types: celui de Tillabéri de grande taille (64,98 ± 6,77 cm au garrot et 68,27 ± 5,91 cm à la croupe) à oreilles et cornes longues (20,3 ± 3,08 et 14,75 ± 4,43 cm respectivement) et de Tahoua également de grande taille (60,71 ± 10,81 cm au garrot et 62,46 ± 10,65 cm à la croupe) à oreilles et cornes courtes (12,32 ± 3,62 cm et 9,95 ± 4,01 cm respectivement). Mots-clés: caractéristiques phénotypiques, quantitatifs, chèvre du sahel, Niger Summary Found throughout the dry regions of Africa, the Sahel goat represents about 80% of Niger s goats. In accordance with national breeding policy in Niger and with a view to better planning management of that genetic resource at local and national levels, its phenotypic characterization has been conducted in the northwestern part of the country from July 2011 to May Based on the voluntary participation of farmers, investigations to characterize the quantitative parameters (height at withers and rump, body length, chest girth, chest depth, hip width and length of horns and ears) focused on 443 goats (77% female, 39% adult) distributed in 145 farms and 60 sites (7,38 ± 6,87 animals/site) in the Tillabéri, Tahoua and Niamey regions. Biometric indexes analysis coupled with analysis of variance and multivariate analysis (ACP, AFD) showed that the study area goat has four morphological subtypes (Tillabéri, Tahoua, Niamey and an intermediate Niamey-Tillabéri). Due to the coexistence of several breeds (local and exotic) in Niamey and consequences like genetic mixing, the Sahel goat can correspond to two subtypes: a large one (64.98 ± 6.77 and ± 5.91 cm at withers and rump respectively), long ears and horns (20.3 ± 3.08 and ± 4.43 cm) in Tillabéri and another large one (60.71 ± and ± cm at wither and rump), shorter ears and horns (12.32 ± 3.62 and 9.95 ± 4.01 cm) in Tahoua. Keywords: Phenotypic characteristics, Niger, Sahel goat, quantitative features Soumis: 27 Le juin 2013; admis: 7 Le janvier 2014 Introduction Ressource animale caractéristique des zones sèches, la chèvre du sahel est très répandue dans tous les pays du Correspondence à envoyer à: H. Marichatou Faculté d Agronomie, Université Abdou Moumouni, BP 10960, Niamey, Niger. Tel: (00227) ,. Fax: (00227) ,. Adresse électronique: maricha@refer.ne / marimani_m@ yahoo.fr sahel, au Sénégal, Mauritanie, Mali, Burkina Faso, Niger, Tchad, Nigéria (Dumas, 1980; Wilson, 1991; Meyer, 2001). Sa présence est également signalée dans les régions humides d Afrique occidentale et centrale au Togo, en Côte d Ivoire, au Cameroun (Lauvergne et al., 1993; Meyer, 2001). Au Niger, la chèvre du Sahel représente environ 80% du cheptel caprin (Rhissa, 2010) et est élevée dans toutes les 21

24 22 H. Marichatou et al. régions du pays. Cependant, à l image des autres pays voisins tels que le Tchad, le Burkina Faso (Dumas, 1980; Lauvergne et al., 1993; Gnanda, 2008; Traoré, 2010), la chèvre du sahel est peu connue au Niger. L objectif visé dans ce travail est d établir les caractéristiques phénotypiques quantitatives de la chèvre du sahel au Niger. En effet, une étape importante dans la connaissance des ressources génétiques animales, la caractérisation phénotypique est selon FAO (2012), essentielle pour la planification de la gestion des ressources zoogénétiques aux niveaux local, national, régional et mondial. Les paramètres phénotypiques quantitatifs étant directement corrélés aux paramètres de production (FAO, 2012), leur connaissance est importante pour la maîtrise et l amélioration des performances de production des animaux. (département de Téra et Ouallam) et la Commune Urbaine de Niamey (Figure 1). Les travaux étant conduits en milieu réel, l adhésion des éleveurs a constitué un facteur déterminant pour la collecte des données. Ainsi, au niveau des exploitations où l accord de principe a été acquis, un certain nombre de têtes est aléatoirement isolé pour les mensurations nécessaires. Au total, 145 exploitations et 443 caprins (dont 77% de femelles, 22,1% âgés de 0 à 1 an, 16,5% de 1 à 2 ans, 22,3% de 2 à 3 ans, 18% de 3 à 4 ans et 20,3% de plus de 4 ans) répartis en 60 sites, ont constitué l échantillon de caractérisation. Le Tableau 1 présente le nombre de sites et de caprins par région de l échantillon de l étude et le Tableau 2 la répartition selon les tranches d âge de l échantillon d étude. Matériel et Méthodes Echantillon d étude Inscrite dans le cadre des investigations de caractérisation phénotypique de la chèvre du sahel au Niger, l étude a été conduite cumulativement à celle de la caractérisation phénotypique qualitative (Mani et al., sous presse), de juillet 2011 à mai 2012 dans la partie nord-ouest du Niger, notamment, les régions de Tahoua (départements de Tahoua, Abalak et Tchintabaraden), Tillabéri Indicateurs de caractérisation En Afrique, la classification des races caprines a été évolutive, plusieurs combinaisons de critères, généralement adoptées chez les bovins, ont été utilisées, une bibliographie conséquente en ce sens a été établie par Bouchel et Lauvergne (1986). Pour atteindre les objectifs assignés à notre étude, les indicateurs quantitatifs retenus de la FAO (FAO, 1986; FAO, 2012) sont: la hauteur au garrot (HG), la longueur du corps (LC), le tour de poitrine (TP), la largeur du bassin (LB), la hauteur à la croupe (HC), la profondeur du thorax (PT), la longueur des oreilles (LO) et la longueur des Figure 1. Zone d étude.

25 Caractérisation de la chèvre du sahel au Niger 23 Tableau 1. Nombre de sites et de caprins par région de l échantillon d étude. Régions Niamey Tillabéri Niamey Total Nombre de sites(*) Nombre de caprins Moyennes/site 6,9 ± 2,9 13,2 ± 8,1 3,78 ± 4 7,38 ± 6,67 Maximum/site Minimum/site (*) Site = village ou quartier. cornes (Lc). Les mensurations (Figure 2) sont effectuées à l aide d une règle, d un mètre ruban et d un compas. L âge des animaux est déterminé par examen de la dentition, l échelle présentée par Hamito (2009) a été utilisée à cet effet tout en considérant que chez l espèce caprine, l animal atteint l âge adulte à partir de 4 paires d incisives (Bouchel et al., 2006). Les données morpho-biométriques ont permis d établir les indices biométriques tels que définis par Lauvergne et al. (1993) et Bourzat et al. (1993). Indice de gracilité sous-sternal (IGS): rapport du vide sous-sternal [différence entre hauteur au garrot (HG) et profondeur du thorax (PT)] sur la profondeur du thorax (PT). IGS = (HG PT)/PT Indice auriculaire thorax (IAT): longueur de l oreille (LO) sur la profondeur du thorax (PT). IAT = LO/PT Analyses statistiques Les données recueillies ont fait l objet d analyse descriptive et d une comparaison des moyennes (ANOVA) au logiciel SPSS (version 17.0). Des analyses multivariées notamment l Analyse en Composante Principale (ACP) et l Analyse Factorielle Discriminante (AFD) ont été effectuées au logiciel XLSTAT Il faut cependant noter que la synthèse des informations de l analyse descriptive sous forme de tableaux et figures a été faite au tableur Excel. Tableau 2. Configuration selon l âge et le sexe de l échantillon d étude. Tranches d âge Mâle Femelle Total N % N % N % 0 1 an 52 11,7% 46 10,4% 98 22,1% 1 2 ans 20 4,5% 53 12,0% 73 16,5% 2 3 ans 19 4,3% 80 18,1% 99 22,3% 3 4 ans 6 1,4% 77 17,4% 83 18,7% >4 ans 5 1,1% 85 19,2% 90 20,3% Total ,0% ,0% ,0% N = Nombre de caprins caractérisé par tranche d âge. L ACP vise à représenter graphiquement les relations entre individus par l évaluation de leurs ressemblances, ainsi que les relations entre variables par l évaluation de leurs liaisons (Martin, 2004). Ces principes sont également pris en compte en AFD (Baccini et Besse, 2007). Selon Martin (2004), deux individus se ressemblent, ou sont proches, s ils possèdent des valeurs proches pour l ensemble des variables. Cette définition sous entend une notion de proximité qui se traduit par une distance. Ainsi, la distance entre deux individus X i et Y j est définie par: d 2 (X i, Y j ) = p k=1 (X ik Y jk ) 2 Par ailleurs, deux variables sont liées si elles ont un fort coefficient de corrélation linéaire. Sur un échantillon de taille n, le coefficient de corrélation linéaire (dit de Bravais-Pearson ou de Pearson) entre deux variables X (x 1, x 2,...x i ) et Y (y 1, y 2,...y i ) est donné par l équation suivante rapportée par Martin (2004) et Rakotomalala (2012). ni=1 (x i x). (y i y) r = n i=1 (x i x) 2 n i=1 (y i y) 2 Plus le coefficient de corrélation est proche de 1 (en valeur absolue), plus la relation est forte (négativement si r < 0 et positivement si r > 0). Si r = 0, il ya absence de corrélation. Graphiquement, les variables sont représentées dans l espace (de dimension n) par des vecteurs. On considère les données centrées réduites par ligne afin de faciliter l analyse du nuage d individu car permettant de ramener l origine des axes au centre de gravité G correspondant à l individu moyen de la population (Martin, 2004). X (x 1 x,..., x n x); Y(y 1 y,..., y n y); Le cosinus de l angle α entre ces vecteurs n est autre que le coefficient de corrélation, Cos(α) = r (formule ci-haut). Résultats Analyse descriptive des données Paramètres biométriques selon la région, le sexe et les tranches d âges Le Tableau 3 présente les moyennes et écarts types des différents paramètres biométriques étudiés selon les régions et le sexe. Il ressort trois classements décroissants selon les régions: Tillabéri, Tahoua, Niamey pour HG et HC; Tillabéri, Niamey, Tahoua pour TP, PT, LO et Lc; Tahoua, Tillabéri, Niamey pour LC et LB.

26 24 H. Marichatou et al. Figure 2. Mensurations réalisées [hauteur au garrot (a), hauteur à la croupe (b), profondeur du thorax (c), tour de poitrine (d), largeur du bassin (e), longueur des cornes (f)]. Cependant, l analyse des variances suivie de test post hoc, ont montré que les différences entre les populations caprines des trois régions sont statistiquement significatives (P < 0,05) pour HG, LC, LB, HC et LO. En outre, les chèvres de Tahoua se distinguent statistiquement de celles de Tillabéri par le TP et de plus de celles de Niamey par la PT et les chèvres de Tillabéri se distinguent statistiquement de celles de Tahoua et Niamey par la LC. Le Tableau 4 présente les moyennes et écarts types des différents paramètres biométriques selon les tranches

27 Caractérisation de la chèvre du sahel au Niger 25 Tableau 3. Moyennes et écarts types moyens (cm) des paramètres biométriques selon les régions Régions Sexes N HG TP LC LB HC PT LO Lc Tahoua M 28 59,34 ± 13,17 61,15 ± 14,38 67,95 ± 18,16 20,13 ± 5,67 60,45 ± 12,82 23,99 ± 3,23 13,02 ± 2,94 9,3552 F 74 61,22 ± 9,83 62,59 ± 15,10 70,10 ± 17,29 21,05 ± 5,25 63,22 ± 9,69 24,15 ± 5,26 12,05 ± 3,83 10,18 ± 3,69 Total ,71 a ± 10,81 62,2 a ± 14,85 69,51 a ± 17,47 20,8 a ± 5,36 62,46 a ± 10,65 24,10 a ± 4,77 12,32 a ± 3,62 9,95 a ± 4,01 Tillabéri M 49 61,9 ± 8,27 62,73 ± 9,24 58,18 ± 7,23 13,21 ± 2,27 65,78 ± 7,38 29 ± 3,25 19,46 ± 3,57 14,02 ± 5,31 F ,78 ± 6,1 68,54 ± 8,69 62,76 ± 8,17 14,32 ± 4,71 68,92 ± 5,3 30,37 ± 5,05 20,51 ± 2,91 14,94 ± 4,17 Total ,98 b ± 6,77 67,34 b ± 9,1 61,82 b ± 8,18 14,09 b ± 4,34 68,27 b ± 5,91 30,09 b ± 4,76 20,3 b ± 3,08 14,75 b ± 4,43 Niamey M 25 52,08 ± 8,41 57,56 ± 7,03 42,16 ± 5,63 10,12 ± 1,71 54,64 ± 8,18 26,64 ± 3,83 13,8 ± 2,31 8,67 ± 4,35 F 78 58,65 ± 8,62 68,58 ± 9,59 50,10 ± 9,31 12,98 ± 2,67 60,81 ± 7,72 30,55 ± 4,66 14,91 ± 2,4 11,56 ± 5,02 Total ,05 c ± 8,98 65,90 ab ± 10,18 48,17 c ± 9,2 12,29 c ± 2,76 59,32 c ± 8,23 29,60 b ± 4,76 14,64 c ± 2,42 10,85 a ± 4,1 Moyennes générales M ,79 x ± 10,56 61,03 x ± 10,58 56,94 x ±14,46 14,3 x ± 5,06 61,58 x ± 10,32 27,05 x ± 3,97 16,34 x ± 4,36 11,4 x ± 5,49 F ,16 y ± 8,21 67,26 y ± 10,85 61,46 y ± 12,94 15,43 x ± 5,34 65,83 y ± 7,86 29,08 y ± 5,62 17,46 y ± 4,68 13,18 y ± 4,75 Tout sexe ,15 ± 8,99 65,82 ± 11,09 60,42 ± 13,42 15,17 ± 5,29 64,85 ± 8,66 28,61 ± 5,35 17,2 ± 4,63 12,77 ± 4,98 HG: hauteur au garrot, TP: tour de poitrine, LC: longueur du corps, LB: largeur du bassin, HC: hauteur à la croupe, PT: profondeur du thorax, LO: Longueur des oreilles, Lc: Longueur des cornes. Les moyennes des régions portant des lettres différentes (a, b ou c) sur la même colonne sont statistiquement dfférentes (P < 0,05). d âge. Il résulte que tous les paramètres varient avec l âge. L analyse des variances fait ressortir des différences statistiquement significatives à certaines tranches d âges. Les variations d HG, HC LC, TP et Lc d une part et LO et LB d autre part sont apparues statistiquement semblables. En effet, les différences sont significatives entre les tranches d âge à l exception des tranches d âge successives à partir de 2 ans pour HG et Lc. Comparativement à ces derniers, les variations sont significatives pour TP entre 2 3 ans et 3 4 ans et non significatives pour LC entre 1 2 ans et 3 4 ans puis, 2 3 ans et > 4 ans et pour HC entre 2 3 ans et > 4 ans. S agissant de LB et LO cependant, les variations sont non significatives à l exception de la tranche d âge 0 1 an et les tranches > 3 ans pour LB et en plus entre 1 2 ans et > 4 ans pour LO. La Figure 3 présente l évolution des différents paramètres biométriques selon le sexe et les tranches d âge. Il ressort que chez le mâle, les valeurs maximales sont atteintes à l âge de 2 3 ans pour les paramètres tels que HG, LC, LB, HC, et la PT, les autres paramètres (TP, LO et Lc) continuant leur croissance jusqu à l âge de quatre ans et au delà. Chez la femelle par contre, tous les paramètres étudiés augmentent selon l âge jusqu au-delà de 4 ans. En termes de comparaison selon le sexe, la Figure 3 permet de noter selon les tranches d âge: les mâles sont supérieurs aux femelles à toutes les tranches d âge pour la Lc; les mâles sont supérieurs aux femelles pour HG, TP, LC, HC, PT, et inférieurs aux femelles pour LO pour la tranche d âge 0 1 an et inversement pour la tranche d âge 1 2 ans; à2 3 ans, au moment où les paramètres sont globalement à leur optimum chez les mâles, ces derniers sont supérieurs aux femelles pour tous les paramètres; au-delà de trois ans, les mâles sont supérieurs aux femelles pour HG et LO et inférieurs pour TP, LC et PT. Afin de mieux apprécier la répartition selon les régions d une part et les paramètres directeurs de cette répartition d autre part, il a été abordé l analyse des indices biométriques mettant en exergue la hauteur au garrot, la profondeur de la poitrine et la longueur des oreilles et l analyse multivariée mettant en jeux tous les paramètres morpho-biométriques en relation avec les paramètres qualitatifs dont la région et le sexe. Indices biométriques Les indices biométriques moyens sont: Indice de gracilité sous sternal IGS = 1,21 ± 0,34 Indice Auriculo-thorax IAT = 0,44 ± 0,14 Les valeurs moyennes indiquent que les caprins dans la zone d étude sont globalement de type longipes (IGS > 1), aux oreilles longues. Cependant, ces moyennes cachent de grandes disparités. La Figure 4, présente les valeurs moyennes de ces indices selon les départements de la zone de l étude. Il ressort que ces indices varient selon les départements. L IGS varie approximativement avec la

28 26 H. Marichatou et al. Tableau 4. : Tableau 4: Variation selon l âge des paramètres biométriques Ages N HG TP LC LB HC PT LO Lc 0 1 an 98 53,79 ± 9,2 (a) 57,16 ± 8,72 (a) 50,87 ± 14,08 (a) 13,32 ± 5,42 (a) 57,14 ± 8,94 (a) 25,57 ± 4,54 (a) 15,63 ± 3,84 (a) 9,18 ± 4,53 (a) 1 2 ans 73 60,21 ± 5,82 (bc) 61,93 ± 7,9 (bc) 59,23 ± 9,08 (b) 15,09 ± 4,82 (ab) 63,24 ± 5,95 (bc) 27,91 ± 3,49 (ab) 16,14 ± 4,3 (ab) 11,66 ± 3,38 (bc) 2 3 ans 99 63,66 ± 7,43 (cd) 66,11 ± 10,73 (c) 62,47 ± 12,04 (bd) 15,62 ± 4,37 (ab) 66,60 ± 7,49 (cd) 28,77 ± 4,89 (b) 17,33 ± 4,69 (ac) 12,78 ± 4,48 (cd) 3 4 ans 83 66,21 ± 7,62 (de) 70,63 ± 9,81 (d) 64,39 ± 12,77 (bd) 15,93 ± 4,39 (b) 68,39 ± 7,78 (d) 29,75 ± 4,87 (bc) 18,40 ± 5,18 (bc) 14,57 ± 4,41 (de) > 4 ans 90 67,44 ± 6,42 (e) 73,66 ± 9,11 (d) 65,85 ± 12,31 (d) 16,09 ± 6,65 (b) 69,37 ± 5,99 (d) 31,27 ± 6,54 (c) 18,54 ± 4,44 (c) 16,1 ± 4,68 (e) Total ,15 ± 8,99 65,82 ± 11,09 60,42 ± 13,42 15,17 ± 5,29 64,85 ± 8,66 28,61 ± 5,35 17,2 ± 4,63 12,77 ± 4,98 Sur la même colonne, une moyenne portant au moins une lettre (a, b, c ou d) portée par une autre, ne lui est pas statistiquement différente (P < 0,05). longueur des pattes (HG PT) alors que les variations de l IAT sont imputables à l évolution des deux facteurs (LO et PT). Ainsi, dans la zone d étude, les chèvres sont de type brèvipes à oreilles moyennes dans la région de Niamey, longipes à oreilles longues dans la région de Tillabéri (Téra et Ouallam) et longipes à oreille relativement courtes dans la région de Tahoua (Tahoua, Abalak). Analyse multi-variée des paramètres biométriques Analyse en composante principale L ACP fait ressortir (Tableau 5), les corrélations entre les paramètres morpho-biométriques étudiés en variables principales mais aussi entre eux et la région et le sexe en variables supplémentaires. Conformément au principe de Kaiser (1960), les deux premiers axes F 1 et F 2 de l ACP sont les seuls dont les valeurs propres sont supérieures à 1, ils résument en outre environ 67,8% des observations. Pour cela, et du fait que l essentiel des variables principales mais aussi supplémentaires ont leurs cosinus carrés élevés sur ces axes, le premier plan factoriel (F1xF2) s adapte bien pour l interprétation des résultats de l analyse. Il ressort sur ce plan factoriel (Figure 5) que toutes les variables (principales et supplémentaires) sont bien représentées. Cette figure permet en outre de constater une forte corrélation positive entre HG, HC, Lc, TP, LC, PT et LO. Ce groupe de variable fortement corrélé est en corrélation approximativement nulle avec LB. Il ressort que ce groupe de variable est bien caractéristique des chèvres de la région de Tillabéri alors que celles de Tahoua sont bien caractérisées par LB, LC, HG, HC et celles de Niamey par PT, LO, Lc et TP. En outre, cette figure permet de lire que globalement à tout âge compris tous les paramètres sont en corrélation positive avec le sexe femelle et négative avec le sexe mâle. Analyse Factorielle Discriminante L Analyse Factorielle Discriminante (AFD) contribue à affiner les résultats de l ACP. En considérant comme variables dépendantes la région et le sexe et comme variables explicatives l âge et les huit paramètres morphobiométriques, les deux premiers axes factoriels expliquent 100% de la variabilité. Il ressort de cette analyse (Figure 6), que les facteurs les plus discriminants mieux représentés sur les deux axes, autrement dit qui permettent de mieux caractériser les chèvres, sont HG, HC, LC, Lc, LO, LB et PT. La matrice de confusion (Tableau 6) montre que les individus sont bien classés en moyenne dans 92,55%, trente trois (33) individus soit 7,45% étant reclassés. Il ressort de la répartition des individus dans le plan factoriel (Figure 7) que dans les régions de Tahoua et Tillabéri, les individus sont bien discriminés sur les deux axes contrairement à la région de Niamey où les individus sont représentés principalement sur l axe F 1. Cette répartition fait également ressortir que les populations de chèvres dans les trois régions ne sont pas totalement distinctes. Il existe en faible proportion des

29 Caractérisation de la chèvre du sahel au Niger 27 Figure 3. Evolution selon l âge et le sexe des paramètres morpho biométriques étudiés ( M: mâle; F: femelle). sous groupes intermédiaires entre les régions. Cependant, conformément à l ACP, l AFD (Figures 6 et 7) montre que dans la majorité, les populations caprines se distinguent d une région à une autre, les chèvres dans la régions de Tahoua bien discriminées par l importance de LC et LB, celles de la région de Tillabéri par l importance de HG et HC, PT, Lc et LO, tandis que celles de Niamey sont mieux représentées par TP et PT. Figure 4. Répartition des indices biométriques selon les départements de la zone d étude.

30 28 H. Marichatou et al. Tableau 5. Matrice de corrélation (Pearson (n)) Variables Âges HG TP LC HC PT LB L0 Lc Tah Til Ny F M Âges 1 HG 0,48 1 TP 0,52 0,72 1 LC 0,36 0,62 0,56 1 HC 0,44 0,94 0,75 0,66 1 PP 0,33 0,48 0,49 0,17 0,50 1 lb 0,16 0,18 0,10 0,41 0,12 0,05 1 LO 0,22 0,52 0,37 0,22 0,58 0,42 0,23 1 Lc 0,47 0,63 0,53 0,40 0,65 0,58 0,13 0,54 1 Tah 0,07 0,09 0,18 0,37 0,15 0,46 0,56 0,55 0,31 1 Til 0,09 0,34 0,15 0,11 0,43 0,30 0,22 0,73 0,44 0,59 1 Ny 0,03 0,31 0,00 0,50 0,35 0,10 0,30 0,30 0,21 0,30 0,59 1 F 0,35 0,20 0,24 0,14 0,21 0,16 0,09 0,10 0,15 0,06 0,06 0,02 1 M 0,35 0,20 0,24 0,14 0,21 0,16 0,09 0,10 0,15 0,06 0,06 0,02 1,00 1 Les valeurs en gras sont différentes de 0 à un niveau de signification alpha = 0,05. Figure 5. Cercle de corrélation ACP.

31 Caractérisation de la chèvre du sahel au Niger 29 Figure 6. Cercle de corrélation AFD. Discussion Plusieurs études de caractérisation des caprins (Lauvergne et al., 1993; Zeuh et al., 1997; Lanari et al., 2003; Traoré et al., 2006; Bouchel et al., 2006; Traoré et al., 2008; Verma et al., 2010; Ebegbulem et al., 2011; Halima et al., 2012) ont fait usage seuls ou associés à d autres, des indicateurs utilisés dans le cadre de cette étude. L échantillon étudié est composé majoritairement des Tableau 6. Matrice de confusion pour l échantillon d estimation (région) de \ Vers Niamey Tahoua Tillabéri Total % correct Niamey ,47% Tahoua ,18% Tillabéri ,64% Total ,55% femelles (77%) conformément à l image générale des troupeaux enquêtés (les mâles sont minoritaires et généralement jeunes), mais aussi traduit l importance socioéconomique de l élevage caprin dans la zone d étude. En effet, les mâles sont abattus à bas âges pour des circonstances de cérémonies (mariage, baptême, tabaski ou autres fêtes) mais aussi de sacrifice. Ils constituent également une épargne facilement mobilisable en cas de besoin pécuniaire et les femelles sauf en cas de forces majeures sont gardées pour le renouvellement du troupeau. Ce constat a été relaté au Nigéria, au Burkina, au Botswana, en Ouganda (Samuel et Salako, 2008; Traoré et al., 2006; Katongole et al., 1996; Semakula et al., 2010). Les moyennes des données biométriques obtenues sont inférieures à celles présentées par Wilson (1991) pour la chèvre «Arabe» du Tchad et les chèvres «Maure» du Mali pour HG, PT, LO et Lc, les chèvres «Toureg» et

32 30 H. Marichatou et al. Figure 7. Répartition des individus dans le plan factoriel (F1-F2). «Sahel» du Mali pour HG et PT, mais semblable à la chèvre «Sahel» du Mali pour la LO. Par ailleurs, ces moyennes biométriques obtenus sont supérieures à celles des chèvres sahéliennes, soudanienne (chèvre Djallonké) et sahélo-soudanienne (Mossi) pour les paramètres HG, PT, LC, LO, Lc et HC (Traoré et al., 2008; Samuel et al., 2008; Abdulmojeed, Adebowale et Ikhide, 2010; Ebegbulem et al., 2011) mais se rapprochent des données de la chèvre «Red Sokoto Goat» obtenues par Abdulmojeed et al. (2010). Les indices biométriques varient selon les localités (régions et départements), l IGS variant de 0,94 à 1,67 et l IAT de 0,35 à 0,52. Etant donné que l IGS évolue avec la longueur des pattes et l IAT dépend des deux grandeurs conjugués, LO et PT, les variations des indices biométriques indiquent des distinctions entre les populations caprines des trois régions étudiées par la longueur des pattes et des oreilles et la profondeur du thorax. Les variations de l IGS observées sont semblables aux variations rapportées sur les populations caprines du Sénégal (Bouchel et al., 2006). Cependant, l IGS moyen obtenu est semblable a l IGS des populations caprines du Nord Cameroun mais inférieur à celui des populations caprines du Nord Tchad (Bourzat et al., 1993) et intermédiaire entre les deux sous populations du sud Tchad (Zeuh et al., 1997). Egalement, cet IGS moyen se rapproche de celui de la population caprine toutes races confondues du Burkina Faso (Traoré et al., 2008), mais spécifiquement inférieur à celui de la chèvre sahélienne et supérieur à ceux des chèvres sahélo-soudanienne et soudanienne. Cela suppose que la population caprine de la zone d étude se rapproche du point de vue longueur des pattes à celles du Burkina Faso et du Nord Cameroun. L IAT quant à lui, est inférieur à ceux des populations caprines de ces deux entités. Cette faiblesse de l IAT ne traduit pas que la LO et PT sont inférieures bien au contraire, mais tient au fait que l évolution de l IAT est apparue indépendante d un seul des deux facteurs du fait des écarts entre les valeurs des deux paramètres (LO approximativement moitié de PT). Les résultats de l analyse en composante principales font ressortir qu indépendamment de l âge, le sexe femelle est en corrélation positive avec tous les paramètres biométriques étudiés. Cela peut être inhérent aux écarts d effectifs (femelle 77%). L analyse des variances a indiqué que les paramètres biométriques varient statistiquement selon l âge et l analyse selon les tranches d âges a fait ressortir globalement que les mâles sont

33 Caractérisation de la chèvre du sahel au Niger 31 supérieurs aux femelles avec quelques exceptions selon les tranches d âges et les paramètres. En effet, selon Vigne, Peters et Helmer (2002), le sexe est le facteur le plus influant sur les dimensions chez les chèvres, les mâles âgés de plus de 12 mois sont toujours plus grands que les femelles; ensuite vient l environnement se traduisant par une diminution des dimensions des chèvres des zones froides et humides vers les zones arides, alors que l âge semble avoir peu d impact sur les dimensions des chèvres. Samuel et Salako (2008) et Semakula et al. (2010), étudiant respectivement les caractéristiques biométriques de la chèvre Djallonké (West African Dwarf Goat) au Nigéria et la chèvre Mubende en Ouganda, arguent que l âge et le sexe sont des facteurs influençant significativement le poids corporel et les paramètres biométriques. Les femelles chez la chèvre Djallonké sont trouvées supérieures aux mâles dans les tranches d âge de 0 à 2 ans selon les résultats de Samuel et Salako (2008) et inversement pour toutes les tranches d âge chez la chèvre Mubende selon les résultats de Semakula et al. (2010). Cependant, les résultats d Ebegbulem et al. (2011) indiquent chez la chèvre Djallonké au Nigeria que les mâles sont supérieurs aux femelles pour HG, HC, TP, LO, LC et LB. L AFD et l ACP, confirment les résultats des tests statistiques et de l analyse des indices biométriques, autrement dit la distinction entre les populations caprines de trois régions. L AFD révèle une quasi distinction et homogénéisation respectivement inter et intra région pour les populations caprines étudiées. Les tests statistiques et les analyses multivariées (ACP et AFD) permettent de distinguer dans la zone d étude quatre (4) sous populations distinctes de chèvres en fonctions des paramètres morpho-biométriques étudiés. Une sous population de Tillabéri caractérisée par de HG, HC, LO, TP, et Lc élevés mais LB et LC faibles par rapport aux autres sous populations, une sous population de Tahoua avec LB et LC élevées, HG et HC intermédiaires et TP, PT, LO, Lc faibles par rapport aux autres, une sous population de Niamey avec PT, LO moyens et des valeurs faibles pour les autres paramètres et une sous population intermédiaire entre la sous population de Tahoua et celle des Niamey avec TP et Lc moyens. Des différences statistiquement significatives sur le plan phénotypique qualitatif ont également été rapportées (Mani et al., sous presse) entre les populations des trois régions. Les valeurs moyennes à faibles observées pour les chèvres de Niamey sont inhérentes au caractère urbain de son élevage, qui fait d elle un melting-pot où est rencontré une diversité de races notamment, les races locales (chèvre Rousse, chèvre du sahel), sous régionale (chèvre Djallonké, chèvre Mossi) et leurs croisés. Les individus intermédiaires peuvent être inhérents à la proximité, en effet, Niamey est circonscrite dans la région de Tillabéri qui, elle-même partage une longue frontière avec la région de Tahoua. Conclusion Aux termes de cette étude, il ressort que l espèce caprine dans la zone d étude est diversement repartie selon les régions en fonction des paramètres quantitatifs étudiés. Trois écotypes spécifiques aux trois régions et un type intermédiaire Niamey-Tillabéri ont été distingués. Les valeurs moyennes à faibles des paramètres morphobiométriques des caprins de Niamey est la résultante de l hétérogénéité zoo-génétique de ce troupeau avec la présence de la diversité génétique de la sous région et son corolaire le brassage génétique. De ce fait, mise à part la population hétéro-génétique de Niamey, l on retiendra que la chèvre du sahel se présente sous deux types phénotypiques dans la zone d étude. Une population type de la région de Tillabéri et celle type de la région de Tahoua. Les caractéristiques sur le plan morphobiométrique sont: grande taille (en moyenne 64,98 ± 6,77 cm et 60,71 ± 10,81 cm au garrot et 68,27 ± 5,91 cm et 62,46 ± 10,65 cm à la croupe respectivement à Tillabéri et Tahoua), oreilles et cornes longues dans la région de Tillabéri (en moyenne respectivement environ 20,3 ± 3,08 cm et 14,75 ± 4,43 cm) et courtes dans la région de Tahoua (environ respectivement 12,32 ± 3,62 cm et 9,95 ± 4,01 cm en moyenne). Le tour de poitrine, la longueur du corps, la largeur du bassin et la profondeur du thorax sont en moyenne respectivement 62,2 ± 14,85 et 67,34 ± 9,1 cm; 69,51 ± 17,47 et 61,82 ± 8,18 cm; 20,8 ± 5,36 et 14,09 ± 4,34 cm et 24,10 ± 4,77 et 30,09 ± 4,76 cm respectivement à Tahoua et Tillabéri. Cependant, afin d affiner la caractérisation, il est fondamental de savoir qu en est-il sur les plans génétiques et zootechniques entre ces deux sous types? En outre, eu égard à l omniprésence de la chèvre du sahel dans toutes les Régions du Niger, il s avère important d étendre les travaux de caractérisation sur les autres régions du pays. Remerciements Les auteurs remercient le partenaire financier qui est l Agence Universitaire de la Francophonie (AUF), précisément son Bureau Afrique de l Ouest et le CNF de Niamey. Ils remercient également le Programme de Productivité Agricole en Afrique de l Ouest (PPAAO) qui a contribué matériellement et financièrement aux activités sur les caprins du Sahel. Références bibliographiques Abdulmojeed, Y., Adebowale, E.S. and Ikhide, G.I Comparative multivariate analysis of biometric traits of West African Dwarf and Red Sokoto goats. Tropical Animal Health and Production, Springer. Baccini, A., et Besse, P Exploration Statistique. Institut de Mathématiques de Toulouse UMR CNRS C5219. Laboratoire de Statistique et Probabilités. Institut National des Sciences Appliquées de Toulouse Toulouse cedex 4., 122.

34 32 H. Marichatou et al. Bouchel, D., Lauvergne, J.J Le peuplement de l Afrique par la chèvre domestique. Revue d Elevage et de Médecine Vétérinaire des Pays Tropicaux 49 (1): Bouchel, D., Sow, R.S., Bibe, B., Tixier-Boichard, M., Lauvergne, J.J., Poivey, J.P., Rognon, X Caractérisation et cartographie des ressources génétiques caprines du Sénégal à l aide d indices phanéroptiques, d indices morphobiométriques et de marqueurs moléculaires: méthodologie et résultats préliminaires. Rencontres Recherches Ruminants 13: 257. Bourzat, D., Souvenir, P.Z., Lauvergne, J.J. et Zeuh, V Comparaison morpho-biométrique de chèvres au Nord Cameroun et au Tchad. Revue d Elevage et de Médecine Vétérinaire des Pays Tropicaux 46 (4): Dumas, R Contribution à l étude des petits ruminants du Tchad. Revue d Elevage et de Médecine Vétérinaire des Pays Tropicaux 33 (2): Ebegbulem, V.N., Ibe, S.N., Ozung, P.O. and Ubua, J.A Morphometric trait characteristics of West African dwarf goats in Abia state, south east Nigeria. Continental Journal of Agricultural Science 5 (2): 1 6. FAO Animal genetic resources data banks. Descriptor lists for cattle, buffalo, pigs, sheep and goat. FAO. Rome, Italia. Animal Production and health. 59 (2): 151. FAO Phenotypic characterization of animal genetic resources. Food and Agriculture Organization of the United Nations. Rome, Italia. Animal Production and Health Guidelines 11: 142. (accessible at Gnanda, I.B Importance socio-économique de la chèvre du Sahel burkinabé et amélioration de sa productivité par l alimentation. Thèse de Doctorat Unique. Université Polytechnique de Bobo-Dioulasso, Burkina Faso, 198. Halima, H., Michael, B., Barbara, R. and Markos, T Phenotypic characterization of Ethiopian indigenous goat populations. African Journal of Biotechnology 11 (73): Hamito, D Goat breeds of Ethiopia: A guide for identification and utilization. Ethiopia Sheep and Goat Productivity Improvement Program (ESGPIP). Technical Bulletin 23: 11. Kaiser, H.F The application of electronic computers to factor analysis. Educational and Psychological Measurement 20, Katongole, J.B.D., Sebolai, B. and Madimabe, M.J Morphological Characterization of the Tswana goat. In: S.H.B. Lebbie and E. Kagwini (eds.), Small Ruminant Research and Development in Africa. Proceedings of 3rd Biennial Conference of the African Small Ruminant Research Network, UICC, Kampala, Uganda, 5 9. Lanari, M.R., Taddeo, H., Domingo, E., Perez-Centeno, M. and Gallo, L Phenotypic differentiation of exterior traits in local Criollo Goat Population in Patagonia, Argentina. Archiv fur Tierzucht Dummerstorf 46 (4): Lauvergne, J.J., Bourzat, D., Souvenir, Z.P., Zeuh, V. et Ngo Tama, A.C Indices de primarité de chèvres au Nord Cameroun et au Tchad. Revue d Elevage et de Médecine Vétérinaire des Pays Tropicaux 46 (4): Mani, M., Marichatou, H., Issa, M., Chaïbou, I., Sow, A., Chaïbou, M., Sawadogo, J.G. Caractéristiques phénotypiques de la chèvre du sahel au Niger par analyse des indices de primarité et des paramètres qualitatifs., Animal Genetic Resources. (sous presse). Martin, A., L analyse de données. ENSIETA Réf.: (En ligne) Accès internet: Meyer, C Races d animaux d élevage en Afrique intertropicale et méditerranéenne: Les caprins. Mémento de l Agronome. CIRAD-EMVT, France, 22. Rakotomalala, R Analyse de corrélation: Étude des dépendances - Variables quantitatives. Version 1.0. Université Lumière Lyon 2. (En ligne) Accès internet: Analyse_de_Correlation.pdf. Rhissa, Z Revue du secteur de l élevage au Niger. Ministère de l Elevage, des Pêches et des Industries Animales, Niger, 115. Samuel, F.O.K. and Salako, A.E Body measurement characteristics of the West African Dwarf (WAD) Goat in deciduous forest zone of Southwestern Nigeria. African Journal of Biotechnology 7 (14): Semakula, J., Mutetikka, D., Kugonza, R.D. and Mpairwe, D Variability in Body Morphometric Measurements and Their Application in Predicting Live Body Weight of Mubende and Small East African Goat Breeds in Uganda. Middle-East Journal of Scientific Research 5 (2): Traoré, A Caractérisation des ressources génétiques caprines du Burkina Faso a l aide d indices morpho-biométriques et de marqueurs moléculaires. Thèse de Doctorat Unique. Université de Ouagadougou, Burkina Faso, 110. Traoré, A., Hamidou, H.T., Adama, K., Luis, J.R., Ivan, F., Isabel, A., Sangaré, M., Bouchel, D., Jean Paul, P., Dominique, F., Sawadogo, L. and Goyache, F Multivariate analyses on morphological traits of goats in Burkina Faso. Arch. Tierz, Dummerstorf 51 (6): Traoré, A., Tamboura, H.H., Kaboré, A., Yaméogo, N., Bayala, B. et Zaré, I Caractérisation morphologique des petits ruminants (ovins et caprins) de race locale Mossi au Burkina Faso. GRI : Verma, N.K., Dixit, S.P., Aggarwal, R.A.K., Dangi, P.S. and Joshi, B.K Phenotypic and genetic characterization of Sangamneri goat breed. Indian Journal of Animal Sciences 80 (11): Vigne, J-D., Peters, J. and Helmer, D First Steps of Animal Domestication: New archaeozoological approaches. 9th ICAZ Conference, Durham The First Steps of Animal Domestication (eds J-D Vigne, J. Peters and D. Helmer), Wilson, T.R Small ruminant production and the small ruminant genetic resource in tropical Africa. FAO. Rome. Animal Production and Health Paper, 194. Zeuh, V., Lauvergne, J.J., Bourzat, D. et Minvielle, F Cartographie des ressources génétiques caprines du Tchad du Sud-Ouest: Hauteur au garrot (HG), profondeur de thorax (PT) et indice de gracilité sous-sternale (IGs). Revue d Elevage et de Médecine Vétérinaire des Pays Tropicaux 50 (3):

35 Animal Genetic Resources, 2014, 54, Food and Agriculture Organization of the United Nations, 2014 doi: /s Morphological, reproductive and productive characteristics of Sudanese native chicken C.E. Wani 1, I.A. Yousif 2, M.E. Ibrahim 3, H.H. Musa 4 and K.M. Elamin 5 1 Department of Animal Production, Faculty of Veterinary Science, University of Bahr Elgazal, Wau, South Sudan; 2 Department of Genetics and Animal Breeding, Faculty of Animal Production, Khartoum North, University of Khartoum, Sudan; 3 Institute of Endemic Diseases, University of Khartoum, Sudan, Khartoum; 4 Department of Microbiology, Faculty of Medical Laboratories, University of Khartoum, Sudan, Khartoum; 5 Department of Poultry Production, Faculty of Animal Production, University of Khartoum, Sudan, Khartoum North Summary A structured questionnaire was administered to 40 households in each of the three localities (Bahri, Dilling and Abu-neama). Morphological characteristics of 900 chickens were physically examined. Data analysis showed that the average flock size per household was 16.7 and that it varied significantly among localities. The flock structure study showed that 46.1 percent were chicks and growers, 34.6 percent hens and percent cocks. Regarding the ecotypes, the flock was composed of 75.7, 8.8 and 15.5 percent large Beladi (LB), bare neck (BN) and Betwil (BT) dwarf, respectively. The Sudanese indigenous chicken are characterized by wide phenotypic variation. The plumage colour frequencies were mixed colour (37.8 percent), light brown (16.5 percent), black (13.1 percent), grey (3.5 percent) and white (6.2 percent). Feathering did not show a well-defined pattern; however, the few birds whose feathering pattern could be identified were crown feather (7.0 percent), partridge (3.5 percent), barred (3.5 percent), laced (3.1 percent) and frizzled (0.01 percent). The average adult body weight was ± 125 g in males and ± 70 g in females. Hens reached sexual maturity at 6.5 months and layed 13.2 eggs/clutch in clutches per year. The average egg weight was 40.8 ± 1.6 g and the hatchability rate under natural incubation was 71.8 percent. The BN ecotype had significantly (P < 0.01) poorer maternal care (57.7 percent hatchability) compared with the other types (LB and BT). Keywords: Sudan, indigenous chickens, morphological variation, phenotypic characteristics Résumé Se facilitó un cuestionario estructurado a 40 hogares en cada uno de los siguientes tres municipios: Bahri, Dilling y Abu-neama. Se examinaron las características morfológicas de 900 aves. El análisis de los datos mostró que el número medio de aves por hogar fue de 16.7, aunque varió de manera significativa entre localidades. El estudio de la estructura de los gallineros reveló que el 46.1 por ciento eran pollitos y pollos en crecimiento, el 34.6 por ciento gallinas y el por ciento gallos. En lo que concierne a los ecotipos, los gallineros se componían en un 75.7 por ciento, en un 8.8 por ciento y en un 15.5 por ciento de Gran Beladi, Cuello Pelado y Betwil Enana, respectivamente. Las gallinas autóctonas sudanesas se caracterizan por una amplia variación fenotípica. En el plumaje, se dieron los siguientes colores: color mixto (37.8 por ciento), marrón claro (16.5 por ciento), negro (13.1 por ciento), gris (3.5 por ciento) y blanco (6.2 por ciento). El plumaje no presentó un patrón bien definido; no obstante, en las pocas aves en que se pudo identificar un patrón para el plumaje, éste era con copete en un 7.0 por ciento, perdiz en un 3.5 por ciento, barrado en un 3.5 por ciento, ribeteado en un 3.1 por ciento y rizado en un 0.01 por ciento. El peso corporal medio de los adultos fue de ± 125 g en los machos y de 1187 ± 70 g en las hembras. Las gallinas alcanzaron la madurez sexual a los 6.5 meses y pusieron 13.2 huevos por nidada, con 3 a 3.5 nidadas por año. El peso medio de los huevos fue de 40.8 ± 1.6 g y la tasa de incubabilidad en incubación natural fue de un 71.8 por ciento. El ecotipo Cuello Pelado mostró un instinto materno significativamente menor (57.7 por ciento de incubabilidad, P < 0.01) que el de los otros ecotipos (Gran Beladi y Betwil). Mots-clés: Soudan, poules autochtones, variation morphologique, caractéristiques phénotypiques Resumen Un questionnaire structuré a été fourni à 40 foyers dans chacune des trois localités suivantes: Bahri, Dilling et Abu-neama. Les caractéristiques morphologiques de 900 volailles y ont été examinées. D après l analyse des données, le nombre moyen d oiseaux par foyer a été de 16.7, avec une variation significative entre localités. L étude de la structure des poulaillers a mis en évidence que le 46.1 pour cent des oiseaux étaient des poussins et des poulets en croissance, le 34.6 pour cent des poules et le pour cent des coqs. Pour ce qui est des écotypes, les poulaillers se composaient à 75.7 pour cent, 8.8 pour cent et 15.5 pour cent de Grand Beladi, Cou Nu et Betwil Naine, respectivement. Les poules autochtones du Soudan se caractérisent par une grande variation phénotypique. Les suivantes couleurs ont été observées dans le plumage: mélange de couleurs (37.8 pour cent), marron clair (16.5 pour cent), noir (13.1 pour cent), gris (3.5 pour cent) et blanc (6.2 pour cent). Bien que le plumage n ait pas présenté un motif bien défini, les Correspondence to: C.E. Wani, Department of Animal Production, Faculty of Veterinary Science, University of Bahr Elgazal, South Sudan. yousifi_2002@yahoo.com; tel.:

36 34 C.E. Wani et al. motifs suivants ont été reconnus sur le plumage d un petit nombre d oiseaux: à huppe (7.0 pour cent), perdrix (3.5 pour cent), barré (3.5 pour cent), dentelé (3.1 pour cent) et frisé (0.01 pour cent). Le poids corporel moyen des adultes a été de ± 125 g chez les mâles et de 1187 ± 70 g chez les femelles. Les poules ont atteint la maturité sexuelle aux 6.5 mois et elles ont pondu 13.2 œufs par couvée, avec 3 à 3.5 couvées par an. Le poids moyen de l œuf a été de 40.8 ± 1.6 g et le taux d éclosion, sous conditions naturelles d incubation, a été de 71.8 pour cent. Compte tenu d un taux d éclosion plus bas (57.7 pour cent, P < 0.01) par rapport aux autres écotypes (Grand Beladi et Betwil), les poules de l écotype Cou Nu sembleraient avoir un plus faible instinct maternel. Palabras clave: Sudán, gallinas autóctonas, variación morfológica, características fenotípicas Submitted 5 September 2013; accepted 11 March 2014 Introduction Poultry meat and egg production accounted for more than 28 percent of the total animal protein produced worldwide in 1997 (Tadelle et al (2003)). It is estimated that by the year 2020, the proportional contribution of poultry to the total animal protein will rise to 40 percent with the major increase being in the developing world (Delgado et al., 1999). About 1.3 billion chickens are found in Africa producing approximately 1.7 and 2.1 million metric tonnes of egg and meat respectively of which 80 percent came from indigenous stock (FAO, 2006). In Sudan, there are no reliable statistics regarding the contribution of the conventional poultry sector. However, it was estimated to constitute about 70 percent of the total production (Suleiman, 1996). Conventional poultry plays an important role in the socio-economic life of the rural community in Africa as well as in Sudan; generating incomes and contributing to food security (Gueye, 2002). Moreover, poultry raising is preferred by small holders because of the low input costs, high adaptability and resistance to diseases (Wimmers et al., 2000). Information about the flock structure of the Sudanese indigenous chicken is lacking. The average flock size of local Malawi chickens was reported to be 12.9 ± 5 birds per household (HH) (Gondwe and Wollny, 2004). In comparison, large flocks per HH (26 and 22 birds/hh) were reported for Pakistan and Senegal native chickens, respectively (Missohou, Dieye and Talaki, 2002; Javed et al., 2003). The Sudanese indigenous chicken breed was classified into three types as described by Desai (1962). The large Beladi (LB) type is the most common, available in Northern, Central, Western and Southern Sudan. The birds are of good size (adult weighing 2½ 3.0 Ibs) with a small crushed comb, plenty of plumage of many colour variations. The bare neck (BN) type is smaller than the LB, characterized by its featherless neck and many colours in plumage. The home land of BN type is Southern Sudan and it was reported to be more resistant to endemic diseases compared with other types (Wimmers et al., 2000). The Betwil (BT) type is a small dwarf bird, mainly found in the Nuba Mountains region of South Kordofan State, Sudan. It has a small compact body (adult bird weighing 1½ 2.0 Ibs) with tiny black legs. Recently in Sudan, the expansion of modern poultry farming in the home lands of indigenous chickens threatened their existence. Therefore characterization of these valuable indigenous animal resources using genetic and phenotypic methods for the purpose of conservation has become very crucial. However, limited research has been conducted to characterize the local fowl of Sudan. In the present study, morphological, productive and reproductive characteristics of Sudanese indigenous chickens were investigated. Materials and methods Study sites This study was conducted in three different localities in the Sudan, namely El Dilling, Abu-Neama and Khartoum North (Bahri). El Dilling locality is situated in South Kordofan State an area which lies 800 km south west of Khartoum (Capital of Sudan) at latitude N and longitude E with rainfall of mm. The area of South Kordofan State is covered by the Nuba Mountains that occur as a series of isolated ranges covering an area of about km 2. Abu-Naema locality in Sinnar State (Central Sudan) is an area located 400 km South of Khartoum at latitude N, longitude E and altitude 445 m above sea level. Bahri locality is in Khartoum State and includes a group of villages situated east of the River Nile. The area is semi desert or poor Savannah. Most of the people in these three localities are either farmers or livestock herders. Questionnaire administration A structured questionnaire was administered to 120 HHs from the three localities, 40 for each locality. Data captured in the questionnaires were on chicken management practices such as flock structure, housing, feeding and marketing. Production data included age at start of laying, clutch size, hatchability as well as live weights and egg weight. A total of 900 chickens were characterized through physical examinations that covered plumage colour and pattern, skin colour, shank length, comb colour and shape. The questionnaire and phenotypic characterization were supported by group discussions among villagers.

37 Characteristics of Sudanese native chicken 35 Table 1. Flock size in different localities. Locality/district Number of HHs Mean ± S.E. Range (birds) El-Dilling b ± Abu-Neama b ± Bahri a ± Total Means with the same letter are not significantly different (P < 0.05). HH, hens per household. Table 2. Flock size and structure of Sudanese indigenous chicken ecotypes. Flock by age Data analysis Descriptive statistics were used to analyse phenotypic characterization data. The frequencies of phenotypic characteristics among the chicken ecotypes were calculated by dividing the number of birds having the trait by the total number of the birds examined. Percentages were used to calculate the prevalence of a trait within each ecotype. The degree of association between prevalence of the trait and chicken ecotypes was tested using Chi-square at the 95 percent confidence level. SPSS Version 10 statistical package (SPSS Inc., Chicago, and III, USA) was used for data analysis. Results LB (n = 681) BN (n = 80) BT (n = 139) Total (n = 900) Percentage kept by HH Cock to hen ratio 1:2 1:1.2 1:1.2 1:1.8 Chicks and growers (%) Hens (%) Cocks (%) LB, large Beladi; BN, bare neck; BT, Betwil; HH, hens per household. Flock size and structure The average flock size per HH was ranging from 2 to 36 chickens/hh, the average varied significantly (P < 0.05) according to locality. Bahri locality scored the highest number of chickens per HH with an average of ± 1.29, followed by El Dilling (15.74 ± 1.3) then Abu- Neama locality (14.61 ± 1.24) (Table 1). The data collected on the Sudanese indigenous chicken flock structure Plate 1. Mix colour feather LB chicken. showed that 46.1 percent were chicks and growers, 34.6 percent hens and 19.3 percent cocks. On the basis of ecotype, the flock was composed of 75.7, 8.8 and 15.5 percent LB, BN and BT, respectively, and the average cock hen ratio was 1:1.8. Most of the HHs tend to keep more females than males and in all ecotypes surveyed, the chicks and growers formed the majority of the flocks (Table 2). Flock management practice The results revealed that 70 percent of the HHs provide night shelters to the chicken flocks they are keeping and it was observed that housing structures were made from wood or bamboo poles. The non-housed chicken mostly perched around farmer s houses (2 percent) or on the trees near the farmer s homesteads (3 percent) and laid eggs in nests placed around the vicinity or in the kitchen (20 percent) or store (5 percent) (Figure 1). Chicken were found to be scavengers for a variety of feed stuffs that included cereals, weeds, seeds, insects, worms and various herbs. Some farmers usually supplement their chicken with whole cereals once every morning. In all the areas surveyed, HHs do not keep records on feed supplement, survival and mortality of chickens. They also do not vaccinate their chickens nor use veterinary drugs for treatment or prophylaxis. The majority of the interviewed HHs (98.2 percent) declared that women were the owners of chicken with their children assisting them for management (cleaning chicken house, providing supplementary Figure 1. Chicken house provision.

38 36 C.E. Wani et al. Plate 2. Crown feather Large beladi chicken. (Source: Faculty of Animal Production, U of K, Sudan, 2011) food and water and selling eggs and chickens). However, men were responsible for decision-making (95 percent), giving permission for selling of chicken or eggs as well as home consumption or slaughter for special ceremonies. Morphological characteristics Approximately, 37.8 percent of chickens had mixed colour plumage (Plates 1, 4 and 7), with the highest percentage Plate 4. Mix colour feather Bare Neck chicken. (Source: Faculty of Animal Production, U of K, Sudan, 2011) (40.3 percent) in LB (Plate 1) and the lowest (33.5 percent) in BN (Plate 4), followed by 16.5 percent light brown (Plate 3) and 13.1 percent black (Plate 5) plumage. Feathering did not have a well-defined pattern in the majority of chicken sampled (89.9 percent). Of the patterns that could be described was partridge 3.5 percent, barred 3.5 percent, laced 3.1 percent and frizzle feather (Plate 9) 0.01 percent (Table 3). The majority of chickens (91.1 percent) that include most of the BT (Plates 6, 7 and 8) and LB (Plates 1 and 2) were well covered with feather over all the body, while 8.9 percent were BN chickens without feather in the neck region (Plates 3, 4 and 5). There were four shank colours including black 38.5 percent, white 29.6 percent, yellow 20.4 percent and greenish 11.4 percent. On the other hand, four types of combs were observed and the most frequent was single comb (48.6 percent), crush or rudimentary comb (42.3 percent), strawberry comb (5.9 percent) and rose comb (3.2 percent). Plate 3. Brown colour feather Bare Neck chicken. (Source: Faculty of Animal Production, U of K, Sudan, 2011) Plate 5. Black colour feather Bare Neck chicken. (Source: Faculty of Animal Production, U of K, Sudan, 2011)

39 Characteristics of Sudanese native chicken 37 Plate 6. Red colour feather Betwill cock. (Source: Faculty of Animal Production, U of K, Sudan, 2011) Plate 9. Frizzle Sudanese chicken. (Source: Faculty of Animal Production, U of K, Sudan, 2011) The majority of the chickens possessed combs with red colour (88.0 percent) and the rest had pale red combs (12.0 percent). The results revealed that Sudanese chicken had a high frequency of white skin colour (91.4 percent) while yellow skin colour constituted a small percentage (8.6 percent). On the other hand, a well grown wattle was common in the majority of chickens (89.9 percent), whereas few lacked wattles or had rudimentary wattles (10.1 percent). However, it was observed that mainly males had projected wattles. Similarly, 70.7 percent of the sampled chickens had rudimentary spur while 29.3 percent (mainly males) had large spurs. The ear lobes were Plate 7. Mix colour feather Betwill cock. (Source: Faculty of Animal Production, U of K, Sudan, 2011) Table 3. Plumage pattern and colour characteristics in percent for different Sudanese indigenous chicken ecotypes. Morphology Trait LB (n = 681) BN (n = 80) BT (n = 139) Overall (n = 900) Plate 8. Light brown feather betwill. (Source: Faculty of Animal Production, U of K, Sudan, 2011) Plumage colour Plumage pattern Mix colour Black Red Brown Light brown White Grey Total Partridge Barred Laced Frizzle feather Undefined Total LB, large Beladi; BN, bare neck; BT, Betwil.

40 38 C.E. Wani et al. Table 4. Morphological characteristics in percent for different Sudanese indigenous chicken ecotypes. Morphology Trait LB (n = 681) BN (n = 80) BT (n = 139) Overall (n = 900) Comb Crush pattern Single Strawberry Rose Comb colour Red Pale Skin colour White Yellow Shank colour White Black Yellow Greenish Earlobe White colour Red Head feather Crown feather Normal Wattle Presence Rudimentary Spur Large 29.3 Rudimentary 70.7 LB, large Beladi; BN, bare neck; BT, Betwil. mostly not prominent and they were mainly red (64.4 percent) and white (35.6 percent) in colour. Crown head feather chicken is known in Sudanese colloquial language as (Abu-Guja). These had a relatively low frequency among the Sudanese indigenous chickens sampled (6.9 percent), while the rest had normal feather on the head (Table 4). The results revealed that the BT chicken ecotype possessed the shortest shank length (4.5 ± 0.47 cm), showing a significant difference (P < 0.05) when compared with LB (7.9 ± 1.0 cm) and BN (7.7 ± 0.93 cm). Reproductive performance Results revealed that LB and BN pullets reached sexual maturity at 6 months of age, whereas BT pullets reached sexual maturity at 7 months of age (Table 5). Moreover, most of the hens laid on average eggs/hen/clutch, with BN being (14.0 eggs/hen/clutch) significantly different (P < 0.01) from LB and BT. The majority of hens produced clutches of eggs per year with the inter-clutch interval ranging from 2.0 to 3.5 months. On the other hand Table 5. Reproductive performances of the Sudanese indigenous chicken ecotypes. Ecotypes LB BN BT Total Age at maturity (month) 6 b 6 b 7 a 6.3 No. of eggs laid/clutch 13.2 b 14.0 a 13.0 b No. of clutches/year Inter-clutch interval (month) No. of eggs incubated 9.87 a 9.9 a 9.27 a 9.68 No. of chicks hatched 7.7 a 5.2 b 7.27 a 6.72 No. of eggs wasted 2.17 b 4.7 a 2.0 b 2.96 Hatchability (%) a 57.7 b 79.2 a 71.8 No. of chicks weaned 5 b 2 c 6 a 4.3 Hatch to wean period (months) Means with the same letter are not significantly different (P < 0.05). LB, large Beladi; BN, bare neck; BT, Betwil. hatchability was 78.34, 57.7 and 79.2 percent in LB, BN and BT, respectively, with BN being significantly lower (P < 0.01) than the other two types. Hatch to rearing period averaged 2.5 months. These results indicated that there were highly significant differences (P < 0.01) among the chicks reared by the respective ecotype hens up to weaning, with the LB exhibiting excellent maternal care and the BN being the poorest in that respect (Table 5). Body weight The average male body weights of LB, BN and BT chicken were ± 312.0, ± and ± g, respectively, with an overall average of ± 125 g. While females corresponding body weights were ± 53.5, ± 98.7 and ± g with an overall average of ± 45 g. Sexual dimorphism was evident among Sudanese chicken types with males being significantly (P < 0.05) heavier than the females in the three ecotypes. However, comparisons among ecotypes revealed that the BT ecotype was significantly lighter (P < 0.05) than the other two ecotypes which were almost similar in their body weight (Table 6). Egg characteristics Average egg weights of LB, BN and BT chickens were 41.3 ± 3.1, 42.4 ± 1.5 and 38.8 ± 5.1 g, respectively. BT eggs were significantly (P < 0.05) smaller than the eggs of the other two ecotypes, LB and BN. The majority (95 percent) of the Sudanese chicken lay white eggs and the Table 6. Average body weights (g) for the mature Sudanese indigenous chicken ecotypes. Types LB BN BT Total No Male body weight a ± a ± b ± ± 125 Female body weight a ± a ± b ± ± 45 Means with the same letter are not significantly different (P > 0.05). LB, large Beladi; BN, bare neck; BT, Betwil.

41 Characteristics of Sudanese native chicken 39 Table 7. Egg characteristics of the Sudanese indigenous chicken ecotypes. Trait LB (n = 60) BN (n = 60) BT (n = 60) Total (n = 180) Egg weight (g) 41.3 a ± a ± b ± ± 1.6 Egg shell colour (%) White Light brown Yolk colour (%) Yellow White LB, BN, BT, large Beladi, bare neck and Betwil, respectively. Means with the same letter are not significantly difference (P > 0.05). rest 5 percent lay light brown eggs. Yolk was 100 percent yellow in all the ecotypes (Table 7). Discussion The flock structure of Sudanese indigenous chicken observed in this study was 46.1 percent chicks and growers, 34.8 percent hens and 19.3 percent cocks. Obviously, the HHs tend to keep females more than males. Males are usually sold and are slaughtered in social and traditional functions. The average flock size of Sudanese indigenous chickens was birds/hh which was higher than that of Malawian indigenous chicken flock size of 12.9 birds/ HH (Gondwe and Wollny, 2004), but lower than that of Pakistan indigenous chickens which amounted to 26 birds/hh (Javed et al., 2003) and Senegal indigenous chickens, 22 birds/hhs (Missohou, Dieye and Talaki, 2002). About 91 percent of village chicken in Sudan are kept by women suggesting gender bias in raising chicken at the village level and that they are a source of women empowerment. This finding is in agreement with that of Kitalyi (1998) and McAinsh et al. (2004). In the surveyed areas, chicken are mainly used for HH consumption, as well as for the purpose of generating income. This observation is in agreement with the conclusion drawn by Gueye (2002). The observation that all chickens scavenged for a variety of feeds is similar to the results reported by Maphosa et al. (2004) and Muchadeyi et al. (2004). However, this method of feeding does not optimize the utilization of the limited feed resources since younger and weaker birds have to compete with mature birds and other scavenging farm animals. This could be one of the reasons for the observed reduced number of weaned chicks coupled with other factors such as predators and the lack of proper housing structures for the birds. Seventy percent of the HHs construct night shelters only and the non-housed birds are either perched on trees or elsewhere around the farmer s homestead. No records were kept for production activities and disease management, which makes the village chicken more vulnerable to diseases and predators. The inheritance of qualitative traits, such as plumage colour, plumage pattern, skin colour, egg colour, shank length and comb type, is controlled by a single/few pairs of major genes. Such traits usually influence the preferences of consumers. In the present study, we found a high variability in plumage colour within and between the three identified Sudanese indigenous chicken ecotypes; these variable colourations provide aid for camouflage against predators particularly in the rural areas. However, with the modern trends of poultry production the importance of plumage colour goes beyond the need for camouflage. White feathering has become a desirable characteristic in poultry breeding particularly in broilers because it eliminates dark feather pigment deposits in the skin and is consequently preferred for the clean appearance of carcass and cut-up parts. Some important alleles such as frizzle and BN genes which are genetically conserved for their special utility in tropical environments were found to exist among the Sudanese indigenous chickens even though they are at low frequencies. There was a high diversity in colour and type of combs and earlobes observed between and within the Sudanese indigenous ecotypes. The most frequent comb type was single comb, it was more common among LB and BN chickens, followed by crush comb type which was more common among BT chickens. According to Nesheim, Austic and Card (1979), the size and colours of combs and wattles are associated with gonad development and secretion of sex hormones. Large wattle and long legs are important morphological traits that allow better heat dissipation in the hot tropical environment. On the other hand, the short tiny legs and the crush combs of BT chickens may be important for adaptation to the mountainous nature of their home land in the Nuba region of South Kordofan State. Generally, the Sudanese indigenous chickens are characterized by small body size. However, the sub normal body size of the BT chicken is an indication to the presence of dwarfism, which is due to a sex-linked recessive gene that tends to reduce body and egg size. This trait is promising in the sense that feed requirements for maintenance are lowered and thereby the efficiency of feed utilization is increased (Bordas and Merat, 1984; Merat, 1990). The age at the start of laying for Sudanese indigenous chicken varied from 180 to 210 days, it is higher than that of Egyptian indigenous chicken (Hanafi and Labban, 1984) and higher than that of the modern strains, such as

42 40 C.E. Wani et al. Rhode Island Red and Rhode Island White (Simeonovova et al., 1989). This late age of sexual maturity may be due to variation in management practices, environmental conditions and genetic potential (El-Zubeir, Salih and El-Amin, 1991; Aganga et al., 2000). The number of eggs per clutch in the present study is very close to that reported for native chickens from other areas such as South Darfur region (Beladi chicken) (9 13 eggs/clutch), Morocco (12 eggs/clutch) and Bangladesh (10 16 eggs/ clutch) (Wilson, 1979; Sazzad, 1986; Benabdeljelil and Arfaoui, 2001). However, the low rate of egg production in the indigenous chickens may be associated in part with the behaviour of broodiness. Hatchability in BN chicken was found to be lower than that of the other ecotypes, and this might possibly be due to the presence of the Na allele in the genotype of the embryos. Horst (1980) mentioned that there was a 13 percent reduction of embryonic survival associated with the Na allele. In addition, it might also be due to poor hatchability. BN hens also seem to have low numbers of weaned chicks, although it was proved by Horst (1980) that there was no difference in post embryonic mortality of naked neck chicken compared with normally feathered birds at different ages. All these aspects of poor performance of BN chickens classify them as poor mothers. The hatchability of the indigenous Sudanese chicken under natural incubation was moderately low when compared with that of White and Brown Hisex eggs hatched under artificial incubation in Sudan (El-Agraa, 1988). In general, a rural-oriented breeding strategy to produce a dual purpose chicken seems to be vital in improving the productive and reproductive potential of the indigenous chickens for the benefit of rural communities. This could be achieved through implementing a selective breeding programme for several generations, to be followed by limited cross-breeding involving improved tropical or exotic breeds that proved to be adaptive to tropical environmental conditions. The newly improved genotypes can be tested, propagated and disseminated to rural HHs accompanied with concomitant withdrawal of the existing unimproved flocks. This strategy must be supported by conducting an active extension programme to increase the awareness of HH members of the importance of raising these genetically improved flocks under relatively improved management. Conclusions The Sudanese indigenous chicken ecotypes exhibited wide phenotypic variation. Characterization information is important for implementing conservation and improvement programmes for sustainable utilization. Estimates of production performance obtained from the extensive rural systems were significantly lower than those obtained under relatively improved managerial conditions. This necessitates exerting efforts to improve management systems, particularly feeding and health care in rural areas. Several qualitative traits of economic value were reported for the Sudanese indigenous chicken populations which can be helpful for preservation and future commercial exploitation. The field survey in this study revealed that the local chickens of Sudan are available in abundance in their homelands with the exception of the BT ecotype in the Nuba Mountains which seemed to be endangered as a result of conflicts in the region and the ensuing displacement of communities. This ecotype requires special consideration for propagation and conservation. Acknowledgements Authors would like to acknowledge the financial support from the University of Bahr Elgazal, South Sudan and the Ministry of Higher Education and Scientific Research, Sudan. Thanks are also extended to the staff, technicians and postgraduate students of the Faculty of Animal Production, University of Khartoum and Faculty of Veterinary Science, University of Bahr Elgazal who offered us technical assistance to conduct this research. References Aganga, A.A., Omphile, U.J., Malope, P., Chabanga, C.H. & Motsamai, L.G Traditional poultry production and commercial broiler alternative for small-holder farmer in Botswana. Livest. Res. Rural Dev., 12(4): 1 8. Benabdeljelil, K. & Arfaoui, T Characterization of Beladi chicken and turkey in rural poultry flocks of Morocco: current state and future outlook. Anim. Genet. Resour. Inf., 31: Bordas, A. & Merat, P Correlated responses in a selection experiment on residual feed intake of adult Rhode-Island Red cocks and hens. Ann. Agric. Fenn., 23: Delgado, C., Rosegrant, M., Steinfeld, H., Ehui, S. & Courbois, C Livestock to 2020: the next revolution. Food Agriculture and the Environment Discussion Paper 28. Rome, FAO. Desai, D.K The status importance and development of poultry keeping in the Sudan. Sudan J. Vet. Sci. Anim. Husb., 3: El-Agraa, S.M Fertility and hatchability of Hisex brown parent layers under control environment in Sudan. In Symposium of Exotic White and Brown Breeds Under Sudan Condition, League of Arab States, Arab Organization for Agricultural Development (AOAD), Khartoum, October 1988, pp El-Zubeir, E.A., Salih, M.E. & El-Amin, H Early reproductive characteristics for Leghorn pullets reared on diet containing sorghum gluten feed. Sudan J. Anim. Prod., 4: FAO Statistical year book Vol. 2. Rome, FAO. Gondwe, T.N. & Wollny, C.B.A With Globalization will low input production system prevail in Malawi? The case study of rural chicken. In Book of Abstract of the 55th Annual Meeting of the European Association for Animal Production. 5 9 September 2004, Bled, Slovenia. Wageningen, Wageningen Academic Publishers, The Netherlands.

43 Characteristics of Sudanese native chicken 41 Gueye, F.H.E Employment and income generation through family poultry in low income food-deficit countries. World Poult. Sci. J., 58: Hanafi, M.S. & El-Labban, A.F.M On estimating genetic parameters of partial egg production records and other related traits in pullets of Dokki-4 chicken produced from triallel mating. Egypt. J. Anim. Prod., 24: Horst, P Genetically perspectives for poultry breeding on improved productive ability to tropical conditions. In 2nd World Congress of Genetics Applied to Animal Production, Vol. 8, pp Javed, K., Farooq, M., Mian, M.A., Durrani, F.R. and Mussawar, S Flock size and egg production performance of backyard chicken reared by rural women in Peshawar, Pakistan. Livest. Res. Dev., 15(11). Kitalyi, A.J Village chicken production system in rural Africa: Household security and gender issues. Animal Production and Health Paper 142. Rome, FAO. McAinsh, C.V., Kusina, J., Madsen, J. & Nyoni, O Traditional chicken production in Zimbabwe. World Poult. Sci. J., 60: Maphosa, T., Kusina, J., Makuz, N.T.S. & Sibanda, S A monitoring study comparing production of village chickens between communal (Nharira) and small-scale commercial (Lancashire) farming areas in Zimbabwe. Livest. Res. Rural Dev., 16(7). Merat, P Pleotropic and associated effects of major genes. In R.D. Crawford, ed. Poultry breeding and genetics, pp Amsterdam, Elsevier. Missohou, M., Dieye, P.N. & Talaki, E Rural poultry production and productivity southern Senegal. Livest. Res. Rural Dev., 14(2): 1 8. Muchadeyi, F.C., Sibanda, S., Kusina, N.T., Kusina, J. & Makuza, S.M The village chicken production system in Rushinga District of Zimbabwe. Livest. Res. Rural Dev., 16(6): Nesheim, C.M., Austic, E.R. & Card, E.L Poultry production. 12th edn, Philadelphia, PA, Lea & Febiger, pp Sazzad, M.H Reproductive performance of Desi (indigenous) hens under scavenging and intensive system of rearing. In Proceedings of the First Annual Livestock Research Workshop. Savar, Dhaka, Bangladesh, Bangladesh Livestock Research Institute, pp Simeonovova, J., Ingr, I., Jerabek, S., Wnterova, J. & Dvorakova, V An evaluation of eggs laid by Rhode Island Red (RIR) and Rhode Island White (RIW) hens. Poult. Abstr., 16(7): 203. Suleiman, M.E.F Egg characteristics, genetic and phenotypic relationships of body weight at various ages in indigenous chicken. University of Khartoum (MSc thesis). Tadelle, D., Millon, T., Alemu, Y. & Peters, K. J Village Chicken Production System in Ethiopia: 2. Use Patterns and Performance valuation and Chicken Products and socio-economic Functions of Chickens. Livestock Research for rural. Wilson, R.T Studies on the livestock of Southern Darfur, Sudan VII. Production of poultry under simulated traditional conditions. Trop. Anim. Health Prod., 11: Wimmers, K., Ponsuksili, S., Hardge, T., Valle-Zarate, A., Mathur, P. K. & Horst, P Genetic distinctness of Africa, Asian and South American Local chickens. Anim. Genet., 31:

44 Animal Genetic Resources, 2014, 54, Food and Agriculture Organization of the United Nations, 2014 doi: /s Phenotypic characterization of indigenous chicken ecotypes in the north Gondar zone, Ethiopia Addis Getu 1, Kefyalew Alemayehu 2 and Zewdu Wuletaw 3 1 Department of Animal Production and Technology, Faculty of Agriculture, Woldia University; 2 Department of Animal Production and Technology, College of Agriculture and Environmental Sciences, Bahir Dar University; 3 Sustainable Land Management Organization (SLM), Bahir Dar, Ethiopia Summary An exploratory field survey was conducted in north Gondar zone, Ethiopia to identify and characterize the local chicken ecotypes. Seven qualitative and 12 quantitative traits from 450 chickens were considered. Chicken ecotypes such as Naked neck, Gasgie and Gugut from Quara, Alefa and Tache Armacheho districts were identified, respectively. Morphometric measurements indicated that the body weight and body length of the Naked neck and the Gasgie ecotypes were significantly (P < 0.01) higher than the Gugut ecotypes except in shank circumstances (circumference). Sex and ecotypes were significant (P < 0.01) sources of variation for both body weights and linear body measurements. The relationship of body weight with other body measurements for all ecotypes in both sexes was highly significant (r = 0.67, P < 0.01). Some traits like the spur length (r = 0.64, P < 0.01) for males and (r = 0.59, P < 0.01) for females of the Naked neck chickens are significantly correlated with body weight. Therefore, highly correlated traits are the basic indicators for the estimation of the continuous prediction of body weight of chicken. Identification and characterization of new genetic resources should be employed routinely to validate and investigate the resources in the country. Keywords: chicken ecotypes, (Naked Neck), Gasgie, Gugut, north Gondar zone Résumé En la zona Norte del distrito de Gondar (Etiopía), se llevó a cabo un estudio exploratorio de campo con el que se pretendía identificar y caracterizar los ecotipos locales de gallina. Se consideraron siete parámetros cualitativos y doce cuantitativos en una muestra de 450 gallinas. Se identificaron ecotipos de gallina tales como Cuello Pelado, Gasgie y Gugut en los distritos de Quara, Alefa y Tache Armacheho, respectivamente. Las medidas morfométricas indicaron que el peso y la longitud corporal de los ecotipos Cuello Pelado y Gasgie eran significativamente (P < 0.01) mayores que los del ecotipo Gugut, con la excepción de la circunferencia de los tarsos. El sexo y el ecotipo resultaron ser fuentes significativas (P < 0.01) de variación, tanto para el peso corporal como para las medidas lineales del cuerpo. Para todos los ecotipos y en ambos sexos, el peso corporal estuvo relacionado, de manera muy significativa (r = 0.67, P < 0.01), con el resto de medidas corporales. En las gallinas de Cuello Pelado, algunos rasgos, como la longitud de los espolones (r = 0.64, P < 0.01 en los machos y r = 0.59, P < 0.01 en las hembras), presentaron una correlación significativa con el peso corporal. Por tanto, los parámetros altamente correlacionados son indicadores básicos para estimar la evolución del peso corporal de las gallinas. La identificación y caracterización de nuevos recursos genéticos debería realizarse de manera rutinaria para validar y conocer los recursos presentes en el país. Mots-clés: ecotipos de gallina, Cuello Pelado, Gasgie, Gugut, Norte del distrito de Gondar Resumen Dans la zone Nord du district de Gondar (Éthiopie), une étude d exploration a été menée sur le terrain pour identifier et caractériser les écotypes locaux de poule. Sept traits qualitatifs et douze quantitatifs ont été évalués sur un total de 450 poules. Des écotypes tels que Cou Nu, Gasgie et Gugut ont été identifiés dans les districts de Quara, Alefa et Tache Armacheho, respectivement. Les mesures morphométriques ont indiqué que le poids corporel et la longueur du corps étaient significativement (P < 0.01) plus élevés chez les écotypes Cou Nu et Gasgie que chez l écotype Gugut, exception faite du tour des tarses. Le sexe et l écotype ont été des sources de variation significatives (P < 0.01) aussi bien pour le poids corporel que pour les mesures linéaires du corps. Pour tous les écotypes et chez les deux sexes, des corrélations hautement significatives (r = 0.67, P < 0.01) ont été mises en évidence entre le poids du corps et les autres mesures corporelles. Chez l écotype Cou Nu, certains traits, tels que la longueur des ergots, ont été significativement corrélés avec le poids corporel, tant pour les mâles (r = 0.64, P < 0.01) que pour les femelles (r = 0.59, P < 0.01). Ainsi, les traits hautement corrélés sont des indicateurs de base pour estimer l évolution du poids corporel des poules. L identification et la caractérisation de nouvelles ressources génétiques devraient se faire systématiquement pour valider et connaître les ressources présentes dans le pays. Palabras clave: écotypes de poule, Cou Nu, Gasgie, Gugut, Nord du district de Gondar Submitted 6 September 2013; accepted 27 February 2014 Correspondence to: Addis Getu, Department of Animal Production and Technology, Faculty of Agriculture, Woldia University. addisgetu2002@yahoo.com 43

45 44 A. Getu et al. Introduction Ethiopia takes the lead in livestock population and is a gateway of domestic animal migration from Asia to Africa, thus rolled for widespread distribution and huge population size in the country (Halima, 2007). Poultry contribute socioeconomic roles in food (security), generation of additional cash and income, etc. (Kondombo, 2005; Salam, 2005). Therefore, almost all rural and many peri-urban families keep a small flock of scavenging chickens (Jens et al., 2004). In Ethiopia, the population of chickens is estimated at about 49.3 million of which 97.3, 0.38 and 2.32 percent are indigenous, hybrid and exotic breeds, respectively (CSA, 2011). Indigenous chickens have a good potential to adapt in different agro-ecological conditions throughout their habitual management system (Tadelle and Alemu, 1997). Local chickens are non-descriptive type and show a large variation, which might be attributed to their widespread distribution (Tadelle, Alemu and Peters, 2003; Halima, 2007; Fisseha, Abera and Tadelle, 2010). Indigenous chickens are underestimated because of their poor performance. To this effect they have been neglected and little attention has been given to them by researchers, development workers and policy-makers (Tadelle, 2003). Some researchers (Tadelle, 2003; Halima, 2007; Dana et al., 2009) have made phenotypic and genetic characterization of indigenous chicken in some parts of Ethiopia. Poultry production and market system were studied in Southern Ethiopia by Mekonnen (2007); characterization of poultry productivity and market system by Bogale (2008) and genetic parameters on Horro chickens for weights and egg production traits was conducted by Dana, van der Waaji and Johan (2010). However, comprehensive genetic resources identification in the remote districts of Northern Gondar zone in general and identification, and characterization of new local chicken ecotypes in particular were not studied. Therefore, the objective of this study was to identify and characterize the new local chicken ecotypes in the north Gondar zone of Ethiopia. Materials and methods Description of the study area The study was conducted in three districts of the north Gondar zone (Quara, Alefa and Tache Armacheho) of Ethiopia (Figure 1). The altitude of the north Gondar zone ranges from 528 to 4620 m mean sea level (MSL) with annual rainfall of mm and temperature ranging from 44.5 to 10 C. Quara district is located in the western part of the north Gondar zone between and 12 21N latitude and between and E longitude. It is 1123 km from Addis Ababa and 324 km from Gondar town and at an altitude between 528 and 654 m above MSL. The annual temperature ranges from 25 to 44 C with annual rainfall range of mm (CSA, 2011). Alefa district is located in the southwest of Gondar town and 909 km from Addis Ababa with the temperature of C and annual rainfall of mm. Armacheho district is located 814 km northwest of Addis Ababa and 65 km northwest of Gondar town at an altitude of m above MSL with the temperature of C and annual rainfall of mm (CSA, 2011). Data collection methods In addition to exploratory field survey, semi-structured questionnaires and participatory rural appraisal, focus group discussion, field observation, trait characterization and body measurements were employed to derive the required information. For the morphological and biometrical measurements, all matured chicken ecotypes n = 450, 150 males and 300 females, were measured. Qualitative traits such as plumage size, body shape, comb type, shank colour, skin colour, head shape and eye colour were documented through direct visualization. Whereas measurable traits like body weight (kg), body length, wing span, shank length and circumference, wattle length and width, keel length, spur length, beak length, comb length and width were measured using a spring balance and a measuring tape in cm, measuring to the nearest two digits (FAO, 2011). Data management and statistical technique Data from personal observation and focus group discussions were summarized and synthesized by researchers, whereas, the other quantitative and qualitative data were analysed using SAS software version 9, Particularly, general linear model was used to analyse quantitative traits (SAS, 2002). Tukey s comparison test was used to compare the sub factor brought significant difference. The model was used for body weight and linear body measurement of chickens ecotypes by considering the fixed effects of sex and ecotype. Y ijk = m + A i + D j + AD ij + e ijk, (1) where Y ijk is the observed body weight and linear body measurement of chickens; μ is the overall mean; A i is the fixed effect of ith eco-type (i = 1, 2 and 3); D j is the effect of kth sex ( j = male and i = female); AD ji is the fixed effect interaction of ith eco-type with jth sex; and E ijk is the random residual error. Multiple liner regression analysis was performed to predict the body weight of matured cocks and hens using 11 linear body measurements (independent variables) in each ecotype Y j = b 0 + b 1 X 1 + b 2 X 2 + b 3 X 3 + b 4 X 4 + b 5 X 5 + b 6 X 6 + b 7 X 7 + b 8 X 8 + b 9 X 9 + b 10 X 10 + b 11 X 11 + e ijk, (2) where Y j is the dependent variable or predicted mean body weight of chickens; β 0 is the intercept; X 1, X 2, X 3, X 4, X 5,

46 Chicken ecotypes in Ethiopia 45 Figure 1. Map of the study area (Amhara region), and the three districts where the ecotypes identified are indicated in rectangular shape. X 6, X 7, X 8, X 9, X 10 and X 11 are the independent variables for wing span, body length, shank length, shank circumference, keel length, spur length, beak length, wattle length, wattle width, comb length and comb width, respectively. β 1, β 2, β 3, β 4, β 5, β 6, β 7, β 8, β 9, β 10 and β 11 are partial regression coefficients of the variables X 1, X 2, X 3, X 4, X 5, X 6, X 7, X 8, X 9, X 10 and X 11 and e ijk is the residual error. tips (13.3 percent). About 53 percent of the birds have white skin colour, 66 percent have single combs, 34 percent have rose combs and 70 percent have plain headed facial appearance (Table 1 and Figure 2). The other peculiar features of this ecotype include aggressive behaviour, higher feed intake, good productive and reproductive performance, tolerance to common diseases and higher dressing percentage. Results Naked neck chicken Naked neck chickens are found in a very hot ecological zone of the Quara district (Figure 2) and are maintained under scavenging system with small feed supplementation and sheltered outside the family house (perch). These chickens having predominantly white (28 percent) and red (20 percent) body plumage colours. However, they have heterogeneity and diverse additional plumage colour like red-brownish (0.7 percent), white with red tips (5.3 percent), black with white tips (10.7 percent), black (7.3 percent), multicolour (5.3 percent) and white black red Gasgie chicken ecotype Gasgie chicken ecotype is distributed in the Alefa district and (Figure 3) most households keep this chicken sheltered in the family house during the night, while they spend the day scavenging in the backyards supplemented with grains and food leftovers. The chicken have a predominantly red (32 percent) body plumage colour though have other diverse plumage colours such as white (9.3 percent), redbrownish (9 percent), white with red tips (9 percent), black with white tips (9 percent), black (5.3 percent), multicolour (4 percent) and white black red tips (1.3 percent). About 60.7 percent of the ecotypes are rose combed (Figure 3 and Table 1). Long neck (especially males), short

47 46 A. Getu et al. Figure 2. Typical Naked neck male (right) and female (left) chicken ecotypes. Table 1. Description of dominant body plumage colours, head shape, comb type and skin colour of newly reported indigenous chicken ecotypes (N = 450). Character Attributes Ecotypes by proportions and their associations P-value Naked neck = 150 Gasgie N = 150 Gugut N = 150 Overall (%) Overall N = 450 Cramer s V Plumage colours White black and red ns tips (Kiy Tikur Teterma) (0.67) (1.34) (2.67) Black with white tips (Tikur Teterma) ns (10.70) (8.67) (6.00) White with red tips (Kiy Teterma) ns (5.33) (8.67) (8.00) Grayish mixture (Gebsema)*** (10.70) b (22.00) a (10.00) b Red-brownish (Kokima)** (13.30) a (8.67) a (2.00) b Multicolour (Ambesa)* (5.33) a,b (4.00) b (10.00) a Black (Tikur)*** (7.33) b (5.33) b (22.00) a White (Nech)*** (28.00) a (9.33) b (9.33) b Red (Kiy)*** (18.70) b (32.00) a (30.00) a Head shape Plain (Ebaberas) ns (70.00) a (65.33) a (76.00) a Crest (Gutya) ns (30.00) (34.67) (24.00) Comb type Doublex (V-shape)* NA (9.33) b (24.00) a Single*** (66.00) a (27.33) b (30.00) b Rose*** (34.00) b (60.70) a (38.70) b Pea ns NA (2.67) a (7.33) a Skin colour Yellow*** (66.00) a (24.70) b (67.30) a White*** (23.30) b (66.00) a (20.70) b Black* (4.00) a,b (0.67) a (8.00) b Green* (6.00) a NA (1.33) b Red* (0.67) a (8.67) b (3.33) a Eye colour Orange ns (0.67) a (4.67) a (4.67) a Black ns (6.00) a NA (3.33) a Purl*** (1.33) b NA (3.33) a Red*** (92) b (95.30) a (88.70) b Body shape Triangular*** (40.7) a (15.30) b (30.70) a Blocky*** (58.00) a (31.30) b (49.30) a Wedge*** (1.30) b (5.33) b (20.00) a Shank colour Yellow*** (68.00) b (66.00) a,b (51.30) a White*** (16.70) a,b (14.00) b (26.00) a Black ns (7.33) a (9.30) a (8.00) a Green ns (7.33) (10.00) (8.00) Red* (0.67) b (0.67) b (5.33) a NA, not available and different superscripts within a row indicate significantly different means (P<0.05).

48 Chicken ecotypes in Ethiopia 47 Figure 3. Typical Gasgie male (left) and female (right) chicken types. weaning time, docile and good productive and reproductive performances are the unique features of this ecotype. Gugut chicken ecotype Gugut chickens (Figure 4) are distributed in the Debresina area of Tache Armacheho district. Most of the households keeping these chickens provide separate hanging shelters during night time to protect them from hyperthermia. Dense feathers from the neck, absence of wattle in female, low performance, short beak length, ability to resist endemic disease, small body size, passive and easily exposed to predators are the unique behaviours of the ecotype. These chicken are dominated by white (22 percent) body plumage colour and other diverse plumage colour such as multicolour (10 percent), red (9.3 percent), black (9.3 percent), white with red tips (8 percent), red-brownish (2 percent), black with white tips (2 percent) and white black red tips (2 percent). The investigated chicken ecotype showed additional heterogeneity in quantitative traits (Table 1). Figure 4. Typical Gugut male and female chicken types.

49 48 A. Getu et al. The results indicated that the dominant average plumage colour of newly identified average three local chicken ecotypes were percent red followed by percent white and percent greyish mixture. About 44.4 percent are rose comb type, percent have white skin colour, percent have blocky body shape and percent are plain headed and are the most dominant observable traits of Gugut chickens ecotypes. About 34, 60.7 and 38.7 percent of chicken from Naked neck, Gasgie and Gugut ecotype, respectively, were characterized by rose comb type. The proportion of plain head shape in chicken populations of Naked neck, Gasgie and Gugut was comparable with 70, 65.3 and 76 percent, respectively (Table 1 and Figure 3). This variation could be adaptation fitness to their environment (Dana, 2011). Quantitative traits of (Naked neck, Gasgie and Gugut) chickens A total of 450 adult hens and cocks with 12 measurable parameters such as wing span (WS), shank length (SL), shank circumference (SC), body length (BL), comb length (CL), comb width (CW), wattle length (WL), wattle width (WW), beak length (bl), spur length (sl), keel length (KL) (cm) and body weight (Wt) (kg) for different sexes were considered. The least-squares mean of body weight and body measurements of Naked neck, Gasgie and Gugut chickens with Honestead significant difference comparison tests are presented in Table 2. The overall least-squares mean of wing span, shank length, shank circumference, body length, comb length, comb width, wattle length, wattle width, beak length, spur length, keel length (in cm) and body weight (in kg) were (37.04 ± 0.13), (7.79 ± 0.15), (3.78 ± 0.07), (35.79 ± 0.09), (2.76 ± 0.09), (1.68 ± 0.04), (1.76 ± 0.06), (1.51 ± 0.06), (2.03 ± 0.02), (0.18 ± 0.02), (8.24 ± 0.09) and (1.46 ± 0.01), respectively. Overall sex effect body weight mean squares of male and female chickens were 1.63 ± 0.03 and 1.37 ± 0.02 kg, respectively. Naked neck chicken male body weight (1.78 ± 0.31 kg) is significantly (P < 0.01) higher than the Gugut male chicken 1.40 ± 0.04 kg but not body weight of Gasgie 1.71 ± 0.05 kg. Further, the Naked neck cocks and hens were found to have significantly taller shank length of 9.61 ± 1.03 and ± 1.10 (cm), respectively, than the Gugut. However, shank circumference of Gugut cocks and hens are inversely superior in shank circumferences than Naked neck and Gasgie male and female Table 2. Comparison (LSM ± SE) of body weight (kg) and linear body measurements (cm) and extraction effect of independent variables of the three indigenous chickens. Parameters Sex Naked neck Gasgie Gugut type CV% P-value Overall mean Grand mean Sample size M F Effects and levels LSM ± SE LSM ± SE LSM ± SE LSM ± SE LSM ± SE WS M ± 2.6 a ± 0.42 a ± 0.23 b ** ± 0.24 a ±.13 F ± 2.36 a ± 0.26 a ± 0.18 b ** ± 0.14 b SL M 9.61 ± 1.03 a 7.25 ± 0.10 b 7.37 ± 0.73 b ** 8.08 ± 0.11 a 7.79 ± 0.15 F ± 1.10 a 6.80 ± 0.06 c 7.08 ± 0.05 b ** 7.64 ± 0.07 b BL M ± 2.14 a ± 0.34 a 35.2 ± 0.09 b ** ± 03 a ± 0.09 F ± 2.61 a ± 0.26 b ± 0.21 b ** ± 0.16 b CL M 3.25 ± 0.87 a 3.16 ± 0.12 a 3.08 ± 0.09 a ns 3.16 ± 0.07 a 2.76 ± 0.09 F 2.99 ± 3.68 a 2.28 ± 0.07 b 2.40 ± 0.06 a,b * 2.55 ± 0.13 b CW M 2.11 ± 0.82 a 1.93 ± 0.13 a 2.19 ± 0.05 a ns 2.08 ± 0.07 a 1.68 ± 0.04 F 1.78 ± 0.85 a 1.07 ± 0.06 b 1.59 ± 0.06 a ** 1.48 ± 0.04 b WL M 2.76 ± 0.69 a 2.70 ± 0.14 a 1.83 ± 0.23 b ** 2.43 ± 0.07 a 1.76 ± 0.06 F 2.44 ± 0.80 a 1.84 ± 0.04 b NA ** 1.42 ± 0.07 b WW M 2.76 ± 1.01 a 2.32 ± 0.16 b 1.45 ± 0.09 c ** 2.17 ± 0.09 a 1.51 ± F 2.34 ± 1.03 a 1.19 ± 0.05 b NA ** 1.18 ± 0.07 b bl M 2.42 ± 0.45 a 2.00 ± 0.02 b 1.85 ± 0.10 c ** 2.09 ± 0.03 a 2.03 ± 0.02 F 2.28 ± 0.60 a 1.93 ± 0.0 b 1.78 ± 0.02 c ** 1.99 ± 0.02 b sl M 0.66 ± 0.8 a 0.49 ± 0.10 a 0.17 ± 0.09 b ** 0.44 ± 0.05 a 0.18 ± 0.02 F 0.09 ± 0.32 a 0.08 ± 0.02 a NA * ± 0.18 a SC M 3.58 ± 0.50 b 3.25 ± 0.07 b 3.85 ± 0.03 a ** 4.81 ± 0.18 a 3.78 ± 0.07 F 3.31 ± 0.59 a 3.11 ± 0.03 b 3.38 ± 0.07 a ** 3.27 ± 0.03 b KL M 9.11 ± 1.02 a 9.55 ± 0.15 a 7.62 ± 0.23 b ** 7.51 ± 0.24 b 8.24 ± 0.09 F 8.56 ± 0.87 b 9.27 ± 0.08 a 7.98 ± 0.07 c ** 8.60 ± 0.05 a Wt M 1.78 ± 0.31 a 1.71 ± 0.05 a 1.40 ± 0.04 b ** 1.63 ± 0.03 a 1.46 ± 0.01 F 1.52 ± 0.26 a 1.36 ± 0.03 b 1.23 ± 0.02 c ** 1.37 ± 0.02 b WS, wing span; SL, shank length; BL, body length; CL, comb length; CW, comb width; WW, wattle width; WL, wattle length; KL, keel length; sl, spur length; bl, beak length; SC, shank circumference, in the measurement of cm; Wt, weight (kg); NA, not available; LSM, least-squares mean; SE, standard error and different superscripts within a row indicate significantly different means (P < 0.05).

50 Chicken ecotypes in Ethiopia 49 chickens (Table 3). A non-significant comb length variation between sexes of Naked neck chicken was obtained. While Naked neck and Gasgie cocks had the longest beak length of 2.42 ± 0.45 and 2.00 ± 0.02 cm, respectively, than Gugut cocks 1.85 ± 0.10 (cm). Beak length variation is recorded among ecotypes but not with in ecotype in respective sexes (Table 3). Correlations of body weight and other linear body measurements Live weight was positively correlated (r = 55.5, P < 0.01) with wing span. Body length and spur length in Naked neck were positively correlated, males (r = 0.62, P < 0.01) and females (r = 0.55, P < 0.01). Whereas WL is the highest correlated trait (r = 0.67, P < 0.01) with body weight of Gasgie male chickens. The high correlation coefficients between body weight and other body measurements (P < 0.01) helped to predict body weight of chickens (Table 3). Prediction equation models First WS, BL and WW traits were used as linear regression to predict body weight of chickens. Whereas stepwise multiple regression was considered to predict the dependant variables by considering other traits like sl, SC, CL, bl and KL at a time in the three chicken ecotypes. In linear regression result, the body weight prediction value of Naked neck cocks and hens and Gasgie cocks were 0.40, 0.31 and 0.45, respectively. In addition to liner regression, multiple regression analysis was considered to determine the effects of other body measurements on body weight prediction (Table 4). To increase meat and egg production it requires genetic improvement of body weight of chickens. But proper measurement of this variable is often hard in villages due to lack of weighing scales. Hence, easily measurable linear body measurements are more relevant for chickens body weight prediction at farmers level rather subjectively judging manually. In addition, the present farmers are active at early morning by providing supplementary feed to their chicken before bringing them to the market to increase the temporary body weight of their chickens. Therefore, prediction equation was important. Discussions Analysing the research result evidenced that more than 70 percent of the population of chicken ecotypes in the study area were carrying the Naked neck chicken characteristics. This is new and significantly (P < 0.001) higher in number than reported result in other parts of Ethiopia (7.9 percent; Table 3. Coefficient of correlations between body weight and linear body measurements for female and male in all ecotypes in the study area (N = 450). Traits Variables Sex and ecotype Naked neck Gasgie Gugut type M F M F M F WS N r 0.64** 0.56** 0.35* 0.41** 0.50** 0.39** SL N r 0.18 ns 0.20* 0.26* 0.29** 0.54** 0.08 ns BL N r 0.59** 0.54** 0.49** 0.59** 0.33* 0.50** CL N r 0.31* 0.05 ns 0.54** 0.41** 0.40** 0.35** CW N r 0.15 ns 0.07 ns 0.64** 0.39** 0.37** 0.21 ns WL N r 0.05 ns 0.01 ns 0.67** 0.39** 0.39** NA WW N r 0.05 ns 0.13 ns 0.52** 0.47** 0.49** NA bl N r 0.22 ns 0.01 ns 0.20 ns 0.30** 0.24* 0.22* sl N r 0.48** 0.27** 0.52** 0.28** 0.21 ns 0 SC N r 0.31* 0.13 ns 0.35** 0.18 ns 0.04 ns 0.02 ns KL N r 0.37** 0.28** 0.62** 0.33** 0.23* 0.21* WS, wing span; SL, shank length; BL, body length; CL, comb length; CW, comb width; WL, wattle length; WW, wattle width; bl, beak length; sl, spur length; SC, shank circumference; KL, keel length; N, number of samples and r, correlation coefficients.

51 50 A. Getu et al. Table 4. Prediction equations in multiple regression analysis of body weight on other variables of female and male. Ecotype Male R 2 Female R 2 Naked neck Y = WS 0.40 Y = WS 0.31 Y = WS sl 0.60 Y = WS BL 0.41 Y = WS BL sl 0.65 Y = WS BL 0.10bl 0.47 Gasgie Y = WS 0.45 Y = BL 0.35 Y = WS + 0.2SC 0.53 Y = BL WW 0.46 Gugut Y = WW 0.30 Y = BL 0.25 Y = WW WS 0.44 Y = BL bl 0.31 Y = WS CL WW 0.50 Y = BL CL bl 0.37 Y = WS CL WW KL 0.55 Y = BL CL bl KL 0.40 Y = BL CL + 0.2bl SC KL 0.42 WS, wing span; sl, spur length; BL, body length; SC, shank circumference; WW, wattle length; CL, comb length and KL, keel length, all variables left in the model are significant at the 0.05 level. No other variable met the 0.05 significance level for entry into the model on body weight. Aberra and Tegene, 2011; <2 percent; Dana, 2011), Nigeria (6 percent; Gueye, 1998) and Botswana (3.6 percent; Badubi, Rakereng and Marumo, 2006). The other results on plumage colours of the identified chicken ecotypes are different from the report result from northwest Ethiopian (Halima, 2007). Variations of rose comb types, white skin colour, blocky body shape and plain head types are the dominant visible traits of chicken ecotypes. This result was not in lined with the reported result done at Bure and Fogera districts in the Amhara region and Dale district in Southern Ethiopia (Fisseha, Abera and Tadelle, 2010). This variation could be a breed-specific trait, nutritional status, genotype and reflected adaptation fitness to their environment (Aberra and Tegene, 2011; Dana, 2011). Complete absence of wattle from Gugut females, long neck and early weaning of Gasgie chicken ecotype is the unique character from the previous studies in Ethiopia and elsewhere in the tropics by Halima (2007), Aberra and Tegene (2011) and Dana (2011). Overall body weight of male and female chickens are varied from Ethiopian chickens reported by Dana (2011) which is 1.63 kg for males and 1.27 kg for females. Body weight of Naked neck and Gasgie chickens were higher than chicken in central (Danna, 2011) and northwest Ethiopia (Halima 2007) in the body weight of 1.26 and 0.87 kg for adult male and female, respectively. The frequency of chicken ecotypes carrying the Naked neck gene that we studied was significantly (P<0.001) higher than those reported in other parts of Ethiopia (<2 percent; Dana, 2011), Nigeria (6 percent; Gueye, 1998) and Botswana (3.6 percent; Badubi, Rakereng and Marumo, 2006). Conclusions and recommendations Naked neck, Gasgie and Gugut chickens are newly identified ecotypes from Quara, Alefa and Tache Armacheho district in the northern parts of Ethiopia, respectively. The identified chicken ecotypes had diversified variations in both qualitative and quantitative characters. As an example, phenotypic characterizations like 12 quantitative and seven qualitative traits were considered among the three chicken ecotypes. Heavier adult body weight and longer shank length were measured from the Naked neck, followed by the Gasgie chicken ecotypes. Qualitatively, the Gasgie chicken eco-type had normal feather morphology and others like the Naked neck chicken ecotype is easily distinguished by the complete absence of feather at neck and chest. Whereas Gugut chicken ecotype is characterized by complete absence of wattle in hens, it is the smallest of all and has dunce feather at neck in both the sexes. All these findings indicated that the investigated chicken ecotypes show heterogeneity in most traits considered. Thus, in-depth molecular characterization using genetic markers should be undertaken to confirm the level of genetic variations and relationships among newly identified and other indigenous chicken ecotypes. Acknowledgements This work is funded by the German International (GIZ) project to which we are grateful. We also thank all interviewed village chicken owners, district agricultural office experts, key informants and other participants who helped us to carry out every activity on this research. References Aberra, M. & Tegene, N Phenotypic and morphological characterization of indigenous chicken populations in southern region of Ethiopia. Anim. Genet. Resour., 49: Badubi, S.S., Rakereng, M. & Marumo, M Morphological characteristics and feed resources available for indigenous chickens in Botswana. Columbia, CIPAV, Livestock Research for Rural Development. Bogale, K In situ characterization of local chicken eco-type for functional traits and production system in Fogera district, Amhara regional state. Submitted to the Department of Animal Science, Haromiya University (MSc thesis). CSA Agricultural sample survey 2010/11. Volume 2: Statistical Bulletin 05. Report on Livestock and Livestock Characteristics, Addis Ababa, 21 February.

52 Chicken ecotypes in Ethiopia 51 Dana, N Breeding programs for indigenous chicken in Ethiopia analysis of diversity in production systems and chicken populations. Submitted in fulfillment of the requirements for the degree of doctor at Wageningen University (PhD thesis). Dana, N., Tadelle, D., Elisabeth, H.V. & Johan, A.M Morphological features of indigenous chicken populations of Ethiopia. Animal Breeding and Genomics Center, Wageningen University. Anim. Genet. Resour., 46: Dana, N., van der Waaji, E. & Johan, A.M Genetic and phenotypic parameter estimates for body weights and egg production in Horro chicken of Ethiopia. Submitted to Tropical Animal Health and Production, Animal Breeding and Genomics, Wageningen University, The Netherlands. FAO Phenotypic characterization of animal genetic resources. FAO Animal Production and Health Guidelines No. 11. Rome. (accessible at Fisseha, M., Abera, M. & Tadelle, D Assessment of village chicken production system and evaluation of the productive and reproductive performance local chicken ecotype in Bure district North West Ethiopia. Afr. J. Agric. Res., 5(13): Gueye, E.F Village egg and fowl meat production in Africa. World s Poult. Sci. J., 54: Halima, H Phonotypic and genetic characterization of indigenous chicken populations in Northwest Ethiopia. Submitted to the Faculty of National and Agricultural Sciences, Department of Animal, Wild Life and Grass Land Sciences, University of the Free State, Bloemfontein and South Africa (PhD thesis). Jens, C.R., Anders, P., Charlotte, V., Ainsh, M.C. & Lone, F Keeping of village poultry. A technical manual for small-scale poultry production. Denmark 34. Kondombo, S.R Improvement of village chicken production in a mixed farming system in Burkina Faso. Wageningen Institute of Animal Sciences, Animal Nutrition Group, Wageningen University, The Netherlands (PhD thesis). Mekonnen, G Characterization of smallholder poultry production and marketing system of Dale, Wonsho and Loka Abaya Woredas of southern Ethiopia. Awassa College of Agriculture, Hawassa University (MSc thesis). SAS Statistical Analysis System (SAS), SAS users guide, version 9.1. NC, SAS Institute Inc. Salam, K Improvement of village chicken production in a mixed (chicken ram) farming system in Burkina Faso. Wageningen Institute of Animal Sciences, Animal Nutrition Group, Wageningen University, The Netherlands (PhD thesis). Tadelle, D Phenotypic and genetic characterization of local chicken ecotypes in Ethiopia. Submitted to Humboldt University of Germany (PhD thesis). Tadelle, D. & Alemu, Y Studies on village poultry production systems in the central highlands of Ethiopia. Submitted to Swedish University (MSc thesis). Tadelle, D., Alemu, Y. & Peters, K Village chicken production systems in Ethiopia: use patterns and performance evaluation and chicken products and socio-economic functions of chicken. Livest. Res. Rural Dev., 15(1).

53 Animal Genetic Resources, 2014, 54, Food and Agriculture Organization of the United Nations, 2014 doi: /s Studies on morphometrical traits of Boran bulls reared on two feedlots in Southern Ethiopia Sandip Banerjee 1, Mohamed Beyan Ahmed 1 and Girma Tefere 2 1 School of Animal and Range Sciences, Hawassa University, P.O Box 05, Hawassa, SNNPRS, Ethiopia; 2 Dilla A.T.V.E.T College, Dilla, SNNPRS, Ethiopia Summary The study was conducted on Boran bulls from three age groups (2, 3 and 4 years) reared at two different locations in Southern Ethiopia. The study was conducted to assess the different morphometrical measurements in Boran bulls, to estimate live weight of the bulls using step-down regression equations and also to calculate structural indices for assessment of type for which the breed was developed and thus assist in the selection of the bulls. The feedlots were situated in Meki district and varied both in location and the type of management, with the second feedlot being better managed than the first. The morphometric traits included in the study were height at withers (HW), height at rump (RH), body length (BL), heart girth (HG), head width (WFH), neck circumference (NC), neck length (NL), chest width (CW), rump length (RL), hip width (WH) and flank width (FW) and the body weight (BW). The results indicated that while the skeletal measurements HW, RH and BL did not vary (P < 0.05) between bulls of a particular age across the two feedlots, the other parameters were higher among the bulls reared in the feed lot with wind breaks and situated away from lake Ziway. The results of the step-down regression equations indicated that BW could be assessed using various morphometrical measurements viz. NL, NC, RL, WH and HG. The parameters included varied both across the age groups and locations. The results related to the structural indices indicated that the Boran bulls had posterior alignment and had higher CW than HW. Keywords: Boran bulls, characterization, morphometric Résumé L étude a été menée sur des taureaux Boran de trois groupes d âge (2, 3 et 4 ans) élevés à deux endroits différents dans sud de l Ethiopie. Les parcs d engraissement étaient situées dans le district de Meki et varié tant dans l espace et le type de gestion, avec le deuxième parc d engraissement est mieux géré que la première. Les traits morphométriques inclus dans l étude étaient la hauteur au garrot (HW), la hauteur à la croupe (RH), la longueur du corps (BL), la circonférence de coeur (HG), la largeur de la tête (FMH), circonférence du cou (NC), la longueur du cou (NL ), la largeur de la poitrine (CW), la longueur croupe (RL), la largeur des hanches (WH) et la largeur de flanc (FW) et le poids corporel de (BW). Les résultats indiquent que, bien que la mesure squelettiques HW, RH et BL ne varient pas (P < 0.05) entre les taureaux d un âge donné dans les deux parcs d engraissement, les autres paramètres étaient plus élevés chez les taureaux élevés dans le parc d engraissement avec des brise-vent et situés loin du lac Ziway. Les résultats de l étape vers le bas équations de régression ont indiqué que BW peut être évaluée en utilisant différentes mesures morphométriques à savoir. NL, NC, RL, WH et HG. Les paramètres inclus variée à la fois à travers les groupes d âge et les lieux. Les résultats liés à des indices structurels indiqué que les taureaux Boran eu alignement postérieur et avait CW supérieur à HW. Mots-clés: taureaux Boran, paramètres morphométriques, caractérisation Resumen Este estudio fue llevado a cabo con toros Boran, de tres grupos de edad (2, 3 y 4 años), criados en dos lugares distintos del Sur de Etiopía. El estudio se realizó con el fin de medir diferentes parámetros morfométricos en toros Boran, de estimar el peso vivo de los toros con ecuaciones obtenidas por el método de regresión por eliminación de variables (análisis step-down) y con el fin también de calcular índices estructurales para la determinación del tipo de animal pretendido cuando se desarrolló la raza y, así, contribuir a la selección de los toros. Los corrales de engorde estuvieron ubicados en el distrito de Meki y difirieron tanto en el emplazamiento como en el tipo de manejo, siendo el segundo cebadero mejor gestionado que el primero. Los parámetros morfométricos incluidos en el estudio fueron la altura a la cruz (AC), la altura a la grupa (AG), la longitud corporal (LC), la circunferencia torácica (CT), la anchura de la cabeza (ACz), la circunferencia del cuello (CCu), la longitud del cuello (LCu), la anchura del pecho (AP), la longitud de la grupa (LG), la anchura de la cadera (ACd), la anchura de los flancos (AF) y el peso corporal (PC). Los resultados indicaron que, si bien las medidas esqueléticas AC, AG y LC no variaron significativamente, para una misma edad, entre los toros de los dos cebaderos, los valores de los otros parámetros fueron más elevados en los toros criados en el corral de engorde con cortinas cortaviento, situado a gran distancia del lago Ziway. Los resultados obtenidos con las ecuaciones de regresión señalaron que el peso corporal podía ser estimado usando algunas medidas morfométricas, en concreto LCu, CCu, LG, ACd y CT. Los parámetros Correspondence to: Sandip Banerjee, School of Animal and Range Sciences, Hawassa University, PO Box 05, Hawassa, SNNPRS, Ethiopia. sansoma2003@yahoo.co.in and sansoma2003@gmail.com; tel.:

54 54 S. Banerjee et al. considerados se vieron afectados tanto por el grupo de edad como por la ubicación del cebadero. En cuanto a los índices estructurales, los resultados mostraron que los toros Boran presentaban alineamiento posterior y que tuvieron mayor AP que AC. Palabras clave: toros Boran, parámetros morfométricos, caracterización Submitted 24 August 2013; accepted 27 February 2014 Introduction Livestock contribute significantly to poverty alleviation of a major section of society, especially in the developing world. Studies by Hoffmann (2010) indicate that livestock are an important component of livelihood of resource challenged societies. CSA (2012) indicate that Ethiopia has about million cattle and the livestock production system can be classified into three types of production systems: the lowland pastoral and agro pastoral system; the highland crop livestock production system; and a very small proportion of urban and peri-urban livestock production system. As indicated by ESPSLMMP (2010) the livestock sector contributes about 20 percent of the Gross Domestic Product, supporting the livelihood (directly or indirectly) of about 70 percent of the natiońs population and representing 11 percent of the annual foreign earnings. According to ESPSLMMP (2011), Ethiopia earned about million USD from export of tonnes of meat and head of livestock, and the revenue was enhanced by 69 percent in 2011 over the previous fiscal year. The study also shows that the earnings were mostly due to the sale of live animals (70 percent), while the rest was from the sale of chilled carcasses. Among the different classes of livestock, the sale of cattle on the hoof accounted for 46 percent of the total livestock exported. Studies by NEPAD CAADP (2005) indicated that most of the cattle raised for export purpose are bred in the lowlands situated below m a.s.l which covers about 60 percent of the total land of the country. These lowlands are situated in the Eastern, Southern and Western parts of the Central highlands. The report also indicates that these regions house 20 percent of the nation s cattle, 25 percent sheep and 75 percent of the goat population. The majority of the residents in these areas directly or indirectly depend on livestock or livestock products to meet their daily needs. Cattle play a multifarious role in the life of the Agrarian Society: their dung is used as manure, they are also used for most of the agriculture-related work, from ploughing to transportation of the agriculture produce to the market, with many other intermediate activities in between (Felleke and Geda, 2001; Little et al., 2001). As indicated by Solomon (2001), Borena plateau of Southern Ethiopia is the home of the best-known cattle breed from this country. The Boran cattle are the most sought after breed for consumption within the country and for the export purpose (LMA 2001). The pastoralists usually grade their livestock using informal methods of quality estimation such as feeling the loin area or by visual estimates alone. Weighing scales are rare and they are mostly inaccurate due to lack of maintenance. Therefore under such constraints, the pastoralists are usually unable to receive a fair price of their livestock. Linear measurements provide a scientific basis to describe the biological variations between breeds and also for animals within a breed and thus can serve as a basis for measuring the performance, productivity and carcass characters (Kayastha et al., 2011; Kugonza et al., 2011). Studies by Gatesy and Arctander (2000) indicated that morphological descriptions are useful tools for distinguishing livestock breeds and strains. These traits are also used to evaluate the breeding goals of livestock (Zechner et al., 2001). Accordingly, Mwacharo et al. (2006) indicated that the relationship between size and breed can also be fairly well estimated using several morphometrical measurements. Many of the morphometrical traits are highly correlated with the body weight (BW) of livestock and hence can serve as an important predictor for assessment of live weight of animals (Goe, Alldredge and Light, 2001; Mwacharo et al., 2006). Skeletal measurements such as height at withers (HW) and body length (BL) indicate measurements pertaining to size of the livestock, which are better indicators of BW when compared to measurements which are related to muscle development and fat deposition, which are affected by nutrition and management (Kamalzadeh, Koops and van Bruchem, 1998). One of the most commonly used procedures for the estimation of live-weight of cattle is by developing regression equations based on some simple linear measurements as predictors. Studies by Heinrichs, Rogers and Cooper (1992) and Gilbert, Bailey and Shannon (1993) indicated that body measurements in livestock can serve as predictors of body weight. Msangi et al. (1999), Slippers, Letty and De Villerrs (2000), Fouire et al. (2002), Willeke and Dürsch (2002) and Bozkurt (2006) reported that emphasis has shifted from subjective methods like body condition score to objective methods such as the use of linear measurements of different body parts for the estimation of body weight. As indicated by Alderson (1999), the importance of BW alone is of limited importance in the absence of associated traits of type and conformation. Studies by Gilbert and Gregory (1952) indicate that ratios of BW/ HW can serve as an indicator of beef type in Hereford cattle. According to Alderson (1999), indices can serve as useful tools in selection of breeding animals and also for assessment of cattle at

55 Morphometrical traits of Boran bulls reared in Southern Ethiopia 55 maturity. They also serve as an alternative option for the assessment of BW as it incorporates measures, which are correlated with desirable conformation. Calculation of indices can also help potential livestock dealers as the traits considered in the development of the indices are associated with the production traits. As indicated by Salako (2006), the desirable body conformation for traits pertaining to meat production is quite complex and hence cannot be truly assessed taking into account a single linear measurement and is usually assessed using body condition score, which itself is a very subjective measurement and hence prone to errors, besides most of the subjective scores have very low heritability, and hence are rarely accurate in selection of breeding stock. Indices are objective body measurements that are relatively accurate to estimate and are relatively easy to understand. The term type indicates the structure of an animal besides its body form and is hence considered as ideal for understanding the purpose for which the animal has been developed. The aim of this study was to investigate morphometrical traits, assessment of live weight using stepdown regression equations and calculation of structural indices values of Boran bulls aged 2, 3 and 4 years of age and reared at two locations of Southern Ethiopia. Materials and methods The study was conducted at Maki district of Oromia region. The feedlots are situated at a distance of 133 km south of Addis Ababa. The study was conducted on two feedlots, which varied significantly as far as the climate was concerned. The first feedlot is situated near Lake Ziway ( N E) in an area devoid of any vegetation or tree breaks; especially during the night when the temperature was low and uncomfortable for the bulls. The second feedlot is situated 25 km west of the first feedlot. The climate of the second feed lot was favourable for the livestock as there are sufficient trees to serve as wind break thereby preventing the chilly winds from the lake thus making the cattle comfortable during the night (Figure 1). The study was conducted between January and April months of 2012, which is considered as the dry season in the region. The cattle were grouped in three age groups, i.e. 2, 3 and 4 years, respectively in both the feedlots, where they were separated by their ages, the bulls were purchased by estimating their age and body weight. The age of the animals was assessed using the dentition method as suggested by Hammond, Mason and Robinson (1971). The Boran bulls (Figure 2) were purchased from their breeding in Southern Ethiopia tract 6 months prior to the commencement of the study. In the feedlot, the bulls were sprayed with Cypermethrin at 2 ml/l of water and drenched with Albendazole based on their BW and as per the manufacturer s recommendation. The cattle were provided with concentrate supplements, which comprised rice and wheat bran besides hulls from different pulses (haricot bean (Phaseolus vulgaris), lentil (Lens culinaris) andpeas (Pisum sativum) cotton seed cake and common salt the roughage (consisted of teff (Eragrostis tef), wheat (Triticum sp.) and barley (Hordeum vulgare) straw) and water was provided adlib, the proximate analysis of the feed indicated that it contained 19 percent crude protein, 2.5 percent crude fat, 16.5 percent crude fibre on dry matter basis, which was in accordance with the recommendation of Hutcheson (2006) for Boran bulls reared for fattening purpose under Ethiopian condition. The concentrate was provided twice a day and the animals were group fed according to their respective age categories. The BW and linear measurements were taken early in the morning (Figure 3) after overnight fasting to ensure minimum gut content in the bulls. The BW of the bulls was assessed using a platform balance, which was calibrated prior to being used. The linear measurements were taken by one of the authors during the course of the study and were carried out according to the methods suggested by Macjowski and Zieba (1982) and Tolenkhomba et al. (2012). The traits included in the study were: HW, height at rump (RH), BL, heart girth (HG), head width (WFH), neck circumference (NC), neck length (NL), chest width (CW), rump length (RL), hip width (WH), flank width (FW) and BW of the bulls. The methodologies for the Figure 1. Boran bulls reared in the second feed lot. Figure 2. Boran bulls of different age groups.

56 56 S. Banerjee et al. Table 2. Calculations for structural indices. Type of index Calculation Height index Height at withers/body length Rump length index Length of rump/body length Over increase index Height at rump/height at withersx100 1 Height slope Rump height withers height 2 Length index Body length/height at withers 2 Width slope Hip width chest width 2 Body weight index Body weight/height at withers Alderson (1999). 2 Szabolcs et al. (2007). 3 Rohrer (1921). Figure 3. Restraining chute for taking the morphometric measurements. assessment of the measurements are presented in Table 1. The data were analysed statistically using SPPS V 12 for Windows (SPSS, 2003). The data were analysed using descriptive statistics, the effect of feedlots on morphometrical traits and BW of Boran cattle from various age groups were compared using one-way analysis of variance and were considered significant both at P < 0.01 and also P < The prediction of BW taking into account different morphometrical traits was conducted using step down linear regression analysis and the accuracy was assessed using R 2 adj values. The formulae used to calculate the various body indexes are presented in Table 2, height, rump length, over increase was calculated according to the methodology suggested by Szabolcs et al. (2007), whereas the values for height slope, length; width slope were assessed according to the methods suggested by Alderson (1999). The weight index was assessed according to the Table 1. Measurement techniques for the morphometrical traits. Traits Height at withers Rump height Body length Heart girth Width of fore head Neck circumference Neck length Chest width Rump length Width of the hips Flank width Body weight Measurements Distance between highest point of the wither to the ground Distance between the highest point of the rump to the ground Distance between point of the shoulder to the pin bone The circumference of the chest at the widest point The point between the two polls Circumference of the neck at its widest point The length between the atlas vertebrae and the first thoracic vertebrae Width of the thorax at the widest point just behind the elbow The distance between the last thoracic vertebrae and just anterior of the hips The widest point between the two pelvic bones The widest point of the rear flank The weight of the bulls on a calibrated weighing bridge with an error margin of ±25 kg method suggested by Röhrer and cited by Szabolcs et al. (2007). Results and discussions Morphometrical traits of Boran bulls aged 2 years and reared in the two feedlots The values pertaining to the various morphometrical traits of Boran bulls aged 2 years and reared in two feed lots and of different age categories are presented in Table 3. The results from the table indicate that the average HW for the bulls aged 2 years did not vary significantly between the cattle reared in the two feedlots. The HW values as assessed in the study are lower than those reported by Alberro and Haile-Mariam (1982) for Boran cattle, Fasil (2006) for Fogera cattle, Kugonza et al. (2011) for Ankole cattle and Mwambene et al. (2012) for Nkasi cattle. The values are comparable with those reported by Mason and Maule (1960) for Arado, Takele (2005) for Sheko cattle, Getinet et al. (2009) for Ogaden cattle, Fasil (2006) for Gojjam Highland Zebu cattle and Nakachew (2009) for Abigar cattle. The results pertaining to the RH too did not vary significantly across the two locations. The values for the trait are comparable with the figures reported by Nakachew (2009) for Abigar cattle. The results of the study pertaining to BL also did not vary significantly across the two locations. The values for BL are comparable with those reported by Shiferaw (2006) for Kereyu cattle and also Nakachew (2009) for Abigar cattle, Zulu (2008) for Angoni and Tonga cattle and also Cyprian et al. (2012) for Bunaji cattle. HG showed variations across the two locations, with higher values being recorded among the cattle reared in the second feedlot, which may be attributed to favourable climate prevailing in the area. The HG values observed are in accordance with the findings of Alberro and Haile-Mariam (1982) for Boran cattle, Fasil (2006) for Fogera cattle and Gojjam Highland Zebu, Getinet, Ayalew and Hegde (2009) for Ogaden, Kugonza et al. (2011) for Ankole cattle, Mwambene et al. (2012) for Nkasi cattle. The HG values as assessed were however lower than those reported

57 Morphometrical traits of Boran bulls reared in Southern Ethiopia 57 Table 3. Morphometrical measurements of Boran bulls of three age groups and across two feedlots. Traits Age-2 Age-3 Age-4 Feedlot-1 (N = 49) Feedlot-2 (N = 10) Feedlot-1 (N = 181) Feedlot-2 (N = 85) Feedlot-1 (N = 127) Feedlot-2 (N = 253) HW (cm) ± ± ± ± ± ± 5.9 RH (cm) ± ± ± ± ± ± 8.1 BL (cm) ± ± ± ± ± ± 8.0 HG (cm) ± ± 8.0** ± ± 9** ± ± 8.0** WFH (cm) 19.5 ± ± ± ± 2.19* 19.7 ± ± 2.2* NC (cm) 72.4 ± ± 7.4** 75.5 ± ± 8.6** 77.9 ± ± 7.5** NL (cm) 36.6 ± ± ± ± 6.1** 38.1 ± ± 5.6** CW (cm) 54.2 ± ± 3.2* 55.2 ± ± 3.7** 56.0 ± ± 4.1** RL (cm) 45.5 ± 2.3* 42.5 ± ± ± ± ± 3.8 WH (cm) 29.9 ± ± 4.0** 30.2 ± ± 3.7** 30.2 ± ± 3.2** FW (cm) 56.9 ± ± 5.1** 57.7 ± ± 5.8** 58.3 ± ± 4.7** BW (kg) ± ± 23.8** ± ± 52.1** ± ± 39.5** *P < 0.05, **P < HW, height at withers; RH, rump height; BL, body length; HG, heart girth; WFH, width of forehead; NC, neck circumference; NL, neck length; CW, chest width; RL, rump length; WH, width of hips; FW, flank width; BW, body weight. by Zulu (2008) for Barotse, Tonga and Baila cattle. The WFH for the cattle reared in the two feedlots did not vary significantly; the values are similar to those reported by Szabolcs et al. (2007) for Charolais and Red Angus cattle. NC values too varied across the two feedlots, with (P < 0.01) higher values among the bulls raised in the second location, the values being similar to those reported by Phanchung and Roden (1996) for Siri cattle from Bhutan. The average NL indicated no significant difference among the bulls reared across the two feedlots. The results as obtained are similar to those reported by Abdelhadi and Babiker (2009) for Baggara cattle but higher values for the trait were reported by Mason and Maule (1960) for Arado cattle, Nakachew (2009) for Abigar cattle and Aamir et al. (2010) for Kenena cattle. The results pertaining to CW varied (P < 0.05) among the cattle reared in the two feedlots. The results pertaining to the RL were higher (P < 0.05) among the bulls reared in the first feedlot. The RL values as recorded in the study were similar to those observed by Szabolcs et al. (2007) for Aberdeen Angus and Red Angus bulls. The results pertaining to WH, FW and BW too differed significantly (P < 0.01) between the two feedlots with higher values observed in the cattle reared in the second location. The WH values as recorded for the bulls raised in the second location are in accordance with the reports of Gilbert, Bailey and Shannon (1993) for Angus cattle; Suriya, Boonsaen and Innuruk (2011) for Kamphaengsaen (grass-fed) cattle. The FL values as recorded in the study are lower than the values reported by Khalafalla et al. (2011) and Abdelhadi, Babiker and Kijora (2011) for Baggara and Zebu cattle from Sudan, the lower values for the trait may be attributed to both genetic and non-genetic factors affecting the trait. The results pertaining to the BW of the bulls reared in the second site are in close accordance with the reports of Nakachew (2009) for Abigar cattle Mwambene et al. (2012) for Sumbawanga rural cattle and also for Baggara cattle of Sudan as observed by Khalafalla et al. (2011). The results pertaining to the morphometrical traits of Boran cattle (reared in the two feedlots) aged 3 years are also presented in Table 2. The study indicated that the HW values did not vary between the two feed lots and are lower than the values reported by Kugonza et al. (2011) for Ankole cattle, Mwambene et al. (2012) for Nkasi cattle, Mwambene et al. (2012) for Sumbawanga cattle Urban cattle, Mwambene et al. (2012) for Sumbawanga cattle rural, Aamir et al. (2010) for Kenena cattle and Abdelhadi and Babiker (2009) for Baggara cattle. The values, however, being higher than those reported by Mason and Maule (1960) for Arado, Takele (2005) for Sheko cattle, Getinet et al. (2005) for Ogaden cattle, Fasil (2006) for Gojjam Highland Zebu cattle, Nakachew (2009) for Abigar cattle and Kayastha et al. (2011) for Assam cattle. The values as assessed in the study find consonance with the assessment of Alberro and Haile-Mariam (1982) for Boran cattle and Fasil (2006) for Fogera cattle. The average values of RH also did not vary significantly across the two locations; the figures are comparable with the values reported by Nakachew (2009) for Abigar cattle. However, the values are lower than those reported by Szabolcs et al. (2007) for Hungarian Simmental cattle, Hereford cattle and for Aberdeen Angus cattle. The similar trend was also observed for BL with no differences among the cattle reared in the two locations. The findings are in close agreement with the observations of Milla, Mahagoub and Bushara (2012) for Nilotic cattle, Mwambene et al. (2012) for Sumbawanga cattle Urban, Zulu (2008) for Angoni cattle, Zulu (2008) for Tonga cattle, Cyprian et al. (2012) for Friesian cattle X Bunajii cattle. However, the HG varied (P < 0.01) among the bulls reared in the two locations with higher values observed among those raised in the second location. The values as assessed for the first feed lot are in agreement with the observations of Kugonza et al. (2011) for Ankole cattle and Mwambene et al. (2012) for Nkasi cattle, while the results among the

58 58 S. Banerjee et al. cattle reared the second feed lot find consonance with the observations of Alberro and Haile-Mariam (1982) for Boran cattle, Shiferaw (2006) for Kereyu cattle, Mason and Maule (1960) for Arado, Milla, Mahagoub and Bushara (2012) for Nilotic cattle, Takele (2005) for Sheko cattle, Getinet et al. (2005) for Ogaden cattle, Fasil (2006) for Gojjam Highland Zebu cattle and Fasil (2006) for Fogera cattle. The findings indicate that the WFH too varied (P < 0.05) among the bulls reared in the two feedlots with higher values observed among the bulls reared in the second location. The findings are in close accordance with the observations of Szabolcs et al. (2007) for Hereford cattle, Aberdeen Angus cattle, Red cattle Angus cattle, Hungarian Simmental cattle, Lincoln Red cattle, Shaver cattle and Charolais cattle. However, the assessment of Gilbert, Bailey and Shannon (1993) for Hereford cattle and Angus cattle, Tolenkhomba et al. (2012) for Manipuri cattle, Szabolcs et al. (2007) for Limousin cattle, and Blonde d Aquitaine cattle are higher than those observed among the bulls reared in first feedlot but lower than those reared in second location. The findings also indicate differences (P < 0.01) for NC among the bulls reared in the two locations with higher values observed among the cattle reared in the second location, the values as assessed in the study are lower than the observations of Phanchung and Roden (1996) for bulls of Siri cattle from Bhutan. The findings also indicate (P < 0.01) differences for NL between the bulls reared in the two locations. The values being higher among the cattle reared in the second location. The observations (irrespective of the feedlots) are in accordance with the results of Mason and Maule (1960); Arado cattle, Abdelhadi and Babiker (2009); Baggara cattle, Nakachew (2009); Abigar cattle and Aamir et al. (2010); Kenena cattle. Similarly CW values too varied among the cattle reared in the two locations, with higher values observed among the cattle reared in the second location the values as assessed find similarity with the reports of Getinet et al. (2005) for Ogaden cattle. The RL did not show significant differences among the cattle reared in the two feedlots; the values are similar to the results observed by Szabolcs et al. (2007) for Shaver, Charolais, Limousin and Blonde d Aquitaine cattle breeds. While the WH too varied (P < 0.01) between the bulls reared in the two locations with higher values observed among the bulls reared in the second location, the findings are in close agreement with the results of Gilbert, Bailey and Shannon (1993) for Hereford and Angus cattle and also Suriya, Boonsaen and Innuruk (2011) for Kamphaengsaen (grass-fed) cattle; however, higher values for the trait were reported for dairy cattle breeds, Holstein Friesian and Brown Swiss by Ozkaya and Bozkurt (2009). The results pertaining to the average values of FW too find differences among the cattle reared in the two feed lots with higher value being observed among those being reared in the second feedlot. The values of this study being lower than those reported by Khalafalla et al. (2011) for Sudan Baggara and also by Abdelhadi, Babiker and Kijora (2011) for Zebu cattle from Sudan. The results pertaining to the average BW indicate that it varied (P < 0.01) among the bulls reared in the two locations, the values being higher among the bulls raised in the second feedlot. The values as observed among the bulls reared in the first location find similarity with the findings of Shiferaw (2006) for Kereyu, Milla, Mahagoub and Bushara (2012) for Nilotic cattle, Mwambene et al. (2012) for Sumbawanga cattle reared in the urban areas. While BW of the cattle reared in the second location are similar to those reported by Phanchung and Roden (1996) for bulls of Siri breed from Bhutan. The results pertaining to morphometrical traits of bulls of Boran breed and aged 4 years and reared in the two feed lots are also presented in Table 3. The results indicate that the HW were similar among the bulls reared in the two locations, the figures are similar to the observations of Alberro and Haile-Mariam (1982) for Boran cattle, Shiferaw (2006) for Kereyu cattle, Mason and Maule (1960) for Arado cattle, Getinet, Ayalew and Hegde (2009) for Ogaden cattl and Fasil (2006) for Fogera cattle. The HW values are however higher than those reported by Fasil (2006) for Gojjam Highland Zebu, Kayastha et al. (2011) for Assam cattle and also Suriya, Boonsaen and Innuruk (2011) for Kamphaengsaen (grass-fed) cattle. The values for HW as assessed in this study are however lower than the figures reported by Suriya, Boonsaen and Innuruk (2011) for Kamphaengsaen (feedlot-fed) cattle, Abdulmojeed et al. (2010) for Sokoto Gudali cattle, Cyprian et al. (2012) for Bunaji and also Angoni, Baila, Barotse and Tonga cattle by Zulu (2008). The results of the study also indicated that the values of HR also did not vary among the bulls reared in the two locations, the findings are in consonance with the observations of Abdulmojeed et al. (2010) for Sokoto Gudali cattle, Shiferaw (2006) for Kereyu cattle and Nakachew (2009) for Abigar cattle; higher values for the trait were reported by Szabolcs et al. (2007) for Hereford, Aberdeen Angus, Red Angus, Charolais and Limousin breeds. The study shows that BL too was non-significant between locations, the values are in close accordance with the findings of Zulu (2008) for Baila, Tonga and Angoni cattle, the values are however higher than those reported by Tolenkhomba et al. (2012) for Manipuri cattle, Rajendran et al. (2008) for Umblachery cattle, and are lower than those reported by Abdulmojeed et al. (2010) for Sokoto Gudali and Bunaji cattle. The HG values indicated differences (P < 0.01) between the bulls reared in the two locations. The HG values of the bulls raised in the first feedlot find similarity with the observations of Fasil (2006) for Fogera cattle, Alberro and Haile-Mariam (1982) for Boran cattle, Aamir et al. (2010) for Kenena cattle and Mwambene et al. (2012) for Sumbawanga Urban cattle, whereas those raised in the second location find similarity with the observations of Mwambene et al. (2012) for

59 Morphometrical traits of Boran bulls reared in Southern Ethiopia 59 Nkasi cattle, Zulu (2008) for Angoni, Baila and Tonga cattle. The WFH figures indicate the difference (P < 0.05) between the bulls reared in the two locations the results, for those raised in the first location being similar to those reported by Szabolcs et al. (2007) for Red Angus, Aberdeen Angus and Charolais bulls, whereas those reared in the second feedlot are similar to those reported by Szabolcs et al. (2007) for Shaver, Hungarian Simmental and Hereford cattle. The NC too varied between the bulls reared in the two feedlots, the values being higher among the bulls raised in the second location similar values were reported by Phanchung and Roden (1996) for bulls of Siri breed. The results of the study pertaining to the NL too indicate that there is variation for the trait among the bulls reared in the two feedlots, with higher values among the bulls reared in the second feedlot, the values as observed among the bulls from the first feedlot are in close accordance with the observation of Abdelhadi and Babiker (2009) for Baggara cattle, whereas those from the second feedlot are similar to those reported by Nakachew (2009) and Aamir et al. (2010) for Abigar and Kenena cattle, respectively. The NL values as assessed by Mason and Maule (1960), Takele (2005) and Tolenkhomba et al. (2012) are lower than those assessed for the bulls reared in the first feedlot, while higher values for the trait have also been reported by Shiferaw (2006) for Kereyu cattle. The CW too showed difference (P < 0.01) among the bulls reared in the two feedlots with higher value observed among the cattle reared in the second feedlot, the results as assessed for the bulls maintained in the first feedlot is in agreement with the results of Getinet et al. (2005) for Ogaden bulls. The differences may be attributed to differences in muscling among the bulls reared in the two feedlots with lower muscling among the bulls reared in the first feedlot. The effect of several non-genetic factors including the temperature prevailing in the region, which may influence the muscling as a part of the ration is utilized in maintenance of the body temperature, the observations are in accordance with the reports of Young Manzi et al. (2012). The results as assessed for RL indicated no significant differences due to two feedlots. The WH too indicated differences among the two feedlots with higher (P < 0.01) values observed among the bulls reared in the second feedlot. The WH for those reared in the first location are similar to those reported by Gilbert, Bailey and Shannon (1993) for Angus and Hereford bulls, while those reared in the second feedlot are in consonance with the observations of Suriya, Boonsaen and Innuruk (2011) for Kamphaengsaen (grass-fed) bulls, while the value was lower than those of Kamphaengsaen bulls reared on feedlot. The study also indicates that the FW of the bulls reared in the two feedlots too varies significantly, with lower values observed among the bulls reared in the second feedlot, the findings are lower than those reported by Khalafalla et al. (2011), and Abdelhadi, Babiker and Kijora (2011). The BW of the bulls indicate that those reared in the second feedlot had higher (P < 0.01) values for the trait, those reared in the first feedlot was higher than those reported by Shiferaw (2006), Abdelhadi and Babiker (2009), Khalafalla et al. (2011), Milla, Mahagoub and Bushara (2012) and Nakachew (2009). The results of a study by Mwambene et al. (2012) indicated that the BW of Sumbawanga cattle reared in the rural and urban areas was lower than those observed for the bulls reared in the first feedlot, whereas the BW of Nkasi cattle was higher than even the cattle reared in the second feedlot. Effect of non-genetic factors influencing the morphometrical traits and Body weight The term adaptability refers to both genetic and physiological traits associated with an animal s response to both external and internal stimuli (Hafez, 1968). According to Hafez (1968) the term physiological adaptation refers to the capacity of an individual to adjust to the external environment. All animals are genetically adapted to the environment where they have evolved (Pianka, 2000), genetic adaptation is the set of heritable traits that favour the survival of the population in a particular environment. This is all the more true for those animals, which are raised on natural grazing as they are subjected to several stress factors when compared to those raised in the farm premises. The present study indicated that the skeletal traits (HW, RL, BL and RH) did not vary significantly between feedlots for a particular age category, which may be because skeletal traits are highly heritable and hence are least affected by non-genetic factors the observations are in accordance with the findings of Mwambene et al. (2012) and Yunusa, Salako and Oladejo (2013). However, the other non-skeletal traits varied significantly across feedlots for a particular age category. The differences among the bulls due to feedlots are observed in the study may be attributed to the fact that the increase or decrease in live weight is a gross expression of many factors viz. organs, viscera, gut fill and also tissue mass, which are influenced by both genetic and non-genetic factors besides sex of the animal. As reported by Oladimeji et al. (1996), the climate of a region affects feed and water intake and also can influence the utilization of components within the feed consumed. It also influences the heat produced by the animal and thereby the net available energy and ultimately the productivity and phenotypic expression of the growing livestock. The results also indicated that among the cattle aged 2 years the values for HG, NC, CW, BD, WH, FW and BW were significantly higher in cattle raised in the second location, while RL was significantly higher among cattle reared in the first feedlot. The study also indicated that traits such as HG, WFH, NC, NL, CW, BD and BW values were higher among bulls of 3 and 4 years of age raised in feedlot 2, whereas WH and FW for both the categories were higher among the cattle raised in feedlot 1. It was

60 60 S. Banerjee et al. also observed that the reared in the second feedlot were better managed. This may be attributed to adaptability of the cattle to the respective location; the findings pertaining to the effect of adaptability on variation in morpohometrical traits have also been reported by Hafez (1968) and Bonsma (1980). The effect of climate prevailing in the first location may have influenced the physiological factors; hence, the apparently cold climate influenced the muscle development thus affecting the BW but did not influence the skeletal development of the bulls, the findings being in consonance with the observations of Otuma (2004). The findings are in consonance with the observations of several earlier studies in the tropics (Willamson and Pyne, 1978; Pagot, 1992; McDonald et al., 2002; Brito et al., 2004). The muscle development of the cattle is significantly affected by several non-genetic factors. This may be the reason why there were differences among bulls (within an age group) reared in the two feedlots; the observations are in accordance with the findings of Rothouge (2000). The difference in muscle development may also be attributed to the nutritional quality of the feed ingredients and the way they were cultivated, harvested and postharvest methods of processing and storing, the observations are in accordance with the findings of Hunters and Buck (1992) and Bosman (1999). The results of step-down regression equation as presented in Table 4 indicate that NC was considered as the best predictor among the cattle aged 2 years and reared in the first feedlot, whereas the assessment of BW was best assessed using NL values for the bulls reared in second feedlot, this was followed by considering the NL and RL values together. The assessment of BW of the bulls aged 3 years indicates that it was best assessed using HG values which was followed by taking both HG and NC values together, whereas the BW of the bulls reared in the second location were best assessed taking into account the values of WFL and RL together. The results of the regression studies also indicate that BW for the bulls aged 4 years and reared in the first feedlot was best assessed using their HG values, while those reared in the second feedlot was estimated using RL, WFH and FW values taken together, the importance of RL as an indicator of BW is in accordance with the findings of Alderson (1999). The importance of NC, NL and WFH for assessment of BW of beef cattle has also been reported by Hammond, Mason and Robinson (1971). The importance of HG as a reliable tool to assess BW of cattle has also been reported (Branton and Salisbury, 1946; Mullick, 1950; ElKhidir, 1980; Khogali, 1999; Bagui and Valdez., 2007; Abdelhadi and Babiker 2009). According to Afolayan et al., (2006), the morphometrical traits are moderately heritable and have a strong relationship with the growth traits of livestock; however, the relationship vary from breed and also are influenced by several factors such as age, type size, condition and degree of fattening of the animals (Heinrichs, et al., 1992; Yanar et al., 1995; Ozkaya and Bozkurt, 2009). Structural indices The results pertaining to the structural indices of Boran bulls of different age groups and reared in the two feed lots are presented in Table 5. The results indicate that the height index values for the bulls aged 2 years varied across the two feedlots with higher values being observed among the bulls reared in the second location. While the height index remained more or less similar among the bulls aged 3 and 4 years and reared in both the feed lots. The observations are in accordance with the findings of Shackelford, Koohmaraie and Wheeler (1995) who also reported that the skeletal maturity score increased till 2 years of age in cattle and thereafter remained more or less constant. The study also indicated that the RL index did not vary between the bulls across both age categories and feedlots. This can be attributed to Table 4. Step-down regression equations for assessment of body weight using morphometric measurements of Boran bulls of different age groups and reared in two different feedlots. Age Location R 2 adjusted Equations Age-2 Location (NC) (NC) (FW) Location (LN) (LN) (RL) Age-3 Location (HG) (HG) (NC) (HG) (NC) 1.263(HW) (HG) (NC) 1.231(HW) (WH) (HG) (NC) 1.188(HW) (WH) (LN) Location (WFH) (WFH) (RL) (WFH) (RL) (HG) Age-4 Location (HG) Location (RL) (RL) (WFH) (RL) (WFH) (FW) (RL) (WFH) (FW) (NC)

61 Morphometrical traits of Boran bulls reared in Southern Ethiopia 61 Table 5. Morphometrical indices of Boran bulls of different age groups and across the two feedlots. Age (years) Index Height index 1 Rump length index 1 Over increase index 1 Height slope 2 Length index 2 Width Weight 3 slope 2 2 Feedlot Feedlot Feedlot Feedlot Feedlot Feedlot Szabolcs et al. (2007). 2 Alderson (1999). 3 According to Röhrer as cited by Szabolcs et al. (2007). the character of the breed, the RL index values as obtained in the study are higher than those reported by Szabolcs et al. (2007) for different taurine breeds of beef cattle. The results pertaining to the over increase index too did not vary between the bulls reared in the two feed lots and also across the ages, the study indicated that the frame of the Boran bulls sloped towards the anterior part of the body, the RH being higher than the height at withers, the observations are in accordance with the findings of Szabolcs et al. (2007). The results pertaining to the height slope as assessed in the study too collaborated with the findings of the over increase index. The length index indicates that the Boran bulls are proportionately longer bodied in comparison to their height. The width slope index values as obtained in the study indicate that the Boran bulls are narrower at the hind quarters. The weight index values indicate that the values were lower (across all age groups) among the bulls reared in the first feedlot, which may be attributed to the fact that the place was colder than the second feedlot and a part of the energy from the ration was utilized for body maintenance rather than for BW development. The similar results were also reported by Young (1981) who reported that under the similar levels of nutrition the cattle subjected to cold stress have poor productivity than those reared under comfortable environment. Conclusion Hence, it can be concluded from the study that Boran bulls if properly managed can be reared profitably for beef purpose. Linear measurements can be effectively used to assess BW of Boran bulls at all ages. Hence, morphometrical measurements such as NL, NC, RL and WFH of Boran bulls can serve as indicators of selection of sires for higher BW. The results of the indices indicate that the Boran bulls have a higher CW when compared to the hips, which may be a breed character. It can be concluded that Boran bulls can be selected for BW using the regression equations. However, similar studies should be conducted using a large population to further confirm the results. Acknowledgements The authors acknowledge the assistance received from Federal A.T.V.E.T co-coordinator, for the financial support. The Dean and staff members of Dilla A.T.V.E.T College, Head and staff members of School of Animal and Range Sciences, Hawassa University. References Aamir, H.M., Babiker, S.A., Youssif, G.M. & Hassan, Y.A Phenotypic characterization of Sudanese Kenana cattle. Res. J. Anim. Vet. Sci., 5: Abdelhadi, O.M.A., Babiker, S.A Prediction of zebu cattle live weight using live animal measurements. Livestock Research for Rural Development. Volume 21 Article 133. Retrived August 2, 2013 from Abdelhadi, O.M.A., Babiker, S.A. & Kijora, C Estimation of zebu cattle carcass weight using body measurements. Livestock Res. Rural Dev., 23: Article #12. Retrieved August 23, 2013 (available at Addis Ababa, Ethiopia. Abdulmojeed, Y., Kingsley Omogiade, I., Hadiza Salihu, H., Matthew, W. & Samuel, A Multivariate analysis of phenotypic differentiation in Bunaji and Sokoto Gudali cattle. Acta Argiculturae Slovenica. 96(2): Afolayan, R.A., Adeyinka, I.A. & Lakpini, C.A.M The estimation of live weight from body measurements in Yankasa sheep. Czech J. Anim. Sci., 51: Alberro, M. & Haile-Mariam, S The indigenous cattle of Ethiopia. Part I. FAO World Anim. Rev., 41: Alderson, G.L.H The development of a system of linear measurements to provide an assessment of type and function of beef cattle. Anim. Genet. Resour. Inf. FAO, 25: Bagui, N.J.G. & Valdez, C.A Live weight estimation of locally raised adult purebred Brahman cattle using external body measurements. Philip. J. Vet. Med., 44: Bonsma, J Livestock production: a global approach. Cape Town, RSA, Tafelberg Pub. Ltd., p Bosman, D.J Selecting cattle for functional efficiency. In M. M. Scholtz, L. Berg & D.J. Bosman, eds. Beef breeding in South Africa. Commemorating 40 years of beef cattle production testing, , pp Irene, Agriculture Research Council Animal Improvement Inst.

62 62 S. Banerjee et al. Bozkurt, Y Prediction of body weight from body size measurements in Brown Swiss feedlot cattle fed under small-scale farming conditions. J. Appl. Anim. Res., 29: Branton, C. & Salisbury, G.W The estimation of the live weight of bulls from heart girth. J. Dairy Sci., 29: Brito, A.F.C., Silva, A.D.E.F., Barbosa, R.T. & Kastilec, J.P Testicular thermoregulation in Bos indicus bulls: relationship with scrotal, testicular vascular cone and testicular morphology and its effect on semen quality and sperm production. Theriogenology, 61: CSA Central Statistics Agency of Ethiopia. Report on Livestock and Livestock Characteristics, Agricultural Sample Survey (2003 E.C). Statistical Bulletin No.532, Vol. II. March Addis Ababa Ethiopia. Cyprian, A., Gerald Nwachi, A.K.P.A., Pano, B.P., Finangwai Hosea Istifanus & Danbab, A.B Comparative evaluation of linear Udder and body conformation traits of Bunaji and Friesian X Bunaji cows. World J. Life Sci. Med. Res., 2(4): El Khidir, O.A A note on prediction of live weight of growing Kenana heifers from linear body measurements. Sudan J. Vet. Sci. Anim. Husbandry, 21(2): ESPSLMMP Focus on Ethiopia s Meat and Live Animal Export, Ethiopia Sanitary and Phytosanitary Standards and Livestock & Meat Marketing Program. Trade Bulletin Issue 1, Addis Ababa. ESPSLMMP Focus on Ethiopia s Meat and Live Animal Export, Ethiopia Sanitary and Phytosanitary Standards and Livestock & Meat Marketing Program. Trade Bulletin 5, July. Fasil, G On-farm phenotypic characterization of cattle genetic resources and their production systems in Awi, East and West Gojjam zones of Amhara region. M.Sc. thesis, submitted to School of Graduate Studies Haramaya University Ethiopia. Felleke, G. & Geda, G The Ethiopian dairy development policy. In: A draft policy document, Addis Ababa, Ethiopia: Ministry of Agriculture/AFRDRD/AFRDT. Food and Agriculture Organization/ SSF. Fouire, P.J., Neser, F.W.C., Oliver, J.J. & Van der Westhizen, C Relationship between production performance, visual appraisal and body measurements of young Dorper rams. South African J. Anim. Sci., 32: Gatesy, J. & Arctander, P Hidden morphological support for the phylogenetic placement of Pseudoryx ngetinhensis with bovine bovids: a combined analysis of gross anatomical evidence and DNA sequences from five genes. Syst. Biol., 49(3): Getinet, M., Ayalew, W. & Hegde, B.P Growth and reproductive performance of Ogaden cattle at Haramaya University, Ethiopia. Ethiopian J. Anim. Prod., 9(1): Gilbert, H.R. & Gregory, P.W Some features of growth and development of Hereford cattle. J. Anim. Sci., 11: Gilbert, R.P., Bailey, D.R.C. & Shannon, N.H Linear body measurements of cattle before and after 20 years of selection for postweaning gain when fed two different diets. J. Anim. Sci., 71: Goe, M.R., Alldredge, J.R. & Light, D Use of heart girth to predict body weight of working oxen in Ethiopian highlands. Livestock Prod. Sci., 69: Hafez, E.S.E Principles of animal adaptation. In E.S.E. Hafez, ed. Adaptation of domestic animals, pp Philadelphia, Lea and Fabiger. Hammond, J. Jr., Mason, I.L. & Robinson, T.J Hammond s farm animals, 4th edition. London, Edward Arnold, p Heinrichs, A.J., Rogers, G.W. & Cooper, J.B Predicting body weight and wither height in Holstein heifers using body measurements. J. Dairy Sci., 75: Hoffmann, I Livestock biodiversity. Rev. Sci. Tech. Off. Int. Epiz., 29(1): Hunters, R.A. & Buck, N Nutritional and climatic limits of beef production in the tropics. In R. Jarrige & C. Beranger, eds. World animal science, C5. Beef cattle production, pp The Netherlands, Elsevier Science. Hutcheson, D Feeding to produce export quality Ethiopian beef requirements and recommendations. Ethiopia Sanitary and Phyto sanitary standards and livestock and meat marketing program (SPS-LMM), p. 18. Kamalzadeh, A., Koops, W.J. & van Bruchem, J Feed quality restriction and compensatory growth in growing sheep: modeling changes in body dimensions. Livestock Prod. Sci., 53: Kayastha, R.B., Zaman, G., Goswami, R.N. & Haque, A Physical and morphometric characterization of indigenous cattle of Assam. Open Vet. J., 1: 7 9. Khalafalla, I.E.E., Atta, M., Eltahir, I.E. & Mohammed, A.M Effect of body weight on slaughtering performance and carcass measurements of Sudan Baggara bulls. Livestock Res. Rural Dev., 23: Article #47. Retrieved August 23, 2013 (available at lrrd.org/lrrd23/3/khal23047.htm). Khogali, A.M The effect of different dietary levels on performance, carcass characteristics and meat quality of the Baggara cattle. (Ph.D. thesis), University of Khartoum, Sudan. Kugonza, D.R., Nabasirye, M., Mpairwe, D., Hanotte, O. & Okeyo, A.M Productivity and morphology of Ankole cattle in three livestock production systems in Uganda. Anim. Genet. Resour. Inf., 48: Little, P.D., Smith, K., Cellarius, B.A., Coppock, C.B. & Barrett, C Avoiding disaster: diversification and risk management among East African herders. Dev. Change, 32(3): Livestock Marketing Authority (LMA) Study on causes of cross-border illegal trades in South, Southwest and Eastern Ethiopia. Addis Ababa, Ethiopia, Market Research and Promotion Department. Macjowski, J. & Zieba, J Genetics and animal breeding part-a. Biological and genetic foundations of animal breeding. Warszawa, Elisever Scientific PWN. Polish Scientific Publishers, pp Manzi, M., Junga, J.O., Ebong, C., & Mosi, R Factors affecting pre and post-weaning growth of six cattle breed groups at Songa Research station in Rwanda. Livestock Research for Rural Development. Volume 24, Article #68. Retrieved March 22, 2014, from Mason, I.L. & Maule, J.P The indigenous livestock of Eastern and Northern Africa technical communication 14. Farnham Royal, UK, CAB (Commonwealth Agricultural Bureaux). Mc Donald, P., Edwards, R.A., Greenhalgh, J.F.D. & Morgan, C.A Animal nutrition, 6th edition. Edinburg, UK, Pearson Edu Ltd, pp. 5 8, Milla, A.P., Mahagoub, M.M.M. & Bushara, I Estimation of live body weight from heart girth, body length and condition score in Nilotic cattle Southern Sudan. J. Anim. Sci. Adv., 2(5): Msangi, B.S.J., Bryant, M.J., Kavana, Y., Msanga, N. & Kizima, J.B Body measurements as a management tool for crossbred dairy cattle at a smallholder farm Ksition. Proc. Tanzania Soc. Anim. Sci., 26:

63 Morphometrical traits of Boran bulls reared in Southern Ethiopia 63 Mullick, D.N The estimation of the weight of cattle and buffalo from heart girth measurements. Indian J. Anim. Nutr., 3: Mwacharo, J.M., Okeyo, A.M., Kamande, G.K. & Rege, J.E.O The small East African shorthorn zebu cows in Kenya. Linear body measurements. Trop. Anim. Health Prod., 38: Mwambene, P.L., Katule, A.M., Chenyambuga, S.W. & Mwakilembe, P.A.A Fipa cattle in the southwestern highlands of Tanzania: morphometric and physical characteristics. Anim. Genet. Resour. Inf., 51: Nakachew, M Characterization of Abigar (Nuer) cattle breed at its production environment in Gambella Regional State, Ethiopia. A thesis Submitted to the School of Graduate Studies Hawassa University, p NEPAD CAADP (New Partnership for Africa s Development Comprehensive Africa Agriculture Development Programme) Ethiopia: investment Project Profile Live Animal and Meat Export Preliminary Options Outline, p. 3. Oladimeji, B.S., Osinowo, O.A., Alawa, J.P. & Hambolu, J.O Estimation of average values for pulse rate, respiratory rate and rectal temperature and development of a heat stress index for adult Yankassa sheep. Bull. Anim. Health Prod., 44: Otuma, M.O Influence of breeding designs and seasonal changes on growth of Nigerian goats and their crosses. Trop. J. Anim. Sci., 7 (2): Ozkaya, S. & Bozkurt, Y The accuracy of prediction of body weight from body measurements in beef cattle. Arch. Tierz, 52: Pagot, J Animal production in the tropics and sub-tropics. London, UK, McMillan Press Ltd., pp. 1 86, Phanchung & Roden, J.A Characterisation of the Siri breed and the Mithun cross Siri in Bhutan. Anim. Genet. Resour. Inf., 20: Pianka, E.R Evolutionary ecology, 6th edition. San Francisco, Addison-Wesley-Longman. Rajendran, R., Raja, T.V., Thiruvenkadan, A.K., Mahalinga Nainar, A. & Thangaraju, P Morphobiometrical characteristics and management of Umblachery cattle from coastal region of Tamilnadu, India. Livestock Res. Rural Dev., 20: Article #40. Retrieved August 23, 2013 (available at raje20040.htm). Rohrer, F Der Index der Körperfülle als Maß des Ernährungszustandes [The index of corpulence as measure of nutritional state]. Münchener medizinische Wochenschrift (in German), 68: Rothouge, A New Ecological perceptions of arid rangelands. Agricola, 11: Salako, A.E Application of morphological indices in the assessment of type and function in sheep. Int. J. Morphol., 24(1): Shackelford, S.D., Koohmaraie, M. & Wheeler, T.L Effects of slaughter age on meat tenderness and USDA carcass maturity scores of beef females. J. Anim. Sci., 73: Shiferaw, G In-situ phenotypic Characterization of Kereyu cattle type in Fentale district of Oromya region. M.Sc. thesis submitted to School of Graduate Studies Haramaya University Ethiopia. Slippers, S.C., Letty, B.A. & De Villerrs, J.F Prediction of body weight of Nguni goats. S. Afr. J. Anim. Sci., 30 (Suppl. 1): Solomon, D Cattle population Dynamics in the Southern Ethiopian Rangelands. Utah State University Pastoral Risk Management Project. Research brief PARIMA. SPSS Statistical package for social sciences. SPSS 12.0 for windows. Chicago, IL, SPSS Inc. Suriya, S., Boonsaen, P. & Innuruk, P Body measurements of male Kamphaengsaen beef cattle as parameters for estimation of live weight. Kasetsart J. (Nat. Sci.), 45: Szabolcs, B., Nagy, B., Nagy, L. & Kiss, B Comparison of body measurements of beef cows of different breeds Arch. Tierz., Dummerstorf, 50(4): Takele, T On farm phenotypic Characterization of Sheko breed of cattle and their habitat in Bench Maji. M.Sc. thesis submitted to School of Graduate Studies Alemaya University, Ethiopia. Tolenkhomba, T.C., Konsam, D.S., Shyamsana Singh, N., Prava, M., Damodor Singh, Y., Ayub Ali, M. & Motina, E Factor analysis of body measurements of local cows of Manipur, India. Int. Multidisciplin. Res. J., 2(2): Willamson, G. & Pyne, W.J.A An introduction of animal husbandry in the tropics, 3rd edition. USA, Longman Inc., p. 2. Willeke, H. & Dürsch, T Prediction of the body weight of Simmental heifers using heart girth measurements. Arch. Tierzucht, 45(1): Yanar, M., Tüzemen, N., Özhan, M., Aydın, R. & Uğur, F Prediction of body weights from body measurements in Brown Swiss cattle. Turk. J. Vet. Anim. Sci., 19: Young, B.A Cold stress as it affects animal production. J. Anim. Sci., 52: Yunusa, A.J., Salako, A.E. & Oladejo, O.A Principal component analysis of the morphostructure of Uda and Balami sheep of Nigeria. Int. Res. J. Agric. Sci., 1(3): Zechner, P., Zohman, F., Sölkner, J., Bodo, I., Habe, F. & Marti, E Morphological description of the Lipizzan horse population. Livestock Prod. Sci., 69(2): Zulu, D.N Genetic Characterization of Zambian native cattle breeds. M.Sc. thesis, submitted to Virginia Polytechnic Institute and State University, p. 68.

64 Animal Genetic Resources, 2014, 54, Food and Agriculture Organization of the United Nations, 2014 doi: /s Analysis of the body structure of Djallonke sheep using a multideterminant approach Peter T. Birteeb 1, Sunday O. Peters 2 and Michael O. Ozoje 3 1 Department of Animal Science, University for Development Studies, P. O. Box TL 1882, Tamale, Ghana; 2 Department of Animal Science, Berry College, Mount Berry, GA 30149, USA; 3 Department of Animal Breeding and Genetics, University of Agriculture, Abeokuta, Nigeria Summary This study aimed at using a multivariate approach to describe the body structure of Djallonke sheep in northern Ghana and to determine which approach explains better the variation in body composition. Live weight (LW) and linear body measurements including heart girth (HG), neck girth (NG), chest depth (CD), height at withers (HW), rump height (RH), body length (BL) and pin-bone width (PBW) were obtained from 172 sheep aged between two and three years. The fixed effects of sex and age were tested using the general linear model (GLM) while the Nearest Neighbor method of Hierarchical Cluster Analysis was used to group body traits into clusters. Principal Component Factor Analysis was used to describe the variation in body traits where extracted factors were varimax rotated to enhance interpretability. The analysis of variance revealed significant (P < 0.01) differences in the morphological traits of the two sexes with higher values recorded for the male in all traits examined except in PBW, which was insignificant (P > 0.05). Age had no significant influence (P > 0.05) on the body traits. The sheep weighed ± 0.89 kg averagely and had averages of other body measurements to be: ± 1.23, ± 0.79, ± 0.52, ± 0.86, ± 0.87, ± 1.06 and ± 0.23 cm for HG, NG, CD, HW, RH, BL and PBW, respectively. The product moments of correlation were positive and significant (P < 0.05, 0.01; r = ) for all pairs of traits. The body traits were categorized mainly into two clusters with the first cluster comprising the HG, HW, RH and BL while NG, CD and PBW formed the second cluster. The grouping of the traits was slightly different in Factor analysis where two underlying principal components (PC) were extracted to discern the variance structure of the Djallonke sheep. The first principal component which consisted of CD, HW, RH, BL and PBW explained percent and the second, percent thereby giving a maximum of percent generalized variance in body measurements. The traits loaded on the first principal component are closely associated with bone growth hence describing the general body size conformation while the traits (HG and NG) on the second component seem to describe only the thoracic region. It can be concluded that both the Hierarchical Cluster analysis and the Factor analysis grouped body traits similarly but the later is to be recommended because of the additional ability of indicating the amount of variation explained by the developed factors. Keywords: body measurements, cluster, generalized variance, genetic resources, multivariate, West African dwarf Résumé Cette étude a adopté une approche multivariée pour décrire la structure corporelle des moutons Djallonke du Nord du Ghana et pour déterminer quelle méthode permet d expliquer le mieux la variation observée dans la conformation corporelle. Le poids vif (PV) et une série de mesures corporelles linéaires, à savoir le périmètre thoracique (PT), le tour du cou (TC), la profondeur de la poitrine (PP), la hauteur au garrot (HG), la hauteur à la croupe (HC), la longueur du corps (LC) et la largeur de la tubérosité ischiatique (LTI), furent obtenus sur un total de 172 moutons âgés de 2 à 3 ans. Les effets fixes du sexe et de l âge ont été analysés suivant un modèle GLM alors que la méthode du plus proche voisin de l analyse hiérarchique a été utilisée pour regrouper les traits corporels dans des grappes. L analyse factorielle en composantes principales, avec rotation selon la méthode Varimax des facteurs identifiés (ceci a été fait pour améliorer l interprétabilité), a été employée pour décrire la variation observée dans les caractères corporels. L analyse de la variance a décelé des différences significatives (P < 0.01), entre les deux sexes, dans les caractères morphologiques, avec les valeurs obtenues chez les mâles étant les plus élevées pour tous les traits sauf pour la LTI, pour laquelle pas de différences significatives ont été observées (P > 0.05). L âge n a pas eu d effet significatif (P > 0.05) sur les caractères corporels. Les moutons ont pesé en moyenne ± 0.89 kg et les valeurs moyennes des mesures corporelles ont été: ± 1.23, ± 0.79, ± 0.52, ± 0.86, ± 0.87, ± 1.06 et ± 0.23 cm pour PT, TC, PP, HG, HC, LC et LTI, respectivement. Les coefficients de corrélation ont été positifs et significatifs (P < 0.05, 0.01; r = ) pour toutes les paires de caractères. Les caractères corporels ont été regroupés principalement dans deux grappes. La première grappe a compris les mesures PT, HG, HC et LC alors que TC, PP et LTI ont constitué la deuxième grappe. Le regroupement des caractères a été légèrement différent avec l analyse factorielle qui a extrait deux composantes principales sous-jacentes pour percevoir la structure de la variance des moutons Djallonke. La première composante principale, formée par PP, HG, HC, LC et LTI, a expliqué le pour cent et la deuxième le pour cent, donnant ainsi un maximum de pour cent de variance généralisée dans les mesures corporelles. Les traits compris dans la première composante principale sont étroitement liés à la croissance des os, comme quoi ils peuvent être utilisés pour décrire la conformation générale du corps. Par contre, les traits de la deuxième composante principale (PT et TC) semblent décrire seulement la région thoracique. Il peut être conclu que l analyse Correspondence to: P.T. Birteeb, Department of Animal Science, University for Development Studies, Tamale, Ghana. bpetert2000@yahoo.com; tel.: ( + 233) , ( + 233)

65 66 P.T. Birteeb et al. hiérarchique de regroupement et l analyse factorielle ont classé les caractères corporels de manière semblable. Néanmoins, la deuxième méthode est retenue pour la capacité additionnelle à indiquer la quantité de variation expliquée par les facteurs révélés. Mots-clés: regroupement, mesures corporelles, variance généralisée, ressources génétiques, multivariée, moutons Nains d Afrique Occidentale Resumen En este estudio se adoptó un enfoque multivariante para describir la estructura corporal de las ovejas Djallonke del Norte de Ghana y para determinar qué método permite explicar mejor la variación en la conformación corporal. Se tomaron el peso vivo (PV) y una serie de medidas corporales lineales (la circunferencia torácica CT, la circunferencia del cuello CC, la profundidad del pecho PP, la altura a la cruz AC, la altura a la grupa AG, la longitud del cuerpo LC y la anchura de la tuberosidad isquiática ATI) en una muestra de 172 cabezas de ganado ovino, de edades comprendidas entre los dos y los tres años. Se analizaron los efectos fijos del sexo y la edad usando un modelo GLM mientras que se empleó el método del vecino más cercano del análisis jerárquico para agrupar los parámetros corporales en conglomerados. Se recurrió al análisis factorial de componentes principales para describir la variación en los rasgos corporales, siendo rotados, según el método Varimax, los factores identificados, lo cual se hizo con el fin de mejorar la interpretabilidad. El análisis de la varianza sacó a la luz diferencias significativas (P < 0.01), entre los dos sexos, en los rasgos morfológicos, siendo mayores en los machos los valores de todos los parámetros examinados, con la excepción de la ATI, para la cual no se dieron diferencias significativas (P > 0.05). La edad no tuvo un efecto significativo (P > 0.05) sobre los parámetros corporales. Los animales pesaron de media ± 0.89 kg y los valores medios de las medidas corporales fueron: ± 1.23, ± 0.79, ± 0.52, ± 0.86, ± 0.87, ± 1.06 y ± 0.23 cm para CT, CC, PP, AC, AG, LC y ATI, respectivamente. Los coeficientes de correlación fueron positivos y significativos (P < 0.05, 0.01; r = ) para todos los pares de caracteres. Los parámetros corporales fueron agrupados principalmente en dos conglomerados. El primer conglomerado comprendió las medidas CT, AC, AG y LC mientras que CC, PP y ATI formaron el segundo conglomerado. La agrupación de los parámetros corporales fue ligeramente diferente en el análisis factorial en que se extrajeron dos componentes principales subyacentes para discernir la estructura de la varianza del ganado ovino Djallonke. El primer componente principal, formado por PP, AC, AG, LC y ATI, explicó el por ciento y el segundo el por ciento, dando así un máximo del por ciento de varianza generalizada en las medidas corporales. Los parámetros comprendidos en el primer componente principal están estrechamente relacionados con el crecimiento de los huesos, de ahí que sirvan para describir la conformación general del cuerpo, mientras que las medidas CT y CC, del segundo componente, aparentemente sólo describen la región torácica. Se puede concluir que el análisis jerárquico de conglomerados y el análisis factorial agruparon los parámetros corporales de un modo similar. No obstante, se recomienda el segundo análisis por la capacidad adicional de indicar la cantidad de variabilidad explicada por los factores identificados. Palabras clave: conglomerado, medidas corporales, varianza generalizada, recursos genéticos, multivariante, oveja Enana de África Occidental Submitted 2 December 2013; accepted 5 March 2014 Introduction Sheep have long been recognized as major contributor to protein supply and family income in most developing countries. The choice of breed in a locality is influenced by the production objectives and the environmental conditions, with increased productivity as the underlying objective. There is a considerable potential for increased sheep production, if proper management is employed (Birteeb et al., 2012), and this demands a look at their genetic resources. At present, many countries are losing their genetic resources which may have lasting effects on food security and sustainable development especially in the light of global warming changes (Hoffman, 2010). This is why the Food and Agriculture Organization of the UN deemed it necessary to promote the conservation of farm animal diversity with the view that humankind may need to keep this specific genetic biodiversity to face future (unknown) challenges such as changes in demand for livestock products, spread of new diseases, reducing environmental impact and climate change (FAO, 2011). In general, Animal Genetic Resources (AnGRs) in developing countries are being eroded through the rapid transformation of the agricultural systems in which the main cause of the loss of indigenous AnGRs is the indiscriminate introduction of exotic genetic resources before proper characterization, utilization and conservation of indigenous genetic resources (Yakubu, Salako and Abdullah, 2011). Nevertheless, AnGRs conservation is receiving keen attention of late as livestock researchers explore classical multivariate methods to study and characterize livestock populations genetically and phenotypically especially through the use of quantitative information, particularly body measurements. Linear body measurements have become very useful in livestock research, as alternative body measurements and indices estimated from various combinations of conventional and non-conventional body parameters not only provide superior guide to weight but also served as indicators of type and function in domestic animals (Salako, 2006a). Body weight is a major characteristic associated with animal breed types. Precise measurement of weight is by using weighing scales

66 Djallonke sheep body structure analysis 67 or bridges. However, proper weighing scales are neither readily available to nor affordable by smallholder livestock farmers, but even if they were, it would be inconvenient and a huge task to carry and assemble those weighing scales each time to weigh animals especially during management practices. According to Traoré et al. (2008), the definition of the breed as the operational unit for the assessment of livestock diversity all over the world, necessitates the characterization of local domestic animal populations in developing countries. The characterization of African small ruminant populations through the use of classical multivariate methods in analyzing morphological traits to assess variations within and between populations will play a major role in the maintenance of autochthonous genetic resources (Birteeb et al., 2012). Factor analysis is a multivariate technique for examining the interrelationships among a set of variables that are correlated. It deals with the reduction of a set of observable variables in terms of a small number of latent factors (Rencher, 2002). Principal component factor analysis has been effectively used to study cattle (Yakubu, Ogah and Idahor, 2009a; Pundir et al., 2011), sheep (Riva et al., 2004; Salako, 2006b; Yakubu, Salako and Abdullah, 2011; Birteeb et al., 2012), and even poultry (Ogah, Alaga and Momoh, 2009; Yakubu, Kuje and Okpeku, 2009b; Ogah, 2011). Cluster analysis is another multivariate technique in which a search is made for patterns in a data set by grouping the (multivariate) observations into clusters, with the goal of finding an optimal grouping for which the observations or objects within each cluster are similar, but the clusters are dissimilar to each other (Rencher, 2002). In grouping the observations, the similarities between all pairs of observations are considered. Cluster analysis is similar to factor analysis in the following ways: (1) neither the number of groups nor the groups themselves are known in advance; (2) when clustering variables, the similarity is based on a correlation (matrix). Among the existing sheep breeds, the Djallonke (West African dwarf) sheep is known to be trypano-tolerant hence has become a dependable asset to the local farmers in West Africa (Wafula et al., 2005). It is well adapted to and produces under the local environmental conditions and so contributes significantly to the African small ruminant genetic resources. Despite the availability of this breed of sheep, detail information on its characteristics is limited. Hence this study was undertaken to provide a detailed description of the Djallonke sheep breed as well as determine the multivariate approach that explains better the variation in its body composition. Materials and methods Management of study animals A total of 172 (28 males and 144 females) Djallonke (West African Dwarf) sheep, reared at the Pong-Tamale Livestock Breeding Station, were used for this study. Pong-Tamale is located at latitude 9 40 N and longitude 0 49 W and is about 32.7 km north of Tamale, the Northern Regional capital in Ghana (Encarta, 2009). The sheep were managed semi-intensively. They were housed in properly constructed pens throughout the night and sometimes during the day when there was the need to restrict their movement. Feed and water were provided for the animals ad libitum throughout the year. Conventional disease and pest control regimes were practised. The Breeding Station practises open nucleus breeding programme, where occasionally, outstanding sires from different farms are selected and brought to the station for improvement. Among the progenies, very superior rams are retained on the station for breeding while all yearling rams are supplied to farmers. The breeding sires and/or rams (which were all >2 years) included in this study, were second and third generation progenies of sires that were selected from different smallholder farms for improvement at the station. Hence the animals under study were quite varied in terms of pedigree. Data collection The sex and age of each sheep were noted. Ages of animals were identified using their birth records and categorized into two groups (2 years and 3 years). Eight biometric traits which include live weight (LW), heart girth (HG), neck girth (NG), chest depth (CD), height at withers (HW), rump height (RH), body length (BL) and pin-bone width (PBW) were measured on each animal. LW was measured with a hanging scale while all linear body dimensions were measured using a measuring tape. The traits were defined and measured in accordance to Birteeb et al. (2012) as follows: HG: The circumference around the chest just behind the front legs and withers. NG: This is the circumference around the neck near the withers. CD: The distance from the backbone at the shoulder to the brisket between the front legs HW: The distance from the surface of a platform on which an animal stands, to the withers of the animal. RH: This is the distance from the surface of a platform to the rump. BL: The distance from the base of the tail to the base of the neck. PBW: It is the distance between the outer edges of the major hip bones on the right and left sides. Statistical analysis The fixed effects of sex and age on the biometric traits were tested using general linear model (GLM). The

67 68 P.T. Birteeb et al. model was defined as: Y ij = m + S i + A j + (SA) ij + 1 ij, where Y ij is the individual biometric trait; μ the overall mean; S i fixed effect of the ith sex (i = ram, ewe); A j the fixed effect of the jth age ( j = 2 years old, 3 years old); (SA) ij the interaction effect of ith sex and jth age; ε ij the random error associated with each biometric trait [ε N (0, δ)]. A paired sample t-test was used to compare the heights (wither and rump) and BL. The Nearest Neighbor method of Hierarchical Cluster Analysis was used to group the various body traits into clusters. Pearson s coefficients of correlation (r), among the various traits, were estimated, which were then used for the principal component factor analysis. Anti-image correlations, Kaiser Meyer Olkin (KMO) measures of sampling adequacy and Bartlett s Test of Sphericity were computed to test the validity of the factor analysis of the data set. The factor matrix was subjected to varimax rotation to enhance interpretability of the extracted factors. All analyses were done using the appropriate programmes in SPSS version 17 (SPSS, 2008). Results The analysis of variance of LW and linear body measurements of Djallonke sheep are presented in Tables 1 and 2. The biometric traits were significantly (P < 0.01) influenced by sex, with the rams (males) being generally superior to the ewes (females) in all the biometric traits under consideration except for the PBW where the females were larger (P > 0.05) in absolute terms (Table 2). However, age did not have any significant influence on the LW and linear body traits as all the body traits of the 2 year-old sheep were statistically similar (P > 0.05) to those of the older sheep, even though the later recorded slightly higher values. Sex by age interaction was significant (P < 0.05) only for RH (Table 1). The S.E. values revealed that HG was the most varied trait, whereas the animals were more consistent by the PBW irrespective of their sex or age. A t-test between the HW and RH revealed that the animal was significantly (t = 29.9; P < 0.001) higher at the withers by 1.15 cm (Table 3). Also, the sheep were significantly taller than longer. The cluster analysis revealed the formation of two distinct clusters from the linear body traits of which the first cluster comprised four traits (Figure 1). The first agglomeration occurred between HW and RH, which then combined with BL and finally HG to form the first cluster. The second cluster resulted from the agglomeration of NG, CD and PBW. Pearson s correlation coefficients between all pairs of the various body traits are shown in Table 4. Even though LW was moderately ( ) correlated with all the body conformation traits, the highest coefficient (0.99) was obtained between RH and HW. All correlations between pairs of traits were highly significant (P < 0.01) except that between NG and PBW, which was only significant at (P < 0.05). The KMO measure of sampling adequacy, which was significantly in the middle (0.791) supported the use of the correlation matrix of the data set for factor analysis. Bartlett s test of sphericity (chi-square = ; P < 0.001), coupled with the high communalities ( ), which represent the explained variance, and the determinant (0.001) obtained from the correlation matrix, all provided support for the validity and reliability of the principal component factor analysis. The estimated factor loadings extracted by factor analysis, eigenvalues and variation explained by each factor, are presented in Table 5. After a varimax (orthogonal) rotation of the component matrix, two principal component factors were extracted. The variance of a variable was partitioned into a common portion communality shared with some or all of the other variables, which showed that percent of the variation in conformation traits were brought about by the underlying factors (PC s). The first principal component which was associated with five linear body traits (CD, HW, TH, BL and PBW), explained about percent of the generalized variance in the body measurements with moderate (0.651) to high (0.838) emphasis given to these body measurements. This factor can be considered as a generalized conformation or size of the sheep. The second factor comprised the HG and NG and accounted for percent of the generalized variance, Table 1. Analysis of variance for morphological traits in Djallonke sheep. Source of variation Mean squares and level of significance** DF LW HG NG CD HW RH BL PBW Sex ** ** ** 57.00** ** ** ** 1.09 Age Sex age * Residual **Highly significant (P < 0.01); *significant (P < 0.05); all other traits are not significant. DF, degrees of freedom; LW, live weight; HG, heart girth; NG, neck girth; CD, chest depth; HW, height at Withers; RH, rump height; BL, body length; and PBW, pin-bone width.

68 Djallonke sheep body structure analysis 69 Table 2. Least-square means (±S.E.) of live weight (kg) and linear body measurements (cm) of Djallonke sheep as affected by sex and age. Traits Sex Age Overall Male Female 2 years 3 years LW ± 1.76 a ± 0.29 b ± ± ± 0.89 HG ± 2.43 a ± 0.41 b ± ± ± 1.23 NG ± 1.55 a ± 0.26 b ± ± ± 0.79 CD ± 1.02 a ± 0.17 b ± ± ± 0.52 HW ± 1.70 a ± 0.28 b ± ± ± 0.86 RH ± 1.71 a ± 0.29 b ± ± ± 0.87 BL ± 2.09 a ± 0.35 b ± ± ± 1.06 PBW ± 0.45 ns ± 0.08 ns ± ± ± 0.23 a,b Means within the same row having different superscripts differ significantly (P < 0.05/0.01) between the two sexes. ns = not significant S.E., standard error; LW, live weight; HG, heart girth; NG, neck girth; CD, chest depth; HW, height at Withers; RH, rump height; BL, body length; and PBW, pin-bone width. Table 3. Comparison of selected linear traits (cm) of Djallonke sheep. Paired traits Mean difference t-statistic P-value HW RH <0.000 HW BL <0.000 RH BL thereby bringing the total variance accounted for to percent. Discussion Production of the autochthonous Ghanaian sheep breeds is directly associated with the concept of sustainable agriculture (Birteeb et al., 2012). For centuries, the Djallonke sheep has been an integral component of African AnGRs as it is kept and reared throughout the continent. This breed has varied names and exhibits biometric variations across geographic regions, hence could be referred to as a traditional population (FAO, 2011). Genetic variation is fundamental for livestock populations to adapt to varying environments and to respond to artificial selection; therefore, any conservation and development scheme should start from assessing the state of variation in the population (Toro et al., 2011). The superiority of males over females in this study was in agreement with the works of Birteeb et al. (2012), Legaz et al. (2011), Carneiro et al. (2010) and Yakubu and Akinyemi (2010) who attributed such differences to sexual dimorphism. Sexual dimorphism is thought to develop post-weaning because of faster mass gain by males during the age of 1 2 years (Festa-Bianchet et al., 1996). These same researchers suggested that males might have a longer season to amass (gain) body mass each year throughout their lives, while females divert annual resources into reproduction, rather than body mass. The insignificant influence of age on any of the body traits in this study was analogous to the findings of Kunene, Nesamvuni and Fossey (2007) in Zulu sheep where all matured (2 and 3 years) sheep recorded similar body traits, suggesting that at maturity sheep exhibited negligible zoometric trait variations as a result of age difference. Hence the growth and development of Djallonke sheep reach a peak after 2 years of age and then seem to remain almost constant afterwards. The report by Traoré et al. (2008) that the Djallonke sheep is generally small in size was confirmed in this study since the sheep herein was much smaller when compared to Zulu sheep in South Africa (Kunene, Nesamvuni and Fossey, 2007) and Yankasa, Uda and Balami sheep in Nigeria (Yakubu and Ibrahim, 2011). Figure 1. Dendrogram indicating the agglomeration of body traits into clusters.

69 70 P.T. Birteeb et al. Table 4. Phenotypic correlations among body traits of matured Djallonke sheep. Trait LW HG NG CD HW RH BL HG 0.70** NG 0.65** 0.59** CD 0.61** 0.54** 0.45** HW 0.62** 0.61** 0.50** 0.71** RH 0.60** 0.58** 0.48** 0.73** 0.99** BL 0.61** 0.40** 0.44** 0.61** 0.62** 0.64** PBW 0.52** 0.35** 0.18* 0.43** 0.46** 0.46** 0.48** **Highly significant (P < 0.01); *significant (P < 0.05). LW, live weight; HG, heart girth; NG, neck girth; CD, chest depth; HW, height at Withers; RH, rump height; BL, body length; and PBW, pin-bone width. Nevertheless, they were comparable to Sudan-Sahel sheep of Burkina Faso (Traoré et al., 2008) and Karayaka sheep of Turkey (Cam, Olfaz and Soydan, 2010). The animal is a typical tropical breed, being higher (taller) than long except that it slopes slightly from the withers towards the rump just like the Nigerian indigenous sheep (Yakubu and Ibrahim, 2011). Since the Djallonke sheep is a traditional population (FAO, 2011) in sub-saharan Africa, it may be biologically favoured in its ability to adapt to changes in management, climate, nutrition and marketing. The significant phenotypic correlations observed herein were similarly reported by Legaz et al. (2011), Cam, Olfaz and Soydan (2010) and Salako (2006a, 2006b). Slightly higher values were earlier reported by Afolayan, Adeyinka and Lakpini (2006). High correlations indicate the interrelationships between/among the traits and such knowledge is very useful in breeding and management practices of livestock, as selection for a given trait directly favours other positively associated traits but disfavours negatively associated traits. Nevertheless, selection for positively or negatively associated traits is influenced by the production objectives, selection goals and even the socio-cultural demands or conditions of the farmer. Correlations among body traits also serve as the basis for employment of further multivariate techniques Table 5. Varimax rotated factor loadings and communalities of body traits of Djallonke sheep. Trait Principal components PC1 PC2 Communalities HG NG CD HW RH BL PBW Eigenvalue Explained variance (%) Cumulative variance (%) LW, live weight; HG, heart girth; NG, neck girth; CD, chest depth; HW, height at Withers; RH, rump height; BL, body length; and PBW, pinbone width. (Yakubu, Salako and Abdullah, 2011; Birteeb et al., 2012) like Principal Component, Factor and Cluster analyses in quantifying trait variability and characterizing breeds. The agglomeration of the traits into two clusters suggests that body conformation of Djallonke sheep could be explained in two dimensions. Explicitly the traits on the first cluster are associated with bone growth hence give information about the general body conformation of the sheep. The second cluster seemed to indicate body thickness which may result from muscle development and/or fat deposition. However, the amount of variation attributable to each cluster could not be estimated. The measure of sampling adequacy, KMO was significantly high but slightly lower than 0.85 reported for Uda sheep (Salako, 2006b). Recently Birteeb et al. (2012) and Yakubu, Salako and Abdullah (2011) reported KMO s of 0.91 and 0.92 for sheep in Northern Ghana and Nigeria, respectively. High KMO values were equally reported for cattle (Yakubu, Ogah and Idahor, 2009a; Pundir et al., 2011). The KMO test shows whether the partial correlations among variables are low. In this study, the high KMO obtained implied that the proportion of the variance in the body traits/measurements caused by the underlying factors is high hence true factors existed and the data are factorable. The extraction of two PC factors in the Factor Analysis is comparable to the results of the Cluster Analysis where traits were grouped into two clusters. Except the replacement of HG by the PBW, the association of major linear measures (HW, RH and BL) with the first PC was analogous to the trait clustering in Figure 1. The similarity of the results of these two procedures suggests that variations in body composition of Djallonke sheep are best explained in two dimensions. The result of this study is similar to earlier reports on Djallonke sheep (Birteeb et al., 2012) and Uda sheep (Salako, 2006b; Yakubu, Salako and Abdullah, 2011) in which two underlying factors each were extracted to explain the generalized variance in body traits. However, the respective total explained variances of 87.19, and 78.7 percent by those earlier reports were higher than that obtained in this study. In studying Bergamasca sheep, Riva et al. (2004) extracted two and four factors for young and adult sheep, respectively, in the sedentary

70 Djallonke sheep body structure analysis 71 flock, and three factors each for both age groups in the transhumant flock. In a similar study in cattle, Pundir et al. (2011) extracted three latent factors to explain a total variance of percent. Again in the White Fulani cattle of Nigeria, two factors were extracted in the age group of years to account for percent of total variation, while four factors extracted in the age group of explained 86.4 percent of the total variance (Yakubu, Ogah and Idahor, 2009a). Except the lower coefficient of HG, the high loadings of body height (RH and WH), BL and PBW on the first factor in the present study is very much comparable to the report of Pundir et al. (2011) in Kankrej cows although the later study was on a different species. Most researchers (Riva et al., 2004; Salako, 2006b; Yakubu, Ogah and Idahor, 2009a; Pundir et al., 2011; Yakubu, Salako and Abdullah, 2011) have described the first factor as a representation of the general body size. In this Djallonke sheep, the first factor equally describes the general body size, while the second seems to indicate the size of the thoracic region. Both principal components could play a role in the ranking of the animals, and thus provide an opportunity to select the animals based on a group of variables rather than on isolated traits (Yakubu, Salako and Abdullah, 2011). Such knowledge will be valuable in obtaining animals with highly coordinated bodies from fewer traits amidst many body measurements, as a result of exploiting underlying factors. In recent times, linear body measurements have become very useful in livestock studies especially breed characterization. Analyses of morphometric variables that are easy to measure make it possible to explore areas such as the structure of breeds, the degree of variability between various populations, the harmony of morphological models and the definition of morphological models for a given breed (Birteeb et al., 2012). Conclusion It may be concluded that the Djallonke sheep is generally small in body size. Both the Cluster and Factor analyses yielded very similar results, thus categorizing linear body measurements into two groups each (clusters and factors). This implies that the linear body measurements considered herein can best be grouped into two main dimensions, and factor analysis approach is preferable because of its ability to quantify the explained variance in body composition. Finally, an easy way to describe the Djallonke sheep is by considering the linear measurements on the first factor. Acknowledgement The authors are grateful to the staff of the Pong-Tamale Livestock Breeding Station especially the Manager, Mr Ibrahim Shahadu and Head of the Sheep Unit, Mr Lawrence Dartay for providing the experimental animals and support during the data gathering process. References Afolayan, R.A., Adeyinka, I.A. & Lakpini, C.A.M The estimation of live weight from body measurements in Yankasa sheep. Czech J. Anim. Sci., 51: Birteeb, P.T., Peters, S.O., Yakubu, A., Adeleke, M.A. & Ozoje, M.O Multivariate characterisation of the phenotypic traits of Djallonke and Sahel sheep in Northern Ghana. Tropic. Anim. Health Prod., 45: Cam, M.A., Olfaz, M. & Soydan, E Body measurements reflect body weights and carcass yields in Karayaka sheep. Asian J. Anim. Vet. Adv., 5: Carneiro, H., Louvandini, H., Paiva, S.R., Macedo, F., Mernies, B. & McManus, C Morphological characterization of sheep breeds in Brazil, Uruguay and Colombia. Small Rum. Res., 94: Encarta (2009). World Atlas, Geography. Microsoft Student with Encarta premium. FAO Draft guidelines on phenotypic characterization of animal genetic resources, Commission on Genetic Resources for Food and Agriculture, 13th Regular Session, July, 2011, Rome (available at Festa-Bianchet, M., Jorgenson, J.T., King, W.J., Smith, K.G. & Wishart, W.D The development of sexual dimorphism: seasonal and lifetime mass changes in bighorn sheep. Can. J. Zool., 74: Hoffman, I Climate change and the characterization, breeding and conservation of animal genetic resources. Anim. Genet., 41: Kunene, N., Nesamvuni, E.A. & Fossey, A Characterization of Zulu (Nguni) sheep using linear body measurements and some environmental factors affecting these measurements. S. Afr. J. Anim. Sci., 37: Legaz, E., Cervantes, I., Perez-Cabal, M.A., de la Fauente, L.F., Martinez, R., Goyache, F. & Gutierrez, J.P Multivariate characterisation of morphological traits in Assaf (Assaf.E) sheep. Small Rum. Res., 100: Ogah, D.M Assessing size and conformation of the body of Nigerian indigenous turkey. Slovak J. Anim. Sci., 44: Ogah, D.M., Alaga, A.A. & Momoh, M.O Principal component factor analysis of the morphostructural traits of Muscovy duck. Int. J. Poult. Sci., 8: Pundir, R.K., Singh, P.K., Singh, K.P. & Dangi, P.S Factor analysis of biometric traits of Kankrej cows to explain body conformation. Asian Aust. J. Anim. Sci., 24: Rencher, A.C Methods of multivariate analysis, 2nd edition. Canada, Wiley-Interscience, John Wiley & Sons, Inc. Publication, pp Riva, J., Rizzi, R., Marelli, S. & Cavalchini, L.G Body measurements in Bergamasca sheep. Small Rum. Res., 55: Salako, A.E. 2006a. Application of morphological indices in the assessment of type and function in sheep. Int. J. Morphol., 24: Salako, A.E. 2006b. Principal component factor analysis of the morphostructure of immature Uda sheep. Int. J. Morphol., 24: SPSS Statistical package for social sciences. Chicago, IL, SPSS Inc. Toro, M.A., Meuwissen, T.H.E., Fernandez, J., Shaat, I. & Maki-Tanila, A Assessing the genetic diversity in small farm animal populations. Animal, 5:

71 72 P.T. Birteeb et al. Traoré, A., Tamboura, H.H., Kabore, A., Royo, L.J., Fernandez, I., Álvarez, I., Sangare, M., Bouchel, D., Poivey, J.P., Francois, D., Toguyeni, A., Sawadogo, L. & Goyache, F Multivariate characterization of morphological traits in Burkina Faso sheep. Small Rum. Res., 80: Wafula, P.O., Jianlin, H., Sangare, N., Sowe, J.M., Coly, R., Diallo, B. & Hanotte, O Genetic characterization of West African Djallonke sheep using microsatellite markers. The role of biotechnology. Turin, Italy, Villa Gualino, 5 7 March, pp Yakubu, A. & Akinyemi, M.O An evaluation of sexual size dimorphism in Uda sheep using multifactorial discriminant analysis. Acta Agric. Scand., Sect A Anim. Sci., 60: Yakubu, A. & Ibrahim, I.A Multivariate analysis of morphostructural characteristics in Nigerian indigenous sheep. Ital. J. Anim. Sci., 10: Yakubu, A., Ogah, D.M. & Idahor, K.O. 2009a. Principal component analysis of the morphostructural indices of White Fulani cattle. Trakian J. Sci., 7: Yakubu, A., Kuje, D. & Okpeku, M. 2009b. Principal components as measures of size and shape in Nigerian indigenous chickens. Thai J. Agric. Sci., 42: Yakubu, A., Salako, A.E. & Abdullah, A-R Varimax rotated principal component factor analysis of the zoometrical traits of Uda sheep. Arch. Zootec., 60:

72 Animal Genetic Resources, 2014, 54, Food and Agriculture Organization of the United Nations, 2014 doi: /s Variation in qualitative traits in Bhutanese indigenous chickens T. Tashi and N. Dorji Department of Animal Science, College of Natural Resources, Royal University of Bhutan, Lobesa, Punakha, Bhutan Summary The objective of the study was to determine the phenotypic variation among four Bhutan indigenous chickens (BIC) namely Seim (SM), Phulom (PL), Yuebjha Narp (YN) and Khuilay (KL). A total of 120 chickens (KL, 30; YN, 30; PL, 30; SM, 30) were included for the present study. Based on descriptive statistic analysis, the results illustrate that there are three types of comb and single type predominant. The Bhutanese chickens do not have feather on the shank and the head type is plain. The common colours of earlobe in four indigenous birds are white and red. White skin and shank colour rank high among BIC groups; however, in SM the yellowish skin colour is observed high. The predominant feather colour of four indigenous chickens is pure black in YN (48.28 percent), white in KL (30.00 percent), reddish brown in SM (33.33 percent) and PL (23.33 percent). In addition, breast, back and neck feather followed the similar pattern of colour. Among BIC populations, KL and YN strains present the greatest and lowest plumage colour diversity, respectively. For example, eight different colours of breast feather are surveyed in KL and PL birds. Therefore, considering the variation in feather colourings, KL and PL seems to have more diversity in morphology but, it requires further studies. On the other hand, YN appears to have less diversity requiring for appropriate development of conservation strategies in Bhutan. Moreover, the large variations present the possibility of making improvement among BIC. Keywords: indigenous chickens, phenotypic variation, diversity, conservation Résumé Le but de l étude a été de déterminer la variation phénotypique entre quatre poules autochtones du Bhoutan, à savoir Seim (SM), Phulom (PL), Yuebjha Narp (YN) et Khuilay (KL). Cette étude a compris un total de 120 poules (KL, 30; YN, 30; PL, 30; SM, 30). Basés sur une analyse statistique descriptive, les résultats ont décelé qu il existe trois types de crêtes, dont le type simple est le type prédominant. Les poules bhoutanaises n ont pas de plumes sur les tarses et leur tête est du type simple. Les couleurs les plus courantes des oreillons sont, chez les quatre poules autochtones, le blanc et le rouge. La peau et les tarses sont blancs chez la plupart des poules bhoutanaises. Néanmoins, de nombreux spécimens SM présentent une peau de couleur jaunâtre. La couleur prédominante du plumage est le noir pur chez YN (48.28 pour cent), le blanc chez KL (30.00 pour cent) et le marron rougeâtre chez SM (33.33 pour cent) et chez PL (23.33 pour cent). En outre, le plumage de la poitrine, du dos et du cou suivent un schéma de couleur similaire. Parmi les différentes populations de poules bhoutanaises, les lignées KL et YN sont celles qui présentent, respectivement, la plus grande et la plus faible diversité chromatique dans leur plumage. Ainsi, par exemple, huit couleurs différentes ont été relevées pour le plumage de la poitrine des poules KL et PL. Par conséquent, d après la variation chromatique du plumage, les lignées KL et PL semblent être celles ayant la plus grande diversité morphologique. Toutefois, des études plus approfondies s avèrent nécessaires. D un autre côté, la lignée YN semble être celle qui présente la plus faible diversité, comme quoi des stratégies devraient être mises en place au Bhoutan pour assurer la conservation de cette lignée. De plus, les grandes variations observées entre les différentes poules bhoutanaises ouvrent la possibilité de procéder à des projets d amélioration génétique. Mots-clés: poules autochtones, variation phénotypique, diversité, conservation Resumen El objetivo del estudio fue determinar la variación fenotípica existente entre cuatro gallinas autóctonas de Bhután, concretamente entre Seim (SM), Phulom (PL), Yuebjha Narp (YN) y Khuilay (KL). En este estudio, se incluyeron un total de 120 gallinas (KL, 30; YN, 30; PL, 30; SM, 30). Con base en un análisis estadístico descriptivo, los resultados mostraron que se dan tres tipos de cresta, de los cuales predomina el tipo sencillo. Las gallinas butanesas no tienen plumas sobre los tarsos y la cabeza es de tipo sencillo. En las cuatro gallinas autóctonas, los colores de orejilla más comunes son el blanco y el rojo. Por lo general, el blanco es el color de la piel y de los tarsos en las gallinas butanesas. No obstante, muchos ejemplares SM presentan piel de color amarillento. El color predominante en las plumas es el negro puro en YN (48.28 por ciento), el blanco en KL (30.00 por ciento) y el marrón rojizo en SM (33.33 por ciento) y en PL (23.33 por ciento). Asimismo, las plumas del pecho, del dorso y del cuello siguen un patrón cromático similar. Entre las distintas poblaciones de gallinas butanesas, las variedades KL y YN son las que presentan, respectivamente, la mayor y la menor diversidad cromática en su plumaje. Así, por ejemplo, se han registrado ocho colores diferentes para el plumaje del pecho en las gallinas KL y PL. Por tanto, de acuerdo con la variación en el colorido de las plumas, KL y PL son probablemente las variedades que mayor Correspondence to: N. Dorji, Department of Animal Science, College of Natural Resources, Royal University of Bhutan, Lobesa, Punakha, Bhutan. s: nepdorpion186@yahoo.com; nedup@cnr.edu.bt 73

73 74 T. Tashi and N. Dorji diversidad morfológica presentan. Sin embargo, se precisan más estudios. Por otro lado, parece que la variedad YN es la que menor diversidad presenta, con lo que se hace necesario un adecuado desarrollo de estrategias para la conservación de esta variedad en Bhután. Por último, las grandes variaciones observadas entre las distintas gallinas butanesas abren la posibilidad de hacer mejora genética. Palabras clave: gallinas autóctonas, variación fenotípica, diversidad, conservación Submitted 2 July 2013; accepted 23 January 2014 Introduction In developing countries, the majority of poultry products come from traditional production practices. Indigenous chickens continue to be most popular due to their adaptability to harsh conditions and their resistance against local diseases (Ajayi, 2010; Dorji and Gyeltshen, 2012). This indigenous bird manifests a great deal of variation which is due to genetic and environmental factors (Olori and Sonaiya, 1992). Thus, they are reservoirs of genetic materials for genetic studies, improvement, preservation and conservation (Dorji, Duangjinda and Phasuk, 2012). Poultry production has been playing crucial roles in sustaining and supplementing a cheap protein to the poor (Dorji and Gyeltshen, 2012). Importantly, indigenous chickens are kept for their socio-cultural importance (Dorji and Gyeltshen, 2012; Dorji, Duangjinda and Phasuk, 2012). For example, society believes in consuming eggs and meat during pregnancy and after delivery. The traditional sub-sector consists of poorly uncharacterized indigenous chickens (Dorji and Gyeltshen, 2012). Recently, Food and Agriculture Organization of the United Nation (FAO) Domestic Animal Diversity Information System (DADIS) listed ten indigenous strains of chicken in Bhutan. Dorji, Duangjinda and Phasuk (2012) conducted genetic study on four strains of Bhutanese local chicken based on the economic importance to the rural society and their popularity. Dorji, Duangjinda and Phasuk (2012) documented for greater genetic diversity of Khuilay (naked neck) and lowest for Yuebjha Narp (black chicken). Moreover, they did mention that the phenotypic variation was likely to be highest for Khuilay but they did not quantify. Therefore, the objective of the study was to access diversity of Bhutanese local chicken based on qualitative trait. Materials and methods Sample sizes Considering the agro-ecology, socioeconomic significance of chicken production and population of indigenous chickens, Tsirang was selected and this district has been recognized as important genetic resources of poultry in the country. Thus, farmers are being trained related to conservation activity of traditional birds. The study was based on four Bhutanese indigenous chicken comprising Khuilay (KL = 30), Yuebjha Narp (YN = 30), Phulom (PL = 30) and Seim (SM = 30). Data collection and analysis The data captured parameters which are visually judged for variation in the phenotypic traits such as colour of feather, shank, ear lobe and shank, comb type and head type (FAO, 2012). Descriptive statistics of SPSS 16 was used to analyze the data. Results and discussion Feather colour The plumage colour of local chickens available was reddish brown constituting the maximum proportion of plumage colour in SM and PL varieties with and percent, respectively. An utmost plumage colour was white (30.00 percent) and black (48.28 percent) in KL and YN, respectively. Similarly, a predominant colouring as reddish brown (20.00 percent), black (30.00 percent) and reddish (30.00 percent) in neck feather was observed for PL, YN and SM, respectively. Neck feather was more varied in PL population. On contrary, YN contains the least variation of neck feather colours. The naked neck gene is a dominant gene and is responsible for general reduction of feather over body surface and total loss of feather in neck region (Merat, 1986; Ige et al., 2012). In terms of back feather colouration, the highest percentage was surveyed in SM, KL, PL and YN was reddish brown (40.00 percent), white (23.33 percent), brownish (20.00 percent) and black (46.67 percent), respectively. The same colour domination was examined for breast feather colourings in SM (reddish brown, percent), KL (white, percent), PL (brownish, percent) and YN (black, percent). Importantly, plumage, neck feather, back and breast feather colour were found to be more varied in KL and PL populations (Table 1). This probably explains the greater diversity with respect to feather colourations (Bhuiyan et al., 2005; Faruque et al., 2010). On the other hand, lowest feathers colour was observed in YN indicating the least diversified among the four Bhutan indigenous chickens (BIC) populations agreeing with Dorji, Duangjinda and Phasuk (2012) findings. Their report was based on FAO recommended

74 Variation in qualitative traits in Bhutanese indigenous chickens 75 Table 1. Feather colour in four Bhutanese indigenous chicken populations (percent). Trait Seim Khuilay Phulom Yuebjha Nap Characteristic Percent Characteristic Percent Characteristic Percent Characteristic Percent Plumage Reddish brown White Reddish brown Black Reddish black Brownish black Barred Black/dark green Reddish dark green Black 6.67 White Black/white Brownish Reddish/brown 6.67 Black Black/brown Reddish brown, Barred 6.67 Brownish black Black/dark green/ 6.90 black lines brown Reddish black Brownish 6.67 Brownish brown Reddish black 6.67 White, yellowish 3.33 brown Reddish 6.67 Brownish white 3.33 Neck feather Back feather Breast feather Reddish Reddish brown Black Reddish black Brownish Black/dark green Brownish White Brownish Reddish brown Reddish Black/brownish Brownish black Whitish black Black/white 6.67 Multiple Black White 3.33 Brownish yellow 3.33 White black lines 6.67 Reddish 3.33 Reddish/dark green 3.33 Brownish black 6.67 Yellowish black 3.33 Reddish brown White Brownish Black Reddish/black Brownish black Whitish black Black/dark green Multiple Reddish brown Reddish brown Black/brownish Brownish 6.67 Black/white Brownish black Black/white 3.33 Reddish/dark green 6.67 Black/dark green White Black 3.33 Reddish 6.67 Reddish 6.67 Reddish 3.33 Brownish 6.67 Yellowish white 3.33 Brownish/dark 3.33 Black 3.33 green Reddish brown White Brownish Black Brownish Brownish White Light black Reddish Black Reddish brown Black/dark green Reddish black Reddish Whitish black Brownish/black 6.67 Brownish white 6.67 Brownish/black 6.67 Reddish Black/white 6.67 Multiple 6.67 Black/white 6.67 Brownish black Brown spot 3.33 Brownish yellow 3.33 Yellowish/white 3.33 Black 6.67 Reddish/brown 3.33 Brownish white 6.67 molecular tools. Moreover, the frizzling and naked neck genes is considered for increased feed efficiency, growth rate, egg production and disease tolerance (Ajayi, 2010; Egahi, Dim and Momoh, 2013). Moreover, genetic distance between broiler and KL was sufficiently close (Dorji, Duangjinda and Phasuk, 2012). Comb type and earlobe Only three types of comb (rose, pea and single) were found in all four indigenous chickens of Bhutan (Table 2). The finding agrees with the reports of Daikwo, Okpe and Ocheja (2011) and Ige et al. (2012) on native chickens of Nigeria. Egahi et al. (2010), Dana et al. (2010) and Badubi, Rakereng and Marumo (2006) reported more than three comb types contradicting with present study. Nevertheless, the common comb occurring was single comb type as examined for Nigerian local birds (Egahi et al., 2010; Apuno, Mbap and Ibrahim, 2011; Daikwo, Okpe and Ocheja, 2011). Parallel result was also obtained from native chickens of Botswana (Badubi et al., 2006) and Bangladeshis (Bhuiyan et al., 2005; Faruque et al., 2010; Uddin et al., 2011). This indicates that the bird have been favoured by hot climatic conditions whereby large comb, such as single comb allows for efficient heat regulation (Apuno, Mbap and Ibrahim, 2011). On the other hand, pea type was the most frequently sampled for Ethiopian traditional birds (Dana et al., 2010). Except for the SM, all the BIC birds have ear lobes (Table 2). Earlobe colour was dominated by red (50.00 percent) in SM, white (46.67 percent) in KL, white and red (36.67 percent) and white (43.33 percent) in YN flocks. The dominant white colour was also reported in Nigerian (Egahi et al., 2010; Ige et al., 2012), Bangladeshis (Faruque et al., 2010) and Ethiopian (Dana et al., 2010) indigenous chicken. Furthermore, Faruque et al. (2010) documented that the red and white earlobe colour dominants in naked neck population and is in line with KL populations.

75 76 T. Tashi and N. Dorji Table 2. Shank, skin and earlobe colour, comb type, heat type and ptilopody among four indigenous chickens of Bhutan (percent). Trait Seim Khuilay Phulom Yuebjha Nap Characteristic Percent Characteristic Percent Characteristic Percent Characteristic Percent Shank colour Yellowish White White Black White Yellow Reddish Ash Ash Black Yellowish white Light red Yellowish white Ash Yellow White Black 6.67 Light red Black 6.67 Pale yellow 3.33 Yellowish white 3.33 Ash white 6.67 Skin colour Yellowish Reddish White White White White Red Light red Light red Reddish white Yellowish Yellowish Red 6.67 Yellowish Yellowish white Ash 6.67 Pale yellow 6.67 Reddish ash 3.33 Reddish white 6.67 Yellowish white 3.33 Ear lobe Present Present Present Present Absent 6.67 Absent 0.00 Absent 0.00 Absent 0.00 Earlobe colour Red White White White White Red Red Red Yellowish white Reddish white 3.33 Reddish white Ash Ash 3.33 Ash Ash black 6.67 Yellowish 3.33 Yellowish white 3.33 Comb type Single Single Single Single Rose Rose Pea Pea Pea Pea Rose Rose Head shape Plain Plain Plain Plain Crested 0.00 Crested 0.00 Crested 0.00 Crested 3.33 Ptilopody Absent Absent Absent Absent Present 0.00 Present 0.00 Present 0.00 Present 0.00 Ptilopody and head type Throughout the study crested head and ptilopody were not found in all BIC populations (Table 2). The absence of feather on shank aligns with the finding of Halima (2007) in Northwest Ethiopian local birds. Faruque et al. (2010) also observed that the shank feather phenomenon was absent in three genotypes of Bangladeshis native chicken. On contrary, lower percent of ptilopody among Nigerian local birds was examined (Daikwo, Okpe and Ocheja, 2011). Skin and shank colour The skin colour ranged from white to red among the studied BIC populations. However, the prime skin colour in SM, KL, PL and YN was yellowish (46.67 percent), reddish (33.33 percent), white (30 percent) and (33.33 percent), respectively (Table 2). Daikwo, Okpe and Ocheja (2011) also reports multiple shank colour for Nigerian traditional birds. This was contradicting with the Ethiopian local bird s shank colour with white and yellow (Dana et al., 2010). Nevertheless, high percent colour of shank follow the same pattern. Shank with white colour accounted the most in KL (26.67 percent) and PL (33.33 percent), yellowish (36.67 percent) in SM and black (36.67 percent) in YN. Faruque et al. (2010) also mentioned that the white shank colouring dominants in Bangladeshis chickens. This contradicts with Mancha (2004) whose finding in Plateau chicken was pink, dark ash, ash and light yellow. Genes that affect plumage colour also affect shank colour (Crawford, 1990).The presence of large variations further confirms that higher diversity in KL and PL as suggested by Daikwo, Okpe and Ocheja (2011). Conclusion Bhutanese indigenous chicken has their place in contributing to the genetic collection because of their hardiness and aptitude to survive under low input conditions. Development of conservation strategies is very important to make genetic resources available for meeting the future unpredicted breeding requirement. Among four studied chicken strains in Bhutan, single comb type predominates and there was no feather on shank and head. Based on feather colours, Kuilay seems to be more diversed while Yuebjha Narp reflect lower morphological variation. It is further suggested to conduct morphometric study on Bhutanese chicken populations to confirm our claims. Acknowledgements We would like to thank individual farmers for their active participation during this study.

76 Variation in qualitative traits in Bhutanese indigenous chickens 77 References Ajayi, F.O Nigerian indigenous chicken: a valuable genetic resource for meat and egg production. Asian J. Poultry Sci., 4: Apuno, A.A., Mbap, S.T. & Ibrahim, T Characterization of local chickens (Gallus gallus domesticus) in Shelleng and Song local government areas of Adamawa state, Nigeria. Agric. Biol. J. N. Am., 2(1): Badubi, S.S., Rakereng, M. & Marumo, M Morphological characteristics and food resources available for indigenous chickens in Botswana. Livestock Res. Rural Dev. (available at lrrd.org/lrrd18/1/badu18003.htm). Bhuiyan, A.K.F.H., Bhuiyan, M.S.A. & Deb, G.K Indigenous chicken genetic resources in Bangladesh: current status and future outlook. Animal Genet. Res. Inf. (AGRI), 36: 73 84, FAO, Rome. Crawford, R.D Poultry breeding and genetics. New York, Elsevier Press, pp Daikwo, I.S., Okpe, A.A. & Ocheja, J.O Phenotypic characterization of local chickens in Dekina. Int. J. Poultry Sci., 10(6): Dana, N., Dessie, T., van der Waaij, L.H. & van Arendonk, J.A.M Morphological features of indigenous chicken populations of Ethiopia. Animal Genet. Res., 46: Dorji, N. & Gyeltshen, T Characterisation of family poultry production in Haa and Mongar districts of Bhutan. Livestock Res. Rural Dev. (available at Dorji, N., Duangjinda, M. & Phasuk, Y Genetic characterization of Bhutanese native chickens based on an analysis of Red Junglefowl, domestic Southeast Asian and commercial chicken lines. Genet. Mol. Biol., 35(3): Egahi, J.O., Dim, N.I., Momoh, O.M. & Gwaza, D.S Variations in qualitative traits in the Nigerian local chicken. Int. J. Poultry Sci., 9: Egahi, J.O., Dim, N.I. & Momoh, O.M The effect of plumage modifier genes on egg quality indices of the Nigerian local chicken. J. Agric. Vet. Sci., 2(2): Faruque, S., Siddiqee, N.U., Afroz, M.A. & Islam, M.S Phenotypic characterization of native chicken reared under intensive management system. J. Bangladesh Agric. University, 8: Food and Agriculture Organization Phenotypic characterization of animal genetic resources. FAO Animal Production and Health Guidelines 11, Rome. Halima, H.M Phenotypic and genetic characterization of indigenous chicken populations in Northwest Ethiopia. University of Free State, Bloemfontein, South Africa. (Ph.D. thesis). Ige, A.O., Salako, A.E., Yakubu, A. & Adeyemi, S.A Qualitative traits characterization of Yoruba and Fulani ecotype indigenous chickens in derived Savannah zone of Nigeria. Int. J. Poultry Sci., 11(10): Mancha, Y.P Characterization of local chickens in Northern part of the Jos Plateau. Animal Production Programme, School of Agriculture, ATBU, Bauchi. (Ph.D. thesis). Merat, P Potential use of Na (naked neck) gene in poultry production. Int. J. Poultry Sci., 42: Olori, V.E. & Sonaiya, E.B Effect of length of lay of Nigerian indigenous chicken on their egg composition and shell quality. Nig. J. Anim. Prod., 19: Uddin, A.H., Ali, A., Aktar, Y. & Khatun, M.A Geographical distribution, classification, characterization and conservation of different native chicken varieties of Bangladesh. Bangladesh Res. Publ. J., 5(3):

77 Animal Genetic Resources, 2014, 54, Food and Agriculture Organization of the United Nations, 2014 doi: /s Genetic polymorphism of α S1 casein in Guéra and Sahel goat D. Traoré, Y. Sanogo, R. Fané, A. Touré, O. Cissé and A.H. Babana Laboratoire de Recherche en Microbiologie et biotechnologie Microbienne (LaboREM-biotech), Faculté des Sciences et Techniques (FST), Université des Sciences, des Techniques et des Technologies de Bamako (USTTB), Mali Summary The α S1 casein gene is one of the lacto protein genes, which are involved in the milk synthesis. The objective of this study was to identify genotypes of B and E alleles at the locus α S1 casein in the Guéra and Sahel goat herds in Mali. These herds were located in the Regional Center of Agronomic Research and the Ségala village in Kayes, Mali. A total of 101 blood samples (53 for Guéra goat and 48 for Sahel goats) were collected and analysed. Alleles B and/or E, located on exon 19, of the α S1 casein gene were amplified using allele specific amplification polymerase chain reaction and analysed using 2 percent agarose gel electrophoresis. Results from this analysis identified three genotypes: BB, BE and EE with frequencies 0.77, 0.17 and 0.06 for the Guéra goats. However, only the BB genotype was identified in the Sahel goats. The frequencies of alleles B and E in the Guéra goats were 0.86 and 0.14, respectively, with an observed and expected heterozygotes of 0.08 and 0.12, respectively. Keywords: allele, α S1 casein, Guéra goat, Sahel goat Résumé Le gène α S1 caséine fait partie du groupe de gène des lactoprotéines intervenant dans la synthèse du lait. L objectif de notre travail était d identifier les génotypes liés aux allèles B et E au locus α S1 caséine chez deux races caprines, la chèvre Guéra et du Sahel au Mali. Les analyses ont porté sur le gène α S1 caséine d un troupeau de chèvre Guéra du Centre Régional de Recherche Agronomique de Kayes au Mali et d un autre troupeau, la chèvre du Sahel en milieu rural dans un village (Ségala) près de Kayes au Mali. Des amorces spécifiques ont été utilisées pour l identification des allèles B et/ou E du gène α S1 caséine (exon 19). Au total, 101 échantillons de sang (53 de la chèvre Guéra et 48 de la chèvre du Sahel) ont été prélevés et le gène α S1 caséine a été amplifié par la méthode de l allèle spécifique polymerase chain reaction (AS-PCR). La révélation des allèles a été faite après électrophorèse sur un gel d agarose 2 percent en utilisant le bromure d éthidium. Trois génotypes, (BB, BE et EE) avec des fréquences respectives de 0,77; 0,17 et 0,06 ont été observés chez la chèvre Guéra. En revanche, un seul génotype (BB) a été identifié chez la chèvre du Sahel. Les fréquences des allèles B et E au niveau du troupeau Guéra étaient respectivement 0,86 et 0,14. L hétérozygotie observée et attendue était de 0,08 et 0,12. Mots-clés: allèles, α S1 caséine, chèvre Guéra, chèvre du Sahel Resumen El gen de la caseína α S1 es uno de los genes, codificantes para proteínas lácteas, implicados en la síntesis de la leche. El objetivo de este estudio era identificar los genotipos de los alelos B y E del locus de la caseína α S1 en los rebaños de cabras Guéra y Sahel de Malí. Estos rebaños se encontraban en el Centro Regional de Investigación Agronómica y en el pueblo de Ségala, en la región de Kayes de Malí. Se tomaron y se analizaron un total de 101 muestras de sangre (53 de cabras Guéra y 48 de cabras Sahel). Los alelos B y/o E, situados en el exón 19, del gen de la caseína α S1 fueron amplificados de forma específica usando la reacción en cadena de la polimerasa y analizados por electroforesis en gel de agarosa (2%). Los resultados de este análisis permitieron identificar tres genotipos: BB, BE y EE, con unas frecuencias de 0,77, 0,17 y 0,06 para las cabras Guéra. Sin embargo, tan sólo el genotipo BB fue identificado en las cabras Sahel. Las frecuencias de los alelos B y E en las cabras Guéra fueron de 0,86 y 0,14 respectivamente, con una heterocigosis observada y esperada de 0,08 y 0,12, respectivamente. Palabras clave: alelo, caseína α S1, cabra Guéra, cabra Sah Submitted 7 August 2013; accepted 12 November 2013 Introduction Correspondence to: D. Traore, Laboratoire de Recherche en Microbiologie et biotechnologie Microbienne (LaboREM-biotech), Faculté des Sciences et Techniques (FST), Université des Sciences, des Techniques et des Technologies de Bamako (USTTB), Mali. dtraore@laborem-biotech.com Small ruminants and especially goats play an important role in the economies of rural smallholder families in Mali. Goat milk is an important source of protein for rural populations in the region of Kayes and particularly 79

78 80 D. Traoré et al. in the fight against child malnutrition. In recent years, the Institute of Rural Economy of Mali through the Programme Small Ruminants conducted research activities to introduce the Guéra goats in rural areas with a view to improve milk production of local goats. The Guéra goats widely found in Mauritania was introduced at the Agricultural Research Station in Kayes Mali in This goat breed is raised for its high milk performance (Traoré, Nantoumé and Diarra, 2005). According to the Mauritanian farmers, the origin of this breed is the Canary Islands (Spain). Information about its genetic potential in Mali is sparse. The present research will contribute to a better understanding of the genetic characteristics of Guéra and Sahel goats, through analysis of α S1 casein gene. In the past few years, α S1 casein gene in ruminants and especially in goats has been used as a tool for genetic characterization (Ramunno et al., 2001; Angiolillo et al., 2002; Feligini et al., 2005). This gene is part of the lactoproteins (α S1, α S2, β, κ casein, α lactalbumin and β lactoglobulin), which has been used in several studies to identify genotypes of animals and their links to milk performances (Boulanger, Grosclaude and Mahe, 1984; Grosclaude et al., 1987). Today, several alleles of these genes have been identified and characterized in different species and breeds. For the α S1 casein gene in goats, seven alleles (A, B, C, D, E, F and O) with four levels of casein synthesis: 3.6 g/l for strong alleles (A, B and C), 1.6 g/l for the allele E, 0.6 g/l for low alleles (D and F) and 0 in goats with the allele null (0), were identified (Grosclaude et al., 1987). In Alpine goats, it has been found that the goats with genotype AA produce significantly less milk than those with genotypes AE, AF, EE and FF, the highest milk yield being observed in goats with EE genotype (Barbieri et al., 1995). At birth, kids with genotype AA had an average weight significantly lower than those with AE and EF genotypes. Alleles A, B and C are recognized to be the ancestral types carried nucleotide substitutions, whereas allele E is from an insertion of 458 bp nucleotides at the last exon of the gene (Martin, 1993). F allele is characterized through epitasis abnormalities in form of deletions of 37 amino acid residues (Barbieri et al., 1995). Studies on genetic characterization of casein gene complex in West African goats (Red Sokoto, West African Dwarf Nigeria, West African Dwarf Cameroon and Borno) were carried out by Caroli et al. (2007). The four casein genes α S1, β, α S2 and κ were typed at the DNA level. A polymerase chain reaction (PCR) single-strand conformation polymorphism method was implemented for the identification of CSN1S1*F allele simultaneously with A/0(1), B/E, N and the new allele (CSN1S1*B ). The investigations showed a high frequency of CSN1S1*B in all the population. Vacca et al. (2009) investigated on genetic diversity at the four casein genes in the Arbi goat, a native goat breed from Tunisia. The most frequent allele at the CSN1S1 locus was B, followed by A and 01. Analysis on α S1 gene were carried out in a dairy goat herd (Saanen and Alpine) in Southeastern region of Brazil (Soares et al., 2009). Three different regions of the α S1 casein gene (CSN1S1) were investigated to determine the frequencies of major alleles for null, low, intermediate and high milk protein expression. Allele E was the most frequent in both Saanen (0.35) and Alpine (0.48) breeds followed with F allele (0.30). Information on casein genes by Guéra goat is rare. The relatively high milk yield in Guéra goat, 1 3 litres, in 142 days of lactation length (Traoré, Nantoumé and Diarra, 2005), could be better understood through molecular genetics analysis. The objective of the investigations was to analyse the genetic polymorphism of α S1 casein in Guéra and Sahel goats and to determine the genetic diversity in the studied populations. The present study was focused on identifying the B and E alleles of the CSN1S1 gene using allele specifics primers. Materials and methods Materials Descriptive characteristics of the breeds Guéra goat In this study, 53 Guéra goats were randomly sampled from the experimental herd of the Regional Center for Agronomic Research in Kayes (Mali). The breed is characterized by a great diversity in coat colour (uniform: white, brown, black, etc. or pied). It is distinguished through its long hair. The average weight of mature males and females are 44 and 36 kg, respectively (Traoré, Nantoumé and Diarra, 2007). Sahel goat During the study 48 Sahel goats were also randomly sampled in farmer s herds in Ségala in the region of Kayes (Mali). This breed is raised in the Sahelian zone of West Africa. It is found from Mali to Chad. It is kept for its milk and meat. The Sahel goat appears in various coat colours: uniform white, brown, black or pied. The average weight of mature males and females are 35 and 25 kg, respectively (Nantoumé, Traoré and Diarra, 2005). Methods Sampling and DNA extraction Blood samples from 53 Guéra goats and 48 Sahel goats were collected using confetti paper type Whatman 903 for molecular genetic analysis. DNA was extracted from blood on confetti paper using the Truett et al. (2000) protocol. For alkaline lysis, a solution

79 Genetic polymorphism of α S1 casein in Guéra and Sahel goat 81 of 25 mm NaOH, 0.2 mm disodium EDTA, ph 12 was used. The precipitation of DNA molecules was obtained with absolute ethanol in the presence of potassium acetate (8 M, ph 7.4). A solution of neutralization (40 mm Tris HCl, ph 5) was also used to stabilize DNA. Final DNA was obtained using 70 percent ethanol by washing. DNA concentrations ranging between 18 and 58 ng/μl measured with a spectrometer (Eppendorf Bio Photometer) at 260 and 280 nm, were used for PCR. Allelic specific amplification by PCR for identification of B and E alleles For the PCR, 50 ng of genomic DNA was amplified in a reaction volume of 30 μl using 150 nm primer, 50 mm KCL, 10 mm Tris HCl (ph 8.3), 2.5 mm MgCl 2, 200 μm dntps and 1.25 U Taq polymerase (Invitrogen). Primers were selected according to the study of Feligini et al. (2005) for the identification of B alleles (90 bp) and/or E (550 bp) at the locus CSN1S1 (exon 19). The accession number of gene at the Genbank is X The thermocycler thermal-type PTC-200 (M) Research was used for amplification with the following programme: initial denaturation at 94 C for 3 min, denaturation at 94 C for 45 s by 25 cycles, hybridizing at 59 C for 45 s, extension at 72 C for 1 min, final extension at 72 C for 5 min and the conservation at 4 C. The PCR products were migrated in 2 percent agarose gel (BDH electran), for 60 min at 100 V. Ethidium bromide (1 μm) was used to reveal the bands. Statistical analysis The frequencies of B and E alleles were calculated using the formula: pb = [2(BB) + (BE)]/2N, qe = [2(EE) + (BE)]/2N, where pb and qe are the frequencies of B and E alleles, BB and EE are the genotypes and N is the total number. The heterozygous rates (observed and expected) were evaluated assuming that the herds were in genetic equilibrium of Hardy Weinberg. The observed heterozygosity (H o )=1/N Σ Hi, where N is the total number of loci, H i is the heterozygosity at the locus i. The expected heterozygosity (H e )=1 (f1 2 +f fk 2 )=1 Σ f 2, where f are frequencies f1, f2,...,fk. Total heterozygosity (H T )=1 (f1 2 T +f2 2 T +...fk 2 T )=1 Σ f 2 T, where, f T is the total frequency. The genetic diversity index was calculated using FST (H T H e )/H T,whereF ST is the genetic diversity indice, H T is the total heterozygosity and H e is the expected heterozygosity. Results Three genotypes (BB, BE and EE) were observed in Guéra goats samples (Figure 1a and b), whereas only genotype (BB) was observed in Sahel goat samples (Figure 2). In the Guéra goat, homozygous BB (41 animals) was the most frequent genotype (0.77) followed by the heterozygous BE (nine animals) with 0.17 and the homozygous EE (three animals) with In Sahel goat, all animals (48) were homozygous BB. The genotypic and allelic frequencies in the two breeds are shown in Tables 1a and 1b. The frequency of allele B in the two breeds and that of E is given in Table 2. Different heterozygosity rates were evaluated. The genetic diversity index (F ST ) in the two breeds was low (0.07). Discussion Previous studies (Missohou, Talaki and Maman Laminou, 2006) on the casein gene on African goat breeds including those of West Africa have shown that the allelic frequencies were low and the most identified allele were A and B (strong allele). Further investigations (Caroli et al., 2007) on the casein group genes using four breeds or varieties of goats in Africa, namely the Sokoto red goat (Nigeria), the Borno goat (Nigeria), the Nigerian and Cameroon dwarf goat, showed that the A and B alleles were more frequent with a predominance of B allele (0.37). A new allele, named CSN1S1*B, was identified with a high frequency in all populations, ranging from (West African Dwarf Cameroon) to (Borno). The intermediate E allele was also observed with a very low frequency of 0.01 in Sokoto red goats. In our study, the results obtained in the Sahel goats were expected. They indicate the non-introduction of exotic breeds in the studied herd in Ségala (Kayes). In contrast, the Guéra goats introduced in Africa for several decades, from Spain was subject of many questions relating its wide range of milk performances observed at the Figure 1. (a) Amplification of the α S1 casein gene (E 19) in 2 percent agarose gel for Guéra goat; M: marker 14; lane 1: EE lanes 2 7: BB lane C: negatif control. (b) Amplification of the α S1 casein gene (E 19) in 2 percent agarose gel for Guéra goat; M: marker 14; lanes 6 et 7: BE lanes 5; 8; 9; 10; 11: BB line C: negatif control.

80 82 D. Traoré et al. Table 2. Observed, expected and total heterozgosity at the locus α S1 casein (exon 19) in Guéra and Sahel goat Heterozygoty observed (H o ) expected (H e ) Guéra goat: 0.17 Guéra goat: 0.24 Sahel goat: 0.00 Sahel goat: 0.00 Both breeds: 0.08 Both breeds: 0.12 Total (H T ) = 0.13 Figure 2. Amplification of the α S1 casein gene (E 19) in 2 percent agarose gel for Sahel goat; M: marker 14; lanes 1 à 6: BB; lanes 7; 8; 9; 10: negatifs lane C: negatif control. Agricultural Research Center in Kayes (Mali). The study made on the Guéra goat breed by Traoré, Nantoumé and Diarra (2005) showed an average daily milk quantity of 1.9 litres ranging from 1 to 3 litres, with a total milk production of 262 litres in 142 days. The milk protein percentage was estimated to 3.9 percent (Traoré, Nantoumé and Diarra, 2005). It appears that the Guéra goat carried the allele (E) with a frequency of This allele is widely identified in some European dairy goats (Feligini et al., 2005) such as the Alpine goat and in the Red Sokoto at a low frequency (Caroli et al., 2007). The allelic frequency of E in our study is higher than those reported by Missohou, Talaki and Maman Laminou (2006) with Studies on Spanish dairy goats, especially those of the Canary Islands breeds showed a high frequency of the allele E (Jordana et al., 1996). A phylogeny relationship between the Guéra goat and Spanish goat breed was not investigated in the present study. However, the relatively high milk production of Guéra could be explained by the presence of the allele E in this breed. The genetic diversity observed was low, due to relatively low heterozygosity Table 1a. Genotype frequencies at the locus α S1 casein (exon 19) in Guéra and Sahel goat Guéra goat Sahel goat Genotypes Number Frequencies Number Frequencies BB BE EE rate. The absence of the E allele in Sahel goats is largely responsible for these low values (0.07). It would be interesting to carry out further studies including other populations of Sahel goat in Mali to confirm the absence of E allele. Allele B was predominant in both breeds. The methodology used in the present study cannot identify the difference between alleles B and B of CSN1S1 like by Caroli et al. (2007). More investigations using other methodology (PCR SSCP, PCR RFLP and DNA sequencing) are necessary to better genetic characterization of Malian goat breeds. Conclusions Genetic diversity is very important for animal breeding. Our study revealed polymorphism in α S1 casein gene in Guéra goat. Allele B and the intermediate allele E were found in this breed. For the Sahel goat, no polymorphism was identified, B was the unique allele. The heterozygosity rates and the genetic diversity index (F ST ) in the two breeds were low. The presence of allele E in Guéra goats could explain the relatively high milk production of this breed and its origin in the Canary Islands. The information obtained for Guéra goat regarding the α S1 casein gene might be used in genetic improvement schemes of local goats in Mali. Acknowledgements The authors express their gratitude to Institut d Economie Rurale du Mali for the help and support of this research. Table 1b. Allele frequencies at the locus α S1 casein (exon 19) in Guéra and Sahel goat Allele Breed B E Guéra goat Sahel goat both References Angiolillo, A., Yahyaoui, M.H., Sanchez, A., Pilla, F. & Folch, J.M Characterization of a new genetic variant in the caprine κ-casein gene. Journal of Dairy Science 85: Barbieri, M.E., Manfredi, E., Elsen, J.M., Ricordeau, G., Bouillon, J., Grosclaude, F., Mahe, M.F. & Bibe, B Effet du locus de la caséine alpha S1 sur les performances laitières et les paramètres génétiques des chèvres Alpines. Genetic Selection Evolution 27:

81 Genetic polymorphism of α S1 casein in Guéra and Sahel goat 83 Boulanger, A., Grosclaude, F. & Mahe, M.F Polymorphisme des caséines α s1 et α s2 de la chèvre (Capra hircus). Genetic Selection Evolution 16: Caroli, A., Chiatti, F., Chessa, S., Rignanese, D., Ibeagha-awemu, E.M. & Erhardt, G Characterization of the casein gene complex in West African goats and description of a new αs1-casein polymorphism. Journal of Dairy Science 90: Cisse, M., Fall, Y. & ly, I Performances laitières et état nutritionnel des chèvres du Sahel conduites sur parcours naturels: relations avec la croissance des chevreaux. In Proceedings of the Third Biennial Conference of the African Small Ruminant Research Network UICC, Kampala, Uganda, 5 9 December y1994. Feligini, M., Frati, M., Cubrik, V.C., Brambilla, A., Parma, P., Curik, I., Greppi, G.F. & Enne, G Caprine alpha S1 casein polymorphism, characterization of A, B, E and F variants by means of various biochemical and molecular techniques. Food Technology Biotechnology 43(2): Grosclaude, F., Mahe, M.F., Brignon, G., Di stasio, I. & Jeunet, R A Mendelian polymorphism underlying quantitative variations of goat αs1 casein. Genetic Selection Evolution 19(4): Jordana, J., Amills, M., Diaz, E., Angulo, C., Serradilla, J.M. & Sanchez, A Gene frequencies of caprine α S 1-casein polymorphism in Spanish goat breeds. Small Ruminant Research 20: Martin, P Polymorphisme génétique des lactoprotéines caprines. Lait 73: Missohou, A., Talaki, E. & Maman Laminou, I Diversity and genetic relationships among seven West African goat breeds. Asian-Australasian Journal of Animal Sciences 19(9): Nantoumé, H., Traoré, D. & Diarra, C.H.T Mise au point de techniques d amélioration des productions de lait, de viande et de laine des petits ruminants. Rapport final de Recherche, 11 e Commission Scientifique du Comité National de Recherche Agricole, Novembre 2005, Bamako. Ramunno, L., Longobardi, E., Pappalardo, M., Rando, A., Di Gregorio, P., Cosenza, G., Mariani, P., Pastore, N. & Masina, P An allele associated with a non-detectable amount of alpha S2 casein in goat milk. Animal Genetics 32: Soares, M.A.M., Rodrigues, M.T., Mognol, G.P., Ribeiro, L.F.C., Silva, J.L.C. & Brancalhão, R.M.C Polymorphism of alphas1-casein gene in a dairy goat herd in the southeastern region of Brazil. Revista Brasileira de Zootecnia 38(6). Traoré, D., Nantoumé, H. & Diarra, C.H.T Amélioration de la production laitière par l introduction de la chèvre Guéra. Rapport final de Recherche, 11 e Commission Scientifique du Comité National de Recherche Agricole, Novembre 2005, Bamako. Traoré, D., Nantoumé, H. & Diarra, C.H.T La chèvre Guéra. Institut du Sahel; les Monographies Sahéliennes ISBN: Truett, G.E., Heeger, P., Mynett, R.L., Truett, A.A., Walken, T.A. & Warmen, M.L Preparation of PCR quality mouse genomic DNA with hot sodium and tris (Hotshot). Biotechnics 29 (1): Vacca, G.M., Ouled Ahmed Ben Ali, H., Carcangiu, V., Pazzola, M. & Dettori, M.L Genetic structure of the casein gene cluster in tne Tunisian native goat breed. Small Ruminant Research 87(1):

82 Animal Genetic Resources, 2014, 54, Food and Agriculture Organization of the United Nations, 2014 doi: /s Estimación de heredabilidad de la curva de crecimiento en el borrego de raza Chiapas en México A.C. Méndez-Gómez 1, R. López-Ordaz 2,4, M. Peralta-Lailson 1, R. Ulloa-Arvizu 2, P. Pedraza-Villagómez 1, F.J. Ruiz-López 3, J.M. Berruecos-Villalobos 2, C.G. Vásquez-Peláez 2 1 Centro Universitario de Investigación y de Transferencia de Tecnología, Universidad Autónoma de Chiapas, Campus III, San Cristóbal de las Casas, México; 2 Facultad de Medicina Veterinaria y Zootecnia, Universidad Nacional Autónoma de México, México D.F.; 3 Centro Nacional de Fisiología y Mejoramiento Animal, Instituto Nacional de Investigaciones Forestales y Agropecuarias, SAGARPA, Ajuchitlán Qro, México; 4 Dirección actual: Universidad Autónoma Metropolitana, Unidad Xochimilco, México D.F. Resumen Se utilizaron los registros genealógicos y de pesos mensuales de 790 ovinos de la raza Chiapas nacidos de 1991 a 2004 con objeto de caracterizar la curva de crecimiento y estimar componentes de (co)varianza de los estimadores (A, k y b) de la función de Gompertz. La estructura del pedigrí consistió por 790 ovinos, incluyendo 45 sementales y 379 madres. La estimación de los componentes de (co) varianza de cada uno de los estimadores de crecimiento se realizó utilizando un modelo animal univariado con la metodología de máxima verosimilitud restringida (REML). Los valores de heredabilidad estimados fueron: A (peso asintótico) 0.21 ± 0.06; b (constante de integración) 0.25 ± 0.07; k (tasa de madurez) 0.16 ± 0.06; edad a la inflexión se estimó como (ln(b)/k) 0.24 ± 0.07 y velocidad de crecimiento absoluto estimada como [k*(0.368*a)* ln(a/0.368*a)] 0.22 ± Los machos muestran una velocidad de crecimiento 24% más rápida que las hembras (P < 0.05), siendo el crecimiento absoluto de 59 ± 2 g y 44 ± 2 g por día para machos y hembras respectivamente. La raza Chiapas es un borrego de tamaño pequeño con lenta velocidad de crecimiento debido probablemente a su adaptación a las condiciones ambientales donde es utilizado. La curva de crecimiento puede ser modificada a través de selección. Palabras clave: heredabilidad, función de Gompertz, curvas de crecimiento, crecimiento absoluto, borrego Chiapas Summary Genealogical and live body weight monthly records from 790 sheep of the Chiapas breed born between 1991 and 2004 were used to characterize the growth curve and estimate (co)variance components for the parameters (A, k and b) of the Gompertz function. The pedigree structure consisted of 790 sheep, including 45 rams and 379 ewes. Estimation of (co)variance components for each growth parameter was achieved using a univariate animal model with the restricted maximum likelihood (REML) method. Estimated values of heritability were: A (adult weight) 0.21 ± 0.06; b (integration constant) 0.25 ± 0.07; k (maturity rate) 0.16 ± 0.06; age at the inflexion estimated as (ln(b)/k) 0.24 ± 0.07; and absolute growth rate as [k*(0.368*a)* ln(a/0.368*a)] 0.22 ± The growth of males was 24% (P < 0.05) faster than that of females; adult weight was 25 ± 0.7 kg for males and 23 ± 0.7 kg for females; absolute growth was 59 ± 2 g and 44 ± 2 g per day for males and females respectively. The Chiapas breed of sheep is a small animal with a slow growth curve, probably due to its adaptation to the environmental conditions where it lives. Growth curve can be modified through selection. Keywords: heritability, Gompertz function, growth curves, absolute growth, Chiapas sheep Résumé Archives généalogiques ont été utilisés et 790 pesos mensuels course de moutons Chiapas nés de 1991 à 2004 afin de caractériser la courbe de croissance et composants estimation de (co)estimateurs de la variance (A, K et b) de la fonction Gompertz. La structure pedigree composé de 790 moutons, dont 45 étalons et 379 mères. Estimation des composantes de (co)variance de chacun des estimateurs de croissance a été réalisée en utilisant un modèle animal univariée méthodologie REML (MVR). Les valeurs d héritabilité estimées sont les suivantes: A (poids asymptotique) 0,21 ± 0,06, b (constante d intégration) 0,25 ± 0,07, k (taux de maturité) 0,16 ± 0,06; âge d inflexion a été estimée à [(ln (b) / k] 0,24 ± 0,07 et le taux de croissance absolue estimé que [k*(0,368*a)* ln(a/ 0,368*A)] 0,22 ± 0,07. mâles montrent un taux de croissance de 24% plus vite que les femelles (P < 0,05), avec la croissance absolue de 59 ± 2 g à 44 ± 2 g par jour pour les mâles et les femelles. Chiapas est une race de moutons de faible taux de croissance faible sans doute en raison de leur adaptation aux conditions environnementales où est utilisé. courbe de croissance peut être modifié par sélection. Mots-clés: héritabilité, la fonction de Gompertz, courbes de croissance, croissance absolue, moutons Chiapas Correspondencia: Dr Carlos Vásquez, Facultad de Medicina Veterinaria y Zootecnia, Universidad Nacional Autónoma de México, México D.F. carlosgv@unam.mx 85

83 86 A.C. Méndez-Gómez et al. Introducción En la región de los Altos de Chiapas, México, se estableció desde el siglo XVI en la población indígena un sistema de producción ovina que prevalece casi sin cambios hasta nuestros días. Esta población de borregos tiene su origen de las razas Churra, Manchega y Lacha y es conocida como raza Chiapas, con tres biotipos: Blanca, Negra y Café (Figura 1) (Pedraza et al., 1992; FAO, 2009). Estos ovinos tienen importancia cultural, social y económica por la producción de lana en la población Tzotzil de origen maya (Perezgrovas y Castro, 2000). Son animales rústicos con partos anuales generalmente sencillos; el empadre se realiza con monta natural teniendo presente al macho durante todo el año, a pesar de existir una marcada estacionalidad con partos al término de la época de lluvias (octubre noviembre). La trasquila se realiza cada 6 meses (abril octubre). Los rebaños son pequeños de entre 8 a 10 animales por comunero y son alojados en corrales de madera pequeños fijos o móviles. Los ovinos son llevados a pastorear en zonas comunales y a pie de carretera donde consumen pastos (Sporobolus sp., Stipa ichu, Avena fatua, Cynodon dactylon, Pennisetum clandestinum, Tripogandra spp., Trifolium amabile, Melilotus alba) y plantas nativas (Eupatorium mairetianum, Eupatorium ligustrinum, Salvia spp., Phytolaccaceae spp., Rulus trilobus, Sambucus mexicana). El agua de bebida es proporcionada de forma individual a cada animal con una cubeta procurando evitar que los animales beban de los aguajes o cauces de ríos a fin de evitar parasitosis. Cuando los dueños tienen oportunidad suplementan con rastrojo de maíz. Los animales enfermos son tratados con herbolaria tradicional (Pedraza et al., 1992; Perezgrovas y Castro, 2000). Por su respuesta reproductiva a estas condiciones ambientales se le considera a la raza de borregos Chiapas como un recurso zoogenético de la región. Actualmente, la Universidad Autónoma de Chiapas conserva un núcleo de esta raza con las tres variedades, con apareamientos controlados exclusivamente dentro de estos biotipos, con el objetivo de caracterizar, evaluar y apoyar la producción ovina de la región. El borrego Chiapas es una raza pequeña con pesos al nacimiento de 2.5 kg y de 26 a 30 kg en la edad adulta en hembras y machos respectivamente (Perezgrovas y Pedraza, 1984; Pedraza et al.,1992; Perezgrovas y Castro, 2000; López et al., 2012); sin embargo, se desconoce la curva y velocidad de crecimiento. López et al. (2012) estimaron heredabilidades de 0.09 a 0.34 para peso al nacimiento y de 0.09 a 0.16 para peso ajustado al destete (90 días), con una correlación genética entre estos pesos de 0.81 ± 0.18, sin que exista información en edades y pesos posteriores. La estimación de la curva de crecimiento del nacimiento hasta la edad adulta permite identificar etapas en el desarrollo de esta raza, pudiendo establecer programas de manejo y alimentación así como estimar la tasa de crecimiento absoluto, la tasa relativa de maduración y peso adulto (Fitzhugh, 1976), e identificar animales genéticamente sobresalientes a las condiciones ambientales. Diferentes funciones no lineales han sido utilizadas para describir y ajustar la curva de crecimiento en diferentes especies, entre las cuales se encuentra la función de Gompertz que tiende a un mejor ajuste con respecto a otros modelos (Özdemir y Dellal, 2009; Forni et al., 2009; Renne et al., 2003; Delgadillo et al., 2009, Martínez et al., 2010). Esta función presenta dos parámetros con interpretación biológica: A supone que el peso tiende a un valor final o asintótico con el tiempo (peso adulto) y k es la tasa de crecimiento la cual disminuye monotónicamente a medida que aumenta el peso hacia la madurez, mientras que el parámetro b está relacionado con las condiciones iniciales (Blasco, 1999). El objetivo de este estudio fue caracterizar la curva de crecimiento utilizando la función de Gompertz desde el nacimiento hasta la edad madura y la estimación de las (co) varianzas de los componentes de la función en ovinos de la raza Chiapas que permitan ayudar a establecer sistemas de manejo, alimentación y genéticos en esta raza. Materiales y métodos Se utilizó la información de los corderos nacidos entre 1991 y 2004, del rebaño de borrego Chiapas localizado en el Centro Universitario de Investigación y Transferencia de Tecnología de la Universidad Autónoma de Chiapas (CUITT-UNACH) ubicado en el municipio de Teopisca, Figura 1. Ovino de la raza Chiapas en sus variedades Blanca, Negra y Café. (a) Variedad blanca Borrego Chiapas (Icsat en lengua Tzotzil) Blanco con manchas negras alrededor de los ojos, el hocico y las orejas (b) Variedad negra o Chamula (Saciol en lengua Tzotzil) Negro de piel y lana, con manchas blancas en la región frontal de la cabeza y la punta de la cola. (c) Variedad Cafó (Mesha en lengua Tzotzil) El color de la lana es blanco cremoso con tres variedades marrón, negro y dorado. Fuente: Adaptado de Perezgrovas y Castro, 2000.

84 Estimación de heredabilidad de la curva de crecimiento en el borrego Chiapas de latitud norte y de longitud oeste y a una altitud de 1780 msnm. El clima es templado subhúmedo C(w 2 ) (w) (clasificación de Köeppen), la temperatura media anual oscila entre los 14 y 18 C con heladas de noviembre a febrero y una precipitación pluvial de 1200 mm anual de abril a octubre (Nahed, 1999). El CUITT-UNACH, tiene una extensión de 50 ha de praderas; 10 ha son dedicadas para la siembra de maíz y casi una ha para instalaciones donde se localizan los corrales rústicos construidos con madera, y techo de lámina de zinc, sala de ordeño, bodega y oficina. Los borregos se mantienen en un sistema mixto con pastoreo diurno de 6 a 8 horas en pradera compuesta principalmente de zacate kikuyu (Pennisetum clandestinum) con encierro nocturno complementada con mazorca de maíz, harina de soya y sales minerales. El manejo sanitario del rebaño consistió en una rotación de principios activos para el control de enfermedades gastroentéricas y respiratorias. En la época de empadre se formaron lotes de 10 hembras por cada semental, con una duración de 60 días de inicios de noviembre a finales de diciembre de 1991 a 1997; a partir de 1998 hasta 2004, las hembras fueron inducidas al estro con el empadre a principios de marzo; los nacimientos fueron desde agosto a finales de septiembre. El destete se realizó cuando los corderos tenían 3 meses de edad aproximadamente. Para cada uno de los corderos se registró la identificación del padre, de la madre, el número de parto del cual provenían, tamaño de la camada, año de parto, sexo, fecha de nacimiento, edad y peso semanal del nacimiento al destete, y mensual del destete hasta los 24 meses de edad. Eliminando del análisis aquellos registros que contaron con menos de 5 pesajes posdestete quedando un total de 790 animales y 15,067 mediciones. lineal (GLM, SAS, 1993) incluyendo los efectos fijos y las interacciones dobles, quedando en el modelo final los efectos de año de nacimiento, tamaño de la camada, edad de la madre y sexo (P < 0.05); mientras que las interacciones dobles se eliminaron al no presentar significancia estadística (P > 0.15). La estimación de los componentes de (co)varianza de los parámetros de crecimiento predichos a partir de la función de Gompertz se realizó utilizando un modelo animal univariado bajo la metodología de máxima verosimilitud restringida (REML) con el programa ASReml versión 2.0 (Gilmour et al., 2006) a partir de un modelo mixto representado como Y = Xb + Za + e donde Y es el vector N 1 observaciones de los estimadores de los parámetros A, b y k del modelo de Gompertz, peso y edad a la inflexión, o velocidad de crecimiento absoluto; β es el vector de efectos fijos ambientales de número de parto, tamaño de la camada, sexo; X es la matriz de incidencia que relaciona los efectos ambientales con las características de crecimiento; a es el vector del valor genético aleatorio de los animales; Z es la matriz de incidencia que relaciona los efectos aleatorios de los animales con las observaciones; e es el vector de efectos aleatorios residuales. Los supuestos en el modelos fueron E[y]=Xβ ye[e]=0y [ Var a ] [ ] = As2 a 0 e 0 I N s 2 e donde N es el número de registros de las características estudiadas; A es la matriz de relaciones genéticas aditivas entre los animales; I N es la matriz de identidad de orden N. Análisis estadístico La caracterización de la curva de crecimiento se realizó utilizando la función de Gompertz (Proc NLIN, SAS, 1993), calculando para cada animal los estimadores de los parámetros de la función y t = Ae be kt donde y t representa el peso corporal a la edad t (en días); A es el valor estimado del peso asintótico cuando la edad tiende a infinito, interpretado como peso adulto o maduro; b es la constante de integración; k es la relación y/a que representa la tasa de crecimiento (tasa de madurez); y e es el número de Euler La relación (ln(b)/k) estima la edad a la inflexión, y ky ln (A/y) la velocidad de crecimiento absoluto al punto de inflexión, donde el valor de y se tomó como el peso a la inflexión que es Para el cálculo de los efectos ambientales sobre los estimadores de la curva de crecimiento se utilizó un modelo Resultados En el Cuadro 1 se presentan las medias de cuadrados mínimos de los efectos ambientales sobre los estimadores de la curva de crecimiento. El tamaño de la camada y número de parto del que provenían los animales, no influyeron en A (P > 0.05), mientras que el valor estimado en los machos fue de 25.4 kg (7.9%) mayor (P < 0.05) que en las hembras (24.8 kg). El parámetro b fue afectado por todos los efectos fijos incluidos en el modelo (P < 0.05). El parámetro k no se vio influenciado por tamaño de la camada (P > 0.05); los machos tuvieron una mayor tasa de madurez (24%) que las hembras (P < 0.05), mientras que los corderos provenientes de madres primíparas tuvieron el valor más bajo (0.0056), lo que representa del 80 a 87% de lo alcanzado por los corderos provenientes de hembras multíparas (P < 0.05). El peso a la inflexión fue alcanzado 30 días antes en los corderos de parto sencillo comparados con los de parto múltiple (P < 0.05); los machos alcanzaron el peso a la inflexión 8% más rápido que las hembras (P < 0.05); los corderos de

85 88 A.C. Méndez-Gómez et al. Cuadro 1. Medias de cuadrados mínimos y errores estándar para las características de crecimiento de acuerdo al tamaño de la camada, sexo y número de parto en borregos de la raza Chiapas Efectos N Peso adulto (kg) (A) Constante de integración (b) Tasa de madurez (k)* Peso de inflexión (kg) Edad de inflexión (días) Velocidad de crecimiento absoluto (g/día) Tamaño de la camada Sencillo ± 0.5a 1.89 ± 0.02a 65 ± 2a 9.2 ± 0.1a ± 2a 55 ± 4a Múltiple ± 1.2a 2.10 ± 0.06b 56 ± 6a 8.7 ± 0.4a ± 7b 48 ± 4a Sexo Macho ± 0.7a 2.09 ± 0.04a 67 ± 4a 9.3 ± 0.2a ± 2a 59 ± 2a Hembra ± 0.7b 1.90 ± 0.04b 54 ± 4b 8.6 ± 0.2b ± 3.b 44 ± 2b Número de Parto Primíparas ± 0.7a 2.03 ± 0.04a 56 ± 4a 9.1 ± 0.2a ± 3a 47 ± 2a Segundo ± 0.8a 1.91 ± 0.04b 64 ± 4b 8.7 ± 0.3a ± 3b 53 ± 2b > ± 0.8a 2.08 ± 0.04a 62 ± 4b 9.1 ± 0.2a ± 3b 55 ± 2b * ln(b)/k ky ln (A/y), tomando y = 0.368*A a, b, c Valores con diferente literal dentro del efecto son medias diferentes (P < 0.05) segundo parto tuvieron la menor edad a la inflexión (107 días), que representa un 13.7% menos que los corderos de primíparas y 10 % menos de las de tres o más partos (P < 0.05). La Figura 2 muestra la curva de crecimiento y la tasa de crecimiento absoluto por sexo; la mayor tasa de crecimiento de los machos fue de 59 g/día (punto de inflexión), mientras que en las hembras la ganancia fue de 44 g/día (P < 0.05). Por otro lado, las crías de hembras de primer parto tuvieron una tasa de crecimiento absoluto menor que los corderos de dos o más partos (P < 0.05). En general, los factores tamaño de la camada, número de parto de la hembra y sexo del cordero, fueron significativos en la variación de la curva de crecimiento del ovino Chiapas. En el Cuadro 2 se presentan los componentes de varianzas aditiva y fenotípica así como los valores de heredabilidad estimados de los parámetros de la función de Gompertz siendo para A (peso adulto) 0.21 ± 0.066, b (Constante de integración) 0.25 ± y k (tasa de madurez) 0.16 ± Figura 2. Pesos estimados con la funcion de Gompertz y velocidad de crecimiento absoluto (V.A.) para hembras y machos en el borrego de la raza Chiapas. La estimación de heredabilidad para peso a la inflexión fue 0.21 ± 0.066; para edad a la inflexión, 0.24 ± y para velocidad de crecimiento absoluto, 0.22 ± Discusión Con respecto al peso adulto, los valores encontrados en este estudio son inferiores a lo observado en razas españolas Churra, Manchega y Lacha (Pedraza et al., 1992), así como con otras razas locales en otras latitudes (Lewis et al., 2002; Topal et al., 2004; Lambe et al., 2006; Malhado et al., 2009; Braga et al., 2006; Ozder et al., 2009; Kubuc y Eyduran, 2009; Darkiran et al., 2010; Solomon et al., 2010), pero semejantes a lo informado por Perezgrovas y Pedraza (1984): 25 y 28 kg en hembras y machos respectivamente en esta misma raza de ovino Chiapas, siendo esto indicativo de que la población de ovino de raza Chiapas no ha sido seleccionada para características de crecimiento, ya que es una raza que es destinada para la producción de lana (Perezgrovas y Castro, 2000) y que el sistema de producción permanece prácticamente sin cambios (López et al., 2012). El modelo de Gompertz para peso maduro (A), mostró una subestimación de 1.5 kg aproximadamente, la cual ha sido también observada por otros autores (Lambe et al., 2006; Malhado et al., 2009), posiblemente debido a la cantidad de tejido graso, que es muy variable a lo largo del estado adulto del animal (Blasco, 1999), y a la disponibilidad de forraje a través del tiempo. También hay referencias que mencionan que la subestimación del peso maduro se debe únicamente al uso de datos realizados en la etapa temprana (Lambe et al., 2006). El peso adulto en los machos fue 7.9% superior al de las hembras, siendo un porcentaje menor a lo observado en

86 Estimación de heredabilidad de la curva de crecimiento en el borrego Chiapas 89 Cuadro 2. Componentes de varianzas y heredabilidades estimadas (h 2 ± e.e) para las características de crecimiento en borregos Chiapas Característica Varianzas h 2 ± e.e Aditiva Residual Fenotípica Constante de integración (b) ± Peso adulto (A) ± Tasa de madurez (k) ± Peso inflexión ± Edad de inflexión ± Velocidad de crecimiento absoluto ± razas de carne como la Suffolk (Lewis et al., 2002). El hecho de que no existió efecto del tamaño de la camada y número de parto (P > 0.05) es explicado debido a que el peso adulto del animal está muy alejada de los efectos maternos que existen en el ambiente uterino y en la lactancia (López et al., 2012), los resultados de este trabajo coincide con algunos estudios (Malhado et al., 2009), aunque otros autores han registrado un efecto del número de parto sobre el peso maduro de sus crías (Bathaei y Leroy, 1996; Lewis et al., 2002; McManus et al., 2003). El parámetro b se considera como una constante de integración sin significado biológico particular y con alta correlación con el parámetro A (Blasco, 1999); de tal modo que la expresión y 0 = A exp (-b), puede ser interpretada como un estimador del peso al nacimiento; esta función estimó el peso al nacer de 3.14 kg para machos y de 3.51 kg para hembras, mostrando una sobrestimación en corderos de la raza Chiapas; Perezgrovas y Castro, (2000) y López et al. (2012) observaron pesos al nacer de 2.15 y 2.27 kg para hembras y machos de esta misma raza. Este sesgo se puede deber a la inconsistencia de la función. En la mayoría de los estudios no se presenta el valor encontrado de b, sin embargo, hay informes en razas del medio oriente (Topal et al., 2004) y brasileñas (Malhado et al., 2009) que presentan valores alrededor de 2, semejante a la encontrado en el borrego Chiapas por López et al. (2012) y en este estudio. En este estudio, los valores de la tasa de madurez son menores hasta en un 50% de los registrados en otras razas ovinas locales (Topal et al., 2004; Kubuc y Eyduran, 2009; Malhado et al., 2009) y solo ligeramente inferiores a los encontrados en la West African Dwarf (Gbangboche et al., 2008), indicando que la raza de ovinos en la raza Chiapas es de lento crecimiento, lo que puede estar asociado a su adaptabilidad al medio ambiente de la zona. Como se esperaba, los machos presentaron una tasa de madurez 1.24 veces mayor que la de las hembras (P < 0.05); resultados similares se han observado en razas locales brasileñas (Malhado et al., 2009). El peso y la edad a la inflexión son los puntos donde la ganancia de peso llega al máximo y posteriormente tiende asintóticamente a cero (Figura 2). En la función de Gompertz este punto es fijo; el peso a la inflexión representó el 36.78% del peso a la madurez (A). Como en borregos Chiapas tienen valores de A menores que los encontrados en otras razas, consecuentemente el peso alainflexión también resultó menor. Sin embargo, como el valor de k también fue menor, dando resultado que la edad a la inflexión fuera superior a lo que presentan las otras razas (Lewis et al., 2002; Topal et al., 2004; Lambe et al., 2006; Malhado et al., 2009; Braga et al., 2006; Ozder et al., 2009; Kubuc y Eyduran, 2009; Darkiran et al., 2010; Solomon et al., 2010). La velocidad absoluta de crecimiento en el borrego Chiapas se presenta por arriba de los 50 g/día. Los machos tuvieron una velocidad absoluta de crecimiento 1.34 veces mayor que las hembras (P < 0.05) y 8 días antes (P < 0.05); los corderos provenientes de parto simple mostraron una velocidad de crecimiento similar que aquellos proveniente de parto múltiple (P > 0.05). La mayor velocidad de crecimiento se da en la fase predestete, donde el cordero es dependiente del ambiente materno dado básicamente por la producción de leche. Las bajas tasas de crecimiento observadas después del destete indican que el manejo nutricional de estos animales no es el óptimo. La estimación de componentes de varianza para los parámetros de la función de Gompertz en este estudio para la raza Chiapas mostraron un rango entre 0.16 y Con respecto al peso adulto o maduro (A), Stobart et al. (1986) estimaron valores de 0.57 con una edad a la madurez de cinco años en razas Rambouillet, Targhee y Columbia, superiores a lo observado en este estudio (0.21) a los dos años de edad. Lewis et al. (2002), utilizando animales de la raza Suffolk, estimaron una heredabilidad de 0.37 en el mismo parámetro, y Lambe et al.(2006) estimaron heredabilidades de con la función de Richards y con la función de Gompertz para la raza Texel. En estudios con otras razas locales, como la Mehraban (Bathaei y Leroy, 1996), Scottish Blackface (Lambe et al., 2006) y Horro (Solomon et al., 2010), los valores de heredabilidad estimados variaron entre 0.29 a En cuanto a la tasa de madurez (k), el valor de heredabilidad estimado en este estudio fue similar a lo reportado en razas Santa Inés y Horro (Braga et al., 2006; Solomon et al., 2010), mientras que Lambe et al. (2006) estimaron una heredabilidad de 0.06 en la raza Scottish Blackface; en contraste, Bathaei and Leroy (1996) encontraron un valor de 0.45 en la raza Mehraban Irani y Stobart et al de 0.32 en razas Rambouillet, Targhee y Columbia.

87 90 A.C. Méndez-Gómez et al. En cuanto al parámetro b, que es considerada como una constante de integración que está ligado a las condiciones de inicio de la curva (Blasco, 1999), en la raza Suffolk, Lewis et al. (2002) estimaron un valor de 0.38, superior a lo encontrado en este estudio. La forma en el crecimiento en ovejas se modifica con la edad y su forma depende de efectos de raza y condiciones ambientales como es la disponibilidad de alimento y manejo. La función de Gompertz se ha sugerido como adecuada, cuando no existen limitantes en las condiciones ambientales (Lewis et al., 2002). Los resultados indican que es posible modificar a través de selección la forma de la curva Selección directa en los parámetros A o b, tendrían como resultado un incremento en el peso adulto así como en los pesos corporales al nacer e indirectamente en la producción de lana y la resistencia a parásitos (Safari and Fogarty, 2003). Sin embargo, tiene el inconveniente de incrementar el intervalo generacional. La selección sobre la tasa de madurez (k) o la velocidad de crecimiento, podría mejorar peso adulto, aunque habría que evaluar su impacto en la eficiencia dentro de este sistema de producción de los Altos de Chiapas. Conclusiones El borrego de la raza Chiapas muestra una curva de crecimiento lento, el cual puede atribuirse a la adaptación al medio ambiente. La variabilidad genética aditiva encontrada a partir de la función de Gompertz en el crecimiento de esta raza de ovinos Chiapas, indica que es posible modificar la curva de crecimiento del borrego Chiapas pudiendo ser utilizada la estimación de velocidad o crecimiento máximo absoluto como criterio de selección. Agradecimientos Se agradece al Consejo Nacional de Ciencia y Tecnología (CONACYT-México) por la beca otorgada a la primera autora para realizar estudios de Maestría en Ciencias (FMVZ-UNAM); Proyectos UNAM-PAPIIT IN , UNAM-PAPIIT IN y SAGARPA-CONACYT 2004-CO1-111/A1. Esta investigación fue posible gracias al convenio de colaboración entre la Facultad de Medicina Veterinaria y Zootecnia de la Universidad Nacional Autónoma de México y el Centro Universitario de Investigación y Transferencia de Tecnología de la Universidad Autónoma de Chiapas. Declaración de interés Los autores declaran que no existen conflictos de interés en la presentación de este artículo. Referencias Abegaz, S., Van Wyk, J.B. & Oliver, J.J Estimation of genetic and phenotypic parameters of growth and productivity of Horro sheep. Archiv für Tierzucht (1): Bathaei, S.S. & Leroy, P.L Growth and mature weight of Mehraban Iranian fat-tailed sheep. Small Ruminant Research22: Blasco, A La descripción del crecimiento. Informe Técnico Ocasional No. 6. Universidad Politécnica de Valencia. Departamento de Ciencia Animal. Braga, L.R.N., Vasques, V.L.C., Oliveira, L.A.M., Sousa, P.J.R. & Oliveira, A.A Parâmetros genéticos de características estimadas da curva de crescimento de ovinos da raça Santa Inês. Revista Brasileira de Zootecnia 3: Darkiran, I., Koncagul, S. & Bingol, M Growth characteristics of indigenous norduz female and male lambs. Journal of Agricultural Science 16: Delgadillo, C.A.C., López, O.R., Montaldo, V.H.H., Berruecos, V.J. M., Luna, A.E., Shimada, M.A. & Vásquez, P.C.G Caracterización de la curva de crecimiento del ciervo rojo (Cervus elaphus scoticus) en el centro de México. Técnica Pecuaria en México 47: FAO Domestic Animal Diversity Information System, fao.org/dad-is/. Food and Agriculture Organization of the United Nations, Rome. Fitzhugh Jr., H.A Analysis of growth curves and strategies for altering their shape. Journal of Animal Science 42: Forni, S., Piles, M., Blasco, A., Varona, L., Oliveira, H.N., Lôbo, R.B. & Albuquerque, L.G Comparison of different nonlinear functions to describe Nellore cattle growth. Journal of Animal Science 87 (2): Gbangboche, A.B., Glele-Kakai, R., Salifou, S., Albuquerque, L.G. & Leroy, P.L Comparison of non-linear growth models to describe the growth curve in West African Dwarf sheep. Animal (7): Gilmour, A.R., Gogel, B.J., Cullis, B.R. & Thompson, R ASReml User Guode Release 2.0 VSN International Ltd, Hemel Hempstead, HP1 1ES, UK. Kubuc, M. & Eyduran, E The determination of the best growth model for Akkaraman and German Blackheaded Mutton x Akkaraman B1 crossbreed lambs. Bulgarian Journal of Agricultural Science 15: Lambe, N.R., Navajas, E.A., Simm, G. & Bünger, L A genetic investigation of various growth models to describe growth of lambs of two contrasting breeds. Journal of Animal Science 84: Lewis, R.M., Emmans, G.C. & Dingwall, W.S A description of the growth of sheep and its genetic analysis. Animal Science 74: López, O.R., Olivera, I.V., Berruecos, J.M., Peralta, L.M., Ulloa-Arvizu, R. & Vásquez, C.G Genetic parameters for birth and weaning weights in the local Chiapas sheep breed from Mexico, Revista Mexicana de Ciencias Pecuarias 3(1): Martínez, C.A., Rodríguez, A.P., Jiménez, A. & Manrique, C Descripción matemática de la función de Gompertz aplicada al crecimiento de animales. Revista de la Facultad de Medicina Veterinaria y de Zootecnia 57: Malhado, C.H.M., Carneiro, P.L.S., Affonso, P.R.A.M., Souza, A.A. O. & Sarmento, J.L.R Growth curves in Dorper sheep crossed with the local Brazilian breeds, Morada Nova, Rabo Largo and Santa Inés. Small Ruminant Research 84:

88 Estimación de heredabilidad de la curva de crecimiento en el borrego Chiapas 91 McManus, C., Evangelista, C. & Fernandes, L.A.C Curvas de crescimento de ovinos Bergamasia criados no Distrito Federal. Revista Brasileira de Zootecnia 32: Nahed, T.J Alternativas para el desarrollo de sistemas de producción ovina sostenibles en los altos de Chiapas. Tesis doctorado, Facultad de Medicina Veterinaria y Zootecnia, Universidad Nacional Autónoma de México, México D.F. Özdemir, H. & Dellal, G Determination of growth curves in young angora goats. Tarim Bilimleri Dergisi 15 (4): Ozder, M., Sezenler, T., Onal, A.R. & Ceyhan, A Genetic and non-genetic parameter estimates for growth traits in Turkish Merino lambs. Journal of Animal and Veterinary Advances (9): Pedraza, P., Peralta, M. & Pérezgrovas, R El borrego Chiapas: una raza local mexicana de origen español. Archivos de Zootecnia 41 (extra): Perezgrovas, R. & Castro, G.H El borrego Chiapas y el sistema tradicional de manejo de ovinos entre las pastoras tzotziles. Archivos de Zootecnia 49: Perezgrovas, R. & Pedraza, P Ovinocultura indígena I. Desarrollo corporal del borrego Chiapas. Cuadernos de Investigación 1 (Universidad Autónoma de Chiapas, Tuxtla Gutiérrez, México), pp Renne, U., Langhammer, M., Wytrwat, E., Dietl, G. & Bünger, L Genetic-statistical analysis of growth in selected and unselected mouse lines. Journal of Experimental Animal Science 42: Safari, A. & Fogarty, N.M Genetic parameters for sheep production traits: estimates from the literature. Technical Bulletin 49, NSW Agriculture, Orange, Australia. SAS Institute SAS/STAT User s Guide. In: 6.03 ed., Series SAS/ STAT User s Guide. SAS Institute, Cary, NC. Stobart, R.H., Bassett, J.W., Cartwright, T.C. & Blackwell, R.L An analysis of body weights and maturing patterns in western range ewes. Journal of Animal Science 63: Topal, M., Ozdemir, M., Aksakal, V., Yildiz, N. & Dogru, U Determination of the best nonlinear function in order to estimate growth in Morkaraman and Awassi lambs. Small Ruminant Research 55:

89 Animal Genetic Resources, 2014, 54, Food and Agriculture Organization of the United Nations, 2014 doi: /s Caracterización genética de la población bovina criolla de la Región Sur del Ecuador y su relación genética con otras razas bovinas L. Aguirre Riofrio 1, G. Apolo 2, L. Chalco 2 y A. Martínez 3 1 Universidad Nacional de Loja, Centro de Biotecnología Reproductiva Animal (CEBIREA), Ecuador y aluno Doutorado em Biociência Animal, Universidade de São Paulo, Brasil; 2 Tesistas, MVZ-UNL, 2012; 3 Facultad de Veterinaria, Universidad de Córdoba, España. Resumen Al no contar a la fecha con ningún estudio sobre la variabilidad genética de la población bovina criolla de la Región Sur del Ecuador (RSE), se ha procedido a su análisis y caracterización con fines de disponer de información técnica que permita planificar y desarrollar medidas tendientes a su conservación y utilización sustentable en este duro ecosistema en donde habita. Para ello se analizaron genéticamente 46 animales adultos, de una población de bovinos criollos de 114, agrupados por sus características fanerópticas en cuatro grupos, conocidos en el medio como: Negro Lojano (15 animales), Encerado (15), Colorado (9) y Pintado o Cajamarca (7). En el análisis de la variabilidad genética intrapoblacional, se obtuvo un Na de 228, un número promedio de alelos por locus de 8,14, y un PIC de 0,70 ± 0,10. Todos los 28 marcadores analizados son polimórficos y altamente informativos. Las He y Ho fueron de 0,74 ± 0,0178 y 0,676 ± 0,013, lo que indica la existencia de una alta diversidad molecular en la población estudiada. El HWE determina que el 32 por ciento de los alelos no se encuentran en equilibrio. El G st fue de 4,5 ± 2,8 por ciento, y el F st de 0,08 por ciento, lo que permite corroborar que no existe diferenciación genética entre los cuatro grupos de esta población. Las medias del F it yf is son de 8,89 y 8,92 por ciento, respectivamente. También se determinó el grado de distanciamiento genético entre esta población criolla con 18 razas taurinas y cebuinas. Los distanciamientos con todas las poblaciones superan el 0,10, las distancias menores están entre Ec:BC (0,1031), Ec:PAJ (0,1132), Ec:NAN (0,1302), Ec:VCA (0,1326) y Ec:FRI (0,1362). El mayor distanciamiento se observa entre Ec:GUZ (0,4725) y Ec: NEL (0,4624). Estos valores indican que los bovinos criollos de la RSE tienen un gran distanciamiento con las razas cebuinas y más bien sus troncos ancestrales están en las poblaciones ibéricas. Los resultados de variabilidad y distancias genéticas permiten disponer de herramientas para trazar estrategias que lleven a un manejo y conservación de la raza Criolla de la RSE y sugieren un manejo independiente de los cuatro grupos faneropticamente diferenciados a ser manejados en los distintos pisos altitudinales de la Región Andina del Ecuador. Palabras clave: Bovino criollo de Ecuador, variabilidad genética, grado de distanciamiento, microsatélites Summary In the absence to date of studies on genetic variability in the Creole cattle of the Southern Region of Ecuador, the analysis and characterization of this population was carried out in order to provide the technical information needed for the planning and development of measures aimed at the population s conservation and sustainable use in the harsh ecosystem these animals inhabit. For this purpose, 46 adult animals, out of a population of 114 Creole cattle, were genetically analysed; on the basis of phaneroptical traits, they were classified into four groups known locally as Negro Lojano (15 animals), Encerado (15), Colorado (9) and Pintado or Cajamarca (7). The analysis of the intra-population genetic variability brought out a Na of 228, an average number of alleles per locus of 8.14 and a PIC of 0.70 ± All 28 markers analyzed were polymorphic and highly informative. The He and Ho amounted to 0.74 ± and ± 0.013, which is an indication of a high molecular diversity in the population studied. The HWE determines that 32 per cent of the alleles are not in equilibrium. The G st amounted to 4.5 ± 2.8 per cent and the F st was 0.08 per cent, which corroborates that there is no genetic differentiation among the four groups of this population. The average values for F it and F is were 8.89 and 8.92 percent, respectively. The degrees of genetic distance between this Creole population and 18 cattle and zebu breeds were also determined. The distance exceeded 0.10 for all populations, with the lowest distances being those between Ec:BC (0.1031), Ec:PAJ (0.1132), Ec:NAN (0.1302), Ec:VCA (0.1326) and Ec:FRI (0.1362). The greatest distances were observed between Ec:GUZ (0.4725) and Ec:NEL (0.4624). These values show that the Creole cattle of the Southern Region of Ecuador are very distant from the zebu breeds, and therefore their ancestors would be more likely to have originated from Iberian populations. The results concerning variability and genetic distance will enable strategies to be devised for managing and conserving the Creole breed of the Southern Region of Ecuador. These results also suggest that the four groups with phaneroptical differences should be managed independently at the different altitudes of the Andean Region of Ecuador. Keywords: Creole cattle of Ecuador, genetic variability, degree of distance, microsatellites Correspondence to: L. Aguirre, Universidad Nacional de Loja, Centro de Biotecnología Reproductiva Animal (CEBIREA), Ecuador. leninaguirrer@yahoo.es 93

90 94 L. Aguirre et al. Résumé Compte tenu du manque d études sur la variabilité génétique des populations bovines créoles de la Région Sud de l Équateur, le but de ce travail a été d analyser et de caractériser ces bovins afin de disposer de l information technique nécessaire à la planification et développement de mesures pour leur conservation et utilisation durable dans le dur écosystème où ils sont élevés. Pour ce faire, 46 animaux adultes d une population de 114 bovins créoles ont été analysés génétiquement. Les bovins créoles sont classés, d après certaines caractéristiques des phanères, en quatre types qui reçoivent localement le nom de: Negro Lojano (15 animaux), Encerado (15), Colorado (9) et Pintado ou Cajamarca (7). Il résulte de l analyse de la variabilité génétique intra-populationnelle que le Na est de 228, le nombre moyen d allèles par locus de 8,14 et le PIC de 0,70 ± 0,10. Il a de même été observé que les 28 marqueurs analysés sont tous polymorphes et très informatifs. La He et la Ho ont été de 0,74 ± 0,0178 et de 0,676 ± 0,013, ce qui indique l existence d une grande diversité moléculaire dans la population étudiée. D après le HWE, le 32 pour cent des allèles ne se trouvent pas en équilibre. Le G st aétéde 4,5±2,8pourcentetleF st de 0,08 pour cent, ce qui permet de confirmer qu il n existe pas de différenciation génétique entre les quatre groupes de cette population. Le F it et le F is ont été en moyenne de 8,89 et de 8,92 pour cent, respectivement. Le degré de d éloignement génétique entre cette population créole et 18 races taurines et de zébu a aussi été déterminé. L éloignement dépasse le 0,10 pour toutes les populations, les plus faibles distances étant celles entre Ec:BC (0,1031), Ec:PAJ (0,1132), Ec:NAN (0,1302), Ec:VCA (0,1326) et Ec:FRI (0,1362). Le plus grand éloignement a été observé entre Ec:GUZ (0,4725) et Ec:NEL (0,4624). Ces valeurs indiquent que les bovins créoles de la Région Sud de l Équateur sont très éloignés des races de zébu, comme quoi leurs ancêtres seraient plutôt parmi les populations ibériques. Les résultats de variabilité et d éloignement génétique vont aider à élaborer des stratégies d élevage et de conservation de la race Créole de la Région Sud de l Équateur. Ces résultats suggèrent de même une conduite indépendante, sur les différents niveaux d altitude de la Région Andine de l Équateur, des quatre groupes à différences phanériennes. Mots-clés: Bovins créoles de l Équateur, variabilité génétique, degré d éloignement, microsatellites Presentado: 13 Marzo 2013; aceptado: 8 Julio 2013 Introducción Según Lucena (1987), fue Sebastián de Belalcázar, procedente de Panamá en 1534, quien arribó al Ecuador en la conquista del Reino de Quito, el que trajo los primeros bovinos Bos taurus, los cuales eran animales con una aptitud a más de la producción de carne y leche, de tracción de carretas, función tan importante para aquella época en las zonas alto andinas con un relieve tan irregular. Estos bovinos criollos han tenido un proceso de adaptación largo de cinco siglos a los diversos climas y ambientes, primero en un proceso de selección natural y luego con la intervención del hombre, en la búsqueda de animales funcionales para estos ecosistemas de montaña. Desde esa época el territorio ecuatoriano estuvo poblado exclusivamente por bovinos criollos (ibéricos naturalizados). Es a partir de los primeros años del 1900 (Pinzón, 1984), en que se inició la introducción de bovinos extranjeros para la producción de carne y de leche por la creencia de que estas razas son la mejor alternativa para los ganaderos, pues los resultados relativamente favorables obtenidos al inicio, fueron atribuidos a las aptitudes de las razas introducidas y no al vigor híbrido que produce el mestizaje. Con el transcurso del tiempo y los cruzamientos desordenados, se han producido procesos de absorción de los animales criollos, y como consecuencia el número de estas poblaciones ha disminuido drásticamente. Sin embargo, a pesar de ello aún se reporta la existencia de ganado bovino criollo en algunos lugares apartados del Ecuador, como es en la RSE limítrofe con el norte del Perú, donde aún se pueden encontrar poblaciones de bovinos criollos en núcleos pequeños y heterogéneos, los mismos que por sus particularidades fenotípicas y genotípicas han logrado sobrevivir y desarrollarse en condiciones complejas, adquiriendo características propias que les permitieron adaptarse y prosperar en este hábitat. Al respecto, De Alba (2009) y Aguirre et al. (2011) manifiestan que estos animales poseen características de adaptación a estos ecosistemas, como: tolerancia al calor y sequedad, con temperaturas sobre los 26 C; estatura pequeña y gran fortaleza, lo que les permite caminar grandes distancias por pedregales y terrenos irregulares en busca de alimento y abrevaderos; resistencia a los parásitos y excelente fertilidad e instinto materno. Los bovinos criollos cumplen un rol importante en la vida de las comunidades campesinas marginales de la RSE, pues constituyen una fuente de sustento y trabajo, su leche, su carne, la venta misma de animales y su utilización como medio de tracción animal benefician estas regiones geográficamente apartadas en donde la mecanización agrícola no es posible y la crianza de otros bovinos sería insostenible. La presente investigación se orientó a realizar un estudio genético con el uso de marcadores moleculares para determinar la variabilidad intrapoblacional y el grado de influencia de las razas ibéricas y cebuinas en esta población de bovinos y promover de esta forma estrategias para el manejo y conservación de este importantísimo material genético de esta parte de los Andes del Ecuador. Materiales y métodos El análisis genético se realizó en 46 bovinos adultos de una población de 114 animales criollos, agrupados por sus

91 Caracterización genética de la población bovina criolla de la Región Sur del Ecuador 95 Figura 1. Bovinos criollos Negro Lojano. características fanerópticas en cuatro grupos: 15 Negros Lojanos (Figura 1, véase las páginas 00 00), 15 Encerados (Figura 2), 9 Colorados (Figura 3) y 7 Pintados (Figura 4), provenientes de los sectores de Purunuma y Nambacola en el cantón Gonzanama, provincia de Loja, Ecuador (S 4 20 / S 4 10 ; W / W ; Figura 5). Los animales seleccionados fueron de ambos sexos y se evitó considerar para el análisis más de un animal proveniente de la misma familia. Se tomaron muestras de pelo de la región terminal de la cola, que fueron identificadas y guardadas herméticamente hasta su procesamiento en el laboratorio del Departamento de Genética Molecular Aplicada de la Universidad de Córdoba, España, en donde se realizó el análisis del ADN por medio de marcadores moleculares amplificados mediante la técnica de reacción en cadena de la polimerasa (PCR). Para la separación por tamaños de los fragmentos obtenidos mediante la PCR se sometieron estos a una electroforesis en gel de poliacrilamida en un secuenciador automático ABI 377XL; el análisis de los fragmentos y la tipificación alélica se realizó mediante los programas informáticos Genescan Analysis y Genotyper 2.5 respectivamente. El panel de 28 microsatélites empleados son: BM1314, BM8125, BM1818, CSSM66, ETH10, INRA32, MM12, TGLA122, BM2113, CRSM60, ETH185, HAUT27, HEL9, HEL13, ILSTS6, INRA23, INRA35, INRA37, INRA63, SPS115, TGLA227, BM1824, ETH225, ETH3, HAUT24, ILSTS011, TGLA53, TGLA126 (FAO, 2011). La caracterización genética se realizó de la siguiente manera: Para el cálculo del PIC de cada microsatélite se aplicó la fórmula de Botstein et al. (1980), empleándose para ello el complemento The Excel Microsatellite Toolkit (Park, 2001), y utilizando el programa MS Excel También se computarizaron los estadísticos F de Wright (1969) y Weir & Cockerham (1984), que corresponden: F=F IT, ɵ =F ST yf=f IS, realizando un análisis factorial de correspondencias mediante el programa Genetix v (Belkhir et al., 2003). Se calcularon las matrices de distancias genéticas Da (Nei & Chesser., 1983); para ello se emplearon los análisis genéticos de 789 individuos pertenecientes a 19 razas taurinas y cebuinas del Consorcio Figura 2. Bovinos criollos Encerado.

92 96 L. Aguirre et al. Figura 3. Bovinos criollos Colorado. Biobovis de la Red CONBIAND Criollos RSE, Retinta, Avileña, Rubia Gallega, Berrenda Colorada, Berrenda Negra, Mostrenca, Pajuna, Negra Andaluza, Canaria, Palmera, Frisona, Hereford, Pardo Suiza, Gyr, Brahman, Sindy, Guzera y Nelore utilizando el programa Populations (Olivier, 2002: see ics.org/tryphon/populations/). Con los valores obtenidos se trazaron árboles de distancias, graficando los mismos con el programa Treeview (Page, 1996) y con la matriz de distancias Da de Nei, se construyeron splitsgraph mediante el algoritmo Neighbor-Net con el programa SplitsTree v 4.10 (Bryant & Moulton, 2004; Huson & Bryant, 2006). Resultados y discusión Variabilidad genética poblacional En la Tabla 1, se observa que la población bovina de la RSE presenta un total de 228 alelos, con una media de 8,14 ± 2,05 y un rango comprendido entre 5 (ILSTS6, INRA63, BM1824,INRA35) y 12 (HEL9, TGLA53) alelos por locus. Estos valores son superiores a los reportados por Villalobos (2010) para las criollas panameñas Guabala y Guaymi, y algunas razas ibéricas tales como Berrenda Colorada, Berrenda Negra, Pajuna, Mostrenca, Retinta, Negra Andaluza, siendo comparables los resultados de Na nuestros, con los obtenidos por Dalvit et al. (2008) en cuatro razas italianas (Na = 8,2). También se determinaron los alelos únicos (Au) en los cuatro grupos poblacionales (datos no mostrados), destacando que 96 alelos son compartidos en los cuatro grupos, siendo el grupo de los Negro Lojanos el que presenta el mayor número de Au (17) y que se correlaciona con una mayor variabilidad alélica, en tanto que los Pintados poseen el menor número de Au (5) y también la menor variabilidad. En general, los marcadores que presentaron los valores más elevados son los más recomendables en caso de asignación de nuevos individuos a esta población. De los 28 microsatélites analizados ninguno mostró alelos con frecuencias superiores a 0,95, por lo cual todos pueden considerarse polimórficos y altamente informativos, exceptuando el HEL13 y ILSTS011, que presentan un PIC medianamente informativo; por tanto el 100 por ciento de estos marcadores son útiles para valorar la variabilidad genética del bovino criollo de la RSE, es de destacar que el PIC Figura 4. Bovinos criollos Pintado.

93 Caracterización genética de la población bovina criolla de la Región Sur del Ecuador 97 Figura 5. Cantón Gonzanama en la provincia de Loja, Ecuador, área de ubicación de los bovinos criollos. medio de esta población (0,7) es ligeramente superior al obtenido en la raza Guaymi panameña y la Limonero venezolana, con 0,69 y 0,65 respectivamente (Villalobos, 2010; Villasmil et al., 2008). Los valores medios de Ho y He en la población criolla de la RSE muestran valores altos de 0,676 y 0,74 respectivamente, existiendo únicamente cuatro marcadores genéticos que presentan un menor grado de heterocigosis: HAUT27, HEL13, ILSTS011 e INRA35, determinando que hay una buena variabilidad genética, si comparamos los mismos con los valores de He y Ho que se han obtenido en estudios de caracterización genética que oscilan entre 0,755 y 0,895 en razas mexicanas (Quiroz, 2007); 0,461 y 0,694 en razas Argentinas (Martínez, 2008); 0,750 y 0,756 en el criollo Casanareño de Colombia (Sastre et al., 2003). Por otra parte Villalobos (2010) cita los estudios hechos en seis razas autóctonas españolas que mostraron valores ligeramente inferiores de 0,564 y 0,570 (Menorquina), 0,590 y 0,512 (Lidia), 0,617 y 0,588 (Pirenaica), 0,681 y 0,666 (Asturiana); igualmente se tiene información reportada en razas autóctonas de Portugal como la Alentejana, valores de 0,677 y 0,638, Mertolenga 0,696 y 0,747, Arauquesa 0,717 y 0,730 (Mateus et al., 2004). En razas de Hungría y Austria como la Carinthian Blond y la Waldviertel Blond, se reportan valores de 0,674, 0,662 y 0,654, 0,634 respectivamente (Baumung et al., 2006). De los microsatélites analizados, nueve no se han mostrado en equilibrio Hardy-Weinberg (P < 0.1). Estos son: BM2113, HAUT27, BM1824, ETH225, ETH3, HAUT24, ILSTS011, INRA35 y TGLA53, lo que representa el 32 por ciento del total de alelos y revela que en estos animales se están operando fuerzas que están cambiando las frecuencias genotípicas ya sea por deriva genética, selección o migración, y que pone de manifiesto que hay una ligera inestabilidad genética en esta población. Estas desviaciones del equilibrio

94 98 L. Aguirre et al. Tabla 1. Índices de variabilidad genética obtenida en los 28 loci analizados de la población criolla de la Región Sur del Ecuador. Locus Na PIC Ho He HWE BM ,64 0,61 0,70 0,23 BM ,70 0,78 0,75 0,91 BM ,62 0,63 0,69 0,92 CSSM ,84 0,85 0,87 0,42 ETH10 8 0,56 0,56 0,59 0,41 INRA ,74 0,67 0,78 0,16 MM12 8 0,63 0,67 0,67 0,98 TGLA ,76 0,70 0,79 0,13 BM ,69 0,55 0,73 0,00* CRSM60 8 0,65 0,76 0,71 0,58 ETH ,72 0,77 0,76 0,92 HAUT27 9 0,75 0,50* 0,78 0,00* HEL13 7 0,51 0,50* 0,54 0,16 HEL9 12 0,74 0,80 0,78 0,19 ILSTS6 5 0,76 0,83 0,80 0,95 INRA ,76 0,70 0,80 0,38 INRA ,71 0,63 0,75 0,10 INRA63 5 0,59 0,72 0,66 0,84 SPS ,74 0,69 0,78 0,36 TGLA ,84 0,89 0,87 0,15 BM ,70 0,76 0,75 0,05* ETH ,78 0,76 0,82 0,02* ETH3 11 0,79 0,72 0,82 0,01* HAUT24 7 0,78 0,62 0,82 0,03* ILSTS ,49 0,42* 0,52 0,01* INRA35 5 0,53 0,35* 0,58 0,00* TGLA ,85 0,77 0,88 0,00* TGLA ,69 0,70 0,74 0,88 Total 228 Promedio ± SD 8,1 ± 2,05 0,7 ± 0,1 0,68 ± 0,01 0,74 ± 0,02 Na: número de alelos por locus; PIC: contenido de información polimórfica; Ho: heterocigosidad observada; He: heterocigosidad esperada; HWE: equilíbrio Hardy-Weinberg. Nota: Se destacan con * los marcadores con menor grado de heterocigosis y en desequilibrio Hardy-Weinberg Hardy-Weinberg (HWE) pueden estar produciéndose aceleradamente debido a factores como los siguientes: en la selección y en el apareamiento de los animales hay intervención del propietario; existe migración de animales por la cercanía geográfica entre las poblaciones; no existe el criterio de tener núcleos de animales fenotípicamente homogéneos. Vale destacar que los microsatélites ETH3, INRA35 y TGLA53 también se encontraron en desequilibrio en los bovinos criollos Patagónicos (Martínez, 2008) y los microsatélites BM1824, HAUT24 y HAUT27 en la población Guabala (Villalobos, 2010). Análisis intrapoblacional En la Tabla 2, se presentan los índices de variabilidad genética intrapoblacional considerando las cuatro subpoblaciones del bovino criollo de la RSE. El valor de G st (coeficiente de diferenciación genética) es entre 0,0091 para el locus BM1818 y 0,115 para el INRA32, y el valor medio para todos los loci representa apenas el 4,5 ± 2,8 por ciento, Tabla 2. Coeficiente de variación genética (G ST ) y Estadísticos F de Wright: coeficiente de consanguinidad dentro de la población (F is ), coeficiente de consanguinidad total (F it ) y coeficiente de variación genética entre las subpoblaciones (F st ), por marcador en la población bovina criolla de la RSE. Locus G st F st (θ) F it (F) F is (f) BM8125 0,03 0,02 0,12 0,13 BM1314 0,03 0,01 0,04 0,03 BM1818 0,01 0,04 0,09 0,13 CSSM66 0,07 0,03 0,02 0,01 ETH10 0,04 0,02 0,06 0,05 INRA32 0,11 0,06 0,14 0,08 MM12 0,08 0,04 0,01 0,05 TGLA122 0,03 0,01 0,14 0,13 BM2113 0,09 0,03 0,26 0,23 CRSM60 0,04 0,00 0,08 0,08 ETH185 0,07 0,03 0,00 0,03 HAUT27 0,03 0,01 0,37 0,38 HEL13 0,02 0,03 0,07 0,09 HEL9 0,01 0,02 0,05 0,03 ILSTS6 0,02 0,02 0,04 0,01 INRA23 0,05 0,00 0,12 0,11 INRA37 0,07 0,02 0,16 0,14 INRA63 0,01 0,03 0,10 0,07 SPS115 0,03 0,01 0,11 0,12 TGLA227 0,02 0,02 0,03 0,01 BM1824 0,02 0,02 0,02 0,00 ETH225 0,02 0,02 0,07 0,09 ETH3 0,03 0,01 0,13 0,14 HAUT24 0,05 0,01 0,24 0,24 ILSTS011 0,09 0,04 0,20 0,17 INRA35 0,06 0,00 0,40 0,41 TGLA53 0,08 0,02 0,11 0,09 TGLA126 0,03 0,02 0,06 0,08 Promedio±SD 0,04±0,03 0,001±0,025 0,09±0,13 0,09±0,12 Nota: Se destacan con * los marcadores con mayor grado de consanguinidad. lo que nos indica que la diferenciación genética entre los cuatro grupos poblacionales es muy pequeña. Esta escasa diferenciación se corrobora con los valores de F st (grado de diferenciación genética entre las subpoblaciones) encontrados en los cuatro grupos fenotípicamente definidos, con un valor medio de F st = 0,0008, valor muy inferior al encontrado en el ganado criollo mexicano con 0,033 (Quiroz, 2007), en los bovinos criollos panameños con 0,068 (Villalobos, 2010) y entre las razas bovinas europeas con 0,068 y 0,07 (Cañón et al., 2001; Jordanaet al., 2003). Vale indicar que si bien no hay diferencia genética entre estas cuatro subpoblaciones, si se los debe manejar como núcleos homogéneamente diferenciados por el color del pelaje, pues ello favorece el bienestar animal y la adaptación que tienen los mismos a los diferentes pisos altitudinales de la Región Andina, en donde el Negro Lojano (Figura 1) y el Encerado (Figura 2) han ido evolucionando para un clima moderado por la altura, absorbiendo y manteniendo más calor de los rayos solares para estas zonas frías y ventosas y no siendo por el color obscuro los más apropiados para las zonas andinas bajas por el stress del calor y la mayor predisposición al ataque de los ectoparásitos (De Alba, 2009). Para estas zonas es más conveniente

95 Caracterización genética de la población bovina criolla de la Región Sur del Ecuador 99 Tabla 3. Matriz de distancias genéticas Da (Nei & Chesser, 1983), sobre 19 poblaciones de bovinos. RS-EC RET AVI RGA BC BN MOS PAJ NAN VCA PAL FRI HER SPA GYR BRH SIN GUZ NEL RS-EC 0 0,1524 0,1580 0,1566 0,1031 0,2379 0,1929 0,1132 0,1302 0,1326 0,2243 0,1363 0,1730 0,1593 0,4289 0,3429 0,4216 0,4725 0,4624 RET 0, ,1153 0,1014 0,1328 0,2456 0,2159 0,1223 0,1466 0,1735 0,2634 0,2143 0,2150 0,1862 0,4798 0,3656 0,4801 0,5165 0,4712 AVI 0,1580 0, ,1186 0,1249 0,2662 0,2221 0,1331 0,1169 0,1804 0,2645 0,1913 0,2441 0,1890 0,5400 0,4307 0,5161 0,5547 0,5246 RGA 0,1566 0,1014 0, ,1343 0,2491 0,2182 0,1256 0,1509 0,1587 0,2646 0,1774 0,1984 0,1789 0,5304 0,3942 0,4881 0,5536 0,5233 BC 0,1031 0,1328 0,1249 0, ,2007 0,1564 0,0673 0,1012 0,1110 0,2079 0,1402 0,1595 0,1363 0,4850 0,3681 0,4426 0,4953 0,4855 BN 0,2379 0,2456 0,2662 0,2491 0, ,2667 0,2057 0,1964 0,2310 0,3029 0,2759 0,2563 0,2370 0,5597 0,4394 0,5490 0,5903 0,5553 MOS 0,1929 0,2159 0,2221 0,2182 0,1564 0, ,1573 0,1814 0,1815 0,2794 0,2342 0,2003 0,2043 0,5389 0,4556 0,5243 0,5846 0,5247 PAJ 0,1132 0,1223 0,1331 0,1256 0,0673 0,2057 0, ,1055 0,1215 0,2177 0,1362 0,1742 0,1466 0,4796 0,3760 0,4771 0,5223 0,4994 NAN 0,1302 0,1466 0,1169 0,1509 0,1012 0,1964 0,1814 0, ,1548 0,2413 0,1807 0,2050 0,1568 0,5115 0,4132 0,5054 0,5347 0,5153 VCA 0,1326 0,1735 0,1804 0,1587 0,1110 0,2310 0,1815 0,1215 0, ,1603 0,1533 0,1813 0,1559 0,5058 0,3921 0,4697 0,5252 0,5006 PAL 0,2243 0,2634 0,2645 0,2646 0,2079 0,3029 0,2794 0,2177 0,2413 0, ,2648 0,2878 0,2693 0,5767 0,4928 0,5118 0,5701 0,5775 FRI 0,1363 0,2143 0,1913 0,1774 0,1402 0,2759 0,2342 0,1362 0,1807 0,1533 0, ,1842 0,1807 0,5580 0,4324 0,5253 0,5856 0,5521 HER 0,1730 0,2150 0,2441 0,1984 0,1595 0,2563 0,2003 0,1742 0,2050 0,1813 0,2878 0, ,1719 0,5394 0,4335 0,5111 0,5602 0,5398 SPA 0,1593 0,1862 0,1890 0,1789 0,1363 0,2370 0,2043 0,1466 0,1568 0,1559 0,2693 0,1807 0, ,5003 0,4022 0,4741 0,5426 0,5033 GYR 0,4289 0,4798 0,5400 0,5304 0,4850 0,5597 0,5389 0,4796 0,5115 0,5058 0,5767 0,5580 0,5394 0, ,1407 0,2330 0,2185 0,2404 BRH 0,3429 0,3656 0,4307 0,3942 0,3681 0,4394 0,4556 0,3760 0,4132 0,3921 0,4928 0,4324 0,4335 0,4022 0, ,1653 0,2048 0,1870 SIN 0,4216 0,4801 0,5161 0,4881 0,4426 0,5490 0,5243 0,4771 0,5054 0,4697 0,5118 0,5253 0,5111 0,4741 0,2330 0, ,2333 0,2664 GUZ 0,4725 0,5165 0,5547 0,5536 0,4953 0,5903 0,5846 0,5223 0,5347 0,5252 0,5701 0,5856 0,5602 0,5426 0,2185 0,2048 0, ,2526 NEL 0,4624 0,4712 0,5246 0,5233 0,4855 0,5553 0,5247 0,4994 0,5153 0,5006 0,5775 0,5521 0,5398 0,5033 0,2404 0,1870 0,2664 0, Abreviación de las razas: RS-EC: Región Sur del Ecuador; RET: Retinta; AVI: Avileña; RGA: Rubia Gallega; BC: Berrenda Colorada; BN: Berrenda Negra; MOS: Marismeña; PAJ: Pajuna o Serrana; NAN: Negra Andaluza; VCA: vaca Canaria; PAL: vaca Palmera; FRI: Frisona; HER: Hereford; SPA: Pardo Suizo; GYR: Gyr; BRH: Brahman; SIN: Sindi; GUZ: Guzera; NEL: Nelore.

96 100 L. Aguirre et al. mantener el grupo de los Colorados (Figura 3) y Pintados (Figura 4). El F it (coeficiente consanguinidad total población) y el F is (coeficiente consanguinidad subpoblaciones) indican que hay nueve marcadores que presentan valores negativos (exceso de heterocigosis); cabe destacar también que marcadores que estén con valores de F it yf is elevados (superiores a 0,1) indican una consanguinidad elevada, sea a nivel de población total o subpoblación. En este caso, son once los marcadores con un F it yf is superior a 0,10 (BM8125, TGLA122, BM2113, HAUT27, INRA23, INRA37, SPS115, ETH3, HAUT24, ILSTS011 e INRA35). De acuerdo a los resultados de estos estadísticos, la media del F it yf is de la población es de 8,89 y 8,92 por ciento respectivamente; valor ligeramente superior al encontrado por Martínez (2008) en los bovinos Patagónicos (0,06579) y muy superior al descrito en los bovinos criollos del Noroeste de Argentina ( 0,00913). Esto revela que los vacunos criollos de la RSE muestran un nivel mayor de homocigosis. Relaciones genéticas Para el cálculo de distancias se trabajó con 19 poblaciones (razas): 13 taurinas europeas, 5 índicas y la población criolla de la RSE. El tamaño de la muestra analizada fue de 789 individuos y 24 locus estudiados; se empleó el cálculo de distancia Da (Nei & Chesser, 1983), y la Tabla 3 muestra los valores de distancia genética de los criollos de la RSE con el resto de poblaciones. Es de destacar que los distanciamientos con las 18 poblaciones de bovinos, superan el 0,10. Las distancias menores están entre RSE:BC (0,1031), RSE:PAJ (0,1132), RSE:NAN (0,1302), RSE:VCA (0,1326) y RSE: FRI (0,1362); el mayor distanciamiento se observa entre RSE:GUZ (0,4725) y RSE:NEL (0,4624). Estos valores indican que los bovinos criollos de la RSE tienen un gran distanciamiento con las razas cebuinas y más bien sus troncos ancestrales están en las poblaciones ibéricas traídas en la conquista de América, destacándose en ello, en su orden: la Berrenda Colorada, Pajuna ó Serrana, la Negra Andaluza, la vaca Canaria y la Frisona (Holstein Friesian), población esta última importada recientemente a la región. En la Figura 6, se puede observar el árbol de distancias genéticas construido mediante el algoritmo Neighbor-Net, de las 19 razas bovinas, analizadas mediante la matriz de distancias Da. Con un valor de replicación del 49 por ciento, se nota la rama distante que separa al grupo (clúster) de razas cebuinas de las taurinas, incluida la criolla de la RSE; así mismo, con un valor de replicación bajo de 39 por ciento, hay una relación genética cerrada entre la Hereford, la Marismeña y la Pardo Suiza formando un clúster; otra rama la conforman la Rubia Gallega, la Retinta y la Avileña. Hay una población separada del resto que es la Berrenda Negra; y con un nivel de remuestreo alto (99 y 68 por ciento), se forma una rama con nodos cortos que dan lugar a una relación genética entre la población criolla de la RS-EC, la Pajuna, la Berrenda Colorada, la Negra Andaluza, la vaca Canaria y la Frisona. Figura 6. Árbol de distancias genéticas entre 19 poblaciones (razas) bovinas: Da, NJ. Abreviación de las razas: NEL: Nelore; GYR: Gyr; GUZ: Guzera; BRH: Brahman; SIN: Sindi; EC:Ecuador; VCA: vaca Canaria; PAL: vaca Palmera; FRI: Frisona; NAN: Negra Andaluza; AVI: Avileña; RET: Retinta; RGA: Rubia Gallega; BC: Berrenda Colorada; PAJ: Pajuna ó Serrana; BN: Berrenda Negra; SPA: Pardo Suizo; MOS: Marismeña; HER: Hereford. En general se puede apreciar que los bovinos criollos de la RSE están muy distanciados genéticamente de las razas cebuínas y están más relacionadas con las razas de origen taurino, especialmente las razas autóctonas españolas; en este análisis también hay que destacar que a excepción de la rama de las razas índicas, prácticamente todos los nodos del árbol están próximos al origen del mismo, lo que indica que las poblaciones representadas son independientes entre sí. Estos resultados filogenéticos son similares a los encontrados por Villalobos (2010) y Martínez (2008) en los bovinos criollos panameños y argentinos respectivamente, lo que demuestran el origen común que tienen los vacunos criollos de América. Conclusiones Los bovinos criollos de la RSE presentan una elevada variabilidad genética, demostrada en el número de alelos por locus, en el PIC y en la heterocigosidad, vale destacar el 32 por ciento de alelos que se encuentran en desequilibrio H-W, lo que demuestra una ligera inestabilidad genética que de no tomarse correctivos conllevaría a una disminución de esta valiosa diversidad.

97 Caracterización genética de la población bovina criolla de la Región Sur del Ecuador 101 Esto se corrobora con los valores de F it yf is encontrados donde se evidencia que existe un exceso de homocigosis que puede deberse a una elevada consanguinidad que supera el 8 por ciento; si bien los valores de G st yf st, indican que no hay una diferenciación genética entre los cuatro grupos de bovinos criollos, faneróptica y fenotípicamente sí se diferencian por lo que convendría manejarlos como núcleos en forma separada por el color del pelaje, pues esto ayuda a un mejor bienestar y desenvolvimiento de los animales de acuerdo a los diferentes pisos altitudinales existentes en la región. En la determinación de las distancias genéticas se aprecia un mayor grado de acercamiento con razas ibéricas, como la Berrenda Colorada, Pajuna, Negra Andaluza, Canaria y Frisona, y un mayor alejamiento genético con las cebuinas; en todo caso los valores de distanciamiento superan el 0,10 en todos los pares en que se comparó a la criolla de la RSE con el resto de razas, lo que permite afirmar que la población bovinacriolla de la RSEpuede ser considerada genéticamente como una raza taurina propia a la que hay que conservar, pues se corre el riesgo a un corto plazodesudesaparición. Agradecimiento Al proyecto Biobovis de la Red CONBIAND, que con su banco de material genético, contribuyó a cumplir con los objetivos propuestos en la presente investigación. Referencias Aguirre, L., Bermeo, A., Maza, D. & Merino, L Estudio fenotípico y zoométrico del bovino criollo de la sierra media y alta de la región sur del Ecuador (RSE). Actas Iberoamericanas de Conservación Animal 1: Baumung, R., Cubric-Curik, V., Schwend, K., Achmann, R. & Soelkner, J Genetic characterization and breed assignment in Austrian sheep breeds using microsatellite marker information. Journal of Animal Breeding and Genetics 123: Belkhir, K., Borsa, P., Chikhi, L., Raufaste, N. & Bonhomme, F Genetix: 4.05 Logiciel sous Windows TM pour la génétique des populations. Laboratoire Génome, Populations, Interactions CNRS UMR Université de Montpellier II, Montpellier, Francia. Botstein, D., White, R.I., Skolnich, M. & Davis, R.W Construction of a genetic linkage map in man using restriction fragment length polymorphism. American Journal of Human Genetics 32: Bryant, D. & Moulton, V Neighbor-Net: an agglomerative method for the construction of phylogenetic networks. Mol. Biol. Evol. 21: Cañon, J., Alexandrino, P., Bessa, C., Carleos, Y., Carretero, S., Dunner, N., Ferran, D., Garcia, J., Jordana, J., Laloe, A., Pereira, A. & Moazami, K Genetic diversity measures of local European beef cattle breeds for conservation purposes. Genetics Selection Evolution 33: Dalvit, C., De Marchi, M., Dal Zotto, R., Gervaso, M., Meuwissen, T. & Cassandro, M Breed assignment test in four Italian beef cattle breeds. Meat Science 80: De Alba, M.J El libro de los bovinos criollos de América (Ed. Papiro Omega, S.A., México), pp , 425. FAO Molecular genetic characterization of animal genetic resources. FAO Animal Production and Health Guidelines. No. 9. Rome. (available at pdf). Huson, D.H. & Bryant, D Application of phylogenetic networks in evolutionary studies. Mol. Biol. Evol. 23: Jordana, J., Alexandrino, P., Beja-Pereira, A., Bessa, I., Cañon, J., Carretero, Y., Dunner, S., Laloe, D., Moazami, K., Sanchez, A. & Ferrand, N Genetic structure of eighteen local south European beef cattle breeds by comparative F-statistics analysis. Journal of Animal Breeding and Genetics 120: Lucena Salmoral, M Sebastián de Belalcázar, Historia 15 (Quórum, Madrid), p Martínez, R Caracterización genética y morfológica del bovino criollo argentino de origen patagónico. Tesis de doctorado, Universidad Politécnica de Valencia, España, pp Mateus,J.C.,Penedo,M.C.,Alves,V.C.,Ramos,M.&Rangel-Figueiredo, T Genetic diversity and differentiation in Portuguese cattle breeds using microsatellites. Animal Genetics 35: Nei, M., & Chesser, R.K Estimation of fixation indices and gene diversities. Ann. Hum. Genet. 47: Olivier, L., Populations Disponible en: Page, R.D.M Treeview: an application to display phylogenetic trees on personal computers. Computer Applications in the Biosciences 12: Park, S.D.E Trypanotolerance in West African cattle and the population genetic effects of selection. Tesis de doctorado, Trinity College, Dublín. Irlanda. Pinzón, M.E Historia de la Ganadería en Colombia. Suplemento Ganadero CEGA, Centro de Estudios Ganaderos y Agricolas. Bogota, Colombia. Vol. 4, p Quiroz, J Caracterización genética de los bovinos criollos mexicanos y su relación con otras poblaciones bovinas. Tesis de doctorado, Universidad de Córdoba, España, p Sastre, H.J., Rodero, E., Rodero, A., Azor, P.J., Sepúlveda, N.G., Herrera, M. & Molina, A Caracterización genética de la raza bovina colombiana criolla Casanare mediante análisis de microsatélites. Revista Colombiana de Ciencias Pecuarias 16 (supl.): 49. Villalobos, A Caracterización genética de las poblaciones bovinas Guaymí y Guabala y su relación con otras poblaciones bovinas mediante microsatélites. Tesis de doctorado, Universidad de Córdoba, España, pp , Villasmil, Y., Roman, R., Yañez, L., Contreras, G., Jordana, J. & Aranguren, J Diversidad genética de la raza criollo Limonero utilizando marcadores de ADN microsatélites. Revista Científica FCV-LUZ 18 (4): Weir, B.S. & Cockerham, C.C Estimating F-statistics for the analysis of population structure. Evolution 38: Wright, S Evolution and the genetics of populations, vol. 2: The theory of gene frequencies. University of Chicago Press, Chicago.

98 Animal Genetic Resources, 2014, 54, Food and Agriculture Organization of the United Nations, 2014 doi: /s Genealogical and population viability analysis of a conservation nucleus of Brazilian Bergamasca sheep H. Carneiro 1, S.R. Paiva 2, H. Louvandini 3, R.M. Miranda 4 and C. McManus 1,4 1 Faculdade de Agronomia e Medicina Veterinária, Campus Universitário Darcy Ribeiro, Universidade de Brasília, Brasília, DF , Brazil; 2 EMBRAPA Recursos Genéticos e Biotecnologia, Parque Estação Biológica, Avenida W5 Norte (Final), Brasília, DF , Brazil; 3 Centro de Energia Nuclear na Agricultura, Universidade de São Paulo, Piracicaba, SP, CEP , Brazil; 4 Departamento de Zootecnia, Universidade Federal de Rio Grande do Sul, Av. Bento Gonçalves, 7712, Porto Alegre, RS, CEP , Brazil Summary A genealogical and a viability analysis was carried out on the 1559 available registers for the Conservation Nucleus of Brazilian Bergamasca sheep of the University of Brasilia farm in Brazil using the ENDOG and the Vortex programs. To run the ENDOG it was used the registered data and for the Vortex it was used information obtained by a questionnaire answered by the curators of the herds. Of the animals registered, 767 had known parents, with significantly more dams known at each generation. The number of pedigrees known has increased over the generations, with higher registration of parents of sires than dams. The Computed Mean Inbreeding calculated by ENDOG was 0.29 percent and mean average relatedness was 1.52 percent. Mean Generation interval was 3.71 years with this values being lower for sires than dams. The population probability of extinction, calculated by Vortex was 17 percent and the average time to extinction was 59 years. Forty-two additional scenarios were created to determine which factors most threaten these populations which were frequency of catastrophes, lack of animal entrance and adult and lamb mortalities, especially adult female mortality. These results indicate that future breeding plans should include exchange of sires between farms to maintain low inbreeding levels and increase genetic variability and upgrade the management to control the mortality rates of animals. Keywords: average relatedness, inbreeding, population modelling, Vortex, Wright statistics Résumé Une analyse généalogique et de viabilité a été réalisée avec les 1559 enregistrements disponibles au Centre de Conservation de la race ovine Bergamasca Brésilienne de l Université de Brasilia (Brésil), en utilisant les logiciels ENDOG et Vortex. Les données des enregistrements ont été utilisées pour travailler avec le logiciel ENDOG alors que l information traitée par le logiciel Vortex avait été obtenue au moyen d un questionnaire auquel ont répondu les personnes prenant soin des troupeaux. Pour 767 des animaux enregistrés, les parents étaient connus, avec le nombre de femelles connues étant significativement plus élevé à chaque génération. Le nombre de généalogies connues a augmenté au fil des générations, grâce à un enregistrement plus fréquent des parents des mâles. La Consanguinité Moyenne a été estimée par ENDOG à 0,29 pour cent alors que la Parenté Moyenne a été de 1,52 pour cent. L Intervalle Générationnel Moyen a été de 3,71 ans, avec l intervalle étant plus court chez les mâles que chez les femelles. La probabilité d extinction de la population a été de 17 pour cent d après Vortex et le temps moyen d extinction a été de 59 ans. Quarante-deux scénarios additionnels ont été créés pour déterminer quels sont les facteurs menaçant le plus ces populations. Ceuxci ont été: la fréquence de catastrophes, le manque de renouvellement des animaux et la mortalité des adultes et des agneaux, notamment la mortalité des femelles adultes. Ces résultats indiquent que les futurs plans de sélection devraient envisager l échange de mâles entre les fermes afin de maintenir bas les niveaux de consanguinité et d accroître la variabilité génétique. Il faudrait, de même, améliorer la gestion des exploitations dans le but de contrôler le taux de mortalité des animaux. Mots-clés: consanguinité, modélisation populationnelle, parenté moyenne, statistiques de Wright, Vortex Resumen Se realizó un análisis genealógico y de viabilidad con los 1559 registros disponibles en el Núcleo de Conservación de la raza ovina Bergamasca Brasileña, de la Universidad de Brasilia (Brasil), usando los programas ENDOG y Vortex. En el programa ENDOG se trabajó con los datos de los registros mientras que en el programa Vortex se usó la información obtenida mediante un cuestionario al que respondieron los encargados del cuidado de los rebaños. De los animales inscritos, se conocían los progenitores de 767, con un número significativamente mayor de hembras conocidas en cada generación. El número de genealogías conocidas ha aumentado a lo largo de las generaciones, gracias a un mayor registro de los progenitores de los machos. La Consanguinidad Media calculada por ENDOG fue de 0,29 por ciento mientras que el Parentesco Medio fue de 1,52 por ciento. El Intervalo Generacional Medio fue Correspondence to: C. McManus, Faculdade de Agronomia e Medicina Veterinária, Campus Universitário Darcy Ribeiro, Brasília DF , Brazil. concepta@unb.br 103

99 104 H. Carneiro et al. de 3,71 años, siendo este dato menor para los machos que para las hembras. La probabilidad de extinción de la población, calculada por Vortex, fue del 17 por ciento y el tiempo medio hasta la extinción de 59 años. Se recrearon 42 escenarios adicionales para determinar cuáles son los factores que, en mayor medida, amenazan a estas poblaciones. Éstos resultaron ser: la frecuencia de catástrofes, la falta de reposición de los animales y la mortalidad de los adultos y de los corderos, en especial la mortalidad de las hembras adultas. Estos resultados indican que los futuros planes de selección deberían contemplar el intercambio de machos entre las granjas para mantener bajos los niveles de consanguinidad e incrementar la variabilidad genética. Asimismo, se debería mejorar el manejo de las explotaciones para controlar las tasas de mortalidad de los animales. Palabras clave: consanguinidad, estadísticos de Wright, modelización poblacional, parentesco medio, Vortex Submitted 11 June 2013; accepted 10 January 2014 Introduction Sheep have been present in Brazil for almost 500 years, but their numbers have not increased to the same extent as other farm animals (Hermuche et al., 2012, 2013). This has been largely due to herd structure (smallholders), perceived use of sheep as a secondary culture and lack of interest from major research groups. Taking into account rapid genetic erosion (Hall and Ruane, 1993; Scherf, 2000) and the importance of Animal Genetic Resources for the future (Notter, 1999), it may not be possible to wait for specific and detailed information on available breeds and ecotypes. The Brazilian Bergamasca is an Italian breed from the Lombardy region, which is widespread mainly in Northern Italy (Dalvit et al., 2009). In Italy, this breed specializes in meat production and is widely used in crossing with other breeds. It is thought to have been introduced in Brazil in the 1930s with Italian immigration (Miranda and McManus, 2000; Paiva et al., 2005). Although once widespread has been substituted for other breeds, especially hair sheep. In Brazil, it is considered a triple purpose breed raised for milk, meat and coarse wool (Miranda and McManus, 2000), and at present is threatened with extinction in the country. More recently, there has been some interest in this breed for milk production to make cheese, which is not traditional in Brazil, but a few herds are left. The rare ARK haplotype has also been described in this breed (Luhken et al., 2007) meaning that it may be useful in studying scrapie resistance in sheep. Embrapa (Brazilian Agricultural Research Corporation) included domesticated and naturalized breeds threatened of extinction in its Program of Research on Genetic Resources in Currently, the conservation of animal genetic resources is carried out by diverse Research Centers of Embrapa, Universities, State Research Companies, as well as private farmers. Currently, much of the naturalized animals are in small and scattered populations. Small populations are unstable, since they are more exposed to stochastic factors that can lead to extinction, such as demographic stochasticity, environmental variation, genetic drift and inbreeding depression (Armstrong, Postiglioni and Gonzáles, 2006). Genetic diversity can be considered the cornerstone of conservation genetics, since it is the basis of the evolutionary potential of species to respond to environmental changes. The most threatened populations are subdivided into different groups in different fragments of habitat, nature reserves or in groups of different breeds or lines, in the case of domestic animals (Toro and Caballero, 2005). This study aims to analyse genealogical information integrated with simulation tools in the risk of extinction and molecular markers of a Conservation Nucleus of this breed to determine the effectiveness of current conservation efforts. Material and methods The flock of Brazilian Bergamasca sheep used in this study is kept on the University of Brasilia (UnB) research farm, is part of the Brazilian National Program for Conservation of Animal Genetic Resources and this is the only one of this breed to participate in this program. The population studied was of animals, which included all Brazilian Bergamasca sheep on the UnB farm born between 1990 and There were 48 percent males and 52 percent females. Data for the population viability analysis were taken from the questionnaire answered by the curator of the flock, the data from which are in Table 1. On the other hand, information for the pedigree analysis were obtained in the animal register. Two analyses were carried out: one looking at the genetic structure of the herd over the 20-year period to determine population parameters (pedigree analysis) and the other simulating future events to evaluate the future of the herd (viability analysis). Pedigree analysis The population genetic parameters were computed using the ENDOG program, version 4.8. The program details can be found in Gutiérrez and Goyache (2005). ENDOG is a computer program for population genetics that performs various demographic and genetic analyses from pedigree information. A number of population and genetic

100 Genealogy and viability analyses for sheep conservation 105 parameters were computed for the reference populations, which consist of animals with both parents known. Parameters included: pedigree completeness level was computed as the proportion of ancestors known per parental generation; the number of equivalent to discrete generations (t) for each individual in a pedigree; the inbreeding coefficient (F), defined as the probability that two alleles at a randomly chosen locus are identical by descent; The average relatedness coefficient (AR), defined as the probability that an allele randomly chosen from the whole population belongs to a given animal; the probability of gene origin was characterized by computing: effective number of founders ( f e ) was computed according to Lacy (1989); effective number of ancestors ( f a ); the founder genome equivalents was calculated from the inverse of twice the average coancestry of the individuals within the population. Effective population size (N e ) was computed following Gutiérrez et al. (2003). Population viability analysis Simulations of interactions between demographic, environmental and genetic factors were carried out using the program Vortex, version A questionnaire was developed to be answered by the curators of the herds linked to the Brazilian National Program for Conservation of Animal Genetic Resources with the information required for program input. According to Miller and Lacy (2005), the Vortex program begins by creating individuals to form a base population and then steps through life cycle events on an annual basis. Stochastic events such as breeding success, progeny number, sex at birth and mortality rate were determined by a predefined probability density function given by the operator (Table 1). Consequently, each iteration of the model gave a different result and by running the model hundreds of times, and a range of possibilities and outcomes were examined. The scenario described in Table 1 was called baseline and all the other scenarios created changed one parameter of it to test the population (Tables 5 and 6). The data including catastrophes described above were informed by the curator of the nucleus. The loss of genetic variability is simulated as the transmission of alleles from parents to offspring in a hypothetical neutral site. For each animal at the beginning of the simulation is assigned two unique alleles at this locus. Then for each individual created during the simulation is randomly assigned one of the alleles from each parent. The program monitors how many alleles remain in the original population, by giving the mean number of alleles remaining within extant population, the average heterozygosity and genetic diversity relative to the initial level. Forty-six Brazilian Bergamasca animals were genotyped using 19 microsatellite markers to provide input allele frequency information, obtained in the Animal Genetics Laboratory in Embrapa Genetic Resources and Biotechnology, Brasilia DF, Brazil (Table 2). Diversity Table 1. Demographic parameters used to simulate the population of sheep Brazilian Bergamasca with Vortex. Input data indexes for microsatellites were calculated using the Molkin package (Gutiérrez et al., 2005). To determine which factors most threaten this population, new scenarios were tested to see the behaviour of the parameters. Forty-two additional scenarios were created, as described in Tables 5 and 6, by changing baseline (Table 1) values, maintaining other variables stable. Results Value Number of iterations Number of years 100 Reproduction system Polygamous Age of first offspring (F/M) 2/2 Inbreeding depression? Yes Maximum number of lambings per year 2 Maximum number of progeny per lambing 3 Sex ratio at birth 50% Maximum age of reproduction 8 Adult female breeding (%) 80 Probability of simple birth 74 Probability of twin birth 23 Probability of triple birth 3 Female mortality age between 0 and 1 (%) 30 Female mortality age between 1 and 2 (%) 5 Female mortality after age 2 (%) 5 Male mortality age between 0 and 1 (%) 30 Male mortality age between 1 and 2 (%) 5 Male mortality after age 2 5 Catastrophe 1: animal attack Frequency (%) 6 Effect on reproduction (%) 80 Effect on survival (%) 90 Catastrophe 2: feeding error Frequency (%) 12 Effect on reproduction (%) 60 Effect on survival (%) 50 Males in breeding pool (%) 5 Initial population 25 Carrying capacity (K) 400 Animals harvested? Sim 100 males per year First year of harvest 5 Last year of harvest 100 Interval between harvests 1 Population supplemented? Sim One male every 2 years First year of supplementation 2 Interval between supplementation 2 Pedigree data on animals were analysed, the herd was classified as commercial using according to Gutiérrez and Goyache (2005) as care has been taken over the years to buy rams where possible to avoid inbreeding. Pedigree Information increased over the generations were analysed, with more information available on dams than sires, with almost 100 percent of dams

101 106 H. Carneiro et al. Table 2. Microsatellite loci information in Brazilian Bergamasca sheep. Locus Number of alleles/locus Allele frequency Inra23A Oar304A MAF214A INRA63A OARHH35A INRA35A OMHC1A ILSTS87A ILSTS05A ILSTS11A MAF65A BM827A OARFCB20A OARCP20A OAR129A INRA172A HUJ616A SRCRSP05A BM6526A known in the present generation (Figure 1). The reference populations were taken as the animals with both parents known. This population will be smaller than that one used to analyse Founders. The effective founder population Size was 229 with 791 animals in the reference population and 215 ancestors contributing to the reference population. The effective number of founders/ancestors for the reference population was 42/34 with 12 animals supplying 50 percent of the ancestors. The Computed Mean Inbreeding was 0.29 percent and Mean AR was 1.52 percent. These numbers are comparable to those for the Galega (Adán et al., 2007) and Mallorquina (Goyache et al., 2010) sheep in Spain and well below the dangerous level (10 percent). As the generations increased, there was little change in inbreeding and AR (Table 1). This was due to careful control of reproduction in the nucleus. Although very low (<0.4 percent) there is an increase over the generations, which should be carefully monitored. There were 2 percent matings between sibs and 1 percent between parent-offspring. These values are summarized in Table 3, per generation. Effective population size per year (Figure 2) has remained relatively stable over the years, varying about 30 animals. The sharp decrease in 1998 was due to an attack by dogs, which killed the whole lamb flock and a large number of ewes. Generation intervals (Table 4) varied from 4.08 (father son) to 2.98 years (Father daughter). This shows a relatively fast turnover of reproducing animals. In terms of mean age of parents when offspring were born these were higher for dams than sires. Alderson (1990) used the genetic conservation index (GCI) to calculate an effective number of founders in the pedigree of an animal. The higher the GCI value the higher the value of an animal for conservation. Mean GCI was 1.36, varying between 0.77 and This has increased steadily over the years with an increment of approximately 0.05 per year (Figure 3). Using sires to analyse differentiation, F is was found to be 0.080; F st = and F it = 0.005, all indices being close to zero. In the present study, when sires were considered subpopulations there is low genetic differentiation (7.7 percent of total genetic variation). Self-coancestry Figure 1. Proportion of parents known per generation in a Conservation Nucleus of Brazilian Bergamasca sheep. P parent, GP grandparent, GGP great grandparent. Figure 2. Effective population size per year of birth in a Conservation Nucleus of Brazilian Bergamasca sheep.

102 Genealogy and viability analyses for sheep conservation 107 Table 3. Inbreeding (F), average relatedness (AR) and effective population size (Ne) per generation in a Conservation Nucleus of Brazilian Bergamasca sheep. F % AR Ne Maximum generations Complete generations was 0.50, inbreeding and Nei distance between sires 0.069, showing that, according to the pedigree analysis, no immediate problems in maintenance of genetic diversity within the herd is seen. Population viability analysis The estimated probability of extinction in population viability analysis was 17 percent and therefore, the possibility of success of 83 percent. From the scenarios that have become extinct, the average time to extinction was 59 years (Table 5). Considering only the deterministic growth of the population, i.e. no inbreeding depression, mating with no restrictions and no entry and exit of animals, the population would grow until it reaches the carrying capacity of the environment (K) and the deterministic growth rate (Det.r) remained stable for the full simulation period. However, the stochastic prediction considering the factors and the stochastic growth rate (Stoch.r) was different (Figure 4). In the herd analysed, population growth was much lower than expected considering only the Det.r (Figure 4). The Table 4. Generation intervals (L), their standard deviations (SD) and standard errors (SE) for a Conservation Nucleus of Brazilian Bergamasca sheep. Interval Standard deviation SE Generation intervals Father Son Father Daughter Mother Son Mother Daughter Total Mean age of parents when offspring born Father Son Father Daughter Mother Son Mother Daughter Total Figure 3. Genetic conservation index for year of birth in a conservation nucleus of Brazilian Bergamasca sheep. difference in expected values between Stoch.r and Det.r is clearly seen. While the first had high variation, the second remained stable throughout the period of the simulation, which supports the idea that the population is highly susceptible to environmental and genetic effects, such as variations in the feed, precipitation levels, inbreeding and genetic drift. When considering the influence of environmental variation, the population did not reach K = 400 (Figure 5). The values of genetic diversity were relatively high, always above 0.8 (Table 5). The overall results did not change significantly when the default program was used (each animal at the beginning of the simulation is assigned two unique alleles and not using real population genetic data (Table 5)), as the genetic diversity values were very similar in both scenarios. However, when calculating the within diversity observed directly from the original microsatellite markers, the following values of diversity was obtained: (1) observed heterozygosity of (SD ); (2) expected heterozygosity of (SD ); (3) mean number of alleles 6.53 (SD 2.29); and fixation index (Fis) of (P > 0.05). The mean observed heterozygosity reported by Vortex is the mean inbreeding coefficient of the population (Miller and Lacy, 2005), in this case, a high inbreeding value of One thousand iterations were needed to ensure stabilization of the results (Harris, Maguire and Shaffer, 1987). This population appeared to be influenced by genetic effects inherent to small populations. This can also be noticed by observing the average number of alleles, which declined to less than half of the initial value for the Figure 4. Difference between the deterministic (Det.r) and stochastic (Stoch.r) growth rates for the period of Vortex simulations in Brazilian Bergamasca sheep.

103 108 H. Carneiro et al. and the lamb mortality in this population is already high (30 percent). Figure 5. Total number of animals achieved at the end of Vortex simulations in Brazilian Bergamasca sheep considering only the stochastic growth rate, i.e. environmental and genetic factors. Figure 6. Total number of alleles obtained at the end of Vortex simulations in Brazilian Bergamasca sheep estimated for the entire period simulated. projected period (Figure 6), indicating a great loss of allelic diversity, which can endanger the long-term population persistence. The results of the simulations (Tables 5 and 6) showed that removing inbreeding from the analysis (Table 5) only reduced the probability of extinction, but the other parameters remained unchanged. This was expected, since the program models inbreeding depression as a reduction in the survival of offspring during the first year of life, When comparing the Effective Number of Founders and Ancestors for the reference population found by ENDOG (42/34) with the initial population size (25) answered by the curator of the herd, it is noticed that the value found by pedigree analysis was higher than the actual initial population. This may be due to error in animal records or the reduction of the herd size due to pressure to reduce the number of animals to increase the commercial herd. To include a value of the initial population obtained in the analysis with ENDOG in the simulation with Vortex, we chose the 12 animals supplying 50 percent of the ancestors to also check the impact of a smaller initial population. This increased the probability of extinction, but did not affect the other parameters (Table 5). The lowest percentage of females breeding each year had a strong impact on the population with a probability of extinction of 100 percent. On the other hand, raising the percentage of males available for breeding increased only the genetic variability of the population and did not affect the other parameters (Table 5). The population seemed to be viable for the simulation period of 100 years, but increasing the simulation time to 500 and years there was a probability of extinction of 99 and 100 percent, respectively (Table 5). Fieberg and Ellner (2000) found that extinction probabilities seem to be an uncertain measure, sensitive to various parameters and data quality. According to these authors, attempting to incorporate density dependence, genetic effects, interactions with other species to the models will inevitability introduce even more uncertainty. Table 5. Result summary obtained from Vortex simulations in Brazilian Bergamasca sheep considering baseline and the alternative scenarios. Scenario r GI female GI male PE Years to first extinction He Ho Alleles Baseline Without inbreeding Using ENDOG output as initial population 20% of females breeding % of males in the breeding pool Without the allelic frequency data Simulated for 500 years Simulated for 1000 years Extreme scenarios Scenario r L female L male PE Years to first extinction He Ho Alleles 70% of lamb mortality % of adult female mortality Higher catastrophe severity Higher catastrophe frequency Smaller initial population % of female lamb mortality % of male lamb mortality r, deterministic growth rate; L, generation interval; PE, probability of extinction; He, expected heterozygosity; Ho, observed heterozygosity; Alleles, estimated final number of alleles.

104 Genealogy and viability analyses for sheep conservation 109 Table 6. Result summary for the scenarios testing alternative management options in Vortex simulations in Brazilian Bergamasca sheep. Scenario r L female L male PE Years to first extinction He Ho Alleles Ten males entering Ten females entering % of catastrophe 1 frequency % of catastrophe 1 frequency % of catastrophe 1 frequency % of catastrophe 1 frequency % of catastrophe 2 frequency % of catastrophe 2 frequency % of female adult mortality % of female adult mortality % of female adult mortality % of female adult mortality % of female adult mortality % of female adult mortality % of male adult mortality % of male adult mortality % of male adult mortality % of male adult mortality % of female lamb mortality % of female lamb mortality % of female lamb mortality % of female lamb mortality % of female lamb mortality % of male lamb mortality % of male lamb mortality % of male lamb mortality % of male lamb mortality % of male lamb mortality r, deterministic growth rate; L, generation interval; PE, probability of extinction; He, expected heterozygosity; Ho, observed heterozygosity; Alleles, estimated final number of alleles. Increasing the mortality of lambs had a strong impact on the population, with a negative growth rate, increase in the generation interval and increased the probability of extinction (Table 5). Separating the mortality of male and female lambs, it is noticed that the major impact is due to the mortality of female lambs, probably due to the role of females in reproduction and the characteristic of the herd of having more females than males. Working with Vortex, Brook et al. (1997) found that small changes in mortality, potentially due to inbreeding depression, minor chronic disease, or new exotic predators, had a profound effect on the population projected future. This same author observed that the mortality of the juveniles was more important in determining the extinction probability than that of the adults. Any increase in mortality also greatly reduced the size of any extant populations. However, in this study the mortality of adult females had a greater impact on the population than the mortality of young females (Table 5). The extreme scenarios, higher mortality of lambs, higher mortality of female lambs (Table 5), affected the population growth rate, the general interval and the probability of extinction, but did not affect the genetic variability, probably because the population was extinct in the simulations before losing the genetic variability. The frequency of the catastrophe seemed to be worse than the severity (Table 5), showing that it is better to have fewer severe catastrophes than several mild disasters. From the conclusions of the extreme scenarios, we attempted to determine the extent to which the population would be still viable by changing some parameters to determine critical points in the management of these animals and suggest measures to be adopted in the conservation of this herd (Table 6). In the first two scenarios, ten male and ten female adults were added to the population every 2 years for 100 years. The entrance of this animals increased genetic variability, greatly influenced the final number of alleles, reduced the probability of extinction, but did not affect the growth rate and the generation interval. The increase in the final number of alleles was expected, since the Vortex considers that all migrants are not related to the initial population and have unique alleles. Another implication of this, according to Brook et al. (1997), as Vortex assumes all supplemented animals are unrelated; inbreeding could not be realistically modelled. Catastrophe 1, animal attack, being less severe (Table 6), even at higher frequencies did not affect the population to a great extent, once the genetic variability was maintained high and did not affect either the growth rate and generation interval. Catastrophe 2, feeding error, being more severe, rapidly affected population. Up to 25 percent did not show much effect, but from 30 percent the

105 110 H. Carneiro et al. population is already in a declining trend and at 35 percent the probability of extinction was 100 percent. The mortality of adult female (Tables 5 and 6) has a greater effect than the mortality of males. At 30 percent mortality of adult females affects the population, but it is still viable. For example, 40 percent adult male mortality does not affect most of the population parameters, but at the same 40 percent of adult female mortality the trend is a decrease in the population growth rate. At 60 percent mortality of adult females, the probability of extinction is 100 percent. The population was more resistant to female lamb mortality than to adult female mortality (Table 6). Up to 50 percent mortality of female lambs, the population was affected, but was still viable. At the 60 percent mortality the population growth rate was negative and only at 80 percent of female lamb mortality that the probability of extinction was 100 percent. On the other hand, the mortality of male lambs did not affect much, since up to 80 percent mortality and the population was still viable. Discussion Pedigree analysis This herd is part of the Brazilian National Program for Conservation of Animal Genetic Resources and the only Brazilian Bergamasca herd in the program. Of the five naturalized breeds of sheep analysed by Paiva et al. (2005), this breed had showed the lowest gene diversity and percentage of polymorphic loci. The lack of sire information is due to multiple ram mating groups, errors in recording and frequent exchange of farm helpers. Nevertheless these numbers are significantly higher than those found by Goyache et al. (2010) and Álvarez et al. (2007) for Mallorquina and by Alfonso et al. (2006) for Latxa Black-Faced sheep breed in Spain. These latter authors note that the lack of genealogies leads to computation of high generation intervals. The effective number of founders/ancestors for the reference population was 42/34 with 12 animals supplying 50 percent of the ancestors. The effective number of ancestors for the reference population of Conservation Nucleus of Galego sheep in Spain was 13 animals with five explaining 50 percent of the ancestors (Adán et al., 2007). The Brazilian Bergamasca Conservation Nucleus has therefore a higher genetic base and a low effective number of ancestors for the reference population. The low number of ancestors, which explain half the genetic variation, indicates a super usage of a low number of sires. This is due to difficulty in finding rams of this breed, distances and cost required to transport these animals when found (usually >1 000 km) as well as funding difficulties for a Federal University sheep herd. Generation intervals are similar to those found for a Conservation Nucleus of Galega (Adán et al., 2007) and Mallorquina sheep (Goyache et al., 2010) in Spain. These are on the lower limit for those calculated by Huby et al. (2003) in six French meat sheep breeds in France but higher than those found by Goyache et al. (2003) for the Xalda of Asturias breed, which is also threatened with extinction. High generation intervals are due to dams and sires remaining in reproduction for a long time. This is due to high lamb mortality (~30 percent per year) mainly caused by harsh dry winters (8 percent relative humidity and temperatures reaching 5 C at night rising to 30 C during the day) and infections caused by Haemonhcus contortus, which are resistant to antihelminthics. Moureaux et al. (1996) reported that when inbreeding using pedigree studies on horse breeds are considered, two groups of horse breeds may be distinguished: one group of international breeds, showing values ranging from 0.81to 2.89 percent, and the other group with small population sizes with values ranging from 2.25 to 14.7 percent. Although the population here was small, the inbreeding is less than found in the international breeds. This may be due to two factors: careful choice of sires in the conservation herd or lack of pedigree information (Sabbioni et al., 2007). Low AR means that an animal shares a low percentage of genes with the rest of the population; these animals could therefore serve to disseminate the breed to other farms or regions in the country. The herd size has not increased and sires are becoming more difficult and expensive to find on the national market, this reduced herd size may bring problems for the conservation nucleus in the coming years. The effective population size is an important parameter to monitor breeds and breeding programs since it is related to inbreeding depression and loss of genetic variability. According to Gandini et al. (2004) Ne is the measure of choice to determine the risk of extinction of a population. The determination of this parameter allows one to infer the rate of inbreeding and loss of genetic diversity within the population. This population kept Ne relatively stable and could recover from the sharp decrease of 1998, but probably with the difficulty of introducing new animals, the Ne could not stay stable for very long. Population viability analysis The Det.r equals the average growth of the population if the flock is large enough to avoid being affected by random or stochastic factors. The Det.r will correctly predict the population development if the birth and mortality rates are constant over time and within the expected values, if there is no inbreeding depression, no limitation or restriction on random mating and has not limiting to the carrying capacity of the environment. However, usually one or more of these assumptions are not respected, thus leading to a slightly lower population growth than expected by the deterministic growth rate (Ballou, Lacy and Miller, 2005).

106 Genealogy and viability analyses for sheep conservation 111 The Stoch.r is calculated for each year of the simulation. Usually, this parameter is lower than Det.r and only shows values close to Stoch.r if the population growth is steady and robust. The Stoch.r will be noticeably smaller than the Det.r if the population is under wide fluctuations due to environmental variation, catastrophes and demographic or genetic instability inherent to the small populations (Frankham, Ballou and Briscoe, 2008). As seen in Figure 4, there is a difference between the Det.r and stoch.r, with a wide variation of the second, supporting the idea of environmental and genetic effects on susceptibility. According to Miller and Lacy (2005), the expected heterozygosity remaining in a simulated population, in the case of this study , is a useful metric of genetic decay for comparison across scenarios and populations. From this result, it is found that the value of heterozygosity calculated by Vortex, using real allele frequency data or not, should be analysed considering only the decrease from the value of 100 percent. We do not suggest using the final value of He to confirm whether the population is threatened or not. The instability present in the population is probably due to inbreeding depression and low allelic diversity (Figure 6), since the housing environment is controlled and does not present major challenges to the animals, unless the mortality of lambs that can be considered high (around 30 percent). The genetic instability is caused possibly by the small number of animals entering the population and the use of a few sires. Population managers should concern with the variation depleting effects of genetic drift, which can be countered by the introduction of occasional immigrants to prevent deleterious inbreeding within population (Lacy, 1987). The extinction probability should not be considered alone and should not be the only factor taken into account when determining conservation strategies. It can be used to indicate the population trend of extinction or persistence and warn if conservation measures must be urgent or not. The reliable predictions of extinction probabilities can be made only for short-term time horizons, 10 to 20 percent as long as the period over which the population has been monitored (Fieberg and Ellner, 2000). These authors stated that the given typical amounts of data available it will rarely be meaningful to estimate a long-term extinction probability, but it may sometimes be possible to reliably estimate a short-term extinction probability. Considering the amount of data available in this study, neither it would be appropriate nor reliable to estimate the probability of extinction for a period longer than 100 years. According to Table 5, it is better to have fewer severe catastrophes than having several mild disasters. Keith et al. (2008), working with plants found that species populations were more viable under the less frequent catastrophe simulations, in this case, fire. The frequency and severity of catastrophes had major effects on both the probability of extinction and the final size of extant populations. When a catastrophe s frequency and severity were increased concurrently, the likelihood of extinction increases dramatically (Brook et al., 1997). Both catastrophes analysed in this study, are easily avoided, improving the feeding control by making a stock for the drought period, training the handlers and protecting the flock from attacks installing fences and arresting the animals at night, period on which most attacks have occurred. These procedures would reduce the frequency of the catastrophes, which seemed to be worse to the population (Table 5). Working with Vortex, Lindenmayer, Lacy and Pope (2000) found that in the absence of animals entering, the population rapidly collapsed. One migrant per generation is an appropriate lower limit to the amount of gene flow that is desirable, but sometimes more than one migrant per generation will be necessary to achieve the desired genetic goals (Mills and Allendorf, 1996). According to these same authors, one migrant per generation is sufficient to avoid the loss of alleles in subpopulations caused by genetic drift and will allow the allele frequencies within subpopulations to respond to local selective pressures. However, deviations from the ideal population structure will tend to compromise the effectiveness of migration relative to expectations, including the social structure and taking into account the relative reproductive success of immigrants. This could mean that more than one migrant per generation may be necessary. The acquisition of new animals for this herd is difficult as there is a lack of pure commercial flocks for purchase or exchange, the distances between farms, priorities for research and bureaucracy in trading animals with a herd belonging to a public university in Brazil. Therefore, it would be very difficult to increase the entry of animals beyond what has already occurred in the analysed period. Based on these results (Table 5), we noticed that the mortality rates of adults and lambs, males and females should always be monitored and controlled to ensure the viability of the population. The lamb mortality rate could be decreased by improving the management, especially taking better care of the new born, such as navel dipping, colostrum, protecting from the cold and wind, dam mastitis control, as well as vaccinating at the appropriate time. However, the mortality of the adult females seemed to be more critical to the population and should be closely monitored by the curators, even though it is already low, about 5 percent. Changes to mortality and fecundity were a dangerous threat to the short-term persistence of the population Brook et al. (1997) studied. Even relatively minor increases in mortality or decreases in fecundity translated into strong effects on population size and extinction probability. Conclusion While no immediate problems are observed in the conservation nucleus, reduced herd size and lack of other herds

107 112 H. Carneiro et al. of this breed could cause problems in the near future. From all of the scenarios, it is concluded that the number of females is critical to sustain the population. The key parameters to be observed in a conservation management were providing the entrance of animals, control the frequency of catastrophes, not so much the severity, and control the mortality of females, especially the adults ones (up to 30 percent). Acknowledgements The authors acknowledge to the CNPq, INCT-Pecuária, FAPDF and Finatec for financial support. Statement of interest No conflict of interest was identified in this study. References Adán, S., Fernández, M., Justo, J.R., Rivero, C.J., Rois, D. & Lama, J Análisis de la información genealógica en la raza ovina ovella galega. Arch Zootec., 56: Alderson, L Genetic conservation of domestic livestock. Wallin, CAB International, US page. Alfonso, L., Parada, A., Legarra, A., Ugarte, E. & Arana, A Effects on geneticvariability of selection against scrapie sensitivity in the Latxa black-faced sheep. Genet. Sel. Evol., 38: Álvarez, I., Fernández, I., Espinhosa, M.A., Payeras, L., Gutiérrez, J. P., Royo, L.J. & Goyache, F Análisis del libro genealógico de la raza ovina mallorquina. In Proceedings SEOC pp (available at Armstrong, E., Postiglioni, A. & Gonzáles, S Population viability analysis of the Uruguayan Creole cattle genetic reserve. Anim. Genet. Resour. Inf., 38: Ballou, J., Lacy, B. & Miller, P Population and Habitat Viability Assessment (PHVA): Briefing Book. I Maned Wolf International Workshop CENAP/IBAMA, Atibaia, SP, Brazil. Brook, B.W., Lim, L., Harden, R. & Frankham, R How secure is the Lord Howe Island woodhen? A population viability analysis using Vortex. Pacific Conserv. Biol., 3: Dalvit, C., De Marchi, M., Zanetti, E. & Cassandro, M Genetic variation and population structure of Italian native sheep breeds undergoing in situ conservation. J. Anim. Sci., 87: Fieberg, J. & Ellner, S.P When is it meaningful to estimate an extinction probability? Ecology, 81: Frankham, R., Ballou, J.D. & Briscoe, D.A Genética e extinção; Conseqüências genéticas do tamanho populacional pequeno. In: Fundamentos de Genética da Conservação, pp Ribeirão Preto, Sociedade Brasileira de Genética. Gandini, G.C., Ollivier, L., Danell, B., Distl, O., Georgoudis, A., Groeneveld, A., Martyniuk, E., Van Arendonk, J.A.M. & Woolliams, J.A Criteria to assess the degree of endangerment of livestock breeds in Europe. Livestock Prod. Sci., 91: Goyache, F., Gutiérrez, J.P., Álvarez, I., Fernández, I., Royo, L.J. & Gomez, E Genetic analysis of calf survival at different preweaning ages in beef cattle. Livestock Prod. Sci., 83: Goyache, F., Fernández, I., Espinosa, M.A., Payeras, L., Pérez-Pardal, L., Gutiérrez, J.P., Royo, L.J. & Álvarez, I Análise demográfico y genetic de la raza ovina Mallorquina. ITEA, 106:3 14. Gutiérrez, J.P. & Goyache, F A note on ENDOG: a computer program for analysing pedigree information. J. Anim. Breeding Genet., 122: Gutiérrez, J.P., Altarriba, J., Díaz, C., Quintanilla, R., Cañón, J. & Piedrafita, J Pedigree analysis of eight Spanish beef cattle breeds. Genet. Sel. Evol., 35: Gutiérrez, J.P., Royo, L.J., Àlvarez, I. & Goyache, F Molkin v. 2.0: a computer program for genetic analysis of populations using molecular coancestry information. J. Hered., 96: Hall, S.J. & Ruane, J Livestock breeds and their conservation: a global overview. Conserv. Biol., 7: Harris, R.B., Maguire, L.A. & Shaffer, M.L Samples sizes for minimum viable population estimation. Conserv. Biol., 1: Hermuche, P., Silva, N.C., Guimarães, R.F., Carvalho Júnior, O.A., Paiva, S.R., Gomes, R.A.T. & McManus, C.M Dynamics of sheep production in Brazil using principal components and maps of auto-organization characteristics. Rev. Bras. Cart., 64: Hermuche, P., Maranhão, R.L.A., Guimarães, R.F., Carvalho Júnior, O.A., Gomes, R.A.T., Paiva, S.R. & McManus, C.M Dynamics of Sheep Production in Brazil. ISPRS Int. J. Geo-Info., 2: Huby, M., Griffon, L., Moureaux, S., De Rochambeau, H., Burge, C. D. & Verrier, E Genetic variability of six French meat sheep breeds in relation to their genetic management. Genet. Sel. Evol., 35: Keith, D.A., Akçakaya, R., Thuiller, W., Midgley, G.F., Pearson, R. G., Phillips, S.J., Regan, H.M., Araújo, M.B. & Rebelo, T.G Predicting extinction risks under climate change: coupling stochastic population models with dynamic bioclimatic habitat models. Biol. Lett., 4: Lacy, R.C Loss of genetic diversity from managed populations: interacting effects of drift, mutation, immigration, selection, and population subdivision. Conserv. Biol., 1: Lacy, R.C Analysis of founder representation in pedigrees: founder equivalents and founder genome equivalents. Zoo Biol., 8: Lindenmayer, D.B., Lacy, R.C. & Pope, M.L Testing a simulation model for population viability analysis. Ecol. Appl., 10: Luhken, G., Buschmann, A., Brandt, H., Eiden, M., Groschup, M.H. & Erhardt, G Epidemiological and genetical differences between classical and atypical scrapie cases. Veter. Res., 38: Miller, P.S. & Lacy, R.C Vortex useŕs manual. A stochastic simulation of the extinction process. Chicago Zoological Society, Chicago. Mills, L.S. & Allendorf, F.W The one-migrant-per-generation rule in conservation management. Conserv. Biol., 10: Miranda, R.M. & McManus, C Desempenho de ovinos Bergamácia na região de Brasília. Rev. Brasil. Zootec., 29: Moureaux, S., Verrier, É., Ricard, A. & Mériaux, J.C Genetic variability within French race and riding horse breeds from genealogical data and blood marker polymorphisms. Genet. Sel. Evol., 28:

108 Genealogy and viability analyses for sheep conservation 113 Notter, D.R The Importance of genetic diversity in livestock populations of the future. J. Anim. Sci., 77: Paiva, S., Silvério, V.C., Egito, A.A., McManus, C., Faria, D.A., Mariante, A.S., Castro, S.R., Albuquerque, M.S.M. & Dergam, J.A Genetic variability of the Brazilian hair sheep breeds. Pesquisa Agropecuária Brasil., 40: Sabbioni, A., Valentino, B., Francesca, T.M. & Paola, S Genetic variability and population structure in the Italian Haflinger Horse from pedigree analysis. Ann. Facoltà Med. Veterin. Parma., 27: Scherf, B.D World watch list for domestic animal diversity, Third edition. Rome, Food and Agricultural Organization of the United Nations. Toro, M.A. & Caballero, A Characterization and conservation of genetics diversity in subdivided populations. Phil. Trans. R. Soc. B, 360:

109 Animal Genetic Resources, 2014, 54, Food and Agriculture Organization of the United Nations, 2014 doi: /s Population structure and genealogical analysis of the Brazilian Crioula Horse F.C. Maciel 1, C.D. Bertoli 2, J. Braccini Neto 2, J.A. Cobuci 2, S.R. Paiva 3 and C.M. McManus 2,4 1 Faculdade de Veterinária, Universidade Federal do Rio Grande do Sul, Porto Alegre, Brazil; 2 Departamento de Zootecnia, Universidade Federal do Rio Grande do Sul, Porto Alegre, Brazil; 3 EMBRAPA Recursos Genéticos e Biotecnologia, Brasília, DF , Brazil; 4 Universidade de Brasília, Brasília, DF , Brazil Summary A genealogical analysis of Crioula horses registered by the Brazilian Association of Crioula Horse Breeders was performed. The state of Rio Grande do Sul is the largest producer of animals with percent of registered animals, of which percent were males and percent females. The inbreeding coefficient was calculated at 0.88 percent and the average relatedness was 0.65 percent in the total population (animals born in Brazil and imported). Inbreeding increased over the generations. An increase was seen in the average genetic conservation index in animals born after the year 1927 (0) until 2010 (8.67 percent). The average generation interval was 10.3 years and the average age of parents when offspring born were 10.5 years, falling in recent years. The effective population size of founders was animals, the number of ancestors contributing to this population was 5086 where 56 ancestors explained 50 percent of the genetic diversity of the breed. Inbreeding is under control in the Crioula horse. The increase in registrations reflects the increased interest from farmers in this breed. Keywords: average relatedness, founding, ancestors, genetic conservation index, inbreeding, statistics of Wright Résumé Une analyse généalogique des chevaux Créoles inscrits à l Association Brésilienne d Éleveurs de Chevaux Créoles a été réalisée. L état du Rio Grande do Sul est le plus grand producteur d animaux avec 89,85 pour cent des animaux inscrits, dont le 45,29 pour cent sont mâles et le 54,71 pour cent femelles. Le coefficient de consanguinité a été estimé à 0,88 pour cent, avec une parenté moyenne de 0,65 pour cent dans la population totale, y compris les animaux nés au Brésil et ceux importés. La consanguinité a augmenté au fil des générations. Un accroissement de l indice moyen de conservation génétique a été observé pour les animaux nés entre 1927 (0) et 2010 (8,67 pour cent). L intervalle générationnel moyen a été de 10,3 ans et l âge auquel les parents ont en moyenne leur premier descendant a été de 10,5 ans, cet âge ayant diminué au cours des dernières années. La taille effective de la population des fondateurs a été de 95,19 animaux. Le nombre d ancêtres ayant contribué à cette population a été de 5086, dont 56 sont responsables du 50 pour cent de la diversité génétique de la race. La consanguinité est sous contrôle dans la race équine Créole au Brésil. L augmentation des inscriptions reflète l intérêt croissant des éleveurs pour cette race. Mots-clés: consanguinité, parenté moyenne, indice de conservation génétique, statistiques de Wright, ancêtres fondateurs, généalogie, cheval créole, Brésil Resumen Se realizó un análisis genealógico de los caballos Criollos registrados en la Asociación Brasileña de Criadores de Caballo Criollo. El estado de Río Grande del Sur es el mayor productor de animales con el 89,85 por ciento de los animales registrados, de los cuales el 45,29 por ciento son machos y el 54,71 por ciento hembras. El coeficiente de endogamia se estimó en un 0,88 por ciento, con un parentesco medio de 0,65 por ciento en la población total, contando tanto con animales nacidos en Brasil como con animales importados. La endogamia aumentó de generación en generación. Se detectó un incremento en el índice medio de conservación genética en los animales nacidos entre 1927 (0) y 2010 (8,67 por ciento). El intervalo generacional medio fue de 10,3 años y la edad media de los progenitores al nacimiento del primer descendiente fue de 10,5 años, habiéndose reducido esta edad en los últimos años. El tamaño efectivo de la población de fundadores fue de 95,19 animales. El número de ancestros que han contribuido a esta población fue de 5086, de los cuales 56 eran responsables del 50 por ciento de la diversidad genética de la raza. La endogamia está bajo control en la raza equina Criolla en Brasil. El aumento de registros refleja el interés creciente de los ganaderos por esta raza. Palabras clave: endogamia, parentesco medio, índice de conservación genética, estadísticos de Wright, ancestros fundadores, genealogía, caballo criollo, Brasil Submitted 19 April 2013; accepted 17 October 2013 Correspondence to: C.M. McManus, Universidade de Brasília, Brasília, DF , Brazil. concepta@unb.br 115

110 116 F.C. Maciel et al. Introduction Brazil has the third largest herd of horses in the world with ~5.9 million animals, second only to China with 7.9 million and Mexico with 6.2 million. The equine generates about direct jobs and if we consider indirect jobs about 3.2 million, according to ESALQ/USP (2006). The Crioula horse of southern Latin America is a direct descendant of the horses brought to the New World by the Spanish and Portuguese conquistadores during the sixteenth century (Rodero, Delgado and Rodero, 1992). Many horses escaped or were abandoned, and returned to a wild state. These were Portugese, Barbe and Spanish (particularly Andaluse) horses. Studies with Uruguayan (Kelly et al., 2002) and Argentinian Crioulas (Mirol et al., 2002) using blood group and protein polymorphism variants that are considered to be breed markers of Spanish Pure-bred and Barb horses were detected in the Crioula breed but some microsatellites and protein polymorphisms alleles were found uniquely in the Crioula horse. Argentinean Crioula horses shared two haplotypes with the Peruvian Paso from Argentina, and the commonest haplotype of the Crioula horses is identical to one of the Andalusian horses. Even when there was substantial subdivision between breeds with highly significant Wright s Fixation Index (F ST ), the parsimony and distance-based phylogenetic analyses failed to show monophyletic groups and there was no clear relationship in the trees between the South American and any of the other horses analysed. Although this result could be interpreted as mixed ancestry of the South American breeds with respect to the Spanish breeds, it is probably indicating the retention of very ancient maternal lineages in the breeds analysed. The South American horse breeds form two main groups, the Crioula types (including Brazilian, Argentinean and Uruguayan) and the other Brazilian breeds (Cothran et al., 1998) and the Crioula is most closely related to Iberian Sorraia, Andalusian, Paso Fino and Lusitano breeds. The pairing of the Sorraia with the Argentine Crioula was probably due to the combination of low variability of the Sorraia, small sample size of the Argentine Crioula and the fact that an Argentine Crioula stallion was used in the regeneration of the Sorraia. According to the Brazilian Association of Crioula Horse Breeders (ABCCC), from 2001 until the end of 2010 there was an increase of percent in applications for registration of temporary records, jumping from in 2001 to in Based on this growth, the Association estimates that this number could reach in the year 2020, an increase of approximately percent over Trade of animals increased 985 percent between 2001 and 2010, rising from $9.2 to $100 million, respectively, and this number could exceed $300 million in the year 2020, according to data from the ABCCC. According Laat (2001), population aspects of horse herds, especially those related to inbreeding and effective size, present peculiarities since the horse herds are generally small, selection does not have well-defined goals and economic characteristics to evaluate progress, and sports prizes, morphology and associated functionality are not objectively measured. Furthermore, the low uses of reproductive biotechnologies such as artificial insemination, embryo transfer and IVF which are prohibited by the Association of Crioula breeders, affect population parameters. Despite its importance, no studies on the level of inbreeding and population structure of the Crioula breed and, therefore, the objective of this study was to analyse the current situation of the breed using population genetics parameters, focusing especially on some aspects of genetic variability and population structure. Material and methods The analysis was based on data from the studbook where the population was animals that included all registered animals (final and interim) by the Brazilian Association of Crioula Horse Breeders (ABCCC) since its foundation up to 8th April 2011 ( males and females). Data were analysed using ENDOG programme to calculate population parameters for the breed. The following parameters were calculated: the pedigree completeness level was computed as the proportion of ancestors known per parental generation (MacCluer et al., 1983); The number of equivalent to discrete generations (t) for each individual in a pedigree (Boichard, Maignel and Verrier, 1997); The inbreeding coefficient (F) (Malécot, 1948); average relatedness coefficient (AR) (Goyache et al., 2003; Gutiérrez, Goyache and Cervantes, 2009). The probability of gene origin was characterized by computing the following parameters: effective number of founders ( f e ) (James, 1977) computed from the genetic contribution of founders to the descendant gene pool of the population (Lacy, 1989); effective number of ancestors ( f a ), defined as the minimum number of ancestors, not necessarily founders, explaining the complete genetic diversity of a population (Boichard, Maignel and Verrier, 1997); the founder genome equivalents (Ballou and Lacy, 1995) obtained by the inverse of twice the average coancestry of the individuals within the population (Caballero and Toro, 2000); Effective population size (N e ) was computed following Gutiérrez, Goyache and Cervantes (2009); F-statistics where a population has a population structure of two levels; one from the individual (I ) to the subpopulation (S) and one from the subpopulation to the total (T ). F-statistics describe the amount inbreeding-like effects within subpopulations F IS, among subpopulations F ST, and within the entire population F IT. In this case Wright s F-statistics are obtained according to Caballero and Toro (2000, 2002); The within-breed coancestry ( f ii ) and the between-breeds coancestry matrix

111 Population structure of the Brazilian Crioula Horse 117 ( f ij ) were computed averaging all pairwise coancestry coefficients of the individuals belonging, respectively, to a given breed i or to two different breeds i and j. Following Caballero and Toro (2000, 2002) the betweenbreeds Nei s minimum distance (D m ) matrix was also computed as D m =((f ii + f jj )/2) f ij where f ii and f jj are the average coancestry within two breeds i and j and f ij the coancestry between two breeds i and j. In the present study, municipalities, farms and sires were considered separate populations; the effective number of founders in an animal s pedigree was calculated using the genetic conservation index, genetic conservation index (GCI; Alderson, 1991): GCI =1/ΣP 2 i where P i is the proportion of genes of founder animal i in the pedigree. Subsets were also analysed using sires with more than 100 offspring (803) giving foals, the fathers of these animals (196) giving foals. Farms (106) and municipalities with more than 500 horses registered were also investigated. Two further reference populations were created families of stallions with more than 500 offspring ( animals FAMILY) and animals registered since 1998 ( animals RECENT). In the latter two populations animals appeared in both. Patterns of inbreeding in the three populations studied (ALL, RECENT, FAMILY). Levels of inbreeding were grouped (0.00, (0.05), (0.10), (0.15), (0.20), (0.25) and >0.25 (0.30). Figure 1. Geographical distribution of the population (Pop) of Crioula horses in Brazil. Results The Crioula horse is present in 23 of the 27 states of the Federative Republic of Brazil (Figure 1). The Crioula horse does not have animals born in states of Alagoas, Maranhão, Rio Grande do Norte and Amapá. The largest producers are the three southern states of the country that together add up to percent, and Rio Grande do Sul, the largest producer with percent, 3.81 percent in Paraná and Santa Catarina with 2.67 percent of the population total. Horses were also imported from Argentina (866), Chile (868) and Uruguay (852) of which ~80 percent were females. While horses from Argentina and Uruguay have been imported since the start of breed registration, those from Chile have been more recent becoming the most important source of importation (Figure 2). Of all registered animals, percent have their pedigrees fully known. The number of known pedigrees has increased over the generations with emphasis on the growth of the ninth generation. In the following, six generations this increased from 4.95 percent of the known pedigrees to percent in the current 15th generation (Figure 3). The increase in the number of records per year is evident (Figure 4). Since the 1970s there has been a steady growth in the breed until the mid-1990s, more precisely in Before 1980, 9.3 percent of all animals were registered. Figure 2. Number of Crioula horses imported per year from Uruguay, Argentina and Chile. In this year, there was a decrease in the number of records of animals to From 1996 to 1998 the number of records was almost stable, and from 1999 there was a Figure 3. Percentage of known generation pedigrees of the Crioula horse in Brazil.

112 118 F.C. Maciel et al. Figure 4. Percentage of records by year of birth for the Crioula horse in Brazil. higher growth than seen in the 1970s and 1980s. From the year 2005, growth has intensified, with an increase in 3042 animals from 2006 to The numbers of records for the years 2010 and 2011 should be disregarded, since the data were collected in April 2011 and most of the animals born in the spring of 2010 onwards had not been through technical inspection on the date on which the data were gathered. Of the animals recorded by the ABCCC, percent are female and percent are males, in a total of observations. As for the month of birth, percent of registered animals are born between September and January, the largest concentration in the months of October and November where we find, respectively, and percent of births (Figure 5). Farms with the largest number of animals were registered for Estância Nazareth (3930), Cabanha Tupambaé (3114), Figure 5. Percentage of records by month of birth for the Crioula horse in Brazil. Cabanha Paineiras (2994), Cabanha St. Angelo (2370) and Estância San Francisco (2252), and these five farms, from a total of 7949, were responsible for 4.5 percent of registered animals. When you consider the 50 largest farms, they, together, represent percent of the animals reared in the country, reaching for a total of animals. The inbreeding coefficient was calculated at 0.88 percent and the AR was 0.65 percent in the total population (animals born in Brazil and imported). With the increase of generations, an increase in inbreeding was observed (Table 1). This increase is evident from the first generation, and especially observed in the fourth generation full generation percent of the animals which have some degree of inbreeding and inbreeding coefficient in this generation was 2.78 percent. There is an increasing trend up to 100 percent of animals in the 15th generation being inbred, Table 1. Inbreeding (F), average relatedness (AR) and effective population size (Ne) per generation in Brazilian Crioula Horse. Generation No. of animals F Mean (%) % Endogamic animals F Mean for endogamic animals % Mean AR Ne Endogamy per generation Endogamy per complete Generation Maximum number of generations traced. 2 Complete generations traced as defined by Gutiérrez, Goyache and Cervantes (2009).

113 Population structure of the Brazilian Crioula Horse 119 Table 2. Average number of generations, inbreeding per generation and effective population size for the Brazilian Crioula horse. Mean generations Increase in endogamy (%) Effective population size All Maximum Complete Equivalent Family Maximum 7, Complete Equivalent which is not complete since the animals can be registered definitively up to 4 years after birth. The number of generations, complete and equivalent, shows that the breed is expanding (Table 2). Large variations in effective herd size per generation are an indication of animals being imported into the herd (for example, from Uruguay, Chile or Argentina). The highest level of inbreeding was 0.41 with 6755 animals with a value >0.1. These animals were on average in the seventh generation, but there were animals in the second generation with high levels of inbreeding. Analysing the full herd, 72 (0.02 percent) animals were from crosses between full-sibs, 3047 (0.89 percent) between half-sibs and 1784 (0.52 percent) between parents and offspring. The mean coefficient of individual relatedness showed an increase in population over the years, especially after the1950s, reaching 2.09 percent in animals born in the year 2010 (Figure 6). An increase in average individual relatedness was found from the 1940s until 1970, the largest observation in 1955 was 3.11 percent. In recent years, there was an average increase of inbreeding about 0.50 percent (Figure 7). An increase in the average individual coefficient of inbreeding was found in animals born after the 1940s, and in 1955 an average rate of 2.95 percent was seen. Figure 7. Genetic conservation index (GCI) and average relatedness (AR) average per year of birth for the Brazilian Crioula horse. After 1970 there was a reduction in the coefficient of average individual inbreeding, and the animals born in the past decade had rates of ~1.50 percent (Figure 7). An increase in the average rate of genetic conservation was seen in animals born after the year 1927 until 2010, which had the highest average GCI, 8.67 percent (Figure 7). The interval between generations (Table 3) was higher for mares than stallions. This indicates that mares are kept longer for reproduction or start reproductive life later. There was no significant genetic differentiation between the states of the federation where animals are raised when considered sub-populations (not shown). Average coancestry within subpopulations was and average coancestry in the metapopulation was Wright parameters show that the inbreeding in the population is greater than expected if matings were performed at random and which occur between the mating subpopulations. The effective population size of founders was animals, expected inbreeding due to an imbalance in the contribution of founders was calculated as 0.53 percent and the mean inbreeding coefficient computed was 1.21 Table 3. Generation interval and the average age of parents when their offspring born for the Brazilian Crioula horse. Relation N L Se Generation interval Sire Son Sire daughter Dam son Dam daughter Total Mean age of parents when offspring born Sire Son Sire daughter Dam son Dam daughter Total Figure 6. Mean individual inbreeding (F) and Average individual change in inbreeding (AF) by year of birth for the Brazilian Crioula horse. N is the number of animals, L is the generation interval and SE is the standard error.

114 120 F.C. Maciel et al. Table 4. Measures of genetic variation and Wright s statistics (F IS, F ST and F IT ) for Brazilian Crioula Horse. Parameter All Recent Family Mean coancestry within Subpopulations Autocoancestry Endogamy Nei s Distance Mean Coancestry in Metapopulation F IS F ST F IT percent for the population of animals and the base population (animals with one or more parents unknown) was 7901 (Table 4). The ratio of founders to ancestors was 1.21 (91/75) meaning that in general no specific lineages were created within the Crioula population. The reference population (both parents known) was smaller than that used for the analysis of founders. The number of ancestors contributing to this population was 5086 for a reference population of animals, where 56 ancestors explained 50 percent of the genetic diversity of the breed. Animals 9838, 5174, , 5178, 5532, 5497, 5208, 5418, 5439 and 5449 had the largest contribution to the herd, reaching ~25 percent of the total genetic variation. The largest number of offspring per stallion was 1428 (sire ) with four animals (9838, and 5208) with more than 1000 offspring. Animals and are offspring of 9838, and eight stallions had more than 500 offspring (forming the second reference population FAMILY). Animal 9838 had the highest average relatedness. Mare had the most offspring (20), but more than mares had more than ten offspring, showing the prolificacy and longevity of the breed. The total population of animals was divided by their state of birth and these subpopulations were analysed for type (Table 5) and structure (Table 6). Only Commercial and Multiplier herds were seen, with no nuclei or disconnected herds. Most multiplier states (14/ 15) use their own breeding males, buy in animals and sell sires to other states. As for commercial herds, Table 5. Number of herds by type and characteristics for the Brazilian Crioula horse. Type Buy-in sires Use own sires Sell sires No. of states % Sires bought Nucleus No Yes Yes Multiplier Yes Yes Yes Multiplier Yes No Yes Commercial Yes Yes No Commercial Yes No No Isolated No Yes No Table 6. Structure of the Brazilian Crioula Horse herds by state. State 1 Region Type % parents born in state/ number of births % males born in state which are sires/number of sires Acre N Commercial Bahia NE Multiplier Ceará NE Commercial Federal CO Multiplier District Espirito SE Commercial Santo Goiás CO Multiplier Minas Gerais SE Commercial Mato Grosso CO Multiplier do Sul Mato Grosso CO Multiplier Pará N Commercial Pernumbuco NE Commercial Piauí NE Multiplier Paraná S Multiplier Rio de SE Commercial Janeiro Rondônia N Commercial Roraima N Commercial Rio Grande S Multiplier do Sul Santa S Multiplier Catarina Sergipe NE Commercial São Paulo SE Multiplier Tocantins N Multiplier Brazilian States; N, North; NE, Northeast; S, South; SE, southeast; CO, Midwest. percent (7/11) did not use their own sires for reproduction, while percent (4/11) using both purchased and own males for reproduction. Rio Grande do Sul state (RS) had the highest percentage of sires born in the State, which is not surprising as this is the home state for this breed. Even when subpopulations are studied, genetic parameters (F, AR, AF) vary little from the mean (Table 7), showing Table 7. Genetic variation by subpopulation (sires and grandsires with largest number of offspring, farms and municipalities) for the Brazilian Crioula horse. Grandsire Sire Farms Municipalities Number Offspring F AR Gen Max Gen Com Gen Eq AF GCI Genetic conservation index (GCI), average relatedness (AR), mean individual inbreeding (F) and average individual inbreeding (AF), GenMax, maximum number of generations; Gen_Com, common number of generations; Gen_Eq, equivalent number of generations.

115 Population structure of the Brazilian Crioula Horse 121 Table 8. Population data depending on recently born (from 1998) or family of sires with largest offspring numbers (family) in Crioula horses from Brazil. All Recent Family No. of reference animals No. of ancestors Effective number of founders Effective number of ancestors 75 (1.21) 45 (1.38) 35 (1.40) No. of ancestors explaining 50% Base population (one or more unknown parents) 7901 Actual Base Population (one unknown parent = half founder) Effective Population Size of Founders Expected F by unbalancing of founders contribution 0.53% Computed mean F 1.21% Mean Average Relatedness 1.58% Age first becoming parent Regression coefficient of age becoming parent on inbreeding coefficient ± ± ± 1.00 Regression coefficient of age becoming parent on increase in inbreeding coefficient ± ± ± 2.25 Ne from regression birth date Ne from log regression birth date Ne via individual increase in inbreeding ± ± Ne from Log regression on equivalent generations Ne via regression on equivalent generations Ne, effective population size. Figure 8. Percentage of animals registered per year depending on inbreeding level on all (ALL) animals registered, those from the major families (FAMILY) and the present generation (RECENT)Levels of inbreeding (0.00 (0.0), (0.05), (0.10), (0.15), (0.20), (0.25) and greater than 0.25 (0.30). Figure 9. Mean inbreeding (F) of Crioula Horse depending on the date the farm started to register animals, the period of time the farm registered animals and the number of animals registered per farm. that sires are imported into farms as a general rule. GCI was significantly higher than herd average for those stallions that left more offspring, as expected. More recently animals are becoming parents at a relatively younger age. This may be a reflection of the rapid expansion of the breed and the need for breeding females to reproduce (Table 8). Effective population sizes were similar for different subgroups. This is a reflection of the fact that the major sire families produce a large proportion of the recently born animals, reflected by the large number of animals in common in both groups. It can be seen that the more recently farms started to register animals the higher mean F (Figure 8, Figure 9). This is also reflected in the longer the farm has been registering animals and the higher the number of animals registered the lower the mean F.

116 122 F.C. Maciel et al. Discussion Due to their high commercial value, the horses, especially those used for sports, need have their pedigree known, both for registration and purchase. Confirmation of genealogy is extremely important, not only for ensuring the ascendancy of the animals, but also because a pedigree can reliably enable the buyer to identify the origin of genetic problems in the herd and reduce or eliminate them (Coelho and Oliveira, 2008). The study of genetic variability and population structure in some horse breeds using pedigree analysis, alone or in combination with information such as genetic markers, has grown in recent years (Chiofalo et al., 2003; Valera et al., 2005). The results on the distribution of this breed in Brazil were in accordance with expected, with a high concentration of animals in the southern region, since the breed was developed in Southern Latin America, a region known as South American Pampa (Argentina, Southern Brazil and Uruguay). The increase in the use of the Crioula Horse in southeastern states such as Sao Paulo, as well as in Northern and Northeastern Brazil is also seen. This reflects the expansion and distribution of the breed, where the largest breeder (Estância Nazareth) has only 1.2 percent of the animals, unlike other local breeds, such as the Pantaneiro Horse, where the largest breeder has 7.25 percent of animals recorded. The amount of information on the Crioula has increased with the number of generations. These data are consistent with those found in studies with other breeds where the most distant generations have less genealogical information, as this may have been lost over the years or were not recorded (Valera et al., 2005). Vicente, Carolino and Gama (2009), as here, found an increase in complete Lusitano pedigrees over ten generations. The increase in the number of records per year is evident (Figure 2), mainly due to two factors: (a) partnership between the Breeder s Association and the Rural TV Channel, where the most important competition involving the Crioula, Freio de Ouro (which involves several tests, such as breed standard, gait, barrel racing, separating cattle, lassoing, horse control and obedience), is televised across the country and worldwide over the internet. This channel also televises most of the auctions with these horses, so farmers from distant regions can acquire animals without leaving home; (b) projects developed by the ABCCC, such as Vaquejada Project (A northeastern tradition where horses are used to drive cattle in a confined space.), created to insert Crioula horse in this activity in the northeast. The number of registers in the last two years (2010 and 2011) should be disregarded as the animals are subjected to technical inspection until they are nine months old and most of the animals born in this period had not passed inspection at the time of data collection. The greater number of females compared to males can be explained by fact that some breeders do not register males which do not meet their specifications. In domestic animals some individuals assume importance in the origin of the breed and its development (Fletcher, 1945, 1946; Rhoad and Kleberg, 1946; Gazder, 1954) as evidenced in this population. This is due to several factors, such as propaganda, success in auction rings and competitions as well as functional and morphological traits. The concentration of births in the months from September to January is mainly due to the occurrence of marked photoperiod in southern Brazil (Winter, 2007). Most mares show anestrus and transitional phases in summer and autumn and, under natural conditions, the transition ends near or after the equinox, in September and October. The mean coefficient of inbreeding found in Crioula was lower compared to other local breeds. Schurink, Arts and Ducro (2012) found average inbreeding coefficient of while the Crioula horse was found 0.88 percent for the total population (including imported) and 1.21 percent for those born in Brazil. Wolc and Balińska (2010) also found much higher values (5.9, 5.1 and 5.6 percent) on three farms breed Polish Konik. Abrahão et al. (2002) found similar levels of inbreeding (0.8 percent) to those found in Crioula when studying inbreeding in Thoroughbred mares reared in Brazil. Low levels of inbreeding found in Crioula in Brazil can be explained by the large population base, large number of founders and ancestors, and the introduction of animals from other countries (Argentina, Uruguay and Chile), introducing different racial strains. The number of inbred animals found in the full third generation (80.99 percent of the population) was similar to that found by Mota et al. (2006) found that 88.0 percent of inbred animals in Mangalarga horses in the period The results in Crioula show a clear increase in the number of inbred animals, reaching 100 percent at the 15th generation, although there are only two animals in this generation. Inbreeding of 10 percent was estimated for ethnic breeds such as Spanish Purebred (Valera et al., 2005) and Lipizzano (Zechner et al., 2002), and after the closing of the studbook this parameter has aggravated (Vicente et al., 2009). For Mangalarga inbred animals, the mean coefficient of inbreeding was 5.7 percent, with a maximum of 46.9 percent (Costa et al., 2005), similar to those found by Procópio et al. (2003) in the Campolina. The increase of inbreeding (Boichard, Maignel and Verrier, 1997) can be used to derive the effective size of a population. However, this method reflects mainly the long-term effects of selection and, furthermore, is very sensitive to incomplete information from a pedigree. An alternative would be additional parameters based on the probability of gene origin, and the effective number of founders and effective number of remaining genomes are commonly used in wild populations, but less common in studies of domestic animals. However, there are several studies in domestic species and different breeds, such as in cattle (Faria et al., 2002; Vercesi Filho et al., 2002; Pereira et al., 2005; Vozzi et al., 2006, 2007; Hammami

117 Population structure of the Brazilian Crioula Horse 123 et al., 2007; Martínez, García, and Gallego, 2008), in horses (Costa et al., 2005), sheep (Goyache et al., 2003), pigs (Toro et al., 2000) and even donkeys (Gutiérrez et al., 2005), based on some sort of pedigree analysis. Ideal levels of inbreeding range from 0.05 (Nicholas, 1989) to 0.01 (FAO, 2000) per generation, and the Crioula are below these levels. Moureaux et al. (1996) reported that when studying inbreeding in horse breeds, two groups can be distinguished: a group of international breeds, with values ranging from 0.81 to 2.89 percent, and another group with small population sizes with values ranging from 2.25 to 14.7 percent. The high variation in mean individual levels of inbreeding and inbreeding rate seen here between 1940 and 1970 may have been a reflection of the lack of breeding animals, the difficulty in finding stallions of the breed as well as in moving between farms, since the increase in effective herd size came after 1970 (Figure 3). The mean coefficient of relatedness (kinship) increased from the start of breed registration until about the year These results can be explained by selection of animals which showed excellent results in morphological and functional competitions, and therefore many breeders sought to acquire its offspring. Mean coefficient of relatedness provides additional information to explain relations between relatives (Gutiérrez, Goyache and Cervantes, 2009). In conservation genetics, knowledge of relatedness is required to optimize conservation strategies. Coancestry relationship is expressed relative to the base population in which all alleles are not defined as being identical by descent, so that the population coancestry base is zero by definition (Falconer and Mackay, 1996; and Lynch Walsh, 1998). Kinship average here was low compared to other breeds of horse (Zechner et al., 2002; Valera et al., 2005). Valera et al. (2005) found average values of inbreeding and AR of the Andalusian horse population were, respectively 8.48 and percent, while Cervantes et al. (2008) found inbreeding average of 7 percent to Spanish Arabian. The average individual inbreeding and variation in inbreeding increased from the mid-1930s to mid-1970s, and remained relatively constant thereafter. This can be explained by an intense exchange of animals between Brazil, Argentina, Uruguay, and especially Chile, where different strains were incorporated into the local population after the 1970s. When comparing inbreeding in the three populations studied (ALL, FAMILY and RECENT), a clear increase in the percentage of animals with relatively low levels of inbreeding (0.05). The percentage of animals with F > 0.05 has not grown since the late 1990s and although an increase in inbreeding (F > 0.1) was seen in the ALL population in the 1940s to 1980, this has largely disappeared. Higher levels of inbreeding are not seen in RECENT or FAMILY. The percentage of animals with no inbreeding has steadily decreased in all populations and should increase with the closure of the herd book. In Andalusian horses, Valera et al. (2005) found 39.6 effective founders and 27 ancestors, with only six animals responsible for 50 percent of genetic variability in the breed. The situation of Crioula is better, with nearly 95 founders, but this is lower than for Pantaneira (McManus et al. 2013), perhaps due to rapid expansion of the breed in recent years. The small number of founder animals highlights the need for monitoring of inbreeding of the herd studied, control of breeding and introduction of animals that have no direct relationship with the principal ancestors identified in this study. Alderson (1991) used the GCI to calculate an effective number of founders in the pedigree of an animal. The higher the value GCI better it is for use in conservation of a breed. The GCI can be used both by individual breeders as an aid in selecting breeding stock, or within a breed to formulate a comprehensive improvement program. However, the index has limitations, as it is not representative of any concentration of breeding animals for non-founders in subsequent generations and cannot be used without pedigree records (Alderson, 1991). The interval between generations found in Crioula was higher when compared with other breeds such as Costa (2002) and Gonçalves (2010), who evaluated animals Mangalarga Marchador found generation interval of 8.9 and 8.98 years, respectively. This may be due to the prohibition of use of reproductive biotechnologies until 2011, as well as the large number of morphological and functional competitions in which these horses compete, means that these horses begin reproductive life at a later age. As natural breeding is still used, many mares exposed to the stallion do not impregnate. Vicente, Carolino and Gama (2009) also found a high generation interval for the Lusitano breed (10.4 years: males 11.2 and females 9.6 years), also possibly due to use after the end of their sporting career. McManus (2013) found lower intervals between generations in the Pantaneiro horse and considered this was because this breed is used on-farm animals instead of sports and therefore enter reproduction earlier. The use of artificial insemination and embryo transfer are beginning to become more widely used in horse populations (2011 in Crioula) but less frequently than in other farm animals (Laat, 2001). The population aspects of horses, especially those related to inbreeding and effective size therefore have an individual character (Costa et al., 2005) and continued monitoring is necessary to avoid future problems. Vinocur et al. (2003) determined the allele frequencies of seven blood group systems and eight protein systems in six herds of Crioula horses raised in Rio Grande do Sul of State. The herds presented a significant component due to isolation (F ST =0.0866, p < 0.01). They found high values for average heterozygosity (0.4631). When all herds were considered in the analysis, the inbreeding level (F IS ) was zero, in line with seen here where values near zero indicate that levels of genetic variability of the flock

118 124 F.C. Maciel et al. are high, since these values indicate an excess of heterozygous animals, both within and paternal lineages (subpopulations) and for the population as a whole (metapopulation). For the F ST index, the value calculated was 0.000, demonstrating that there is no differentiation between the parental strains with the formation of subpopulations. The Crioula is not at risk status (FAO 1998; Database DAD-IS) as opposed to other breeds of naturalized horses (Luís et al., 2005). The integration of herdbooks from other countries with Crioula horses used for crossing with the Brazilian Crioula (such as Uruguay, Argentina and Chile) could help to identify some genetic bottlenecks that may exist but were not seen here. The fact that the smaller, younger farms have higher levels of inbreeding may reflect a policy of the larger farmers to import stallions or sell off inbred animals to those starting up in the business. It may also reflect a lack of breeding policy by the smaller farmers who may have a single stallion. This is also from the case seen with some pig breeds in the USA (Welsh et al., 2010) but to a lesser extent. Conclusion The indices obtained from the genetic data of genealogical record of the Crioula Breed show that in general inbreeding is under control in the total population and the effective population size is not at a critical level. Recent decrease in generation intervals may be a reflection of the rapid expansion of the breed and the need for mares for reproduction. Acknowledgements Thanks are due to the Breeder s Association for the Crioula Horse, INCT Pecuária (CNPq, FAPEMIG, MCT) and CAPES for scholarships. Statement of interest No conflicting interests were identified. References Abrahão, A.R., Mota, M.D.S., Oliveira, H.N. & Madureira, A.P Endogamia em éguas da raça Puro-Sangue Inglês. In Proceedings IV Simpósio da Sociedade Brasileira de Melhoramento Animal, Campo Grande, Brazil, pp Alderson, G.L.H A system to maximize the maintenance of genetic variability in small populations. In L. Alderson, ed. Genetic conservation of domestic livestock, pp Wallingford, UK, CAB International. Ballou, J.D. & Lacy, R.C Identifying genetically important individuals for management of genetic diversity in pedigreed populations. In: Ballou, J.D., Gilpin, M. & Foose, T.J. (Eds) Population management for survival & recovery. Analytical methods and strategies in small population conservation. New York: Columbia University Press, p Boichard, D., Maignel, L. & Verrier, É The value of using probabilities of gene origin to measure genetic variability in a population. Genet. Sel. Evol., 29: Caballero, A. & Toro, M.A Interrelations between effective population size and other pedigree tools for the management of conserved populations. Genet. Res., 75: Caballero, A. & Toro, M.A Analysis of genetic diversity for the management of conserved subdivided populations. Conserv. Genet., 3: Cervantes, I., Molina, A., Goyache, F., Gutiérrez, J. P. & Valera, M Population history and genetic variability in the Spanish Arab Horse assessed via pedigree analysis. Livestock Sci., 113: Chiofalo, L., Portolano, B., Liottal, L., Rundo Soteral, A. & Finocchiaro, R Demographic characterization, inbreeding and genetic variability within Sanfratellano population horse from genealogical data. Ital. J. Anim. Sci., 2(Suppl. 1): Coelho, E.G.A. & Oliveira, D.A.A Testes genéticos na eqüideocultura. Revista Brasileira de Zootecnia, 37: Special Number. Costa, M.D Caracterização demográfica e estrutura genética da raça Mangalarga Marchador. Universidade Federal de Minas Gerais, Belo Horizonte, Brazil (Thesis). Costa, M.D., Bergmann, J.A.G., Resende, A.S.C. & Fonseca, C.G Análise temporal da endogamia e do tamanho efetivo da população de equinos da raça Mangalarga Marchador. Arquivo Brasileiro de Medicina Veterinária e Zootecnia, 57: Cothran, E.G., Santos, S.A., Mazza, M.C.M., Lear, T.L. & Sereno, J. R.B Genetics of the Pantaneiro horse of the Pantanal region of Brazil. Genet. Mol. Biol. (available at S ). Escola Superior de Agricultura Luiz de Queiroz ESALQ. Centro de Estudos Avançados em Economia Aplicada da ESALQ Estudo do Complexo do Agronegócio do Cavalo no Brasil. Brasília, CAN, MAPA. Falconer, D.S. & Mackay, T.F.C Introduction to Quantitative Genetics. Harlow, Longman. Faria, F.J.C., Vercesi Filho, A.E., Madalena, F.E. & Josahkian, L.A Estrutura populacional da raça Nelore Mocho. Arquivos Brasileiro de Medicina Veterinária e Zootecnia, 54: Fletcher, J.L A genetic analysis of the American Quarter Horse. J. Hered., 36: Fletcher, J.L A study of the first fifty years of Tennessee Walking horse breeding. J. Hered., 37: Gazder, P.J The genetic history of the Arabian Horse. J. Hered., 45: Gonçalves, R.W Efeito da endogamia nas características morfométricas e reprodutivas dos eqüinos da raça Mangalarga Marchador. Universidade Estadual de Montes Claros, Montes Claros, Brazil. (Dissertation) (available at ). Goyache, F., Gutiérrez, J.P., Fernández, I., Gomez, E., Alvarez, I., Díez, J. & Royo, L.J Using pedigree information to monitor genetic variability of endangered populations: the Xalda sheep breed of Asturias as an example. J. Anim. Breed. Genet., 120: Gutiérrez, J.P., Marmi, J., Goyache, F. & Jordana, J Pedigree information reveals moderate to high levels of inbreeding and a

119 Population structure of the Brazilian Crioula Horse 125 weak population structure in the endangered Catalonian donkey breed. J. Anim. Breed. Genet., 122: Gutiérrez, J.P., Goyache, F. & Cervantes, I Useŕs Guide, ENDOG v4.6, A Computer Program for Monitoring Genetic Variability of Populations Using Pedigree Information, (available at ucm.es/info/prodanim/html/jp_web_archivos/en_us_g_.pdf). Hammami, H., Croquet, C., Stoll, J., Rekik, B. & Gengler, N Genetic diversity and joint-pedigree analysis of two importing Holstein populations. J. Dairy Sci., 90: James, J.W A note on selection differentials and generation length when generations overlap. Anim. Prod., 24: Kelly, L., Postiglioni, A., De Andrés, D.F., Vega-Plá, J.L., Gagliardi, R., Biagetti, R. & Franco, J Genetic characterisation of the Uruguayan Crioula horse and analysis of relationships among horse breeds. Res. Vet. Sci., 72(1): Laat, D.M Contribuição genética de fundadores e ancestrais na raça Campolina. Universidade Federal de Minas Gerais, Belo Horizonte, Brazil. (Dissertation) (available at com.br/produto/contribuicao_genetica_de_fundadores_e_ancestrais_ na_raca_campolina-10151). Lacy, R.C Analysis of founder representation in pedigrees: founder equivalents and founder genome equivalents. Zoo Biol., 8: Luís, C., Cothran, E.G., Oom, M.M. & Bailey, E Major histocompatibility complex locus DRA polymorphism in the endangered Sorraia horse and related breeds. J. Anim. Breed. Genet., 122: Lynch, M. & Walsh, J.B Genetics and analysis of quantitative traits. Sunderland: Sinauer Assocs. Inc., 980p. Maccluer, J.W., Boyce, A.J., Dyke, B., Weitkamp, L.R., Pfenning, D.W. & Parsons, C.J Inbreeding and pedigree structure in Standardbred. J. Hered., 74: Malécot, G Les Mathématiques de l Hérédité. Masson et Cie, Paris Martínez, R.A., García, D., Gallego, J.L Genetic variability in Colombian Crioula cattle populations estimated by pedigree information. J. Anim. Sci., 86: McManus, C Evaluation of Conservation Program for the Pantaneiro Horse in Brazil, Revista Brasileira de Zootecnia, in press. McManus, C., Santos, S.A., Dallago, B.S.L., Paiva, S.R., Saraiva, R. F., Braccini Neto, J., Marques, P.R. & Abreu, U.G Evaluation of conservation program for the Pantaneiro horse in Brazil. Rev. Bras. Zoot. 42: Mirol, P.M., Peral García, P., Vega-Pla, J.L. & Dulout, F.N Phylogenetic relationships of Argentinean Crioula horses and other South American and Spanish breeds inferred from mitochondrial DNA sequences. Anim. Genet., 33: Mota, M.D.S., Prado, R.S.A. & Sobreiro, J Characterization of the Mangalarga horse population in Brazil. Arch. Zoot., 55: Moureaux, S., Verrier, É., Ricard, A. & Mériaux, J.C Genetic variability within French race and riding horse breeds from genealogical data and blood marker polymorphisms. Genet. Sel. Evol. 28: Nicholas, F.W Incorporation of new reproductive technology in genetic improvement programs. In: Hill, W.G., Mackay, T.F.C. (Eds) Evolution and animal breeding. Wallingford: CAB International, p Pereira, M.C., Mercadante, M.E.Z., Albuquerque, L.G. & Razook, A. G Estimativa de ganho genético a partir de diferenciais de seleção e parâmetros populacionais em um rebanho Caracu. Revista Brasileira de Zootecnia, 34: Procópio, A.M., Bergmann, J.A.G., Costa, M.D Formação e demografia da raça Campolina. Arq. Bras. Med.Vet. Zoot. 55: Rhoad, A.O. & Kleberg, J.R The developments of a superior family in the modern Quarter Horse. J. Hered., 37: Rodero, A., Delgado, J.V. & Rodero, E Primitive Andalusian livestock and their implication in the Discovery of America. Arch. Zootec., 41: Schurink, A., Arts, D.J.G. & Ducro, B.J Genetic diversity in the Dutch harness horse population using pedigree analysis. Livestock Sci., 143: Toro, M.A., Rodrigañez, J., Silio, L. & Rodriguez, C Genealogical analysis of a closed herd of Black Hairless Iberian pigs. Conserv. Biol., 14: Valera, M., Molina, A., Gutiérrez, J.P., Gómez, J. & Goyache, F Pedigree analysis in the Andalusian horse: population structure, genetic variability and influence of the Carthusian strain. Livestock Prod. Sci., 95: Vercesi Filho, A.E., Faria, F.J.C., Madalena, F.E. & Josahkian, L.A Estrutura populacional do rebanho Tabapuã registrado no Brasil. Arquivos Brasileiro de Medicina Veterinária e Zootecnia, 54: Vicente, A., Carolino, N. & Gama, L.T Indicadores Demográficos no Cavalo Lusitano. Arch. Zootec., 58(Supl. 1): Vinocur, M.E., Brass, K.E., Rubin, M.I.B. & Silva, C.A.M Genetic variability in the Brazilian Crioula horse breed. Cienc. Rural, 33: Vozzi, P.A., Marcondes, C.R., Magnabosco, C.U., Bezerra, L.A.F. & Lôbo, R.B Structure and genetic variability in Nellore (Bos indicus) cattle by pedigree analysis. Genet. Mol. Biol., 29: Vozzi, P.A., Marcondes, C.R., Bezerra, L.A.F. & Lôbo, R.B Pedigree analyses in the Breeding Program for Nellore Cattle. Genet. Mol. Res., 6: Welsh, C.S., Stewart, T.S. & Schwab, C Pedigree analysis for 5 swine breeds in the United States and the implications for genetic conservation. J. Anim. Sci., 88: Winter, G.H.Z Características reprodutivas sazonais da égua Crioula em um propriedade latitude S no Rio Grande do Sul. Universidade Federal de Santa Maria, Brazil. (Dissertation) (available at Wolc, A., Balińska, K Inbreeding effects on exterior traits in Polish konik horses. Arch. Anim. Breed., 53: 1 8. Zechner, P., Sölkner, J., Bodo, I., Druml, T., Baumung, R., Achmann, R., Marti, E. Habe, F. & Brem, G Analysis of diversity and population structure in the Lipizzan horse breed based on pedigree information. Livestock Prod. Sci., 77:

120 Animal Genetic Resources, 2014, 54, Food and Agriculture Organization of the United Nations, 2014 doi: /s Pedigree and population viability analyses of a conservation herd of Moura pig H. Carneiro 1, S.R. Paiva 2, M. Ledur 3, E.A.P. Figueiredo 3, V.H. Grings 3, F.C.P. Silva 1 and C. McManus 1 1 Faculdade de Agronomia e Medicina Veterinária, Brasília, DF, Brazil; 2 EMBRAPA Recursos Genéticos e Biotecnologia, Brasília, DF, Brazil; 3 EMBRAPA Suínos e Aves, Concordia, SC Summary Pedigree and population viability analyses of a conservation herd of Moura breed from Embrapa Swine and Poultry were carried out. ENDOG software was used for pedigree analysis and VORTEX for population viability analysis. The data were obtained with the curator of the herd and a total of 411 animals born between the years of 1999 and 2011 for the analysis with ENDOG. The input data in VORTEX were obtained from a questionnaire answered by the curator of the herd. The Inbreeding Coefficient (F) was 0.06 percent, the average effective population size (Ne) was 30 animals with an average generation interval of 3.48 years. The extinction probability was 0 percent for a simulation period of 100 years and 3 percent for a period of 500 years. The determinant parameters for the viability of this herd were the entry of animals and prolificacy of the breed. Although the study did not indicate immediate risk of extinction of this herd, it should be constantly monitored with controlled population parameters to prevent the extinction of this last herd of Moura pig breed. Keywords: inbreeding, conservation population modelling, Moura pig Résumé Les analyses de la viabilité de la population et pedigree ont été effectuées sur un harde de la race Moura de l Embrapa Porcins et Oiseaux. ENDOG logiciel a été utilisé pour l analyse de pedigree et VORTEX pour l analyse de la viabilité de la population. Les données ont été obtenues avec le conservateur du harde et un total de 411 animaux nés entre les années 1999 et 2011 ont été analysés avec ENDOG. Les données d entrée dans VORTEX ont été obtenus a partir d un questionnaire auquel ont répondu par le conservateur du harde. Le coefficient de consanguinité (F) était 0,06%, la taille effective de la population moyenne (Ne) était 30 animaux, avec un intervalle de 3,48 ans en moyenne de la production. La probabilité d extinction était de 0% pour une période de simulation de 100 ans, et 3% pour une période de 500 ans. Les paramètres déterminant pour la viabilité de ce harde a été l introduction des animaux et la prolificité de la race. Bien que l étude ne mentionne pas le immédiat extinction, la dernière race de porcs Moura doit être contrôlée en permanence avec les paramètres de population pour éviter l extinction. Mots-clés: endogamie, modèles de conservation de la population, porcs Moura Resumen Se realizó el análisis de pedigrí y de viabilidad de la población de un rebaño de la raza Moura Embrapa Suínos y Aves. Se utilizó el Software ENDOG para el análisis de pedigree y VORTEX para el análisis de viabilidad de la población. Los datos se obtuvieron con el curador de lo rebaño y un total de 411 animales nacidos entre los años 1999 y 2011 se analizaron con ENDOG. Los datos de entrada en VORTEX se obtuvieron de un cuestionario contestado por el curador de lo rebaño. El coeficiente de endogamia (F) fue de 0.06%, el tamaño medio efectivo de la población (Ne) fue de 30 animales, con un intervalo medio de generación de 3,48 años. La probabilidad de extinción fue de 0% para un período de simulación de 100 años y 3% durante un período de 500 años. Los parámetros más importantes para la determinación de la viabilidad de este rebaño fuerón el ingreso de animales y prolificidad de la raza. Aunque el estudio no indica peligro inmediato de extinción, el último rebaño de la raza Moura debe ser monitoreado constantemente con los parámetros de población para evitar la extinción. Palabras clave: Endogamia, modelos de conservación de la población, cerdos Moura Submitted 23 July 2013; accepted 16 October 2013 Introduction Correspondence to: C. McManus, Faculdade de Agronomia e Medicina Veterinária, Campus Universitário Darcy Ribeiro, Brasília, DF , Brazil. concepta@unb.br Pigs were introduced in Brazil since the sixtieth century during the colonization period. The animals probably originated from the Iberian Peninsula (Cavalcanti, 2000). The variety of environment conditions found in farms in 127

121 128 H. Carneiro et al. different regions of Brazil resulted in a wide variety of breeds that for centuries sustained livestock production in the country (Mariante et al., 2009). Naturalized swine produce a higher fat to meat ratio than commercial breeds, so there was little interest in using them for improvement programmes. However, in rural areas these were the preferred animals, not only due to the large production of lard but also because of the taste of their meat. They are adapted to different ecosystems in different regions of Brazil and have a high capacity for survival in unfavourable conditions. The naturalized pig breeds of Brazil are characterized by their toughness, resistance to diseases, low management requirements and feeding as well as a high adaptability. These traits make them an important genetic resource to be used in breeding programmes and research, thus justifying efforts for their conservation (Castro, Albuquerque and Germano, 2002). Eight naturalized Brazilian swine breeds have already been described in the literature: Canastra, Canastrão, Nilo, Piau, Caruncho, Pirapetinga, Moura and Tatu. Currently, there is no standardization in the phenotypic characterization of pig breeds as well as their denomination leading to a possible loss of genetic diversity. On the other hand, these breeds can also be considered as a potential source of new allelic combinations of extreme importance for the future of the Brazilian pig industry (Sollero, Paiva and Faria, 2009). However, the lack of data recording and specifying traits of zootechnical interest coupled with the large distances between farms and the small size of the breeding population makes it difficult to effectively work with naturalized swine herds (Garcia and Barbosa, 2005). Crossbreeding and selection can rapidly modify characteristics of a breed, giving rise to different genetic groups and complicating breed identification. Although molecular techniques have aided in the identification and characterization of individuals and breeds, phenotypic characterization is also necessary. A few studies have been carried out on the diversity of naturalized swine breeds in South American countries and subjective data are frequently missing for the most naturalized breeds of farm animals (McManus et al., 2010). The Moura breed is a primary source of edible fat or lard. In addition, its relatively efficient conversion of food into weight gain makes it a useful cross with modern breeds. These features make this breed a good alternative for subsistence farming in the various regions of Brazil (Castro et al., 2001). The Moura is usually blue roan in colouration, occasionally red roan (Mason, 1996). The breed emerged in southern Brazil (Machado, 1967), although the precise period remains unknown. This breed was widely used in the farms of the region before the introduction of the breeds such as Duroc, Landrace, Large White and others (Fávero et al., 2007). The first initiatives aiming the formation of breeding herds for selection programmes of Moura pigs occurred in the state of Paraná, initially led by the Universidade Federal do Paraná in 1985, and later by the Company Café do Paraná. The first genealogical records of the breed were issued by the Brazilian Association of Swine Breeders (ABCS) in 1990, the year that marked the opening of Brazilian Pig Book (PBB) of the Moura breed. Two other breeders, also in Paraná, started the records of animals since the year The records issued by ABCS stopped in But returned in 2002, when Embrapa Swine and Poultry created a pure herd using animals from the Universidade Federal do Paraná, and this persists until today. One thousand, five hundred and twenty eight (1528) births were registered as pure animals in the ABCS from 2003 to 2011 (ABCS, 2011). Currently, there are records on animals in the municipalities of Teutônia, Roca Sales, Cruzeiro do Sul, Venâncio Aires and Carlos Barbosa. In the state of Santa Catarina, there are reports of the occurrence of the breed in the region of Lages. In Paraná state, the records of occurrence are in the municipalities of Bituruna and Roncador. In addition to these states, boars were sent to the Southern, UNESP (Ilha Solteira-SP), and Noreasthern, Universidade Federal do Piauí (Teresina-PI), regions to compose pure herd. The aim of this study was to monitor the conservation programme of the herd of Moura pigs maintained by Embrapa Swine and Poultry, located in the state of Santa Catarina (Brazil) through pedigree and population viability analysis. Materials and methods Experimental design Animals Pig data from the conservation herd of Moura pig breed maintained by Embrapa Swine and Poultry, located in Concordia, Santa Catarina, Brazil were analysed. The herd comprised a pure breeding herd of about 20 females and ten males of different families. These families are crossed with each other to expand the genetic base for sire and dam replacement. When there is no need to replace animals, females are crossed with boars of MS115 strain, a third-generation light pig developed by Embrapa, and the piglets sold for finishing. The data from all purebred pigs for pedigree and population viability analyses were obtained directly from the curators of the Moura herd. Analysis of pedigree The genetic population parameters were analysed using the ENDOG program, version 4.8 (Gutiérrez and Goyache, 2005). The information included pedigrees of animals born between 1999 and 2011, totalling 411 animals: 255 females and 156 males. Two analyses were performed, one with all the animals of the pedigree and another considering only the animals that

122 Moura Pig genetic resources conservation 129 Table 1. Allele frequencies in the conservation herd of Moura pig from Embrapa Swine and Poultry. Locus Allele frequency Number of alleles per locus S S SW S SW S SW SW S OPN SW SW S S SW S SW are currently part of the herd. For this purpose, animals born after 2006 were used, totalling 364 records. Analysis of population viability The known distribution of allele frequencies was obtained by microsatellite marker analysed in Animal Genetics Laboratory LGA, Embrapa Cenargen, coming from the genotypic analysis of 35 animals (Table 1). Simulations of interactions between demographic, environmental and genetic factors were made in the program VORTEX Population Viability Analysis Software, version 9.99 (Miller and Lacy, 2005). Table 2 shows the demographic parameters used to simulate the Moura pig herd, obtained mainly from the answers of the curators of the conservation herds linked to the National Programme for Conservation of Animal Genetic Resources. The base scenario was used in the VORTEX model. VORTEX models population dynamics as discrete sequential events that occur according to the distribution of the random variables informed. A population was simulated as undergoing a series of events that describe the annual cycle of a sexually reproducing organism: mate selection, reproduction, mortality, age increase each year, migration, among others. At the end of the simulations, the programme provides: (1) the probability of extinction of the population; (2) the mean time to extinction if the population is extinguished in at least 50 percent of the simulations; and (3) the average size and genetic variation within the remaining population. Loss of genetic diversity is estimated by simulating the transmission of alleles from parents to offspring in a hypothetical neutral locus. The programme monitors how original alleles remain in the population, the average heterozygosity and the genetic diversity relative to the initial level. To determine which factors mostly affect the population, 26 new scenarios were created by changing just one of the parameters of the base scenario (Table 3). Table 2. Demographic parameters input in VORTEX used to simulate the population of Moura pigs. Input data in VORTEX Results Value Number of interactions 500 Number of years 100 Reproduction system Polygamous Age of first reproduction (F/M) 1/1 Inbreeding depression? Yes Maximum number of farrowing per year 2 Maximum size of the litter 11/farrowing Mean sexual rate at birth 50% (males) Maximum age of reproduction 5 years Adult females in reproduction (%) 80 Female probability having zero farrowing 2 Female probability having one farrowing 40 Female probability having two farrowing 58 Mortality of females aged between 0 and 1 (%) 8 Mortality of adult females >1 (%) 5 Mortality of males aged between 0 and 1 (%) 8 Mortality of adult males >1 5 Catastrophe 1: not determined by curator Frequency (%) 1 Effect on reproduction (%) 5% Effect of the survival (%) 10% Male reproduction 10 Initial population 12 Support capacity of the environment (K) 120 Animals are removed? Yes 80 males and 50 females per year Animals entering the population? Yes 20 females and three males per year The reference population was formed by parents of known animals. The pedigree information with the identification of parents was 85 percent. The values found for the effective number of founders (fe), effective number of ancestors (fa), average relatedness (AR), coefficient of inbreeding

123 130 H. Carneiro et al. Table 3. Population viability analysis of the Moura herd, considering the base and additional scenarios. r L females L males PE Final population He Ho N alleles Base scenario No entry Two catastrophes % males reproduction Lower prolificacy % females reproduction % mortality of piglets % mortality of piglets % mortality of piglets % mortality of piglets females More severe catastrophe Fewer animals entering Greater frequency catastrophe % mortality of piglets females Greater adult mortality No inbreeding depression No genetic data years Initial population ENDOG Breed lower More animals coming out Most one farrowing per year Lower reproduction. more output and less input More output and less entry of animals Lower reproduction and more output of animals Lower reproduction and less input of animals r, rate of population growth; L, interval between generations; PE, probability of extinction; He, expected heterozygosity; Ho, observed heterozygosity; N alleles, The average number of alleles at the end of the simulation period. (F) and mean index of genetic conservation (ICG) for the reference population with pedigree are listed in Table 4. The coefficient of inbreeding was 0.06 percent, which is considered low for the herd. The reference population (animals with both parents known) was smaller than that used for analysing the founders. From a total of 411 animals in the pedigree, 353 were included in the reference population. The base population, with at least one known parent, had 58 animals. When considering only the current herd, the base population fell to 52 animals. The number of ancestors that contributed to the reference population was 31. Analysing only the current herd, the reference population was 364 animals and 35 ancestors. There are a few generations available for evaluating the Moura herd. This is common in conservation programmes, since the interest in the conservation of these animals is recent, and insufficient to obtain detailed information on Table 4. Population parameters from ENDOG pedigree analysis in Moura pigs. Parameter Full dataset Present herd Effective number of founders (fe) Effective number of ancestral (fa) Average relatedness (AR) 8.38% 8.38% Coefficient of inbreeding (F) 0.06% 0.06% Mean index of genetic conservation (ICG) their genealogy. The average number of complete generations observed was 1.32, the average maximum number of generations observed was 2.64 and the average equivalent complete generations observed was The interval between generations of the conservation herd from the Moura pigs and average ages at birth are shown in Table 5. In the pedigree analysis, there were 27 records of animals born in The next birth was registered only in 2004, indicating gaps in the registry of pedigree. The generation interval in Moura pigs was obtained in the simulation with VORTEX. The interval between generations refers to the average value at the end of the 100 Table 5. Interval between generations of the conservation herd from the Moura pigs (in years) and average ages at birth. Generation interval Full population Present population Sire offspring (male) Sire offspring (female) Dam offspring (male) Dam offspring (female) Overall mean Average ages at birth Boar offspring (male) Boar offspring (female) Dam offspring (male) Dam offspring (female) Overall mean

124 Moura Pig genetic resources conservation 131 years simulation and was close to the reality of the species. The generation interval in male was 1.77, in female 1.77 and the average size of the final population was The fixation index (F-statistics) describes the level of heterozygosity in the population. The F statistics shows how inbreeding influences within the subpopulation (F IS ), among subpopulations (F ST ) and in the whole population (F IT ). The values found for the Moura herd were: F IS = , F ST = 0, and F IT = The zero value indicates no population subdivision. The Ne averaged by year of birth of the animals was 30 and decreased in the period analysed. The herd size did not increase, probably due to lack of market for the products, the absence of replacements and to the structural difficulties in maintaining a larger herd. This population size may cause problems in the viability of this herd. Factors threatening the herd and possible management strategies for conservation were tested in the new scenarios in the PVA to determine the major threats. The average final herd size was very close to the environmental carrying capacity (K), indicating the growth capacity of the herd, considering the rearing conditions. The probability of extinction was 0 based on 500 iterations performed in the VORTEX, indicating no extinction of the herd. The herd increased rapidly to environmental carrying capacity (K) with an average deterministic growth rate (Det.r) of Figure 1 shows the population growth comparison between deterministic (Det.r) and stochastic (Stoch.r) simulation. The Det.r assumes no inbreeding depression, no mating restrictions and no entry or exit to the animal population. Therefore, it simulates how the herd would react without any fluctuation due to chance or environment. The Stoch.r considers the occurrence of environmental events at random. Figures 2 and 3 show the comparison of the evolution of inbreeding and the final average number of alleles, at different scenarios, respectively. Figure 4 shows the comparison of the growth rate, Stoch r. between different scenarios of mortality in piglets and Figure 5 shows the comparison of the evolution of genetic diversity and inbreeding in scenario 500 years. Discussion The parameters that support population viability are the high prolificacy of the breed and the entry of new animals. The number of animals added to the herd was considered high. Evaluating the number of births in each year added to the entry of animals and subtracting the animals from the population, intentionally or by mortality, it was noticed an important increase of animals every year in this herd. Scenarios were created by changing only the factors: prolificacy and fecundity; entry of animals and animals leaving the population. In scenarios with fewer offspring, more animals produced and more than one farrowing per year, the values of the maximum number of offspring, number of animals leaving the herd and number of Figure 1. Comparison between growth rates Det r. and Stoch r. Figure 2. Comparison of the evolution of inbreeding in different scenarios.

125 132 H. Carneiro et al. Figure 3. Comparison of the evolution of the final average number of alleles at different scenarios. Figure 4. Comparison of the growth rate, Stoch r. between different scenarios of mortality in piglets. Figure 5. Comparison of the evolution of genetic diversity and inbreeding in scenario 500 years. farrowings per year changed. The scenario that most impacted the herd was of the maximum number of farrowings per year. The other parameters were not affected in relation to the base scenario. Other scenarios were that there was a decrease in the entry of animals caused the extinction of the herd in all simulations. It was noted, therefore, that maintaining the entry of animals in the herd is important for its persistence. This result corroborated with Fimland (2007), who also noted that the most effective way to maintain genetic variability is the immigration of individuals carrying alleles from different populations. Hamann and Distl (2008) found the 19 immigrants per generation were sufficient to maintain the inbreeding within the acceptable level. The coefficient of inbreeding observed was 0.06 percent (Table 4). Estimating the inbreeding coefficient is complex because it usually requires knowledge of the pedigree of individuals. In addition, the calculation of the coefficient from the pedigree information requires the specification of the degree of relationship between the founders (Hedrick and Kalinowski, 2000). According to Toro Rodriganez and Rodriguez (2000), for the calculation of the inbreeding coefficient, it is preferable to use pedigree information

126 Moura Pig genetic resources conservation 133 instead of using the molecular marker information. Molecular marker can be useful to verify, correct or even complete pedigree data. The pedigree analysis always begins with a founder population, in other words, individual ancestry unknown and presumably unrelated, from which the population is descended from pedigree (Lacy, 1989). When the founder population information is not available (as here), it is assumed that the founders are not related and are not endogamous (Hedrick and Kalinowski, 2000). Sollero Paiva and Faria (2009) observed that the Moura and Monteiro had the lowest genetic diversity among native Brazilian pig breeds in terms of the effective number of alleles and observed heterozygosity. Pedigree analysis is always recommended to allow planning to maximize the effective size of the reproduction in relation to the total population size, and consequently, decrease the variability in the effects of genetic drift (Lacy, 1989). A few generations were available for the evaluation of the Moura herd, and time was insufficient to obtain detailed information of their genealogy. The intervals between generations (Table 5) were considered high for the species compared with studies of Toro Rodriganez and Rodriguez (2000), who found that a generation interval of 2.44 years for Iberian pigs. This is probably because the animals from the Moura herd were not used for commercial purposes and the substitution in the herd only took place when they became ill or died. The herd size has not increased during the time period studied, probably due to the absence of available replacements for the parents and structural difficulties in maintaining a larger herd. There was a small difference between Det.r and Stoch.r, which could indicate low environmental influences on population (Figure 1). However, the answers given by the herd curators were considered very optimistic, which could explain the positive results when the base scenario was analysed alone. Stoch.r will be smaller than Det.r if the population suffers wide fluctuations in environment, catastrophes or genetic instabilities, as well as due to the demographics of small populations (Miller and Lacy, 2005), as seen here. To demonstrate the changing pattern of inbreeding coefficient of the herd, the scenario without animal entry was observed. The maximum value was attained in the first 10 years of the simulation (data not shown). The best performance of this parameter (Figure 2) was when 100 percent of males were available for reproduction. No significant difference was seen between the base scenario and that without the true population genetic data. The inbreeding coefficient is calculated in each year of the VORTEX simulation and, if no true allele frequency is provided, the programme determines a unique allele for each animal in the initial population. The data from the pedigree analysis with ENDOG, even though not very deep, are more accurate. VORTEX considers inbreeding depression solely as increased mortality of animals up to 1 year of age without considering decline in fertility and prolificacy. This may explain the low difference found between the base scenario and that without inbreeding depression, as the mortality rate of piglets reported was low (8 percent). Figure 3 shows that there was no significant difference in the mean number of alleles between the base scenario without inbreeding depression, and that with the analysis of true genetic information from population. The exception was for the scenario of no entry of animals, whose population rapidly went into extinction and, therefore, the final average number of alleles was 0. When analysed the impact of increased probability of reproduction in males and females, was observed that increasing the number of males in reproduction had more impact on the final average number of alleles than the increase in the number of females (data not shown). This can be explained by two factors. The information provided by the herd curator who reported that 80 percent of females and 10 percent males were available for reproduction. The second is the method of calculating these parameters by VORTEX. It was considered that every female of reproductive age has 80 percent chance of reproducing each year. As for the males, this was only 10 percent. Therefore, any increase in male number will have a greater impact than female. Analysing the mortality rates of the base scenario it was noted that a higher mortality rate of piglets affected the rate of population growth, the interval between the generations and the final number of animals. These parameters do not alter the probability of extinction, heterozygosity or number of alleles. The mortality of piglets had a more severe effect on the population than adult mortality, and especially mortality of female piglets (Figure 4). However, mortality of animals up to 1 year informed by the curator was low, indicating that the conservation herd has good management of the offspring. It was observed that increasing the number of catastrophes reduced the rate of growth and the final number of animals in the population, but did not affect the number of alleles and heterozygosity. Changing other scenarios such as the frequency and severity of the catastrophe, had greater effects on the population. The higher frequency of catastrophes had greater impact, with more effect on genetic diversity. The highest frequency of one catastrophe affected the population more than the lower frequency and severity of two catastrophes (data not shown). Thus, it was inferred that minimizing the consequences of frequent catastrophes in the herd conservation is more important than specific effects of more severe catastrophes. When the viability of the population was tested for a period of 500 years, it was noted that the probability of extinction was small (3 percent), but the population was seriously compromised at the end of this period, when looking at the final number of alleles and heterozygosity. The other parameters were not affected. Figure 5 shows the evolution of genetic diversity and inbreeding coefficient in this period, and it is seen that in the first 100 years these parameters remain relatively stable. From the year 100, these

127 134 H. Carneiro et al. indices are seriously affected. This result demonstrates the ability of the pigs to adapt to the negative effects of small population size for a period, but after a certain point, this control is no longer possible and the population is seriously compromised. According to Rangel (2012), a species does not go extinct immediately after the reduction of habitat or climate change beyond the tolerance limit of the animals. Extinction can occur several generations after the initial impact on the population. This initial period, until effective extinction, creates a window of opportunity for conservation, in which it is still possible to ensure the viability of the population (Wearn Reuman and Ewers 2012). Simulations performed for a shorter period of time may not be sufficient to predict what conservation strategy is needed to ensure persistence of the population. A population may seem feasible when the simulation considers only 100 years, but the risk of extinction may increase considerably if a longer period is considered, as observed in Figure 5. Simulations should therefore be over a long period of time (over 100 years) or performed for two different periods and check if the results found in the short-term simulation remain in the long term (Nilson, 2003). Conclusion The characterized Moura herd has low population size. Only a few breeders or producers are interested in this breed, a usual situation in production of Animal Genetic Resources in Brazil. In the present study, this small effective population size may cause problems for the population in the near future, despite the inbreeding coefficient being considered low. The herd curators should continue to import animals from other breeders as this is essential for the viability of this conservation herd, as well as care should be taken to maintain the reproductive rate of the animals. In conservation programmes with small herds of domestic breeds, it is important to evaluate the risk of extinction and which management activities should be prioritized to reduce this risk. However, no specific risk factor is identified in this study. Acknowledgments The authors acknowledge CNPq for scholarships and financial aid and INCT-Pecuária Informação Genético Sanitária da Pecuária Brasileira (CNPq/MCT/FAPEMIG), as well as the farmers and researchers who contributed with data and access to the animals. References ABCS Associação Brasileira de Criadores de Suínos Relatório anual do registro genealógico e provas zootécnicas de suínos. Available at the website Castro, A.M.M.G., Alvares, E.P., Zocoller-Seno, M.C., Neves, M.F. & Passipiéri, M Morphological study of the small intestine of Moura pigs (Sus scrofa Lineaus, 1758) during fetal development. Braz. J. Morphol. Sci. 8: Castro, S.T.R., Albuquerque, M.S.M. & Germano, J.L Census of Brazilian naturalizes swine breeds. Arch. Zootec. 51: Cavalcanti, S.S Suinocultura Dinâmica. Rome, Fep Mvz Ed, Contagem. 494 pp. Fávero, J.A., Figueiredo, E.P., Fedalto, L.M. & Woloszyn, N A raça de suínos Moura como alternativa para a produção agroecológica de carne. Rev. Bras. Agroecol. 2: 1. Fimland, E Genetic diversity and sustainable management of animal genetic resources, globally. Anim. Genetic Resources Inf. 41: Garcia, S.K. & Barbosa, A.S Conservação e Estudo de raças suínas brasileiras na UFMG 12 anos. I Simpósio Brasileiro de Melhoramento Animal, 295. Gutiérrez, J.P. & Goyache, F A note on ENDOG: a computer program for analysing pedigree information. J. Anim. Breed Genet. 122(3): Hamann, H. & Distl, O Genetic variability in Hanoverian Warmblood horses using pedigree analysis. J. Anim. Sci. 86: Hedrick, P.W. & Kalinowski, S.T Inbreeding depression in conservation biology. Annu. Rev. Ecol. Syst. 31: Lacy, R.C Analysis of founder representation in pedigrees: founder equivalents and founder genome equivalents. Zoo Biol. 8: Machado, L.C.P Os suínos. Porto Alegre-RS, Editora A granja Ltda. 622p. Mariante, A.S., Albuquerque, M.S.M., Egito, A.A., McManus, C., Lopes, M.A. & Paiva, S.R Present status of the conservation of livestock genetic resources in Brazil. Livestock Sci. 120: Mason, I.L A World dictionary of livestock breeds, types and varieties, 4th edition. C.A.B International, Wallingford, UK 273. McManus, C., Paiva, S.R.P., Silva, A.V.R., Murata, L.S., Louvandini, H., Cubillos, G.P.B., Castro, G., Martinez, R.A., Dellacasa, M.S.L. and Perez, J.E Phenotypic characterization of naturalized swine breeds in Brazil, Uruguay and Colombia. Braz. Arch. Biol. Technol. 53: Miller, P.S. & Lacy, R.C Vortex User s Manual. A stochastic simulation of the extinction process. Version Brookfield, IL: Chicago Zoological Society. Nilson, T Integrating effects of hunting policy, catastrophic events, and inbreeding depression, in PVA simulation: the Scandinavian wolf population as an example. Biol. Conserv. 115: Rangel, T.F Amazonian extinction debts. Science 337: Sollero, B.P., Paiva, S.R. & Faria, D.A Genetic diversity of Brazilian pig breeds evidenced by microsatellite markers. Livestock Sci. 13: 1 8. Toro, M.A., Rodriganez, L.S. & Rodriguez, C Genealogical analysis of a closed herd of black hairless Iberian pigs. Conserv. Biol. 14: Wearn, O.R., Reuman, D.C. & Ewers, R.M Extinction debt and windows of conservation opportunity in the Brazilian Amazon. Science 337:

128 Animal Genetic Resources, 2014, 54, Food and Agriculture Organization of the United Nations, 2014 doi: /s Evaluación del semen criopreservado de toros Curraleiro Pé Duro, perteneciente al Banco Brasilero de Germoplasma Animal I.S. Montesinos 1, J.O. Carvalho 2, J.V. Malaquias 3, E. Arnhold 1, G.E. Freneau 1, M.A.N. Dode 2, M.C.S. Fioravanti 1 y J.R.B. Sereno 3 1 Universidad Federal de Goiás, CEP Goiânia-GO, Brasil; 2 Embrapa Cenargen, CEP , Brasilia-DF, Brasil; 3 Embrapa Cerrados, CEP , Planaltina-DF, Brasil Resumen Esta investigación tuvo el objetivo de evaluar la calidad del semen criopreservado de 21 toros Curraleiro Pé Duro en banco de germoplasma, categorizando las dosis de los reproductores según parámetros relacionados a la viabilidad espermática, y verificando la capacidad fecundante in vitro de los espermatozoides. Todas las muestras fueron descongeladas y analizadas en cuanto a morfología, cinética, integridad de membrana plasmática e integridad de acrosoma. Para evaluar las tasas de fecundación, muestras de sólo cuatro toros fueron utilizadas. Los resultados indicaron que las dosis de semen Curraleiro Pé Duro, almacenadas en el Banco Brasilero de Germoplasma Animal, mostraron calidad espermática variable y alta capacidad fecundante in vitro, posibilitando delinear estrategias adecuadas de reproducción asistida para la conservación de este recurso genético brasilero en peligro de extinción. Palabras clave: conservación ex situ, espermatozoide, razas criollas Summary This investigation aimed to evaluate the quality of cryopreserved semen from 21 Curraleiro Pé Duro bulls in a germplasm bank, categorizing the doses of the sires according to parameters related to sperm viability, and verifying the fertilizing capacity in vitro of the spermatozoids. All the samples were thawed and analyzed as to morphology, kinetics, plasma membrane integrity and acrosome integrity. Samples of only four bulls were used for evaluating the fertilization rates. The results indicated that the Curraleiro Pé Duro semen doses, stocked in the Brazilian Animal Germplasm Bank, showed variable sperm quality and high in vitro fertilizing capacity, making it possible to design appropriate strategies for assisted reproduction in order to conserve this Brazilian genetic resource, which is in danger of extinction. Keywords: creole breeds, ex situ conservation, spermatozoid Résumé Cette recherche visait à évaluer la qualité du sperme cryoconservés de 21 taureaux Curraleiro Pé Duro dans une banques de gènes, categorisant les doses des reproduteurs selon les paramètres liés à la viabilité du sperme, et de vérifier la capacité fecondante in vitro des spermatozoïdes. Tous les échantillons ont été décongelés et analysés pour la morphologie, la cinétique et la membrane plasmique et intégrité de l acrosome. Pour évaluer le taux de fecondation, des échantillons de seulement quatre taureaux ont été utilisés. Les résultats ont indiqués que les doses de sperme Curraleiro Pé Duro, stokée dans la Banque Bresilienne de Gènes Animal, ont montré de haute qualité du sperme variable capacité de fécondation in vitro, permettant délimiter les stratégies appropriés de reproduction assistée pour la préservation de cette ressource génétique bresilienne en voie de disparition. Mots-clés: créoles races, la conservation ex situ, spermatozoïde Presentado: 25 Agosto 2013; aceptado: 27 Febrero 2014 Introducción El bovino Curraleiro Pé Duro es una raza criolla brasilera, originaria del Bos taurus ibericus, traído por los colonizadores europeos en el siglo XVI, y sometido a constante Correspondencia: I.S. Montesinos, Universidad Federal de Goiás, Brasil. ivansm.vet@gmail.com selección natural y adaptación por más de 400 años. Este taurino tropical representa un tesoro genético, ya que a pesar de no tener niveles superiores de producción, es rústico, resistente al calor, enfermedades y parásitos, y además aprovecha eficientemente pastos de baja calidad. Debido a la importación de razas especializadas con mayor desempeño zootécnico, el Curraleiro Pé Duro fue descartado de los rebaños, lo que casi provocó su 135

129 136 I.S. Montesinos et al. desaparecimiento (Mariante et al., 2009). Gracias a programas de conservación aún existen ejemplares, en el Centro Oeste y Noreste brasilero, registrando animales en el año 2008, distribuidos entre los estados de Pará, Piauí, Tocantins y Goiás (Fioravanti et al., 2011). La conservación consta de dos estrategias: una es la in situ, donde los animales permanecen en ambientes naturales o artificiales, y otra la ex situ, criopreservando en bancos de germoplasma los espermatozoides, ovocitos, embriones y otras células, como fuente de multiplicación, regeneración y distribución de una población animal en extinción (Mariante y Egito, 2002). Según Dode y Rumpf (2002) técnicas de reproducción asistida son utilizadas en conservación animal, destacando la inseminación artificial (IA), transferencia de embriones (TE), producción in vitro de embriones (PIV) e inyección intra-citoplasmática de espermatozoides (ICSI). Desde 1983, Embrapa Recursos Genéticos y Biotecnología (CENARGEN) realiza la conservación ex situ de razas criollas brasileras, criopreservando diferentes células en el Banco Brasileiro de Germoplasma Animal (BBGA) (Mariante y Egito, 2002). Para asegurar el futuro del bovino Curraleiro Pé Duro (Figura 1) se necesitan más estudios enfocando parámetros reproductivos, por lo que objetivamos evaluar la calidad del semen criopreservado de 21 toros de esta raza, perteneciente al BBGA. Material y métodos Esta investigación fue realizada en el año 2011, en las instalaciones del laboratorio de reproducción animal CENARGEN (DF). Se utilizaron pajillas de semen de 21 toros Curraleiro Pé Duro, donadas por el BBGA en número de dos unidades por reproductor, excepto cinco toros, con diez dosis cada uno. Este material genético se criopreservó entre 1994 y 2008, mediante protocolos convencionales desarrollados por el CENARGEN. Las dosis fueron descongeladas por 30 segundos en baño María (37 C), y el semen depositado en un tubo para evaluación in vitro. Figura 1. Toro Curraleiro Pé Duro en el bioma Cerrado. Evaluación de morfología espermática La morfología espermática fue evaluada de acuerdo con Barth y Oko (1989), utilizando microscopio de contraste de fase (Nikon Eclipse E200) con inmersión y aumento de 1000x. Se contaron 200 espermatozoides por muestra, clasificándolos en normales y con defectos (cabeza o cola), siendo expresado en porcentajes. Evaluación de cinética espermática La cinética espermática fue evaluada por el Computer Assisted Sperm Analysis (CASA), modelo IVOS Ultimate 12 de la Hamilton Thorne Biosciences, ajustado (set up) para semen bovino. Se colocó 10 μl de semen en la cámara de Makler, con selección de tres campos de lectura, evaluando: motilidad total (MT, %) y progresiva (MP, %); velocidad de trayecto (VAP, μm/s), rectilínea (VSL, μm/s) y curvilínea (VCL, μm/s); amplitud lateral de cabeza (ALH, μm); frecuencia de batimientos de cola (BCF, Hz); rectilinearidad (STR, %); linearidad (LIN, %) y células rápidas (RPD, %). Evaluación de integridad de la membrana plasmática y acrosoma La integridad de membrana plasmática fue evaluada utilizando diacetato de 6 carboxifluoresceína (C-FDA) y yoduro de propidio (PI) (Harrison y Vickers, 1990). Una muestra de semen (10 μl) fue diluida en solución colorante (40 μl), compuesta por 480 μl de citrato de sodio a 3%; 5 μl de formol salina a 1,6%; 5 μl de IP 0,75 mm; y 10 μl de C-FDA 0,46 mg/ml en DMSO. Los espermatozoides teñidos de rojo se consideraron con membrana plasmática dañada y los de verde con membrana íntegra. Para evaluar la integridad acrosomal se utilizó la conjugación isotiocianato de fluoresceína (FITC) con lectina Peanut aglutinin (PNA) y PI (Klinc y Rath, 2007). Una muestra de semen (10 μl) fue diluida en solución colorante (30 ml), con los mismos componentes usados en la evaluación de membrana plasmática, excepto C-FDA que fue substituido por 10 μl de FITC-PNA (1 mg/ml en PBS). Las células teñidas de rojo se contabilizaron como muertas. En los espermatozoides vivos el acrosoma teñido de verde fue considerado como reaccionado, y el acrosoma sin fluorescencia, íntegro. Ambas evaluaciones se realizaron en microscopio de epifluorescencia (Axiophot Zeiss: filtro de longitud de onda 395/420 nm y 494/517 nm excitación/emisión). Aproximadamente 200 células fueron examinadas en cada muestra y los resultados se expresaron en porcentajes. Evaluación de tasas de fecundación Para evaluar las tasas de fecundación fueron utilizadas dosis de sólo cuatro toros Curraleiro Pé Duro, debido a la escasez de pajillas disponibles. Se utilizó la metodología propuesta por Carvalho et al. (2010), con tres replicas independientes, donde a 18 horas post

130 Evaluación del semen criopreservado de toros Curraleiro Pé Duro 137 inseminación in vitro los posibles cigotos fueron retirados de las gotas de fecundación y transferidos para una gota de 200 μl de tampón fosfato (PBS). Estos fueron desnudados y fijados en solución de etanol y ácido acético glacial (3:1), por 24 horas. Después se lavaron y tiñieron con solución lacmoide a 1% en ácido acético glacial 45%. La lectura de las láminas fue en microscopio de contraste de fase bajo inmersión, con aumento de 1000x (Nikon Eclipse E200). Se utilizó 336 ovocitos, siendo clasificados en no fecundados (presencia de cromatina femenina y ausencia de masculina) y fecundados (presencia de cromatina femenina y espermatozoide en el citoplasma, presencia de cabeza descondensada, pronúcleo, clivaje y polispermia). Análisis estadístico Los datos obtenidos en los exámenes espermáticos de las 21 dosis fueron analizados a través del programa R (R Development Core Team, 2011). Primero se realizó estadística descriptiva (promedio y desvío estándar). Después se escogieron parámetros para la construcción de un dendrograma, mediante el método UPGMA (Unweighted Pair Group Method with Arithmetic Means), utilizando la distancia Euclidiana padronizada. El Cluster formó tres grupos, los cuales fueron comparados en cada examen espermático a través de análisis de varianza y test Duncan, a nivel de significancia del 5%. Las tasas de fecundación y estados de cromatina 18 horas post FIV se compararon por el test Chi-cuadrado. Resultados y discusión Evaluación de la calidad espermática Morfología espermática Considerando 70% como mínimo de espermatozoides normales para semen descongelado bovino (CBRA, 1998), las 21 dosis Curraleiro Pé Duro presentaron buen promedio (81,1% ± 13,3%). Individualmente tres dosis fueron inferiores: DT1 (60,2%); DT10 (37,5%) y DT17 (68,2%). Los defectos mayores (DMA) en una muestra seminal no deben superar el 20% (CBRA, 1998), a pesar de haber encontrado cinco dosis con alto porcentaje (DT1, 34,7%; DT7, 24,3%; DT10, 39,5%; DT16, 22,5% y DT17, 28,1%), el promedio general de DMA era 14,7% ± 9,9% y los defectos menores 4,2% ± 7,2%. Este promedio fue mejor al encontrado por Sousa et al. (2004) en ocho toros Curraleiro Pé Duro (21,1% ± 8,4%). Ciertos DMA superaron el límite de 5% que relata Freneau (2011) para anormalidades individuales: cráter (DT1, 21,8%; DT16, 9,0%); contorno anormal (DT1, 5,6%); gota proximal (DT10, 19,0%); cola fuertemente doblada (DT11, 7,2%); cola fuertemente doblada con gota (DT7, 7,1%; DT17, 15,1%; DT10, 6,5%) y cabeza piriforme (DT20, 6,5%). Britto y Mello (1988) obtuvieron individualmente DMA exacerbados, en el semen de cinco toros Curraleiro Pé Duro, destacando pouch formation (17,1%). Las alteraciones encontradas pueden tener origen en deficiencias nutricionales, interferencia climática y escasa selección de este grupo genético nativo. Además las altas tasas de gota proximal y cola fuertemente doblada con gota indicarían problemas de madurez sexual, epidídimo y fertilidad (Rao, Bane y Gustafsson, 1980). Saacke (2008) caracteriza cabeza piriforme, cráter y pouch formation como defectos no compensables, ya que suelen poseer aberraciones en la cromatina que afectan la fecundación y/o desarrollo embrionario, causando sub-fertilidad espermática. Cinética espermática Según Kathiravan et al. (2011), altos porcentajes de MP y RPD revelan resistencia a criopreservación. Además la MP está correlacionada con MT, y contribuye 62,6% en variaciones de las tasas de fecundación, teniendo como promedio grupal 27,3% ± 9,7% y RPD 35,4% ± 11,4%. En el promedio individual destacaron las dosis DT5 (MP 45%, RPD 55%), DT9 (MP 40%, RPD 48%), DT16 (MP 39%, RPD 47%) y DT19 (MP 31%, RPD 51%); contrario a DT10 (MP 15%, RPD 17%), DT18 (MP 9%, RPD 11%) y DT20 (MP 13%, RPD 17%) que fueron inferiores. Las velocidades VAP, VSL y VCL (μm/s) fueron altas en DT5 (105,6; 88,4; 179,2), DT11 (88,7; 72,1; 156,9), DT19 (87,2; 60,9; 168,4) y DT21 (91,3; 75,0; 156,5). Para ALH (μm) obtuvieron tasas superiores DT8 (7,3), DT14 (7,6) y DT19 (7,7), lo que concuerda con sus bajos valores en BCF (DT8, 20,4 Hz; DT14, 22,4 Hz; DT19, 23,6 Hz), STR y LIN (DT8, 67,0%, 39,0%; DT14, 75,0%, 41,0%; DT19, 70,0%, 38,0%). La DT19 mostró hiperactividad espermática (HE), por sus altos valores de velocidad; ALH superior en el grupo, así como BCF, STR y LIN inferiores. La HE conduce a movimientos vigorosos y desordenados, provocando trayectoria irregulary y curva, situación que sólo ocurre in vivo en el local de fecundación (Mortimer, 1997; Verstegen, Iguer-Ouada y Onclin, 2002); mas in vitro podría favorecer en la penetración del ovocito. La criopreservación origina mudanzas en membrana plasmática, similares a las de capacitación. Esa criocapacitación post descongelamiento talvez sería responsable por la HE detectada (Januskauskas et al., 1999). Integridad de membrana plasmática y acrosoma Se consideraron la membrana plasmática íntegra (MPI) y los espermatozoides vivos con acrosoma íntegro (EVAI), porque reflejan buena conformación estructural y funcional de la célula (Garner et al., 1986; Graham, 2001). En el promedio general MPI obtuvo 36,7% ± 14,2% y EVAI 34,0% ± 14,4%. Los valores más bajos fueron: MPI (DT10, 15,0%; DT14, 21,3% y DT18, 10,5%) y EVAI (DT2, 18,0%; DT10, 12,5%; DT14, 21,5%; DT18, 9,5% y DT20, 22,0%). Para Januskauskas, Johannisson y Rodriguez-Martinez (2003), la integridad de membrana plasmática es la variable con mayor impacto en la fertilidad, caso contrario a Carvalho et al.(2010) que no identificaron alta correlación con la fecundación. Cabe resaltar

131 138 I.S. Montesinos et al. comparación al G2. El G2 obtuvo la mayor tasa de defectos morfológicos; no en tanto cinética, viabilidad y células vivas con acrosoma íntegro aceptables, características que son propias de espermatozoides con defectos no compensables (Saacke, 2008). G3 obtuvo altos porcentajes de morfología espermática como G1, pero la peor cinética, integridad de membrana plasmática y acrosoma, indicando mayor presencia de espermatozoides muertos, ya que según Garner et al. (1986) células con membrana plasmática dañada no poseen movimiento. Figura 2. Dendrograma de las 21 dosis de semen. que la reacción acrosómica prematura disminuye el potencial reproductivo del espermatozoide (Silva y Gadella, 2006). Casi todas las dosis obtuvieron un porcentaje bajo ( 2,0%), para espermatozoides vivos con acrosoma reaccionado, excepto DT17 con 14,5%, que puede estar relacionado a su alta tasa de DMA (28,1%). Parámetros escogidos para construcción de un dendrograma Fueron analizados parámetros que según Amann (1989)son parte de los atributos espermáticos para fecundación (NRM, DMA, MP, RPD, MPI y EVAI). Se construyó un dendrograma (Figura 2) utilizando la distancia Euclidiana, para formar tres grupos (G1, G2 y G3). La DT19 y DT10 resultaron excluidas, por ser consideradas outliers, al presentar valores muy distantes en relación a las demás dosis (Sartorio, 2008). Se sabe que la DT19 mostró HE y DT10 obtuvo los valores más bajos en los exámenes in vitro. Caracterización de los grupos y comparación estadística En los tres grupos formados (Cuadro 1), el G1 alcanzó altos valores en morfología; la mejor cinética; así como viabilidad y tasas de espermatozoides vivos con acrosoma íntegro razonables, mostrando cierta superioridad en Evaluación de las tasas de fecundación Se utilizaron dosis de sólo cuatro toros Curraleiro Pé Duro (DT18, DT19, DT20 y DT21), debido al amplio estoque de pajillas donadas por el BBGA para esos reproductores. El semen presentó valores NRM y DMA, dentro de los padrones CBRA (1998). Sólo la DT18 y DT20 fueron inferiores en cinética y exámenes de fluorescencia. Previo al uso de las dosis para FIV, estas se seleccionaron mediante gradiente Percoll (45:90%), eliminando los espermatozoides con defectos. Después fue evaluada la MP en el microscopio de contraste de fase, tanto pre como post pasaje por el gradiente, siendo que incrementó considerablemente. Rodriguez-Martinez (2005) relata que esa prueba mide la capacidad de migración espermática, seleccionando las células con mayor potencial para ultrapasar in vitro la barrera del Percoll, que simularía el trayecto in vivo hasta el local de fecundación. Una vez tratadas las dosis seminales para FIV, se seleccionó e inseminó in vitro 336 ovocitos, evaluando post 18 horas las tasas de fecundación (Cuadro 2). Las tasas de fecundación fueron altas y próximas, sin diferencia estadística significativa, revelando resultados no esperados antes del Percoll (45:90%). Como fue el caso de DT18 y DT20, que tenían bajos porcentajes para MP (33%; 30%), en el microscopio de contraste de fase, y exámenes de fluorescencia (MPI 10,5%; 28,3% y EVAI 9,5%; 22%), pero post Percoll mostraron buena MP (70%; 53%) y fecundación ovocitaria (74,4%; 68,8%). Según Tangue et al. (2002), las tasas de fecundación obtenidas son aceptables (> 60%), revelando la capacidad espermática in vitro del semen Curraleiro Pé Duro, para ligar y penetrar la ZP del ovocito, así como promover la Cuadro 1. Parámetros espermáticos analizados en los tres grupos formados por el dendrograma. Parámetros espermáticos (%) Grupo 1 (G1) Grupo 2 (G2) Grupo 3 (G3) Promedio DS Promedio DS Promedio DS Espermatozoides normales (NRM) 84,7 a 4,5 66,3 b 5,5 88,9 a 6,8 Defectos mayores (DMA) 12,0 b 5,3 29,0 a 5,2 8,5 b 4,4 Motilidad progresiva (MP) 34,2 a 5,3 25,0 b 4,5 15,2 c 5,2 Células rápidas (RPD) 41,5 a 6,4 34,0 a 7,2 23,2 b 9,1 Membrana plasmática íntegra (MPI) 39,1 a 8,1 41,6 a 9,5 24,9 b 9,5 Espermatozoide vivo acrosoma íntegro (EVAI) 36,9 a 9,0 35,5 a 4,9 22,4 b 8,5 DS: desvío estándar. Promedios con letra diferente (a,b,c) no son iguales, por el Test Duncan al 5%.

132 Evaluación del semen criopreservado de toros Curraleiro Pé Duro 139 Cuadro 2. Tasas de fecundación ovocitaria 18 horas post inseminación in vitro. Dosis Toro NO Fecundado No fecundado fertilización. Este resultado fue algo similar al obtenido por Obando, Luque y Martínez (2003), en las tasas de FIV ( 61,9%) de cinco toros de la raza criolla San Martinero en el trópico colombiano. Se constató en los cigotos, que los estados de cromatina masculina 18 horas post inseminación in vitro presentaron valores estadísticamente diferentes. Con tasas superiores para pronúcleo (> 60%), siendo que la DT19 obtuvo el mayor porcentual, quizás debido a su HE que contribuyó en la FIV, en cuanto la DT20 el menor, y además un valor elevado para cabeza descondensada. Según Eid, Lorton y Parrish (1994), esto revelaría en DT20 una fecundación tardía, posiblemente debido a sus porcentajes de cabeza piriforme (6,5%), defecto espermático no compensable, que podría haber ultrapasado el gradiente de Percoll, y afectaría una futura embriogénesis si fuese desarrollada. Finalmente, los resultados obtenidos sugieren que el G1 sea utilizado en la IA a campo, por tener valores favorables para fecundación in vivo. Diferente de la DT19, DT10, G2 y G3, que deben utilizarse con prioridad, para evitar mayor deterioro de estas dosis y en estricta asociación a biotecnologías reproductivas, como el pasaje por gradiente Percoll y PIV de embriones, que optimizarán los requisitos espermáticos para fecundación, como ya fue verificado, sabiendo que sólo espermatozoides competentes tienen contacto con el ovocito (Barbas y Mascarenhas, 2009). De esa forma se evitará la pérdida de la valiosa variabilidad genética de estas dosis de semen del BBGA, garantizando la máxima producción de becerros y embriones Curraleiro Pé Duro. Conclusiones N % N % P valor X 2 DT , ,6 DT , ,9 0,4 DT , ,2 DT , ,5 NO: número de ovocitos; N: número; X 2 : Chi-cuadrado. La calidad seminal encontrada en las dosis del BBGA presentó variabilidad, mostrando características favorables y desfavorables para fecundación. Esto indica la importancia de la evaluación espermática para determinar el uso adecuado de las dosis seminales, tanto in vivo como in vitro, en asociación a biotecnologías reproductivas que busquen resultados positivos. Permite así la conservación del bovino Curraleiro Pé Duro, a través de programas de recuperación del efectivo poblacional y criopreservación en bancos de germoplasma, tal como viene siendo ejecutado por instituciones de investigación y criadores de esta raza. También surge su posible uso en el cruzamiento industrial con razas comerciales, para otorgar mayor rusticidad y adaptación al trópico brasilero. Referencias Amann, R.P Can the fertility potential of a seminal sample be predicted accurately? Journal of Andrology 10 (2): Barbas, J.P. & Mascarenhas, R.D Cryopreservation of domestic animal sperm cells. Cell and Tissue Banking 10: Barth, A.D. & Oko, R.J Abnormal morphology of bovine spermatozoa. Iowa State University Press, Ames (IA), USA. Britto, C.M.C. & Mello, M.L.S Anomalias de cabeça em espermatozóides de gado Pé Duro. Revista Brasileira de Genética 11 (1): Carvalho, J.O., Sartori, R., Machado, G.M., Mourão, G.B. & Dode, M.A.N Quality assessment of bovine cryopreserved sperm after sexing by flow cytometry and their use in in vitro embryo production. Theriogenology 74: CBRA Manual para exame andrológico e avaliação de sêmen animal. 2da edición, Colégio Brasileiro de Reprodução Animal, Belo Horizonte, Brasil. Dode, M.A.N. & Rumpf, R Produção in vitro de embriões na espécie bovina. Biotecnologia Ciência & Desenvolvimento 26: Eid, L.N., Lorton, S.P. & Parrish, J.J Paternal influence on S-Phase in the first cell cycle of the bovine embryo. Biology of Reproduction 51: Fioravanti, M.C.S., Juliano, R.S., Costa, G.L., Abud, L.J., Cardoso, V.S., Carpio, M.G. & Costa, M.F.O Conservación del bovino Curraleiro: cuantificación del censo y caracterización de los criadores. Animal Genetic Resources 48: Freneau, G.E Aspectos da morfologia espermática em touros. Revista Brasileira de Reprodução Animal 35 (2): Garner, D.L., Pinkel, D., Johnson, L.A. & Pace, M.M Assessment of spermatozoal function using dual fluorescent staining and flow cytometric analyses. Biology of Reproduction 34: Graham, J.K Assessment of sperm quality. En: Proceedings of the 47th Annual Convention of the AAEP, November 2001, San Diego (CA), USA. Harrison, R.A.P. & Vickers, S.E Use of fluorescent probes to assess membrane integrity in mammalian spermatozoa. Journal of Reproduction and Fertility 88: Januskauskas, A., Gil, J., Soderquist, L., Haard, M.G.M., Haard, M. C., Johannisson, A. & Rodriguez-Martinez, H Effect of cooling rates on post-thaw sperm motility, membrane integrity, capacitation status and fertility of dairy bull semen used for artificial insemination in Sweden. Theriogenology 52: Januskauskas, A., Johannisson, A. & Rodriguez-Martinez, H Subtle membrane changes in cryopreserved bull semen in relation with sperm viability, chromatin structure, and field fertility. Theriogenology 60: Kathiravan, P., Kalatharan, J., Karthikeya, G., Rengarajan, K. & Kadirvel, G Objective sperm motion analysis to asses dairy bull fertility using computer-aided system a review. Reproduction in Domestic Animals 46: Klinc, P. & Rath, D Reduction of oxidative stress in bovine spermatozoa during flow cytometric sorting. Reproduction in Domestic Animals 42:

133 140 I.S. Montesinos et al. Mariante, A.S. & Egito, A.A Animal genetic resources in Brazil: Result of five centuries of natural selection. Theriogenology 57: Mariante, A.S., Albuquerque, M.S.M., Egito, A.A., McManus, C., Lopes, M.A. & Paiva, S.R Present status of the conservation of livestock genetic resources in Brazil. Livestock Science 120: Mortimer, S.T A critical review of the physiological importance and analysis of sperm movement in mammals. Human Reproduction Update 3 (5): Obando, H., Luque, O. & Martínez, R Evaluación de la capacidad fertilizante in vitro del semen de toros San Martinero y Brahman. ICA Informa 30 (2): R Development Core Team R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Disponible en Rao, A.R., Bane, A. & Gustafsson, B.K Changes in the morphology of spermatozoa during their passage through the genital tract in dairy bulls with normal and impaired spermatogenesis. Theriogenology 14: Rodriguez-Martinez, H Methods for semen evaluation and their relationship to fertility. En: Anais do 16 Congresso Brasileiro de Reprodução Animal, 1 5 Agosto 2005, Goiânia (GO), Brasil. Saacke, R.G Sperm morphology: Its relevance to compensable and uncompensable traits in semen. Theriogenology 70: Sartorio, S.D Aplicações de técnicas de análise multivariada em experimentos agropecuários usando o software R. Dissertação em Estatística e Experimentação agronômica, Escola Superior de Agricultura Luiz de Queiroz, Universidade de São Paulo, Piracicaba, Brasil. Silva, P.F.N. & Gadella, B.M Detection of damage in mammalian sperm cells. Theriogenology 65: Sousa, A.P.F., Freneau, G.E., Fioravanti, M.C. & Juliano, R.S Estudo da espermatogênese pelo sêmen e a morfologia testicular do Gado Pé Duro. Anais do 1 Congresso de Pesquisa, Ensino e Extensão da Universidade Federal de Goiás, Outubro 2004, Goiânia (GO), Brasil. Tangue, S., Van Soom, A., Sterckx, V., Maes, D. & De Kruif, A Assessment of different sperm quality parameters to predict in vitro fertility of bulls. Reproduction in Domestic Animals 37: Verstegen, J., Iguer-Ouada, M. & Onclin, K Computer assisted semen analyzers in andrology research and veterinary practice. Theriogenology 57:

134 Animal Genetic Resources, 2014, 54, Food and Agriculture Organization of the United Nations, 2013 doi: /s Unique cultural values of Madura cattle: is cross-breeding a threat? T.S.M. Widi 1,2, H.M.J. Udo 1, K. Oldenbroek 2, I.G.S. Budisatria 3, E. Baliarti 3 and A.J. van der Zijpp 1 1 Animal Production Systems Group, Department of Animal Sciences, Wageningen University, PO Box 338, 6700 AH Wageningen, The Netherlands; 2 Centre for Genetic Resources, PO Box 16, 6700 AA Wageningen, The Netherlands; 3 Department of Animal Production, Universitas Gadjah Mada, Jl. Fauna no 3, Kampus Bulaksumur UGM, Yogyakarta, Indonesia Summary In Indonesia, cross-breeding local cattle with European beef breeds is widely promoted to stimulate beef production. This cross-breeding is threatening local breeds that have often different functions, including cultural roles. This study analysed the cultural values of Madura cattle and the effects of cross-breeding on local traditions in Madura Island. Bull racing (karapan) and cow conformation contests (sonok) are traditional cultural events on Madura. Since 2001, cross-breeding with Limousin is allowed. The local government also promotes a conformation contest for cross-bred (madrasin) cattle. Quantitative and qualitative information were collected through participatory approaches involving farmers (n = 97), government officials, community groups, key informants and through direct observation of sonok, karapan and madrasin events. Phenotypic characteristics were collected from 184 cows. The Madura cattle population and production systems are not homogeneous. Four cattle types could be distinguished: karapan, sonok, madrasin and conventional Madura cattle. These cattle were found in three discrete areas, differing in land sizes, cropping and cattle keeping in terms of management practices and importance of specific cultural practices. Sonok and madrasin cows were significantly bigger and had higher body condition scores than karapan cows and the conventional Madura cows in the madrasin area. Madura cattle participating in cultural events were valued at prices that were times higher than Madura cattle not participating in cultural events. Cross-breeding will not directly influence the cultural events or the management practices of Madura cattle in the karapan and sonok areas; however, outside the karapan and sonok areas, crossbreds are rapidly replacing conventional Madura cattle. The present top-down approach towards conservation and breeding strategies has to be turned into bottom-up approaches that consider the needs of the sonok and karapan Madura cattle sub-populations. Monitoring and characterization studies have to collect information at different aggregation levels and have to be aware of the sub-populations. The Madura example shows that the cultural values of livestock can be a main driver for maintaining relatively small populations of local breeds. Keywords: Madura, bull racing, cattle contests, conservation, cultural values Résumé En Indonésie, le croisement des bovins locaux avec des races bovines à viande européennes est largement promu afin d augmenter la production de viande. Cette pratique menace les races locales qui ont souvent différentes fonctions, y compris des rôles culturels. Cette étude a analysé les valeurs culturelles des bovins Madura et l impact des croisements sur les traditions locales sur l île de Madura. Les courses de taureaux (karapan) et les concours de conformation des vaches (sonok) sont des évènements culturels traditionnels en Madura. Depuis 2001, le croisement avec la race Limousine est permis. Le gouvernement local promeut également un concours de conformation pour les bovins croisés (madrasin). Des informations quantitatives et qualitatives ont été recueillies suivant une approche participative impliquant des éleveurs (n = 97), des fonctionnaires des gouvernements, des groupements communautaires et des informateurs clés, et à travers l observation directe des évènements sonok, karapan et madrasin. Les caractéristiques phénotypiques ont été mesurées sur un total de 184 vaches. La population bovine Madura et les systèmes de production ne sont pas homogènes. Quatre types de bovins ont pu être distingués: karapan, sonok, madrasin et les bovins Madura conventionnels. Ces bovins se trouvent dans trois zones distinctes, qui diffèrent par la taille des terres, les cultures et la production animale en ce qui concerne les pratiques d élevage et l importance de pratiques culturelles spécifiques. Les vaches sonok et madrasin ont été significativement plus grandes et ont obtenu des notations de conformation corporelle plus élevées que les vaches karapan et les vaches Madura conventionnelles dans la zone madrasin. Les bovins Madura participant aux évènements culturels atteignent des prix qui sont de 2 à 3,5 fois plus élevés que ceux des bovins Madura qui ne participent pas aux manifestations culturelles. Les croisements n influenceront pas directement les évènements culturels ou les pratiques d élevage des bovins Madura dans les zones karapan et sonok. Cependant, en dehors des zones karapan et sonok, les animaux croisés sont rapidement remplacés para des bovins Madura conventionnels. L actuelle approche descendante dans les stratégies de conservation et de sélection doit être tournée vers une approche ascendante qui tienne compte des besoins des sous-populations sonok et karapan de bovins Madura. Les études de surveillance et de caractérisation doivent recueillir l information à différents niveaux d agrégation et doivent prendre en considération les sous-populations. L exemple Madura montre que les valeurs culturelles du bétail peuvent être un levier important pour le maintien de populations relativement petites de races locales. Mots-clés: Madura, courses de taureaux, concours bovins, conservation, valeurs culturelles Correspondence to: T.S.M. Widi, Animal Production Systems Group, Wageningen University, PO Box 338, 6700 AH Wageningen, The Netherlands. widi.tsm@ugm.ac.id 141

135 142 T.S.M. Widi et al. Resumen En Indonesia, el cruzamiento del ganado bovino local con razas europeas de aptitud cárnica es fomentado ampliamente con el fin de incrementar la producción de carne de bovino. Esta práctica supone una amenaza para las razas locales, que tienen a menudo diversas funciones, incluidos roles culturales. Este estudio analizó los valores culturales del ganado bovino Madura y el impacto del cruzamiento sobre las tradiciones locales en la isla de Madura. Las carreras de toros (karapan) y los concursos de conformación de las vacas (sonok) son acontecimientos culturales tradicionales en Madura. Desde 2001, se permite el cruzamiento con ganado Limousin. El gobierno local promueve también un concurso de conformación para ganado bovino cruzado (madrasin). Se recogió información cuantitativa y cualitativa siguiendo un enfoque participativo que implicaba a ganaderos (n = 97), funcionarios públicos, grupos comunitarios y personas de alto rango, y mediante la observación directa de los actos sonok, karapan y madrasin. Las características fenotípicas fueron evaluadas en 184 vacas. La población de ganado bovino Madura y los sistemas de producción no son homogéneos. Ha sido posible diferenciar cuatro tipos de ganado bovino: karapan, sonok, madrasin y ganado bovino Madura convencional. Estos tipos de ganado se encuentran en tres zonas distintas, que difieren en el tamaño de los terrenos, en los cultivos y en la actividad ganadera en relación con el manejo de los animales y la importancia de prácticas culturales específicas. Las vacas sonok y madrasin son significativamente más grandes y alcanzan puntuaciones de condición corporal mayores que las vacas karapan y las vacas Madura convencionales en la zona madrasin. El ganado Madura que participa en actos culturales alcanza precios entre 2 y 3,5 veces mayores que el ganado Madura que no interviene en estos acontecimientos. El cruzamiento no afectará directamente a los actos culturales o a las prácticas de manejo del ganado bovino Madura en las zonas karapan y sonok. Sin embargo, fuera de las zonas karapan y sonok, los animales cruzados están siendo rápidamente sustituidos por ganado bovino Madura convencional. El actual enfoque verticalista descendente en las estrategias de conservación y mejora debe ser redirigido hacia un planteamiento ascendente que considere las necesidades de las subpoblaciones sonok y karapan de ganado bovino Madura. Los estudios de seguimiento y caracterización deben recoger información de los diferentes niveles de agregación y tomar en consideración las subpoblaciones. El ejemplo de Madura muestra que los valores culturales del ganado pueden ser un factor principal para el mantenimiento de poblaciones relativamente pequeñas de razas locales. Palabras clave: Madura, carreras de toros, concursos de ganado bovino, conservación, valores culturales Submitted 3 May 2013; accepted 5 August 2013 Introduction In Indonesia, population increase is a major driving force for the rising demand for animal products. To satisfy the demand for red meat, the government has implemented a policy to import meat and live exotic livestock. From the 1980s, the government has also promoted an Artificial Insemination (AI) programme using beef breeds from temperate regions, such as Simmental, Limousin and Angus, to improve the beef performances of local cattle. At present, the AI organisation uses almost exclusively exotic semen. Thus, cross-breeding is rapidly progressing. Many publications state that cross-breeding leads to a loss of indigenous breeds and loss of adaptation of livestock to local environments (Anderson, 2003; FAO, 2007). On Indonesia s Madura Island, until 2001, cross-breeding was not allowed to protect the local Madura cattle breed. Madura cattle developed approximately years ago from crossing wild banteng or Bali cattle and zebu (Payne and Rollinson, 1976). Continuous selection by the farmers in Madura created a uniform breed population (Anonymous, 2003). Its coat is reddish-brown with a nonspecific white pattern on the rump and legs (Anonymous, 2003); see Figure 1. Madura cattle are extremely well adapted to the local conditions and traditional management. They are reported to be one of the best draught animals in the world relative to their size (Barwegen, 2004). Since the removal of import restrictions, the local government has allowed crossing with the Limousin breed (which is also preferred by Madura farmers) because of its red colour, similar to the Madura cattle. As a result, the crossbreeding through AI is threatening the survival of the Madura breed (Barwegen, 2004). Currently, the number of cattle in Madura Island is about It is not possible to obtain accurate data about the population of pure Madura cattle as not all regional government offices separate the data in terms of genetic group. Understanding and assessing the cultural values of livestock are important in the implementation of interventions aiming at conservation and utilization of animal genetic resources (FAO, 2012). Madurese people have strong cultural ties to their animals, caring for their cattle as if they are their family members (De Jonge, 1990). The significance of cattle within the culture is also evident from the appearance of bulls in local folk-tales, proverbs and wood carvings depicting bulls, and from the horns on traditional houses. Bull racing (karapan) and cow conformation contests (sonok), are important cultural events on the island (De Jonge, 1990). Cross-bred cattle are kept primarily for meat production. Therefore, cross-breeding may be a threat not only to the existence of purebred Madura cattle, but also for the cultural activities involving cattle in Madura. In Indonesia, Madura cattle offer a unique opportunity to study cultural practices involving cattle and the effect of

136 Unique cultural values of Madura cattle 143 Figure 1. Colour pattern of the rump and legs of Bali versus Madura cattle (source: Widi, T.SM). cross-breeding. This paper aims to analyse the cultural values of Madura cattle and the effect of cross-breeding on these local traditions. Materials and methods Study areas Madura (Figure 2) is a small, densely populated Indonesian island located off the northeast coast of Java Island. It is administered as part of East Java Province and consists of four districts: Bangkalan, Sampang, Pamekasan and Sumenep. Besides the mainland, Madura has some small outer islands. The climate is tropical, exhibiting defined wet (September to February) and dry (March to August) seasons, with an average annual rainfall of mm. Two districts were used in this study (Figure 2). One was Sumenep, within which is Sapudi, an isolated island that is demarcated by the national government as a conservation area for Madura cattle, particularly as a source of karapan cattle. For the study, Nong Gunong, one of two subdistricts, was used. This study area will be described as the karapan area. The second district was Pamekasan; Waru, Pasean, Larangan and Pegantenan sub-districts were used in this study. Waru and Pasean sub-districts are famous as the source of sonok cattle and are wellknown as areas where the sonok contest started. This study area will be described as the sonok area. Larangan and Pegantenan sub-districts were selected by local government as cross-breeding areas owing to the abundance of crop by-products. Cross-bred cattle are expected to have larger body sizes than Madura cattle and, consequently, require more feed. Cross-bred cattle are not used for draught purposes. All the crossbreds resulted from Madura cows being inseminated with semen of Limousin bulls. The farmers named the cross-bred cattle madrasin. This study area will be described as madrasin area. Since 2007 the local government has promoted a conformation contest for madrasin cattle. Madura cattle in the madrasin area will be described as conventional Madura cattle. The three study areas do not only differ in cattle keeping practices, but also differ in cropping practices. In the karapan area, cropping conditions are difficult because of small farm sizes and poor fertility of the soil. In the sonok area, tobacco is the main crop, which results in a relatively good income (Smith, 2011). In the madrasin area, more crop by-products are available mainly because of the large paddy growing areas in this lowland area. Data collection Secondary data collection Secondary data were collected from the Agricultural Department of Indonesia, the Animal Husbandry Office of East Java Province, Pamekasan and Sumenep District offices, government websites, publications and institutional reports. These provided insights into the local situation and policies. Participatory observations Quantitative and qualitative information were collected through participatory approaches involving farmers (cattle-

137 144 T.S.M. Widi et al. Figure 2. Map of Indonesia and Madura Island with the three study areas (source: keepers), government officials, community groups and other key informants and through direct observation of sonok, karapan and madrasin events. A historic overview of the cultural events was based on the secondary data and open interviews with key informants, comprising of persons well-versed in cattle keeping and cultural events, such as heads of cultural event groups, experienced farmers or village elders (three persons per study area). Semi-structured interviews were conducted with three categories of cattle-keepers: 30 karapan farmers, 37 sonok farmers and 30 madrasin farmers. The farmers were interviewed about their background and motivation, technical aspects such as cattle management and the economics of cattle keeping. The input costs and the selling and buying prices of individual cattle in a farm over a period of one year (February 2010 January 2011) were estimated by farmers. The input costs consisted of purchases of forage and supplements, such as maize and rice bran, and additional feeds such as herbs, eggs and palm sugar for animals participating in the cultural events, veterinary services, AI or natural mating, training for and participating in the cultural events and hired labour costs. The interviews were used to evaluate the role of cattle in the livelihoods of Maduranese people, and the special characteristics of cattle that are important in maintaining local traditions. Phenotypic characteristics of Madura and cross-bred cattle Phenotypic characteristics of cows in the study farms were determined by taking height at withers, and estimation of body weight by measuring chest girth, using a measurement tape (FHK Ogawa Seki Co. Ltd, Tokyo, Japan). The age of each animal (in years) was determined by inspecting its teeth. Data analysis The qualitative data were analysed descriptively. The quantitative data were analysed using one-way ANOVA to compare the different areas (Ott and Longnecker, 2001). Ranking was used to determine the level of agreement of farmers motivations for keeping cattle. General Linear Model (GLM) (Ott and Longnecker, 2001) was used to analyse ranking preferences. The smallest score of rank meant the most important motivation and the highest score of rank meant the least important motivation.

138 Unique cultural values of Madura cattle 145 Results The history of cultural cattle events in Madura Island Karapan Karapan has a very long history. Noer and Maduratna (1975) noted that on Sapudi Island, during the twelfth to thirteenth century, Prince Katandur wanted to make the soils more fertile. He introduced ploughing using a pair of bulls. To encourage people to keep cattle, he created a bullracing game, karapan, in which pairs of bulls raced against each other in a field. The equipment used in the race was similar to equipment used in ploughing (Figure 3). Over time, karapan became a favourite game for Maduranese people, organized after the harvest season. These are now held in nearly every village in the eastern sub-districts. In these karapan races, two or three pairs of bulls are raced over a course of m pulling a sledge with a jockey. The winners and losers of the first round are assigned to separate categories, after which the races continue according to the cup system : the winners go on, while the losers drop out. A major competition is held each year in which all the island s districts are represented. Since 1982, races for tourists are organized in western Madura Island. In the 1970s, local government set some rules for karapan (Noer and Maduratna, 1975): (1) Karapan bulls have to be purebred Madura bulls. (2) The minimum height at the withers of the bulls is 120 cm and the maximum 130 cm. (3) The minimum age of the bulls is 2 years. (4) Mistreating bulls during racing is not allowed. (5) The minimum number of bull pairs having to participate in a sub-district is 40. At that time, there were 20 sub-districts in Madura Island, so bulls had to be used every year. Nowadays, some of the rules are not respected because of the limited availability of Madura bulls that fulfil the physical and age criteria for karapan. In 2011, in the biggest racing event, no more than 30 pairs of bulls competed. The total number of bulls actually participating in karapan is about 400 pairs on Sapudi Island and 200 pairs in Madura mainland (Kuswadi, 2010, personal communication). Sonok The key events in the sonok have resulted from the common practice in Madura that, from about 9 am until 1 pm, cattle are tethered between two pillars, locally named tacek, in view of the front veranda of the house. The cattle s forefeet stand on a platform, a piece of wood about 15 cm high, while they are tethered. This tethering practice gives farmers the opportunity to show their great pride in caring for their cattle. In the meantime, farmers clean the cattle and the barn, and undertake other activities, such as foot care and horn shaping. In 1963, a village head became particularly interested in farmers behaviour in caring for their cattle. He showed cattle of good conformation tethered in the front of farmers houses to high officials who were visiting his village. This idea to show cattle daily became a routine activity in the area (Anonymous, 2007) In 1927, Sommerfeld (1927) described pajengan in which farmers bring their cattle, especially female cattle, to be tethered together in a field. In 1967, a local government official organized pajengan activities twice a week during the tobacco harvest season. Pajengan is currently still taking place every month. It gives farmers an opportunity to meet, showing their good cattle and sharing information with the extension officers. Year after year, this event is well supported by farmers (Rudi Haryanto, 2009, personal communication). Key informants mentioned that in the 1980s, to make pajengan more attractive, the competing farmers chose pairs of good females, harnessed to make them walk in pairs and dressed with beautiful adornments. The pairs of cattle, guided by a jockey, have to walk 25 m to reach a finishing line, designed like a gate, while their forefeet step in a harmonious manner. Saronen, traditional music, Figure 3. Karapan race (source: Schultinga, M). Figure 4. Sonok contest (source: Widi, T.S.M).

139 146 T.S.M. Widi et al. is played to accompany the cattle when they walk to the stage. This event was named sonok; see Figure 4. The total number of cattle participating in different sonok events in Madura is about 600 pairs (Rudi Haryanto, 2009, personal communication). Sonok cattle are judged by conformation traits, such as height at withers, colour, body conformation, body condition, health and harmonious walking in a pair. Unlike karapan, there are no winners and losers in this event. Every contestant farmer receives a gift from the officials without exception, and there is dancing accompanied by traditional music. There are juries who provide the audience with information about the farmer, scores of performance during walking and their opinion about the matching appearance of the pairs. Often the jury consists of influential people, for example, the head of the sonok community group, the inseminator and the head of the district. Since there are no winners, there is no gambling. Although there are no prizes, the contest is very important because during the contest day many transactions among farmers take place; farmers can show their cattle and trade them for good prices. The cows that perform well are very popular for breeding. Madrasin In the area where cross-breeding is practiced, farmers, encouraged by the government, have organized the madrasin contest since The criteria are body size and body weight, and, for cows, the conformation of their calves. The cattle are tied up next to each other in the traditional way for Madura cattle: rope goes through the nose of the cattle and is tied between two pillars, and cattle stand with their front legs on a platform. In the madrasin contest, the cattle do not parade and are not harnessed or dressed with adornments; see Figure 5. Responses from the interviews showed that the number of participants and the number of participating cattle are growing every year, but it is less attractive to spectators than the karapan and sonok events. Figure 5. Picture of a winner in madrasin contest (source: Widi, T.S.M). Livestock farming systems Table 1 gives selected household and farm characteristics for the livestock farming systems in the three study areas. Karapan farmers were relatively older than farmers in the other two areas. Most of the farmers have crop production as their main source of farm income. Family size was significantly larger in the madrasin area. More than half of the farmers in the three study areas only completed elementary school. Farmers had a long experience in cattle keeping, about 20 years or more; karapan farmers had significantly more experience than madrasin farmers. Cattle are the main livestock (Table 1). Goats, sheep and chickens are of secondary importance. Even though the land sizes of farmers in Sapudi Island were quite different from those of farmers in the sonok area, the number of cattle was not much different. Farmers kept about three head of cattle. Only in the madrasin area were the numbers of cattle slightly smaller than in the two other areas. Farmers in the sonok area will participate in the sonok contest only if they have pairs of animals that meet the criteria for participation and if they have sufficient cash available to pay for the fee to join the contest. If farmers in the karapan area have a pair of potential racing bulls, they will first enter this pair in a village race. If this pair runs well, they will enter the pair in a regional contest. Again, a farmer will need sufficient cash to join the race and to cover the additional costs. When the farmers have insufficient cash to purchase cattle, the farmers join with other farmers. Around 12 percent of the farmers were sharing cattle (Table 1). The main occupation of the interviewees was farming, although in the madrasin area, 23 percent were traders or businessmen. Motivations for keeping cattle As in other areas in Indonesia, in Madura, cattle-keeping serves various objectives. Table 2 shows the ranking of the motivations for cattle-keeping in the three study areas. Farmers mentioned the following motivations: financial security (saving), income, providing manure, utilization of crop by-products, raising the social status of their owner, cultural events, draught purposes and hobby. Most farmers mentioned saving as the most important motivation. Farmers consider saving in terms of being able to sell cattle to meet unexpected or large expenditures, such as sending children to school, paying hospital bills for a family member, financing a wedding party or a pilgrimage to Mekka. Sonok and madrasin farmers gave income as the second motive. Farmers consider income as the cash they receive regularly from the annual sale of progeny. Karapan farmers ranked manure as the second most important motive for keeping cattle. Farmers in the karapan area depend on their cattle for ploughing. Since

140 Unique cultural values of Madura cattle 147 Table 1. Characteristics of households and farms in the three study areas. Area Karapan (N = 30) Sonok (N = 37) Madrasin (N = 30) Age of household head (year) 51.6 a ± b ± a,b ± 10.7 Land size (ha) 0.31 a ± b ± b ± 0.61 Family size 3.7 a ± a ± b ± 0.4 Experience in cattle keeping (year) a ± a,b ± b ± 2.1 Number of livestock per farm (n) 1 - Cattle 3 (1 13) 3 (1 6) 2 (1 6) -Cows 2(0 6) 2 (0 4) 2 (1 6) - Heifers 0 (0 2) 1 (1 2) 0 (0 2) - Males 0 (0 6) 0 (0 1) 0 (0 4) - Calves 0 (0 2) 0 (0 2) 0 (0 2) - Small ruminants (goat/sheep) 0 (0 6) 0 (0 4) 0 (0 2) - Chickens 0 (0 6) 0 (0 6) 0 (0 10) Cattle ownership (%) - Private Sharing Main occupation (%) - Farming Government official Trader/private business Labourer Crops Maize, cassava and groundnut a,b,c Different superscripts indicate significant differences between areas (P < 0.05). ns Non-significant. 1 Number of livestock is presented in mode and range. tobacco, maize, groundnut and cassava Rice, maize, tobacco and cassava field sizes are small, and land is stony, a bit hilly and dry, cattle are the best option for ploughing. They use both pairs of cows or bulls for ploughing. If needed, they will borrow an animal from a neighbour to match their own animal. Sonok cattle are rarely used for ploughing. Sonok farmers said that they are afraid that ploughing will negatively affect the beauty of their cattle. They only use cattle that are not participating in sonok events for ploughing. Madrasin cattle are never used for ploughing, because of the absence of a hump (needed for the ploughing harness used) and their limited endurance. Another reason is that hand tractors are used extensively in the madrasin area, which is a lowland paddy-growing area. Time allocation for caring for cattle Table 3 gives the time spent caring for cattle in the three different areas. Sonok farmers and their family members spent significantly more time per day managing their cattle than karapan and madrasin farmers. They spent time on cleaning cattle, sunbathing, foot care and horn shaping. In the sonok area, significantly more time was spent on these activities than in the other two areas. Sonok respondents emphasized Table 2. Farmers ranking of motivations for keeping cattle in the three study areas. Role of cattle Area Karapan (N = 30) Sonok (N = 37) Madrasin (N = 30) Average score 1 Rank 2 Average score 1 Rank 2 Average score 1 Rank 2 1. Saving 2.2 a ± a ± a ± Income 5.1 b ± b ± b ± Manure 2.6 a ± c ± b ± Social status 5.9 bc ± c ± c ± Cultural values 6.4 c ± c ± d ± Draught power 5.7 bc ± d ± d ± Utilization of crop byproduct 7.5 d ± de ± de ± or backyard land 8. Hobby 8.4 d ± e ± e ± Different superscripts indicate significant differences between reasons (P < 0.05). 2 Rank = the smallest score of rank (1) means the most important motivation and the highest score of rank (8) means the least important motivation.

141 148 T.S.M. Widi et al. Table 3. Time allocation per farm in caring for cattle in the three study areas. Activity (h/day) Area Karapan (N = 30) Sonok (N = 37) Madrasin (N = 30) 1. Cleaning barn 0.48 a ± b ± a ± Cleaning the cattle, sunbathing, foot caring, horn shape, etc a ± b ± c ± Gathering feed 2.17 a ± ab ± b ± Offering feed ns 1.08 ± ± ± Training 0.30 a ± a ± b ± 0.00 Total time spent (h/day) 7.42 a ± b ± c ± 1.12 a,b,c Different superscripts indicate significant differences between areas (P < 0.05). ns Non-significant. the cleanliness of their cattle and the surrounding environment, since the daily appearance of sonok cattle, in particular during the contests and pajengan meetings, is very important. Madura cattle are very docile and easy to handle. About once a month, farmers pay attention to horn shape and foot caring. Farmers have a special preference for horn shape. They shape the horns when the cattle are around 4 months of age, by punching holes in the two horns and binding them using a wire: this results in small to medium size horns, curving inwards perfectly symmetrically. Most farmers shape cattle s horns, even when the cattle will not be used for cultural events. They check the hoofs of cattle participants in the karapan or sonok contest very regularly. Farmers use simple equipment for foot caring. Madrasin farmers spent the least amount of time on caring for their cattle (P < 0.05). However, per animal, the time allocation was about the same in the three areas. This was because the madrasin farmers kept slightly fewer animals than the other farmers, and forage collection took up a large amount of time for madrasin cattle. Physical characteristics of the cows Table 4 shows that cows in the sonok area were higher and heavier (P < 0.05) than karapan and conventional Madura cows in the madrasin area. Madrasin cows were significantly higher and heavier compared with their conventional Madura herdmates. Sonok cows, however, were comparable with madrasin cows in these measurements. Both sonok and madrasin cows had a significantly higher body condition score compared with karapan cows and conventional Madura cows in the madrasin area. Selling prices, costs and profit estimates The time to sell cattle varies with the purpose of cattlekeeping. In general, farmers sell animals when they need cash, but they also sell surplus animals because they keep a specific number of animals in relation to the feeds available and the time they have available to care for their animals. Farmers sell weaned cattle for fattening or breeding, usually at 4 6 months after weaning, mainly to traders, whereas cattle for cultural events are sold directly to other farmers who are looking for cattle that can participate in the cultural events. Sonok farmers said that sometimes they are looking for cattle whose appearance is well-suited to be paired with one of their own cattle. Once they find a well-matched animal they are willing to offer a good price. Sonok cows are used not only for the cultural event but also for breeding, whereas only excellent karapan bulls are used as breeding bull. Once karapan bulls cannot be used for racing anymore, for instance they are already old or injured, they will be sold for slaughter. During the survey, only two karapan bulls were used both in the cultural event and for breeding. Table 5 presents the comparison of cattle prices and input costs among purposes and areas. The selling prices of animals below 1.5 years indicate that animals used for cultural Table 4. Physical characteristics of Madura and madrasin cows in each study area. No. Physical characteristic Area Karapan Sonok Madrasin Madura (N = 52) Madura (N = 61) Madura (N = 53) Madrasin cross 1 (N = 18) 1. Age (year) ns 4.5 ± ± ± ± Height at the withers (cm) ± 4.9 a ± 6,1 b ± 10.7 c ± 6.2 b 3. Body weight (kg) ± 43.0 a ± 60.4 b ± 89.0 a ± 92.6 b 4. Body condition score (BCS) 2.6 ± 0.5 a 3.9 ± 0.7 b 2.6 ± 0.8 a 3.8 ± 0.7 b a,b,c Different superscripts indicate significant differences between areas (P < 0.05). 1 Madrasin cross: cross of Madura ( ) Limousin ( ).

142 Unique cultural values of Madura cattle 149 purposes fetch much higher prices (P < 0.05) than their herdmates not used for cultural events. Also, the cash inputs are considerably higher (P < 0.05) for animals participating in cultural events. At (relatively) the same age and body weight, Madura cattle used for cultural events were valued at prices that were times higher than cattle not participating in cultural events. Karapan and sonok farmers spent considerably more money on cattle older than 1.5 years, which participate in the cultural activities, 86 and 87 percent, respectively, than on their herdmates with no cultural function. In particular, the additional feed costs were high for cattle participating in the cultural events. The price of madrasin cattle usually depends on body size. Even when an animal wins the madrasin contest, the slaughter price is based on its body weight. In 2009, a 800 kg cross-bred bull won the contest, and a butcher bought it for 22 Million IDR (1 US$ = 9200 IDR), a normal price per kg live weight. The profits per animal aged more than 1.5 years for cattle participating in karapan and sonok were 3 and 2 times higher, respectively, than for their non-cultural herdmates (Table 5). Madrasin cattle did not show significantly higher profit than their conventional Madura herd-mate cattle. Cattle not participating in the cultural event in the sonok area gave a more-or-less similar profit asmadrasin cattle. Discussion The introduction of cross-breeding in Madura has highlighted that the Madura cattle population is not homogeneous. There are three cattle production systems, each in a more-or-less separate area, with now four types of cattle: karapan, sonok, madrasin and conventional Madura cattle. The main source area of karapan cattle is Sapudi Island. Sonok cattle are found mainly in the northern part of Madura Island. Madrasin cattle and the conventional Madura cattle can be found in the rest of Madura Island. This zonation reflects different cattle production systems in terms of cattle management practices and cultural activities. These differences also emphasize the need for different conservation and breeding strategies for the different Madura cattle sub-populations. Synergy of culture and cattle farming in Madura Karapan and sonok events as well as Madura cattle are unique, as markers of the traditions and culture for Maduranese people (De Jonge, 1990; Smith, 2011). The events are famous, not only in Madura itself but also throughout Indonesia and even abroad. Thus, they are also a tourist attraction. The total number of spectators in all events in a year is approximately 20, people. Karapan is much older and better known than sonok, but the enthusiasm for sonok seems to increase year by year. Sonok may become as important for rural tourism as karapan and might be preferred in the future because there are concerns about cruelty to karapan cattle and gambling during racing. Even though there is a rule about not hurting the bulls during racing, in practice, to get the bulls to run faster, people use nails or thorns, causing wounds and then even pour salt or lemon into the wounds (Kuswadi, 2009, personal communication). On Madura, spiritual leaders are already criticizing these practices and the gambling during karapan events. The cultural practices have a marked effect on cattle husbandry. Cattle used for cultural events are well cared for and trained. This gives the owners an overall sense of wellbeing. Most writers about Madurese society emphasize the attention that is given to the animals. The men are said to be more devoted to their animals than to their wives (De Jonge, 1990). Karapan and sonok farmers give intensive daily care and exercise to their animals (about 3 h/day per animal) to improve their physical appearance and performances. They also spend considerably more money (on average 46.5 percent) on better quality feeds for the animals that participate in the cultural events. Alderson (2003) mentioned that breed characteristics that are important for socio-cultural practices generally relate to the appearance of an animal, and traits that define the market value of an animal. Also, in Madura, cultural values have a strong economic component. Most karapan and sonok farmers argued that they like the race or contest because it keeps the prices of cattle high. Karapan and sonok farmers received more than 50 percent higher prices for breeding stock than the normal market prices. Not only the prices of karapan and sonok cattle that participate in the contests were high, but also the overall input costs (Table 5). In general, farmers in the sonok area are relatively well-off due mainly to the income generated from tobacco (Smith, 2011). Part of this income will be invested in their sonok cattle and the participation in sonok events. Effect of cross-breeding on culture and cattle farming Farmers outside the karapan and sonok areas kept conventional Madura cattle before the introduction of Limousin semen. These were relatively small in size. This made it easy for the government to introduce Limousin semen at that time, as the farmers wanted cattle as big as the sonok cattle. Almost all madrasin farmers said that they took more pride in keeping madrasin cattle than conventional Madura cattle. Table 4 shows that cross-breeding increased the body size of cows to about one-third. The body condition score of madrasin cows was also much higher than that of their conventional Madura herdmates. Cross-bred animals were fed better; this resulted in higher feeding costs for madrasin cattle (Table 5). Nevertheless, the selling prices minus the total cash costs for the period an animal is kept, show that the cash profits of madrasin animals were on average, 55 percent higher than from conventional Madura herdmates (Table 5). The economic

143 150 T.S.M. Widi et al. Table 5. Age, period of keeping, selling price, input costs and cash profit per period of keeping of Madura and madrasin cattle based on the age group, area, purpose of keeping cattle and type of cattle (Million IDR) 1. Age group/ Area/category of cattle Karapan Sonok Madrasin Animals in the event Animals not in the event Animals in the event Animals not in the event Madura breed Madrasin cross 2 <1.5 year (N =13: =13; =0) (N = 22: = 10; = 12) (N =12: =0; = 12) (N = 18: =6; =12) (N =4: =0; =4) (N =16: =13; =3) Age (year) 0.60 ± ± ± ± ± ± 0.42 Period of keeping 3 (year) 0.47 ± ± ± ± ± ± 0.30 Selling price (Million IDR) a ± b ± a ± b ± a ± a ± 5.53 Total input cost per period of keeping (Million IDR) 0.90 a ± b ± a ± b ± a ± a ± 0.28 Forage 0.04 ± ± ± ± ± ± 0.18 Additional feed ± 0.69 (88.9%) ± 0.93 (39.6%) 0.06 ± ± ± 0.14 Veterinarian 0.01 ± ± AI/mating Training, racing/contest 0.07 ± 0.08 (7.8%) ± 1.29 (53.0%) Labour ± Profit per period of keeping (Million IDR) 14.2 a ± b ± a ± b ± a ± a ± 5.2 >1.5 year (N =10: =10; =0) (N =6: =0; =6) (N =10: =0; = 10) (N = 11: =0; =11) (N =9: =0; =9) (N =29: =16; =13) Age (year) 2.86 ± ± ± ± ± ± 0.89 Period of keeping (year) 1.03 ± ± ± ± ± ± 0.49 Selling price (Million IDR) a ± b ± a ± a ± a ± b ± 4.89 Total annual input cost (Million IDR) 4.30 a ± b ± a ± b ± a ± a ± 0.40 Forage 0.34 ± ± ± ± ± ± 0.27 Additional feed 3.70 ± 1.5 (86.0%) ± 0.94(50.4%) 0.66 ± ± ± 0.22 Veterinarian ± ± ± ± AI/mating 0.01 ± ± ± ± ± ± 0.02 Training, racing/contest 0.25 ± 0.20(5.8%) ± 0.83 (44.4%) Labour ± 0.18 Total input cost per period of keeping 5 (Million IDR) 4.4 ± ± ± ± ± ± 0.9 Estimated profit per period of keeping 6 (Million IDR) 17.6 ± ± ± ± ± ± 4.7 Annual profit a ± b ± a ± b ± a ± a ± 4.3 a,b Different superscripts indicate significant differences between different categories of cattle between areas (P < 0.05). 1 One US $ = 9,200 IDR. 2 Madrasin cross: cross of Madura ( ) Limousin ( ). 3 Time needed to keep cattle from purchasing or being born until sold. 4 Feeds that are added into the basal diet (forage), such as concentrate, rice bran, maize bran, palm sugar, herbs, etc. 5 Total input cost per period of keeping (Million IDR) for cattle with age group >1.5 years was calculated by multiplying total annual input cost (Million IDR) and period of keeping (year). 6 Profit per period of keeping (Million IDR) was calculated by deducting total input cost per period of keeping (Million IDR) from selling price (Million IDR). 7 Annual profit (Million IDR) = Profit per period of keeping (Million IDR)/period of keeping (year).

144 Unique cultural values of Madura cattle 151 profit estimates in Table 5 ignore the other benefits cattle provide, such as manure, draught power and security. A comprehensive economic analysis is needed to calculate the overall benefits of the different types of cattle keeping (Smith, 2011). Also, cross-breeding contributes to the overall pride resulting from keeping cattle that perform well. The madrasin contest was initiated by the government and seems a topdown initiative. It attracts in particular farmers; no tourists attend. The main motivation of farmers for enrolling their animals in the contest is the expectation that it can result in a fair price for their animals, since all the participants cattle are weighed and the price per animal is based on the actual weight. In general, when the cattle are sold, the prices that traders offer are based on visual appraisal only. Cross-breeding is not only influencing the feeding practices and selling prices of cattle, but also influences the cattle trade flows. Key informants explained that most small-scale slaughter-men in Madura do not like to slaughter cross-bred animals as they cannot handle and sell the products within one day. Therefore, cross-bred slaughter animals are transported as live animals to the abattoir in Surabaya, the nearest major city in Java. Conservation and breeding developments The different farming systems and the different cultural events, and consequently the different types of Madura cattle, show the importance of considering local variation when developing conservation strategies. Government bodies that are concerned with livestock conservation focus primarily on numerical factors (Alderson, 2009). Also, in Indonesia, the inventory of animal genetic resources for the FAO world s animal genetic resources gives only the total number of Madura cattle (Anonymous, 2003). It does not distinguish between different Madura types of cattle. Although Sapudi is designated by the government as an in-situ conservation area for Madura cattle (Anonymous, 2003), until now, no specific conservation strategies have been implemented, except that cross-breeding is not allowed. In the whole of Madura, there are no reliable data on numbers of different cattle. There is an urgency to correct that deficiency, especially in the face of the rapid change in the genetic structure of the island s cattle herd because of the high adoption rate of cross-breeding in the major part of the island. Field observations and the discussions with key stakeholders indicated that the populations of cattle in the karapan and sonok areas are relatively small: about and animals, respectively. However, there is no immediate threat that these populations will be reduced by cross-breeding because of the continuing importance of the cultural values of the karapan and sonok events. It is unlikely that the ban of importing other cattle, in particular crossbreds, in Sapudi Island will be lifted. Nevertheless, the future of the karapan event is at stake. There is no strict adherence to the government rules for the karapan race, and the numbers of bulls that can join a karapan race is going down. Inbreeding is strictly avoided in the sonok area, by circulating the bulls between villages whereas, in the karapan area, which is an isolated island, selection to avoid inbreeding is not done. This may have contributed to the small size of karapan cows. This requires urgent attention, as it can be a threat to karapan as well as the capacity for ploughing on Sapudi Island. AI was introduced in the late 1990s in Sapudi Island, but it was never a success. The main reasons were that infrastructure was poor, the availability of AI personnel was limited, farmers considered AI a sin and an abnormal way of mating their cattle, and farmers never believed that the Madura semen used for AI was from a karapan bull. The phenotypic performances of sonok cattle are comparable with the performances of madrasin cattle. This shows that through good breeding practices, farmers have maintained good performances of sonok cattle. This guarantees successful conservation of Madura cattle in the sonok area in the future. It can be speculated that, in the future, the great majority of the conventional Madura cattle will be crossbreds. Farmers outside the sonok and karapan areas prefer the madrasin cattle over conventional Madura cattle due mainly to their heavier body weights and consequently higher market prices; although farmers said that they are aware that there is a risk that the second and third generations of cross-bred cattle will show poorer performances than the present F 1 crossbreds. Research is needed on the performances of future generations of crossbreds to inform farmers about potential losses in production with continuing cross-breeding. The governmental breeding centre is promoting crossbreeding in the south of Madura and Madura cattle in the north. Unfortunately, it does not distinguish between the sonok and karapan sub-populations. The present topdown approach towards implementing the conservation of Madura breed has to be turned into support approaches that involve the livestock keepers themselves. Assessment of cultural values Gandini and Villa (2003) developed a framework for assessing different cultural contributions of livestock breeds in Europe with the aim of strengthening their conservation. The present study illustrates the complexity behind assessing cultural values of local breeds in countries such as Indonesia. Cultural assessments should not be restricted to identifying cultural functions and estimating market value of cultural products and services; they have to include the effect of cultural functions on management and specific requirements for animals that have cultural value, and the social and economic contribution of

145 152 T.S.M. Widi et al. these animals to the livelihoods of the households concerned. The Madura example shows that the cultural value of a specific breed can be related to a complex mix of areaspecific cultural functions that are associated with different animal performances, management practises and economic benefits. Consequently, participatory research approaches are needed to unravel such complexities, with information collection at animal, farm and regional levels. Conclusions The Madura cattle example shows that the cultural values of livestock can be a main driver of maintaining a local breed. The enthusiasm of people and farmers for maintaining the cultural values attached to Madura cattle, and consequently the high prices for cultural cattle are keys to conserve the Madura cattle within the areas where the cultural events are maintained. The future of sonok cattle is better than karapan cattle in terms of performances, and general preference of people. Cross-breeding will not directly influence the cultural events and management practices of Madura cattle in the areas where the cultural events are important. In the areas where the cultural events are not important, farmers, however, take more pride in keeping madrasin than conventional Madura cattle, mainly because of their growth performance and the resulting financial benefits. Therefore, cross-breeding is an immediate threat to the future of conventional Madura cattle in the main part of Madura. Monitoring and characterization of animal genetic resources studies have to consider cultural values of livestock breeds and how these values interact with management practices, animal performances and economic benefits. Acknowledgements The first author was supported by a PhD grant from the Directorate General of Higher Education, Department of Education, Republic of Indonesia. The first author acknowledges the support of the DGHE, Gadjah Mada University and Animal Production Systems, Wageningen University, the Netherlands, staff members of Livestock Bureaus in Madura, the leaders of sonok and karapan farmers community as well as all farmers. Last but not least, the first author would like to thank the undergraduate students for their support and assistance in collecting the data. References Alderson, L Breed at risk: definition and measurement of the factors which determine endangerment. J. Livestock Sci. 123: sustainable liveli- Anderson, S Animal genetic resources and hoods. J. Ecological Economics. 45: Anonymous National Report on Animal Genetic Resources Indonesia; a Strategy of Development. Department of Agriculture, Jakarta. Anonymous Riwayat singkat tentang kegiatan sapi cangkean/ pajangan dan sapi sono di kecamatan Waru. Kabupaten Pamekasan, Pamekasan. Barwegen, M Browsing in livestock history; large ruminants and the environment in Jawa, In P. Boomgaard, & H. David, eds, Smallholder and Stockbreeders; Histories of Foodcrop and Livestock Farming in Southeast Asia. KITLV Press, Leiden: De Jonge, H Of bull and men: The Madurese aduan sapi. Bijdragen tot de Taal-, Land-en Volkenkunde 146(4): FAO Global Plan of Action for Animal Genetic Resources and the Interlaken Declaration. FAO, Rome (available at docrep/010/a1404e/a1404e00.htm). FAO Phenotypic Characterization of Animal Genetic Resources. FAO, Rome (available at i2686e00.htm). Gandini, G.C. & Villa, E Analysis of the cultural values of local livestock breeds: a methodology. J. Anim. Breed. Genet. 120(1): Mohammad Noer, D. & Maduratna, Ch Kerapan sapi: permainan dan kegemaran rakyat di kepulauan Madura. Kinta, Jakarta. Ott, R.L. & Longnecker, M An introduction to Statisctical Methods and Data Analysis. Duxbury, Thomson Learning, Inc, Pacific Grove, CA. Payne, W. & Rollinson, D Madura cattle. J. Z. Tierzüch Zuchtsbiol 93: Smith, W Ecological anthropology of households in East Madura, Indonesia. Wageningen University, Wageningen (Ph.D. thesis). Sommerfeld, K Das Madurarind, ein hervorragender Typ der Java-Madura-Sumatra-Rasse, eine kombinationszuchtung zweier Rinderarten, des Balirindes (domestizierten Bantengs) und des vorderindischen Zebus und seine wirtschaftliche und sonstige Verwendung und Bedeutung. J. Zeitschrift fur Tierzüchtung und Zuchtungsbiologie 8:

146 Animal Genetic Resources, 2014, 54, Food and Agriculture Organization of the United Nations, 2014 doi: /s Some traditional livestock selection criteria as practiced by several indigenous communities of Southern Ethiopia Sandip Banerjee, Mohammed Beyan and Hiwot Bekele School of Animal and Range Sciences, Hawassa University, Hawassa Ethiopia Summary Livestock husbandry plays an important role in the rural economy of the country in a tangible or intangible way. The livestock are selected for their functional efficiency and also for their adaptability to the particular agro climate. The study was conducted at two selected woredas (districts) of Southern Ethiopia and was based on individual interviews; focus group discussion and questioner survey. The results indicated that livestock are selected based on some traditional methods by the elders of the society. The respondents indicated that they maintained informal pedigree for cattle and horses based on their dam line, while phenotype of an animal viz. presence or absence of hair whorl, udder and teat length, length of the legs, body length played an important role in selecting the bovines. Tail type, presence of wattles, leg length and coat colour were considered while selecting the small ruminants. Chickens with single comb are preferred over the other types. Coat colour played an important role in selection for functional efficiency of cattle. The study indicated that most of these selection attributes had correlation with the production and reproduction efficiency of the livestock as was explained by several authors. It is suggested that further studies should be carried out to validate the selection methods and incorporate them in region-wise livestock selection strategies. Keywords: livestock selection, Southern Ethiopia, traditional methods Résumé L élevage joue un rôle important dans l économie rurale du pays de façon tangible ou intangible. Le bétail sont sélectionnés pour leur efficacité fonctionnelle et aussi pour leur capacité d adaptation au climat agro particulier. L étude a été menée dans deux woredas sélectionnés du sud de l Ethiopie et a été basée sur des entretiens individuels, groupes de discussion sondage de discussion et poser des questions. Les résultats indiquent que le bétail sont sélectionnés sur la base des méthodes traditionnelles par les anciens de la société. Les répondants ont indiqué qu ils maintenaient pedigree informel pour les bovins et les chevaux en fonction de leur lignée maternelle, alors que le phénotype d un savoir sur les animaux. La présence ou l absence de spirale de cheveux, la mamelle et les trayons longueur, la longueur des jambes, la longueur du corps ont joué un rôle important dans la sélection des bovins. Type de queue, la présence de barbillons, la longueur des jambes et la couleur du pelage ont été considérés lors de la sélection des petits ruminants. Poulets avec un peigne simples sont préférés aux autres types. Robe a joué un rôle important dans le choix de l efficacité fonctionnelle du bétail. L étude a indiqué que la plupart de ces attributs de sélection avait corrélation avec la production et l efficacité de reproduction du bétail comme cela a été expliqué par plusieurs auteurs. il est suggéré que d autres études doivent être menées afin de valider les méthodes de sélection et de les intégrer dans les stratégies de sélection du bétail sages de la région. Mots-clés: sélection du bétail, méthodes traditionnelles, Sud d Éthiopie Resumen La ganadería desempeña un papel importante en la economía rural del país ya sea de forma tangible o intangible. El ganado es seleccionado en base a su eficiencia funcional y también de acuerdo con su capacidad de adaptación a unas condiciones agroclimáticas particulares. Este estudio fue llevado a cabo en dos woredas del Sur de Etiopía y para su realización se emplearon entrevistas individuales, grupos focales de discusión y una encuesta. Los resultados indicaron que el ganado es seleccionado por los ancianos del lugar en base a métodos tradicionales. Las personas encuestadas señalaron que controlan, de manera informal, la genealogía del ganado bovino y de los caballos, siguiendo la línea materna. Rasgos fenotípicos del animal como, por ejemplo, la presencia o ausencia de remolinos de pelo, el tamaño de la ubre y la longitud de los pezones, la longitud de las patas y la longitud del cuerpo son criterios importantes para la selección del ganado bovino. El tipo de cola, la presencia de mamellas, la longitud de las patas y el color de la capa son aspectos que se tienen en cuenta para la selección de los pequeños rumiantes. En cuanto a las gallinas, se prefieren aquéllas con cresta sencilla. El color de la capa juega un papel destacado cuando la selección va encaminada a mejorar la eficiencia funcional del ganado bovino. El estudio mostró que la mayoría de estos atributos de selección estaban correlacionados con la eficiencia productiva y reproductiva del ganado, tal y como ya había sido observado por varios autores. Se sugiere que los próximos estudios Correspondence to: S. Banerjee, School of Animal and Range Sciences, Hawassa University, P.O. Box 05, Hawassa, SNNPRS, Ethiopia. sansoma2003@ yahoo.co.in; sansoma2003@gmail.com 153

147 154 S. Banerjee et al. deberían ir dirigidos a validar los métodos de selección para, a continuación, incorporarlos en las estrategias de selección de ganado de la región. Palabras clave: selección del ganado, métodos tradicionales, Sur de Etiopía Submitted 5 September 2013; accepted 27 February 2014 Introduction Livestock husbandry plays an important role in shaping the livelihood among all traditional societies of the world. Since the advent of agriculture, livestock rearing has been an integral part of rural development. The livestock species being reared and the type within a species have been largely selected by rearers keeping in mind the needs and prevailing agro climate in the region. Such a delicate balance between need, adaptation and availability of the livestock species govern the development of a particular type which is further selected both naturally and by the inhabitants to suit their specific requirements. In the last century, many temperate livestock breeds were introduced in Asia, Africa and Latin America which altered the traditional livestock husbandry practices and also the introduction of modern selection criteria has led to the dilution of the traditional attributes of livestock selection in these regions. Result of a study by Mohanty (2004) indicated that selection of livestock has been an age old practice in India and the methods are often made simple by integrating it with the folk traditions, similar observations have also been reported by Hiwot (2013) from Ethiopia, where she recorded folk songs sung by the livestock rearers of Southern Ethiopia indicating the virtues of good cattle, in both cases the authors have observed that the selection methods are mostly based on the phenotypic description of the livestock. As indicated by Jena (2007) indigenous technical knowledge (ITK) is an integral part of the culture and history of local communities. It has evolved through many years of regular experimentation on the day-to-day life and available resources surrounding the community, and is essential for maintenance of the genetic resources for continued survival of the breed/type. Indigenous knowledge has significance in the designing of sustainable farming systems including animal husbandry practices thereby improving the livelihood of rural populations who accept, develop and maintain innovations and interventions (Singh and Kumar, 2012). Ghosh and Sahoo (2011) were in the opinion that ITK is essential for the maintenance of genetic resources (both plant and animal) for the well being, sustenance and development of the community. The present study highlights some of the traditional livestock selection methods as practiced by the inhabitants of two woredas of the Sidama zone of Southern Ethiopia. Materials and methods The study was conducted at Arbegona and Lok Abaya woredas situated in SNNPRS region of Southern Ethiopia. The Arbegona woreda is situated between 6.47 N and 6 82 N latitude E and E longitudes. The agro ecology of the woreda comprises 92percent (highland) and 8 percent mid land although Lok Abaya woreda is situated West of Lake Abaya ( N E) and the major part of the woreda is in the lowlands. These two woredas were purposively selected as they were identified by the Bureau of agriculture to have a large section of agrarian society practicing traditional livestock husbandry practices. The two woredas are in contrast to each other as the former is largely highlands and midlands whereas in the latter a major part is the lowlands; based on the agro climatic conditions the livestock selection methods too were expected to be different. The study was conducted using stepwise purposive selection methods. The respondents were selected from the community based on their history of livestock rearing followed by their association with livestock husbandry methods, furthermore the selection was narrowed down to people who practiced traditional methods of selection for the livestock they owned. Thus, five kebeles (villages) were selected based on their proximity to all weather roads, three villages (Toga, Rooko and Hafurssa Nameto) were selected from Arbegona woreda, whereas two villages (Falka and Abaya Zurya) were selected from Lok Abaya woreda. The study was based on a semi-structured questionaire followed by focus group discussions and individual interviews. The respondents were also requested to practically show the selection criteria they followed for their livestock, all the interviews were video recorded. The data were analysed using simple descriptive statistics (percentage and ranking). Results The results of the study indicated that almost all the respondents were elders (aged above 50 years) and most of them did not have any formal education, these observations are in accordance with the findings of Mesfin and Obsa (1994), Fekadu (2010) and Mishra and Patro (2010) who also reported that the harbourers of traditional

148 Some traditional livestock selection criteria as practiced by several indigenous communities of Southern Ethiopia 155 Table 1. Various reasons for rearing livestock as indicated by the respondents. Purpose of keeping livestock Arbegona Lok Abaya Milk and milk products 1 st 1 st Income source 2 nd 3 rd Pride and insurance 3 rd 2 nd Others (hide, skin, traction, meat etc.) 4 th 4 th 1 denotes the most and 4 the least important criteria. knowledge are the elders among the community. The results from Table 1 indicate the various reasons for which the respondents prefer to rear livestock. The results indicate that the primary reason is to meet their requirements of dairy products. Butter is an important product both for culinary uses like niter kibbeh (clarified spiced butter used in almost all the dishes) and also as a hair moisturiser for women. Moreover, Ayib (Ethiopian cottage cheese prepared from fermented milk) and Ergo (naturally fermented curd) are considered as delicacies in Ethiopian cusine, thus requiring a lot of dairy inputs. The results also indicate that the respondents in Arbegona rear livestock to supplement their income while those of Lok Abaya prefer to rear livestock as a source of tribal pride and insurance against unforeseen calamities. However, the study also indicates that the sale of livestock products such as hide, meat etc., is ranked the lowest as there is a lack of market opportunities for the sale of such value added livestock products. It was also observed during the study that the respondents of Arbegona woreda still barter dairy products for household commodities like salt, the findings are in accordance with the results of Fekadu (1994), from southern parts of Ethiopia and also from Amhara region by Azage et al. (2010). The uses of livestock as a source of income and for fulfilling emergency needs are in accordance with the observations of Iqubal (2013). The results from Table 2 indicate the criteria used by the respondents for selection of livestock. The results also Table 2. Criteria used for selection of livestock by the respondents in the study areas. Criteria A (percent) L (percent) Selection for milking cow and heifer Small head, thin slender neck in cow Long teat, well-structured udder Good milk veins Large, hollow body cavity Thin skin denotes high milk yields 6 66 Fine tail of a cow, not too long Docility, hence good lactation and fertility in cows Hair whorl at the back 75 4 Straightened rear flank Hind loin no flesh Prominent pelvic girdle and tail bone Selection of bull Long penis and short and thick neck denotes good fertility in bull Straight and strong hump Large scrotum Strong voice denotes fertility Broad ear Long tail Lengthy and good temperament Selection of cattle for fattening Straight backbone Straight legs denote good meat production Selection of ox Body length and muscularity Castrated Not very fat Sheep and goat Big udder in goat milk production 78 Long penis and large scrotum denote good fertility Well-developed wattle Flat and broad ear Fat and broad tail in sheep while short for goats Long and thin legs 100 Colour 44 Selection of horse Thin and long leg 89 Aggressive and muscular 78 Long tailed 56 Sire line 49 Broad hoof 87 Selection of poultry Single comb A, Arbegona L, Lok Abaya.

149 156 S. Banerjee et al. Figure 1. Dairy traits were used by the respondents of Lokabaya to select a cow. indicate that while the primordial criteria for the selection of milch cattle as was reported by the respondents from both the studied woredas was the length of the teat and well-structured udder (Figure 1), followed by prominent milk veins in Lok Abaya and hair whorls on the back in Arbegona. The third important criteria as mentioned by the respondents in Lok Abaya was the presence of thin skin among the cattle while those from Arbegona preferred to give importance to straight rear flank and prominent pelvic girdle and tail bones. The study further indicated that the bulls were selected based on the size of their scrotum with larger scrotum denoting better fertility this criterion was similar in both the studied areas, this was followed by the length of the penis where it was considered by the respondents that bulls with longer penis had better fertilizing ability, this was followed by selection of docile bulls (for better handling) and also the bulls with longer body length were considered to have a better libido. The results also showed that the respondents selected bulls with hoarser voice, which they thought denoted muscularity. The selection of fattening livestock depended on the two prime criteria which was consistent in both the studied woredas, the first being straight backbone and legs, while the respondents indicated that an ox is usually selected based on its body length and muscularity, it should be castrated with not much fat, the rankings are consistent; however, the degree varied between the respondents of the two woredas. The result of the study further indicated that the goats are selected based on the presence of wattles, where goats with wattles are preferred over those not having one, the observation was consistent in both the woredas. The respondents of Lok Abaya woreda preferred to select does with good udder capacity, whereas those of the Arbegona woreda preferred to select bucks with longer penis. The respondents of Lok Abaya preferred to select goats with long and thin legs followed by flat and broader ears while it was the primary criteria for selecting goats by the respondents of Arbegona worda. The study further indicates that respondents of both the woredas preferred to rear sheep with fat and broader tails, while for goats short tail was preferred. Selecting small ruminants based on their coat colour was practiced only among the respondents of Arbegona woreda. The study further indicated that horses were reared only in Arbegona woreda, where they were selected based on their long and thin tail and broad hooves followed by their muscularity, tail length and sire line. The study further revealed that the respondents (irrespective of the woredas) preferred single combed chickens. The results as presented in Table 3 indicate the respondent s perception towards selection of livestock based on their coat colour. The study indicates that while cattle with black coat colour is preferred in Arbegona woreda (Figure 2), the respondents of Lok Abaya woreda preferred rearing cattle with light-coloured coat (Figure 3). The results further indicated that the respondents in both the studied locations preferred to rear cows with reddish brown or red colour coats for they believe that such cattle are good milk providers, whereas the respondents in Arbegona woreda prefer to rear cattle with grey colour coats which they consider to have beef with high fat percentage. Discussions The respondents of the present study preferred to rear native livestock breeds over the crossbreds, they indicated Table 3. Selection of cattle based on their coat colour as indicated by the respondents in the study areas. Coat colour (Percent) Remark Arbegona Lok Abaya Black 49 Strong survive the dry season, do not emaciate High market access White (creamy) 80 Can resist the hot environment, shiny appearance High market demand Dalecha (grey) 7 Bure (reddish brown) Can give more milk Red 11 6 Blue 10 4 Have fatty meat

150 Some traditional livestock selection criteria as practiced by several indigenous communities of Southern Ethiopia 157 Figure 2. Dark coat color of cattle reared in the Arbegona Woreda. that the native breeds play a multi-functional role in rural livelihoods, contributing not only to cash products but also to manure, and have a cultural importance. The respondents in the study reported that the crossbred and exotic breeds of livestock are unable to thrive in the area due to lack of proper veterinary facilities and feed availability throughout the year. They are also prone to commonly occurring diseases prevailing in the region, while on the other hand the native breeds have much better tolerance of diseases and are also able to thrive under environmental stress; during the dry season (October March) they are mostly under fed and receive water once in several days. The quality of the feed they receive is also very poor, just enough for their sustenance. The results are in consonance with the observations of Kunene and Fossey (2006) who reported that the farmers in the Natal region of South Africa prefer native (Nguni) livestock (cattle, sheep and goats) over the exotic breeds because of better adaptability and the ability of the native breeds to thrive on locally available feed resources. Similar observations as indicated in this study too have been reported by Raziq and Younas (2007), Raziq, Younas and Kakar (2008), Bavikatte and Jonas (2009), Kakar (2009). Figure 3. Light coat colour of cattle selected at Lok Abaya woreda. The study further indicates that the respondents still prefer the age old phenotypic selection methods and that the better animals are identified based on the coat colour. The coat is correlated with the production traits and other similar phenotypic characters, these observations are similar to those reported by Karthikayan and Gajendran (2005). The phenotypic selection methods (both qualitative and quantitative traits) as observed in the study find similarity with the observations of Ouma, Abdulai and Drucker (2006), Wurzinger et al. (2006), FABRE (2006), Mbuku, Isaac and Alexander (2006), Shiferaw (2006), Kakar (2009), Endashaw et al. (2012) and Shigdaf (2011). The study also indicated that the mating of the livestock is by and large panmictic, this is because most of the mating takes place in the communal grazing lands. The observations are in close agreement with the reports of Shiferaw (2006) for Kereyu cattle reared in the Fental district of Oromia region and also of Agere (2008) for Horro cattle from the Horro Gudru area. The two prevalent methods of livestock selection are the phenotypic selection and also an oral tradition of keeping the pedigree of the calves based on their maternal lineage. Formal pedigree recording in livestock has been a very old tradition in many parts of the globe with reports from Winchester (1977) indicating the presence of such lineage of horses on clay tablets of Mesopotamia. The study also indicates that the respondents were aware of the heritable attributes of the livestock, which included physical, production and behavioural traits. The method of pedigree in the traditional breeding system has also been reported by Mathias, Ilse and Jacob (2005) from other parts of Ethiopia, and also by Aujla, Jasra and Munir (1998), Kunene and Fossey (2006), Mbuku, Isaac and Alexander (2006), Raziq and Younas (2007), Raziq, Younas and Kakar (2008) and Kakar (2009) from different parts of the world. The respondents were well aware of the problems associated with inbreeding in any form and try to avoid the same in cattle and horses as they believe that the offspring born of such breeding have a weak physical conformation, are susceptible to diseases and have retarded growth thereby affecting their survivability. The results presented in Table 2 indicate that the major criteria for the selection of cows is the presence of long teat and well-structured udder, the long teats facilitate milking and studies by Morse et al. (1988) have indicated that cows with longer teat tend to suffer less from mastitis, the reason being that the teat canal serves as an effective barrier towards any external infection, as the teat canal closes within a few minutes of milking and thus shutting off the link between the external environment and the udder tissues, the results, as reported by the respondents, too find similarity with the findings of Kuczaj (2003), Alphonsus, et al. (2010) and Yakabu (2011). However, results of studies by Berry, Harris and Winkelman (2005) and Yamazaki et al. (2009) indicate that cows

151 158 S. Banerjee et al. with long teats tend to hurt themselves while grazing or even while lying down and are therefore more prone to mastitis, whereas Tilki et al. (2005) reported that there was a negative relationship between teat length and milk yield. The presence of well-developed milk veins was also considered as a selection criterion and the results are in accordance with the observations of Bonsma (1980). The respondents in Lok Abaya woreda reported that the cattle with thin skin have better milk producing ability, the observations are in consonance with the results obtained by Bonsma (1980) and Hamid et al. (2000). However, the results of studies by Hammond, Ian and Terence (1971) indicate that cattle with thinner hide tend to be more susceptible to tick borne diseases as thinner skins with high vascularity have blood capillaries near the skin surface, which in one way helps in dissipation of heat but on the other hand are exposed to infections transmitted by stinging pests thereby leading to susceptibility towards many external parasite borne diseases. Kshatriya, Trivedi and Dhami (2009) in their study concluded that both biometry of udder and skin thickness were related to milk yield and therefore could be considered as one of the criteria for the selection of dairy cows. All the above selection criteria are in accordance with the findings of Endashaw et al. (2012) who reported that the Musi and Bodi pastoralists of Ethiopia select cows based on their milking ability, coat colour, fertility and udder size. The docility and fertility are of course the primary concern for the rearers as a cow which is docile can be easily handled by all the family members followed by good fertility ensuring a higher frequency of lactation and also an increase in herd size. The observations are in accordance with the reports of Raziq and Younas (2007), Raziq, Younas and Kakar (2008) and Kakar (2009) from Baluchistan where docility was considered an important criteria for the selection of livestock by traditional camel, cattle and sheep breeders of tribal areas of Pakistan. The traits like hind loin which is devoid of flesh with prominent hip and pin bone, as was reported by the respondents, may be attributed to safe calving and less chances of dystocia, the observations are in accordance with the observations of Bonsma (1980) and Hammond et al. (1971). Result of a study by Peters, Horst and Kleinheisterkamp (1982) indicated that coat colour pigmentation can assist in productive and reproductive ability of various livestock species especially in the humid and sub-humid environment. Thin skin of cattle is correlated with good milking ability especially in the lowland areas, where the climate is usually quite hot. This might be attributed to the low body fat deposition and the fact that most of the energy is utilized for milk production. Studies by Foster, Fourie and Neser (2009) indicated that hide thickness is correlated with rectal temperature, high rectal temperature depresses the let down of milk. Similarly, cattle with thick and dense hair coat can exacerbate the effects of heat stress (Berry and Shanklin (1961), Finch, Bennett and Holmes (1984)), thereby the milk productivity of such animals is expected to be less in the lowland areas besides it may also affect the reproductive efficiency of the cattle. As observed in the study, coat colour too plays a primordial role in the selection of small ruminants by respondents of Arbegona wordea; the observations are in accordance with the findings of Shigdef (2012) who indicated that white and red coat colour are preferred in sheep and goats in the Washera and Ferta districts whereas black is the least preferred one. The selection of goats on the basis of their coat colour too has been reported by Mbuku, Isaac and Alexander (2006) by pastoralists of Kenya. The preferences of selecting cattle based on their coat colour as reported by the respondents have been presented in Table 3. The results indicate that coat colour is considered an important criteria for selection of cattle especially for their potential to adapt and production capacity. Coat colour as a selection criterion for cattle has also been well documented in the folklores of India as reported by Mohanty (2004). The respondents in this study indicated that coat colour is related to the adaptability and influences the disease tolerance of the livestock, they also reported that coat colour also influences the ability of the livestock to ward away the evil eye, meat of animals of a particular coat colour are considered to be tasty and certain coat colours have more market demand than the others. This might be a fallout, as an animal with a particular coat colour has better ability to tolerate the environmental variations; these findings are in close agreement with the results presented by Ebozoje and Ikeobi (1998) and Sanusi (2008) who reported better survivability. The observations are in consonance with the findings of Tekleyohannes et al. (2012) from Ethiopia and Adedeji et al. (2012) from Nigeria. Results of a study by Bonsma (1980) indicated that at higher altitudes there are short wave or ultraviolet rays which are tolerated well by those with dark coat colour, like black and brown whereas the cattle with white coat colour tend to suffer from skin burns associated with ultraviolet rays. Although in the lowlands the infrared or long wavelength rays predominate which is reflected by the animals having white coat colour, similar observations too have been reported by Finch and Western (1977); Olson, Lucena and Chase Junior (2003). Studies by Makokha et al. (2006) indicate that animals with dark coat colour attract the Tse Tse flies and are therefore not preferred in the lowland areas such as Lok Abaya and places infested by stinging insects, which in this case is at Lok Abaya woreda. Similarly the results as obtained in the study finds consonance with the observations of Stewart (1953), Hutchinson and Brown (1969), Finch (1986) and Hansen (1990) who reported that lighter coat colour reflect the sunlight and hence beneficial for cattle reared in the lowlands. This result goes in agreement with the finding of Ndumu et al. (2008) who reported that the farmers prefer red colour in Ankole longhorn cattle over other

152 Some traditional livestock selection criteria as practiced by several indigenous communities of Southern Ethiopia 159 coat colours, Shiferaw (2006) too reported high ranking for white colour preference for Kereyu cattle owners, the findings are in consonance with the present study for cattle selected in the Lok Abaya woreda. The influence of coat colour on milk yield has also been reported in folklores by Mohanty (2004) who indicated that cattle with black coat colour are supposed to be better in milk yield, the importance of feather colour in selection of chickens have also been reported by Kassahun (2013) from Ethiopia and also by Karma (2005) from Bhutan and Banerjee (2012) from India. The hair whorl is also one of the important aspects to be selected according to the respondents. It is also known that a cow with high hair whorl tends to reach puberty earlier and that they also have high total milk and total solid yield (Young, McNaughton and Spelman, 2007). Results of a study by Broucek et al. (2007) also indicate that animals with more hair whorl may grow faster than animals with a middle or low hair whorl; which can be correlated to the early development of mammary gland. According to the results of a study made by Bharadwaj, Sastry and Yadav (2011), coat colour, shape of head and face, tail switch, shape of udder and teats and skin thickness were some of the traits which were associated with milk production. The short and straight femur as reported by the respondents (straight thigh) can serve as an indication of its good grazing ability and endurance while grazing or reaching the watering points. The results are in accordance with the observations of Kedija (2007) who reported that farmers preferred cattle with long and straight thighs in the Mieso district of Ethiopia. As indicated by the selection of bulls based on penile length, the farmers correlated long penis with good fertility as they indicated that if the bull had a short penis there will be problems associated with its ability to breed, the observations are in accordance with the results obtained by Bruce (2012), Bonsma (1980) and Hammond, Ian and Terence (1971). The indication of long penis and fertility may be associated with the proper function of the cremaster muscle which is responsible for the movement of the penile sheath and therefore frequent urination and libido are correlated with the high level of androgen in bulls; Kedija (2007) from his study in Meiso district too reported that the farmers selected bulls with long penis. In addition, scrotal circumference is a measure of the daily and total sperm production of a bull. Studies agree that neck conformation in bulls is very important by which a good bull should have a thick crest over his neck as well as a short neck. Studies by Hammond, Ian and Terence (1971) too indicate that short neck in bulls are a good indicator of libido and high testosterone levels, as well as larger scrotal circumference (which relate to higher levels of hormones) and wider, more muscular shoulders. Short-necked bulls tend to sire wide-rumped (more depth in the hindquarters) steers and daughters that mature early, higher levels of testosterone are also associated with thick bones especially those of scapulae and metatarsus. Results of studies by Hammond et al., (1971) indicate that bulls with thick shoulders have large capacity in the ribcage (where heart and lungs are located) also indicating masculinity in bull, thicker and coarser muscles associated with thick shoulders, the observations are in accordance with the reports of Bonsma (1980). Large and welldeveloped scrotum and testes is also a good indicator for fertility amongst male livestock, hoarse voice is usually associated with androgen activity and is an indicator of muscularity as indicated by Hammond, Ian and Terence (1971), broad and long ear is a breed character and it is usually associated with high vascualrity, which may be related to better dissipation of body heat and thereby better feed efficiency and also grazing ability. Long tail is usually associated with an animal s ability to get rid of external parasites especially flies and other stinging insects which if the animal is able to overcome can be correlated with good health of the livestock as a whole. The results pertaining to selection of livestock for fattening indicates that those with flat spinous process and long femur are better for fattening, a common indicator for welldeveloped longissimus muscle, and a long and thick femur is an indicator for an animal s ability to walk long distances for fodder and also a thick bone provides large surface area for the attachment of muscles, thus muscle bundles are large and thick under such circumstances, the observations are in accordance with the results obtained by Hammond, Ian and Terence (1971) and Kedija (2007). The selection of small ruminants and equine with long legs is a criteria to select animals which have endurance in travelling long distances. Results of studies by Hafez (1968) indicate that longer femur and metatarsus are adaptive means for livestock reared in hot climates, long legs in one hand help the livestock travel long distances besides they help protect the vital organs against the solar radiation from the hard soil surface. Fat tail of sheep is a breed character but among the animals of a particular breed those having fatter tails are able to tolerate drought better as they are able to have enough reserves in the form of fat, even the livestock with fat deposits have usually better feed conversion efficiency as excess of energy in feed is converted to fat deposits. Broad hoof in horses is associated with its stability as a pack animal especially in the highlands of Ethiopia where the equines are generally employed as pack animals and hence strong and broad hoof is associated with stronger planter cushions and tendons and cartilages which are in accordance with the observation of Hafez (1968). The selection criteria for small ruminants (sheep and goats) as observed in the present study agree with the finding of Estifanos (2010) in the Umbulo watcho water shed of Southern Ethiopia, the presence of wattle as a selection

153 160 S. Banerjee et al. criteria for goats as observed in the study are in accordance with the findings of Odubote (1994b) who reported that goats with wattles are docile and thermoregulatory mechanism of the goats with wattles are better than those without wattles. Results of a study by Casu, Boyazoglu and Lauvergne (1970) indicated that ewes with wattles had better fecundity when compared to the non-wattled. Although Richard et al.,(1990) reported that there might be a genetic linkage between the presence of wattles and disease tolerance with Alphine and Saanen goats. The goats sporting wattles had significantly lower faecal output of strongyle eggs than those without wattles indicating a form of genetic tolerance to helminthoses. The result, as assessed in the present study pertaining to the selection of goats having wattles and of a particular coat colour, is also in accordance with the observation of Adedeji et al. (2012). Conclusion The results from the present study indicate that the respondents in the studied areas practice selection of livestock based on some predefined physical attributes most of which have proper scientific justifications and are best suited to their agro-climatic conditions. Hence, these criteria can be included in developing selection mechanisms which can be further coupled with modern selection techniques. The age old system of livestock selection using phenotypic markers can be further refined using molecular markers; hence, it is of primordial importance to properly document the oral traditions from the elders who practice them. Acknowledgements The authors acknowledge the financial assistance received from Coordinator NORAD project. The authors also acknowledge the assistance received from the elders and traditional livestock rearers of both the woredas, the authorities of the agricultural bureau, head and staff members of the School of Animal and Range Sciences, Hawassa University. References Adedeji, T.A., Ozoje, M.O., Ojedapo, L.O., Fasoyin, O.A., & Ogundipe, R.I Indirect selection amongst small scale holders of West African Dwarf (WAD) goats based on coat pigmentation and wattle types in Ogbomoso, Nigeria. Iran. J. Appl. Anim. Sci., 2(1): Agere, M Assessment of feed resources, feeding system and milk production in Horro district, western Oromia, Ethiopia. Presented to the School of Graduate Studies of Hawassa University, Hawassa, Ethiopia (M.Sc. Thesis). Alphonsus, C., Akpa, G.N., Oni, O.O., Rekwot, P.I., Barje, P.P. & Yashim, S.M Relationship of linear conformation traits with bodyweight, body condition score and milk yield in Friesian Bunaji cows. J. Appl. Anim. Res., 38: Aujla, K.M., Jasra, A.W. & Munir, M Socioeconomic profile of camel herders in South-western mountainous areas of Pakistan. In Proc. Third Annual Meeting Animal Products under Arid Conditions, vol 2, pp Azage, T., Belete, A., Fekadu, B. & Berhanu, E Cattle milk and meat production and marketing systems and opportunities for market orientation in Fogera woreda, Amhara region, Ethiopia. Improving Productivity and Market Success (IPMS) of Ethiopian Farmers Project, International Livestock Research Institute (ILRI), Addis Ababa, Ethiopia, Working Paper No.19. Banerjee, S Morphological characterization of indigenous chickens of Sikkim and West Bengal, India. Anim. Genet. Res., 51: Bavikatte, K. & Jonas, H Bio-cultural community protocols: a community approach to ensuring the integrity of environmental law and policy. Natural Justice, United Nations Environment Programme (available at Berry, I.L. & Shanklin, M.D Physical factors affecting thermal insulation of livestock hair coats. Missouri Agricultural Experiment Station Research Bulletin Berry, D.P., Harris, B.L. & Winkelman, A.M Phenotypic associations between traits other than production and longevity in New Zealand dairy cattle. J. Dairy Sci., 88: Bharadwaj, A., Sastry, N.S.R. & Yadav, M.S Lactation and mammary system influences on milking behaviour of buffaloes a note. Indian J. Anim. Prod. Manage., 3: Bonsma, J.C Livestock production: a global approach. Cape Town, South Africa, Tafelberg Publishers. Broucek, J., Peter Kisac, S., Tefan, M., Anton, H., Michal, U. & Vladimir, T Hair whorls of holstein friesian heifers and affects on growth and behaviour. Arch. Tierz. Dummerstorf, 50(4): Bruce, E.E Bull, In Male Theriogenology (available at assessed on 3rd September Casu, S., Boyazoglu, J.G. & Lauvergne, J.J The inheritance of wattles in the Sardinian breed of sheep. Annals. Genet. Sel. Anim., 2: Contribution to animal production and management. Review and. Scientific Tech. Off. int. Ebozoje, M.O. & Ikeobi, C.O.N Colour variation and reproduction in the West African Dwarf goats (WAD). Small Rumin. Res., 27: Endashaw, T., Tadelle, D., Aynalem, H., Wudyalew, M. & Okeyo, M Husbandry and breeding practices of cattle in Mursi and Bodi pastoral communities in Southwest Ethiopia. Afr. J. Agric. Res., 7 (45): Estifanos, T Sheep production system and on farm evaluation of dry season Ethiopia. Hawassa university. p.104 (M.Sc. Thesis). FABRE Sustainable farm animal breeding and reproduction A vision for FABRE technology platform. February p. (available at Fekadu, B Present situation and future aspect of milk production, milk handling and processing of dairy products in southern Ethiopia. Farm made milk products in southern Ethiopia: 1. Chemical and microbial quality. Department of Food Science, Agricultural University of Norway. (Ph.D. Thesis). Fekadu, F Ethiopian medicinal plants in veterinary healthcare. A mini-review. Ethiopian e-j. Res. Innov. Foresight, 2(1) Finch, V.A Body temperature in beef cattle: its control and relevance to production in the tropics. J. Anim. Sci., 62:

154 Some traditional livestock selection criteria as practiced by several indigenous communities of Southern Ethiopia 161 Finch, V.A. & Western, D Cattle colours in pastoral herds: natural selection or social preference. Ecology, 58: Finch, V.A., Bennett, I.L., & Holmes, C.R Coat colour in cattle: effect on thermal balance, behaviour and growth, and relationship with coat type. J. Agric. Sci. (Camb.), 102: Foster, L.A., Fourie, P.J. & Neser, F.W.C Effect of heat stress on six beef breeds in the Zastron district: the significance of breed, coat colour and coat type. S. Afr. J. Anim. Sci., 39: 224. Ghosh, P.K. & Sahoo, B Indigenous traditional knowledge. Orissa Review (available at e-magazine/orissareview/2011/jan/engpdf/66-71.pdf). Hafez, E.S.E. (Ed.) Adaptation of domestic animals, Philadelphia, Lea & Febiger. Hamid, M.A., Husain, S.M.I., Khan, M.K.I., Islam, M.N. & Biswas, M.A Skin thickness in relation to milk production in crossbred cows. Pak. J. Biol Sci., 3(9): Hammond, J. Jr., Ian, L.M. & Terence, J.R Hammond s farm animals. U.K., Edward Arnold. Hansen, P.J Effects of coat colour on physiological responses to solar radiation in Holsteins. Vet. Rec., 127: 333. Hiwot, B Assessment of indigenous knowledge in livestock rearing practices at Arbrgona and Lokabaya woreda of Sidama Zone of Southern Ethiopia. Submitted to School of Graduate Studies, Hawassa University, Ethiopia (M.Sc. thesis). Hutchinson, J.C.D. & Brown, G.B Penetrance of cattle coats by radiation. J. Appl. Physiol., 26: Iqubal, M.A Livestock husbandry a sustainable livelihood in Ethiopia. Int. J. Econ. Manage. Soc. Sci., 2(8): Jena, M Community Health Knowledge Register. Tradit., 5: Kakar, A.R Assessing the potential of the indigenous livestock breeds of Baluchistan. A Drynet Science and Technology Expertise. 59 p. (available at ID=144&DISPOP=VRCPR) assessed on 3rd Sep Karma, N., Penjor, D.P., Gurung, R., Arasta, P. & Moran, C Genetic structure of the indigenous chickens of Bhutan. SAAR J. Agric., 3: Karthikayan, S.M.K. & Gajendran, K Traditional technologies in the improvement of breeds of livestock in Tamil Nadu. Indian J Tradit. knowl., 4(3): Kassahun, H Evaluating poultry production system and egg quality parameters of village chickens at Hawassa zuria woreda of southern Ethiopia. Hawassa University, Ethiopia. (M.Sc. Thesis). Kedija, H Characterization of milk production system and opportunity for market orientation: a case study of Mieso district, Oromia region, Ethiopia. Haramaya University. (M.Sc. Thesis). Kshatriya, P.S., Trivedi, M.M. & Dhami, A.J Association of udder biometry and skin thickness with milk yield in Kankrej and Crossbred cows. Indian J. Field Vet., 5(1): Kuczaj, M., Analysis of changes in udder size of high yielding cows in subsequent lactations with regard to mastitis. Electron. J. Pol. Agric., 6(1), #02. (available at media.pl/volume6/issue1/animal/art-02.html). Kunene, N.W. & Fossey, A A survey on livestock production in some traditional areas of Northern Kwazulu Natal in South Africa. Livestock Research for Rural Development. Volume 18, Article #113. Retrieved September 3, 2013, (available at org/lrrd18/8/kune18113.htm). Makokha, S.N., Karugiab, J., Staalc, S., & Kosura, O Valuation of cow attributes by conjoint analysis: a case study in Western Kenya. Contributed paper prepared for presentation at the International Association of Agricultural Economists Conference, August 12 18, Gold Coast, Australia (available at accessed on, September Mathias, E., Ilse, K.-R. & Jacob, W Pastoralists, local breeds and the fight for livestock keepers rights. Prepared for the PENHA 15th Anniversary Conference Pastoralism in the Horn of Africa: Surviving against all odds on Thursday, 29 September London. Mbuku, S.M., Isaac, S.K. & Alexander, K.K Identification systems and selection criteria of pastoral goat keepers in northern Kenya implications for a breeding programme. Conference on International Agricultural Research for Development. University of Bonn, October 11 13: Mesfin, T. & Obsa, T Ethiopian traditional veterinary practices and their possible contribution to animal production and management. Rev. Sci. Tech. Off. Int. Epiz., 13(2): Mishra, D. & Patro, L Ethno veterinary practices among the rural people of Ganjam district Orissa (India) : a case study on some common veterinary ailments. Bioscan., (3): Mohanty, R.B Traditional wisdom on livestock selection and management in folk proverbs of Orissa. Indian J. Tradit. Knowl., 3(1): Morse, D., DeLorenzo, M.A., Wilcox, C.J., Collier, R.J., Natzke, R.P. & Bray, D.R Climatic effects on occurrence of clinical mastitis. J. Dairy Sci., 71: Ndumu, D.B., Baumung, R., Hanotte, O., Wurzinger, M., Okeyo, A. M., Jianlin, H., Kibogo, H. & Sölkner, J Genetic and morphological characterisation of the ankole longhorn cattle in the African great lakes region.genet. Select. Evol., 40: Odubote, I.K. 1994b. Influence of qualitative traits on the performance of West African Dwarf goats. Nigerian J. Anim, Prod., 21: Olson, T.A., Lucena, C. & Chase Junior, C.C Evidence of a major gene influencing hair length and heat tolerance in Bos taurus cattle. J.Anim. Sci., 81: Ouma, E., Abdulai, A. & Drucker, A.G Pastoralists preferences for cattle traits: letting them be heard. pastoralism and poverty reduction in East Africa: Nairobi, Kenya, A Policy Research. Peters, K.J., Horst, P. & Kleinheisterkamp, H.H The importance of coat colour and coat type as indicator of productive adaptability of beef cattle in a sub-tropical environment. Prod. under Arid Conditions., 2: Raziq, A. & Younas, M Socioeconomic profile of camel in Suleiman mountainous region of Balochistan, Pakistan: in Recent trends in camelids research and future strategies for saving camels. In: Proc. International Camel Conference Rajasthan, India, February 16 17, pp Raziq, A., Younas, M. & Kakar, M.A Camel a potential dairy animal in difficult region of Balochistan, Pakistan: in recent trends in camelids research and future review Ethiopian e-journal for research and innovation foresight. Health Issue, 2(1) 48. Richard, S., Cabaret, J. & Cabourg, C Genetic and environmental factors associated with nematode infection of dairy goats in north western France. Vet. Parasit., 36(3 4): Sanusi, A.O Effects of coat colour genes on heat stress and tolerance to Haemonchus contortus among West African dwarf sheep. University of Agriculture, Abeo (M.Sc. Thesis). Shiferaw, G In-situ phenotypic characterization of Kereyu cattle type in Fentalle district of Oromia region, Ethiopia. M.Sc. Thesis. Presented to the School of Graduate Studies of Haramaya University, Haramaya, Ethiopia.

155 162 S. Banerjee et al. Shigdaf, M Performance evaluation of Washera, Farta and their crossbred sheep in the Western highland of Amhara Region, Ethiopia. Bahir Dar University, Ethiopia (MSc. Thesis). Singh, G. & Kumar, J Traditional knowledge on some less known wild edible plants used among Munda tribe of Jharkhand. Ecoscan., 6(3&4): Stewart, R.E Absorption of solar radiation by the hair of cattle. Agric. Eng., 34: 235. Tekleyohannes, B., Jamroen, T., Sayan, T., Girma, A., Asrat, T. & Somkiert, P Purposes of keeping goats, breed preferences and selection criteria in pastoral and agro-pastoral districts of South Omo Zone. Livestock Research for Rural Development. Volume 24, Article #213. Retrieved September 5, 2013, (available at lrrd.org/lrrd24/12/berh24213.htm). Tilki, M., İnal, S., Colak, M. & Garıp, M Relationships between milk yield and udder measurements in brown swiss cows. Türk Veterinerlik ve Hayvancilik Dergisi. 29: Winchester, A.M Heredity: an introduction to genetics. New Delhi, India, Oxford & IBH. Wurzinger, M., Ndumu, D., Baumung, R., Drucker, A., Okeyo, A.M., Semambo, D.K., Byamungu, N. & Sölkner, J Comparison of production systems and selection criteria of Ankole cattle by breeders in Burundi, Rwanda, Tanzania and Uganda. Trop. Anim. Health Prod., 38: Yakabu, A Path analysis of conformation traits and milk yield of Bunaji cows in small holder herds in Nigeria. Agricultura Tropica et Subtropica., 44(3): Yamazaki, T.H., Takeda, A., Nishiura, & Togashi, K Relationship between lactation curve and udder disease incidence in different lactation stages in first lactation holstein cows. Anim. Sci. J., 80: Young, S.K., McNaughton, L.R. & Spelman, R.J Hair whorl patterns are related to age at puberty and milk-production traits in dairy cattle. Proc. N. Z. Soc. Anim. Prod., 67:

156 Animal Genetic Resources, 2014, 54, Food and Agriculture Organization of the United Nations, 2014 doi: /s Phenotypic characterization and description of production systems of autochthonous sheep breeds in Kosovo H. Bytyqi 1, R. Baumung 2, H. Mehmeti 1 and B. Fuerst-Waltl 2 1 University of Prishtina Faculty of Agriculture and Veterinary, Prishtina, Kosovo; 2 Department of Sustainable Agricultural Systems, Division Livestock Sciences, University of Natural Resources and Life Sciences (BOKU), Vienna, Austria Summary Considering about 13.9 percent of total surface under pastures and meadows and big demand for milk and meat, sheep are considered as an important livestock species in Kosovo. Hence, the overall objective of this study was to provide the characterization of autochthonous sheep breeds and sheep production systems in Kosovo. Breed characterization and proper economic management schemes, selection for most economically important traits and best animals in the flock will assist farmers in increasing the feeding efficiency, reproduction and productive traits and thus profit. Including Bardhoka (BAR), Balusha (BAL), Sharri (SHA) and Kosva (KOS) sheep breeds, for a period of 1 year (September 2009 August 2010) the study was concentrated in 20 sheep farms in different regions of Kosovo. The least-squares means show that for milk yield and milk content, breed differences were significant (P < 0.05). The highest average daily milk yield for BAR (0.63 kg) exceeded the milk yield of BAL, KOS and SHA with 0.09, 0.14 and 0.18 kg milk per day, respectively. For milk butterfat and protein SHA ewes obtained the highest rate compared with other three breeds for percent. Milk dry matter among different breeds ranged between 18.4 and 19.3 percent. Voluntary disposal of ewes mainly occurred at lactation 4 and 5. Considering a standard yield at 1st lactation specific to the breeds, the highest production is reached at 3rd lactation. Keywords: breed characterization, economic traits, sheep breed, least-squares mean Résumé Compte tenu du fait que près du 13.9 pour cent de la superficie totale correspond à des pâturages et des prairies et qu il existe une grande demande en lait et viande, les moutons constituent une espèce animale importante au Kosovo. Ainsi l objectif principal de cette étude a été de caractériser les races ovines autochtones et les systèmes de production ovins au Kosovo. La caractérisation des races, la description des schémas de gestion économique les plus pertinents et la sélection des paramètres les plus importants sur le plan économique et des meilleurs animaux du troupeau serviront aux éleveurs pour augmenter l efficience alimentaire, les performances reproductives et productives, et par là même les bénéfices. L étude, qui a duré un an (de septembre 2009 à août 2010) et a été menée sur 20 exploitations ovines de différentes régions du Kosovo, a compris les races ovines Bardhoka (BAR), Balusha (BAL), Sharri (SHA) et Kosva (KOS). Les moyennes par moindres carrés ont décelé des différences significatives (P < 0.05) entre les races pour la production laitière et la composition du lait. La production moyenne journalière de lait la plus élevée de la race BAR (0.63 kg) a dépassé de 0.09, 0.14 et 0.18 kg journaliers de lait la production laitière des races BAL, KOS et SHA, respectivement. Pour ce qui est des taux butyreux et protéique du lait, les brebis SHA ont atteint les pourcentages les plus élevés, qui ont dépassé de 9.4 à 19.9 pour cent les trois autres races. La matière sèche du lait a varié entre 18.4 et 19.3 pour cent selon la race. La plupart des brebis sont retirées du troupeau à la quatrième ou cinquième lactation. Si la première lactation est considérée comme étant la production standard propre aux races, la production la plus élevée est atteinte à la troisième lactation. Mots-clés: race ovine, caractérisation raciale, moyenne par moindres carrés, paramètres économiques Resumen Teniendo en cuenta alrededor de un 13.9 percent de la superficie total bajo pasturas y praderas y gran demanda de leche y carne, ovino son considerados como una importante especie de ganado en Kosovo. Por lo tanto, el objetivo general de este estudio es proporcionar a la caracterización de las razas ovinas autóctonas y sistemas de producción ovina en Kosovo. Sistemas de gestión económica adecuada caracterización de razas y la selección de rasgos económicamente más importantes y sus mejores animales en el rebaño serán ayudar a los agricultores a aumentar la eficiencia de la alimentación, la reproducción y las características productivas y por lo tanto los beneficios. Incluyendo Bardhoka (BAR), Balusha (BAL), Sharri (SHA) y Kosva (KOS) razas ovinas, por un período de un año ( hasta agosto 2010), el estudio se concentró en 20 granjas de ovejas en diferentes regiones de Kosovo. Los medios mínimos cuadrados muestran que para la producción de leche y el contenido de la leche, las diferencias entre razas fueron significativas (P < 0.05). La producción diaria de leche promedio más alto para BAR (0.63 kg) superó la producción de leche de BAL, KOS y SHA, con 0.09, 0.14 y 0.18 kg de leche por día, respectivamente. Para la grasa de mantequilla leche y proteína de ovejas SHA obtener la tasa más Correspondence to: H. Bytyqi, Faculty of Agriculture and Veterinary, Animal Science Department, University of Prishtina, Rr Lidhja e Pejes 10000, Prishtina, Kosovo. hysen.bytyqi@uni-pr.edu 163

157 164 H. Bytyqi et al. alta en comparación con otras tres razas de 9.4 a 19.9 percent. Leche de materia seca entre diferentes razas varió entre 18.4 y 19.3 percent. Entrega voluntaria de las ovejas se produjo principalmente en la lactancia 4 y 5. Considerando un rendimiento estándar en primera lactación específica a la cría, la producción más alta se alcanza a tercera lactancia. Palabras clave: Raza ovina, breed caracterización, por mínimos cuadrados medios, los rasgos económicos Submitted 2 October 2013; accepted 7 January 2014 Introduction Maintaining local sheep breeds is a complex problem. It integrates economic, social, environmental and technical aspects, and involves many different actors. These actors are the direct users, such as farmers and consumers and other indirect users to benefit, such as tourism industry, extension services, etc. Domestic animals make a major contribution to human requirements for food in the form of meat, milk, milk products, eggs, fibre and fertilizer for crops as well as draught power. Approximately breeds drawn from 40 or more animal species are listed in this context (FAO, 2007). Nowadays, it is well known (Dillon et al., 2003; Hayes et al., 2003; Bytyqi et al., 2005) that animals best adapted to their environment, show the best overall trait performance (i.e. survival, production, reproduction, etc.). However, even though Kosovo is a rather small area, most breeds of sheep that originated in the different regions perform very differently and have gone through diverse selection focuses. Breed characterization, including selection and breed performance reactions to environment should thus be a part of all management and breeding decisions in sheep production enterprises. Kosovo has a long tradition in sheep farming, but like other sectors of Kosovo economy it has experienced a period of decline. Sheep production reached a peak during the beginning of the 1960s ( sheep). A dramatic decrease started after 1996 and accelerated during the war declining to a minimum of less than head, in the year Currently, the number of sheep has started to grow again and it is estimated to be of about animals (Ministry of Agriculture Forestry and Rural Development-MAFRD, 2012). Normally, the sheep are kept in the barn from the end of November until the end of March (winter period) independent of breed. The grazing period for sheep normally starts in April, close to farms. Mountain grazing time takes place from end of May until September, where sheep production is solely dependent on grass feed (summer period). After that, sheep graze on low land close to the barns again. Concentrated feed is given to the lactating sheep only during the winter period at the amount of kg per ewe. Small diversity of feed sources, poorly constructed barns, and low level of hygiene make intensive sheep production difficult. Owing to small size of land owned by the sheep farmers, about 1.5 ha land per farm (MAFRD, 2012), cultivation of such feedstuffs is often fragmented in 6 10 plots and located at large walking distances from the house. Following their normal seasonal state of lambing, sheep in Kosovo are seasonally bred (summer) and offspring are born at the time of December and January, mostly. So far, limited attempts were made to change this breeding practice in Kosovo. There is an obvious lack of education and awareness concerning sheep breeding and proper management despite the efforts made by local authorities and some of donor organizations in this regard. Farmers need to be trained to be familiar with breeding schemes specific to sheep farms in Kosovo in order to be qualified to manage their flocks and, in particular, to distinguish productive from non-productive animals. There are not yet, proper breeding and selection schemes that will assist farmers in increasing the feeding efficiency, reproduction and productive traits. Thus, their breeding strategy was focused clearly to the methodical selection criteria. All sheep breeds in Kosovo are classified in the long tail group and share triple purpose breed characteristics, milk meat wool (Bytyqi and Mehmeti, 2006). Milk production is in the focus of all sheep breeds in Kosovo but performance testing does not exist yet. Thus, a survey including detailed milk performance testing was carried out in eight sheep farms to describe the production level of ewes. Additionally, to receive a first insight into growth and non-production traits, 20 sheep farms (including the eight farms with milk recording) were integrated in an investigation regarding various other traits in sheep flocks. Hence, the overall objective of this study was to provide the necessary information regarding characterization of autochthonous of sheep breeds and sheep production systems in Kosovo. Material and methods Sheep production and management characteristics in Kosovo After cattle, the sheep production is the most important segment of the economy in rural households in Kosovo. The sheep production includes various breeds and

158 Production systems of autochthonous sheep breeds in Kosovo 165 categories of sheep. To date, sheep production in Kosovo can be clearly defined as an industry consisting of small and large farms ranging from 40 to ewes producing for home consumption and the market. The sheep production of Kosovo has many characteristics typical for sheep production in developing countries, as it depends almost exclusively on the resources locally available on the farm. Production characteristics Most of the quantity of sheep s milk produced in the Kosovo is turned into cheese, and is rarely consumed as fresh milk. Therefore, the quality of the sheep milk should refer to its ability to be transformed into milk products. Milk is frequently being processed following the traditional way of milking ewes only after the weaning period, which occurs about 90 days after lambing. Very little investments were made in regard to milking and processing infrastructure. Lambs and culling ewes are usually sold as a live animals and no specific commodity has so far been developed to increase added value of sheep meat. There is almost no farm that has any connection to milk or/and processing unit or market chain. The milk is being processed at farm gate and sold locally (green market), just a small proportion of farmers sell their products to retailer shops or gross shop houses. The usual price for cheese ranges from 5 to 6 /kg. Almost, there is no sheep meat process habit among farmers in Kosovo. Mostly, lambs (a part of those kept for breeding) are sold at an age of about 150 days, at spring time before the mountain-grazing period. The price ranges from 3 to 4 /kg live weight. Mature animals are being sold at no specific period at lower prices about 2.5 /kg live weight. Minimum attempts were made to promote sheep products as a tool to increase profitability and production. History of breeds of sheep in Kosovo The sheep population in Kosovo mainly consists of triplepurpose breeds of widely different body sizes and production capacities, i.e. Bardhoka (BAR), Balusha (BAL), Sharri (SHA) and Kosva (KOS). BAR sheep For centuries this sheep breed has been bred in the Southwest of Kosovo, mostly (Figures 1 and 2). The name originates from the word (Albanian language) Bardhe-White, as the fleece colour of this sheep is completely white. These sheep have a share of 19.6 percent of the total number of sheep bred in Kosovo (Table 1). The BAR sheep are kept for triple purpose, milk meat wool. However, through natural and selective breeding this sheep breed became one of the most yielding sheep for milk among the long tails sheep in the Balkan Peninsula (Cinkulov et al., 2008; Bytyqi, 2009). On average, the height at withers for mature animals is 64 cm. Although the grease fleece weight ranges between 2.5and 3.5 kg, it is of very poor quality with an average fibre diameter of Figure 1. Living area for different sheep breeds in Kosovo μm (Mehmeti et al., 2007). The males are horned and the females are polled, usually. This strain is well suited to a grazing production system, calm, easily milked and easy to work with. BAL sheep The area in which these sheep are kept and bred is also in the southwestern part of Kosovo sharing the same region with BAR breed, named Dukagjini plane (Figure 1). The name originates from the word (Albanian language) Bale-Spot. The total number of this breed is about animals bred in Kosovo, and covers about 7.8 percent of the total sheep population (Table 1). As in the case of BAR breed, BAL sheep have also been going through breeding selection with special focus on milk production. This strain is well adapted to natural grazing abilities and easy milking. BAL is bred as triple purpose sheep Figure 2. Bradhoka sheep breed.

159 166 H. Bytyqi et al. Table 1. Breed structure, number of sheep per breed and breed proportion. Sheep breed Number per breed Proportion (%) SHA BAR KOS BAL Others Total Source of data: Identification and registration unit of Kosovo (milk meat wool) of the medium size. The height at whithers for mature animals is on average about 65 cm. BAL sheep have a black head (Figure 3), whereas the colour of the fleece and the legs are white. The grease fleece weight for females and males averages and kg, respectively. The quality of the fibre is coarse with an average fibre diameter of 43 μm. Usually, the males are horned and the females are polled (Bytyqi, 2009). SHA sheep This strain is well adapted to natural hardiness, grazing abilities, and is well adapted to the cold mountain climates. The breeding region for these sheep is mainly in the Sharri Mountain, in the western part of Kosovo, from where the name originates (Figure 1). This breed represents approximately 53.4 percent of the total sheep population (Table 1). SHA sheep are primarily kept for milk production, although quality of carcass and wool gives this strain a triple purpose status. Regarding body size, SHA is one of the smallest breeds in Kosovo. The mean height at withers for mature animals is about 62 cm. SHA sheep have a white colour of the fleece,thefaceandthelegs(figure 4). The grease fleece weight for females and males averages and kg, respectively. The mean fibre diameter is 37 μm (Bytyqi and Mehmeti, 2006). The males are horned and the females are polled, usually. KOS sheep The name of this breed comes from the Kosovo plane, where it has its area of distribution (Figure 1). From the Figure 4. SHA sheep breed. total population of sheep this breed represents about 8.4 percent (Table 1). This strain has advantages of natural hardiness and grazing abilities. The Kosovo sheep is characterized with its triple milk meat wool, purpose. The tail is long; the colour of the head and the legs is black, while the fleece is of white colour (Figure 5). Some individuals can also be found having black or grey spotted faces. The grease fleece weight for males and females averages and kg, respectively. The mean fibre diameter is 37 μm (Bytyqi and Mehmeti, 2006). In accordance to the other sheep breeds of Kosovo males are normally horned while females are polled. The mean height at withers for mature animals is about 62 cm. Other sheep represent about 6.8 percent of the total population, mainly composed by unknown crosses. Data recording in selected herds For a period of 1 year (September 2009 August 2010), eight sheep farms in different regions of Kosovo were included in a study regarding dairy performance. Each sheep breed (SHA, KOS, BAR and BAL), was represented by two farms, respectively. The average herd size was 114, 94, 92 and 143 animals for SHA, KOS, BAR and BAL breeds, respectively. The milking sheep in the flock were selected randomly, resulting in milk test day yield (TDY) samples, which were taken during evening and morning milking. An amount of ml milk was put Figure 3. BAL sheep breed. Figure 5. Kosova sheep breed.

160 Production systems of autochthonous sheep breeds in Kosovo 167 in sterile preserved bottles (Azidol), stored in mobile coolers at a temperature of 4 C and transferred to the laboratory for analyses of milk contents. Recording cards containing information (i.e. animal ID, lambing date, lamb weight at different stages, test day milk yield, stillbirth, lambing characteristics, etc.) were developed and distributed in order to enable farmers to record further traits. Two types of scales (from 1 to 30 kg and 30 to 100 kg) were used to record lamb live weight at different stages (birth, 3, 6 and 12 months of age). The milk yield was measured individually by manual milking. For accurate performance recording, farmers were trained and monitored monthly. In order to receive more information for relevant traits (e.g. milk production, lamb growth rate, lambing characteristics, fertility traits, survival rate, mating type, feeding, price, costs, culling characteristics, etc.), farmers of 20 sheep farms (including the eight previously mentioned farms) were interviewed. Results were processed and stored electronically. Statistical data analysis All results referring to daily milk yield and component traits refer to actually recorded milk samples at the eight farms. All other averages and proportions represent results of interviews made in all 20 farms. For milk and milk composition the JMP-starter business unit program of SAS, (Institute SAS Inc. Sall, Lehman and Creighton, 2004) was used to analyse the data for the least-squares means and significant variances. Variations between different variables were tested using the Duncan test (Steel and Torrie, 1980). The same ewes were sampled for milk production during the whole year. Owing to small number of farms included, the only breed and lactation effects were included in the milk model. Results In Table 2, the results of the descriptive analysis for the four Kosovar sheep breeds in 20 different farms and different disposal reasons are shown. The proportion of ewes in different lactations depended on the percentage of culling for infertility, for involuntary or voluntary reasons. With small variations among different breeds, the ewes voluntary culling mainly occurred at 4th, 5th and 6th lactation, ranging from 4.6 to 26.0 percent. Culling for involuntary reasons occurred in 1st, 2nd and 3rd lactation, respectively and ranged from 2.4 to 7.8 percent. Culling for infertility tends to be constant across all lactations, ranging from 0.1 to 2.0 percent. Most of ewes in lactation >6 were disposed for voluntary reasons. Arithmetic means for milk yield and their standard deviations for lactations are listed in Table 3. The highest yield in the first lactation was recorded for BAR, followed by BAL, SHA and KOS breeds, yielding 0.62, 0.54, 0.42 and 0.41 kg/day milk. For all sheep breeds, highest milk yields were registered for the 3rd lactation with 0.72, 0.65, 0.51 and 0.49 kg/day for BAR, BAL, SHA and KOS sheep, respectively. Comparing milk yield between ewes in the 3rd and those in the 5th lactation, a decrease of about 22.9, 33.8, 33.3 and 20.6 percent for BAR, BAL, SHA and KOS breed becomes obvious. The least-squares means results (not shown) for lactation number differences, were significant (P = ) Based on milk yield results in eight farms and the interviews in 20 sheep farms, results of a descriptive analysis are stated in Table 4. Lactation production length ranged from 190 to 280 for different breeds. Owing to longer production, BAR and BAL ewes produce more milk in their 1st lactation (174.0 and kg) compared with KOS and SHA (82 and 80 kg, respectively). The age at first lambing and minimum days open were about the same in all breeds. Table 2. Proportions for different disposal reasons in the four sheep breeds of Kosovo according to interviews in 20 Kosovar farms. Breed Reason for disposal Proportion in lactations (%) >6 Total BAR Voluntary culling Involuntary culling Infertility Total culling BAL Voluntary culling Involuntary culling Infertility Total culling SHA Voluntary culling Involuntary culling Infertility Total culling KOS Voluntary culling Involuntary culling Infertility Total culling

161 168 H. Bytyqi et al. Table 3. The average daily milk yield and its standard deviation (SD; in brackets) per lactation in the breeds. Breed Average milk yield (SD) in kg/day in lactations >6 BAR 0.62 (0.01) 0.68 (0.02) 0.72 (0.02) 0.61 (0.01) 0.51 (0.01) 0.37 (0.02) 0.37 (0.02) BAL 0.54 (0.02) 0.61 (0.02) 0.65 (0.03) 0.52 (0.03) 0.43 (0.02) 0.36 (0.01) 0.33 (0.01) SHA 0.42 (0.04) 0.47 (0.05) 0.51 (0.05) 0.47 (0.04) 0.34 (0.04) 0.29 (0.04) 0.29 (0.04) KOS 0.41 (0.02) 0.43 (0.04) 0.49 (0.02) 0.43 (0.04) 0.34 (0.02) 0.31 (0.02) 0.31 (0.02) Depending on breed, stillbirth rate was reported to range between 1.8 and 2.5 percent, whereas survival rate between born live and mating for breeding lambs was about 94 percent. The proportion of male lambs kept for breeding for different breeds ranged between 6.8 and 20.0 percent. Proportion of single/twin/multiples born was for BAR (78.5/21.0/0.5 percent), BAL (77.0/20.0/3.0 percent), SHA (84.5/15.0/0.5 percent) and KOS (82.6/16.4/1.0 percent). Survival rate of ewes from the 1st to 2nd lactation was higher than 93.7 percent. Birth weight of lambs varies depending on breeds, type of lambing (single/tweens/triple) and gender, and ranges from 2.5 to 4.2 kg. For all breeds, the body weight is considered to be medium. The mature female individual on average weighs 65, 73, 67 and 65 kg for BAR, BAL, SHA and KOS, respectively. The ram mature live weights are 110.0, 113.0, and 95.0 kg. In Table 5, the least-squares means with corresponding standard errors and level of significance on test-day milk yield records are given. For milk production, breed differences were significant (P = ; Table 5). The highest average daily milk for BAR, exceeded the milk yield of BAL, KOS and SHA with 0.09, 0.14 and 0.18 kg milk per day, respectively (Table 5). Breed showed a significant effect on butterfat content in milk (P < 0.05; Table 5), with the highest percentage obtained by SHA ewes (6.68 percent), followed by KOS, BAL and BAR with 6.23, 6.05 and 5.51 percent, respectively. The SHA sheep also had significantly higher protein content in milk compared with other three breeds (P < 0.05; Table 5). SHA ewes had the highest protein rate on milk (5.49 percent), followed by BAL (5.29 percent), BAR (5.07) and KOS sheep (4.54 percent). For milk dry matter content breed differences were not significant (Table 5). However, BAL sheep produced slightly more (5.02, 4.92 and 3.88 percent) dry matter milk content compared with the other three breeds. Discussion Even though there have been some movements in improving overall management in recent years, sheep farms are managed mostly traditionally and new technologies take place at very minimum in the sheep industry in Kosovo. It should be pointed out that among all four breeds, definition of clear breeding objectives will be a need for the near future. Ewes can survive in the flock till the seventh lactation at maximum. This is a rather empiric selection method, which might be associated with a decrease in milk production and incapability of sheep to graze well (long walking distance) in their later lactations. Therefore, due to tradition, sheep farmers in Kosovo prefer to cull their sheep Table 4. Descriptive statistics per trait in the breeds based on interviews in 20 Kosovar farms. Trait (unit) Breeds BAR BAL SHA KOS Min. lactation length (day) Age at first lambing (day) Min. days open (day) st lactation yield (kg/lact.) Stillbirth rate (%) Single/twin/multiples (%) 78.5/21.0/ /20.0/ /15.0/ /16.4/1.0 Male lambs for breeding (%) Fattening period (day) Male birth weight single/twin/multiples (kg) 4.1/3.5/ /3.3/ /4.0/ /3.6/2.5 Female birth weight single/twin/multiples (kg) 3.9/3.3/ /3.1/ /3.6/ /3.5/2.5 Ewe mature weight (kg) Ram mature weight Survival rate between born live and mating (%) Survival rate ewes from 1st to 2nd lactation

162 Production systems of autochthonous sheep breeds in Kosovo 169 Table 5. Number of observations and least-squares means ± SD for breeds and the traits daily milk yield, as well as fat, protein and dry matter percentage (within a line different letters indicate significant differences, P < 0.05). Traits Breed (N ) BAR BAL SHA KOS Pr > F Milk yield N Milk yield (kg/day) 0.63 ± 0.04 a 0.54 ± 0.05 ab 0.45 ± 0.04 b 0.49 ± 0.05 b Milk contents N Fat (%) 5.51 ± 0.23 b 6.05 ± 0.22 ab 6.88 ± 0.20 a 6.23 ± 0.24 ab Protein (%) 5.07 ± 0.39 ab 5.29 ± 0.36 a 5.49 ± 0.33 a 4.54 ± 0.41 b Dry matter (%) ± 0.73 a ± 0.68 a ± 0.66 a ± 0.71 a voluntarily rather at higher ages than in other, more intensively kept sheep populations (e.g. Fuerst-Waltl and Baumung, 2009). At present, no artificial insemination is used in Kosovo sheep breeding. As a consequence, the breeding ram is supposed to breed about 28 ewes per breeding period. However, due to small populations per breed and narrow breeding replacement stock cycle (exchange of rams between small number of flocks usually grazing together for many years), inbreeding could be one area of concern. Compared with the milk yield at the 1st lactation, the highest production is assumed to be reached at the 3rd lactation. Ewes of BAR, BAL, SHA and KOS breed tend to produce more milk in the 3rd lactation compared with the 1st for about 16, 20, 21 and 5 percent, respectively. The decrease in average daily milk yield from the 3rd to 4th lactation ranged from 7.9 percent for SHA to 20.0 percent for BAL sheep. Consistently, for all breeds lactation further milk yield decreases occur after the 4th lactation. In some dairy breeds, an increase of lactation yields was reported until the 4th or 5th lactation (e.g. Ploumi, Belibasaki and Triantaphyllidis, 1998; Ploumi and Emmanouilidis, 1999; Fuerst-Waltl und Baumung, 2009). No information was available on individuals regarding the milk samples, therefore was not able to include the permanent environmental effect or a stage of lactation effect for daily milk yield. The milk production data show that BAR and SHA ewes produce more persistently all over the lactations compared with other two breeds (BAL and KOS sheep). Therefore, breeding programme of selecting individuals for a better persistent milk production in sheep could be one alternative of increasing profit of farmers of the latter breeds. Besides, milk content traits are important under Kosovar conditions as hardly any sheep milk but mainly cheese is consumed. The BAR breed was observed to produce the highest milk and SHA breed for protein and fat content. However, it is still below other breeds in the Balkan region. For Dalmatian Pramenka, fat and protein contents of higher than 7 and 6 percent were reported (e.g. Matutinovic et al., 2011). As Morand-Fehr et al. (2007) pointed out, sheep production systems based on grazing strongly affect milk composition and thus in turn the cheese produced. While an increased milk yield may be achievable by intensive farming, the typical cheese product may not be producible. Hence, farmers should aim at sufficient milk yield while also including milk content traits in their breeding goals. Apart from the genetics, management measures should be taken in order to guarantee certain standards. For some traits, i.e. lactation length, yield per lactation, there were huge variations between breeds living in Dukagjini valley (BAR and BAL) compared with SHA and KOS breeds managed under other regions in Kosovo. Although not far living distance from one region to another for all four breeds, according to Figure 1, it seems that a better environmental conditions (more fertile soil, irrigation of land) in Dukagjini valley has led BAR and BAL breeds to gradual milk yield improvement, for centuries. Apart from dairy production traits, farmers were also questioned with regard to functional traits in their herds. In generally, stillbirth rate was reported to be low having a positive economic impact on sheep farms in Kosovo. As to other breeds of Pramenka (long tail) sheep in Balkan Peninsula (Porcu et al., 2006), the twinning rate tends to be very low for all Kosovo breeds as well, therefore breeding improvements might take place in this trait, considering in details factors that affect this trait (inbreeding, feeding, etc.). Some other traits (i.e. days open, seasonal lambing, lambfattening period, proportion of male lambs sold for breeding, etc.) are set mainly due to sheep traditional management characteristics in Kosovo. The lambs for meat consumption are sold mostly as live animals and except a few visual indications and observations, no meat grading rate is taking place in Kosovo. Conclusions This is first approach used to get a first insight into the actual breed characterization and management situation

163 170 H. Bytyqi et al. of sheep farms in Kosovo. Sheep production in Kosovo seems to be characterized by being predominately semiextensive. In this environment, focus in breed selection should not only be only on increased production, but also on how well the breeds are fitting to the local environment. The establishment of clear breeding objectives and a recording system to enable breed characterization has been a challenge for all sheep breeders in Kosovo. The present study has considered describing a sheep-breeding system that meets the current farmer demand and is a great contribution for development of the sheep sector and rural economy as a whole. A further research for each trait is required in order to get a maximum profit from different sheep breeds and breeding systems in Kosovo. The sheep farmers showed detailed knowledge of the actual situation, therefore the prerequisites for collecting reliable data are considered to be very good. Acknowledgements The authors acknowledge the support of Austrian Development Agency (ADA and Kosovo Austrian Institutional Partnership (KAIP) project for fanatical support. Special thanks to farmers, KVFA laboratory staff and others who contributed to this research. References Bytyqi, H Within the Kugler Waltrud rare breeds and varieties of the Balkan Atlas. Monitoring Institute. Stuttgart, Germany, Heidehof Foundation, pp Bytyqi, H. & Mehmeti, H Identification and Conservation of Animal Genetic Resourcesin South Eastern Europe. Catalogue of West Balkan Pramenka Sheep Types Breed Bardhoka Strain. Skopje. ISBN , pp Bytyqi, H., Klemetsdal, G., Ødega rd, J., Mehmeti, H. & Vegara, M A comparison of the productive, reproductive and body condition score traits of the Simmental, Brown Swiss and Tyrol Grey breeds in smallholder herds in Kosovo. Anim. Genet. Res. Inf., 37: Cinkulov, M., Popovski, Z., Porcu, K., Tanaskovska, B., Hodžić, A., Bytyqi, H., Mehmeti, H., Margeta, V., Djedović, R., Hoda, A., Trailović, R., Brka, M., Marković, B., Važić, B., Vegara, M., Olsaker, I. & Kantanen, J Genetic diversity and structure of the West Balkan Pramenka sheep types as revealed by micro satelite and mitochondrial DNA analysis. J. Anim. Breed. Genet., 125(6): Dillon, P., Buckley, F., O Connor, P., Hegarty, D. & Rath, M A comparison of different cow breeds on a seasonal grass-based system of milk production. 1. Milk production, live weight, body condition score and DM intake. Livest. Prod. Sci., 83: FAO The state of the World s Animal Genetic Resources for Food and Agriculture in brief, edited by Daffyd Pilling & Barbara Rischkowski. Rome (accessible at a1250e/a1250e00.htm). Fuerst-Waltl, B. & Baumung, R Economic values for performance and functional traits in dairy sheep. Ital. J. Anim. Sci., 8: Hayes, B.J., Carrick, M., Bowman, P. & Goddard, E.M Genotype and environmental interaction for milk production of daughters of Australian dairy sires from test-day records. J. Dairy Sci., 86: Matutinovic, S., Kalit, S., Salajpal, K. & Vrdoljak, J Effects of flock, year and season on the quality of milk from an indigenous breed in the sub-mediterranean area. Small Rumin. Res., 100: Mehmeti, H., Hysen, B., Skender, M. & Imeri, B Breeding strategy for small ruminants in Kosovo. Scientific session. Indigenous breeds traditional products, as potential for economic growth rural areas, Albania, Kukes September. Ministry of Agriculture Forestry and Rural Development Mid-Term Evaluation Report of the ARDP Livestock numbers, pp Morand-Fehr, P., Fedele, V., Decandia, M. & Le Frileux, Y Influence of farming and feeding systems on composition and quality of goat and sheep milk. Small Rumin. Res., 68: Ploumi, K. & Emmanouilidis, P Lamb and milk production traits of Serrai sheep in Greece. Small Rumin. Res., 33: Ploumi, K., Belibasaki, S. & Triantaphyllidis, G Some factors affecting daily milk yield and composition in a flock of Chios ewes. Small Rumin. Res., 28: Porcu, K., Dzabirski, V., Popovski, Z., Andonov, S. & Tanaskovska, B Identification and Conservation of Animal Genetic Resourcesin South Eastern Europe. Catalogue of West Balkan Pramenka Sheep Types Breed Bardhoka Strain. Skopje. ISBN , pp Sall, J., Lehman, A. & Creighton, L JMP start statistics. A guide to statistics and data analysis using JMP and JMP IN software, 3rd edition. USA, SAS Institute, p Steel, R.G.D. & Torrie, J.H Principles and procedures of statistics. New York, McGraw-Hill.

164 Animal Genetic Resources, 2014, 54, Food and Agriculture Organization of the United Nations, 2014 doi: /s Evaluation des pratiques paysannes de conservation in situ du taurin Baoulé au Sud-Ouest du Burkina Faso L.Y. Mopaté 1, M.J-B. Kamuanga 2, S. Hamadou 3 and C-Y. Kaboré-Zoungrana 4 1 Laboratoire de Recherches Vétérinaires et Zootechniques de Farcha (LRVZ), B.P: 433 N Djaména (Tchad), chercheur associé au LERNSE de l UPB, Burkina Faso. mopate.ly@gmail.com; 2 Consulting Agricultural Economist Whispering Wind Lane Manassas, VA (USA). Ancien Responsable de l Unité Elevage et Environnement au Centre International de Recherche-Développement sur l Elevage en zone Subhumide (CIRDES), Bobo-Dioulasso (Burkina Faso); 3 Ancien chercheur à l Unité Elevage et Environnement au CIRDES, Bobo-Dioulasso (Burkina Faso); 4 Laboratoire d études et de recherches des ressources naturelles et des sciences de l environnement (LERNSE), Université Polytechnique de Bobo-Dioulasso (UPB), Burkina Faso Résumé L étude a évalué les pratiques paysannes mises en œuvre par les éleveurs pour assurer la conservation du taurin Baoulé. Elle a porté sur 52% des 636 producteurs recensés, répartis dans 54 villages ciblés qui constituent 7% du total des villages des provinces de Poni et de Noumbiel au Burkina Faso. L enquête transversale et rétrospective doublée des observations des pratiques a été réalisée. Les modalités des fréquences des actes et des coefficients qui y sont affectés ont été transformées en variables quantitatives. L adhésion à un éventuel projet de conservation, à la vie associative et la scolarisation induisent des actes de conduite et de gestion, de soins de santé et des activités sociales positifs à la conservation mais négatifs à la reproduction (croisement). L appartenance à l ethnie Lobi contribue à la conservation du Baoulé. La possession d un attelage favorise les actes de conduite et gestion, de soins de santé et des activités sociales conservatoires mais les actes de reproduction ne contribuent pas à la conservation. De plus, la possession d un troupeau exclusif de Baoulé, la castration d autres mâles non Baoulé, la prise individuelle des décisions de gestion et l appartenance à l ethnie Lobi sont favorables à la conservation du Baoulé. Le bilan global positif des pratiques mises en œuvre en pays Lobi burkinabè montre un faible développement du métissage du Baoulé. Les variables quantitatives obtenues par transformation constituent une nouvelle approche, permettant de quantifier les pratiques de conservation des races animales en milieu naturel. Mots-clés: Pratiques Conservation in situ, Taurin Baoulé, Pays Lobi, Burkina Faso Abstract The study evaluated the small farmers practices implemented by cattle breeders to ensure the conservation of the Baoulé breed. The study involved 52% of the 636 known producers from the 54 targeted villages, which constitute 7% of the total villages of the Poni and Noumbiel provinces in Burkina Faso. The transverse and retrospective survey was carried out and was combined with direct observation. The modalities of the frequencies of various actions and the coefficients that affected them were transformed into quantitative variables. Adhesion to a possible project of conservation, participation in farmers groups and other associations and schooling induce positive acts of conduct, health care and social activities for conservation but have a negative effect on reproduction (cross breeding). Membership of the Lobi ethnic group contributes to the conservation of Baoulé cattle. The possession of draught animals promotes the activities of conduct, management, health care and social activities that are protective but active management of reproduction does not contribute to conservation. Moreover, the possession of an exclusive herd of Baoulé cattle, the castration of males from other cattle breeds, the taking of individual decisions of management and the membership of Lobi ethnic group are all favourable to the conservation of Baoulé cattle. The positive total assessment of the practices implemented in the Lobi region of Burkina Faso shows the limited development of Baoulé cattle crossbreeding. The quantitative variables obtained by transformation constitute a new approach, making it possible to quantify the practices of conservation of the animal breeds in their natural environment (in situ). Keywords: In situ conservation practices, Baoulé taurine, Lobi Region, Burkina Faso Resumen El estudio ha evaluado las prácticas implementadas por los ganaderos para asegurar la conservación del ganado bovino Baoulé. El estudio fue llevado a cabo con el 52 por ciento de los 636 productores censados, repartidos entre 54 pueblos que constituyen el 7 por ciento del total de pueblos de las provincias de Poni y Noumbiel en Burkina Faso. Se realizó una encuesta transversal y retrospectiva pareada sobre las prácticas observadas. Las modalidades de acción y los coeficientes correspondientes fueron transformados en variables cuantitativas. La adhesión a un determinado proyecto de conservación, a la vía asociativa y a la escolarización conlleva acciones de dirección y gestión, de atención sanitaria y de actividades sociales, positivas para la conservación pero negativas para la reproducción (cruzamiento). La pertenencia a la etnia Lobi contribuye a la conservación del ganado Baoulé. La posesión de una yunta favorece las acciones de dirección y gestión y de atención sanitaria y las actividades sociales conservacionistas pero las acciones Correspondence à envoyer à: L.Y. Mopate, Laboratoire de Recherches Vétérinaires et Zootechniques de Farcha (LRVZ), B.P: 433 N Djaména (Tchad), chercheur associé au LERNSE de l UPB, Burkina Faso. Adresse électrorique: mopate.ly@gmail.com 171

165 172 L.Y. Mopate et al. de reproducción no contribuyen a la conservación. Asimismo, la posesión de un rebaño exclusivamente de raza Baoulé, la castración de los machos de otras razas, la toma individual de las decisiones de gestión y la pertenencia a la etnia Lobi ayudan a la conservación del ganado Baoulé. El balance global positivo de las prácticas implementadas en el territorio de los Lobi en Burkina Faso muestra un escaso mestizaje del ganado Baoulé. Las variables cuantitativas obtenidas por transformación representan una nueva aproximación que permite cuantificar las prácticas de conservación de las razas animales en su medio natural. Palabras clave: prácticas de conservación in situ, ganado bovino Baoulé, territorio de los Lobi, Burkina Faso Soumis: 18 L octobre 2013; admis: 10 Le janvier 2014 Introduction En Afrique de l Ouest, notamment au Burkina Faso, les menaces d absorption (par le sang zébu) des taurins à courtes cornes des savanes ont été évoquées par plusieurs auteurs (CIPEA, 1979; Thiombiano, 1993; Rege, Aboagye et Tawah, 1994; Camus, Landais et Poivey, 1981; Lhoste, 1995; Sigué et Kamuanga, 1997; Moazami Goudari et al., 2001). Ces menaces s inscrivent dans un processus global des perturbations climatiques intervenues depuis les années 1970 et Des changements dans les zones pastorales ont été observés, avec un déplacement progressif de l élevage zébu des zones du Nord vers les zones du Centre et du Sud (Maillard et al., 1992). En plus des sécheresses, s y ajoute la pression foncière croissante dans le Nord pour le Burkina Faso (Ouédraogo, 2002). La concentration importante de la biomasse animale dans ces zones constitue une menace pour les races taurines adaptées à leur milieu. En effet, la cohabitation étroite entre zébus et taurins favorise des croisements passifs. De plus, pour accroître les formats des taurins sources de prestige aux yeux des paysans, améliorer l aptitude au travail et le rendement en viande, les agriculteurs des savanes se sont lancés dans des croisements actifs depuis les décennies 1980 et 1990 (Landais, 1983; Chupin, 1994). Par ailleurs, la multiplication des animaux trypanotolérants constitue l une des stratégies d utilisation et d exploitation des parcours infestées par les glossines (Achukwi et Musongong, 2009). Les travaux d identification des pratiques et stratégies paysannes de conservation in situ du taurin Baoulé (Figure 1) au Burkina Faso sont rares. Ce taurin constitue le seul bovin trypanotolérant bien implanté en pays Lobi de ce pays (Maillard et al., 1992; Thiombiano, 1993). Il est admis que l étude et l évaluation des pratiques concrètement mises en œuvre par les agriculteurs s imposent et constituent un passage obligé, pour qui s intéresse aux conséquences matérielles de l activité agricole dans un milieu donné (Landais et Balent, 1993). Il est donc indispensable pour mieux appréhender la reproductibilité de systèmes d élevage du taurin Baoulé dans le Sud-Ouest du Burkina, de mener des investigations dans ce sens afin de mieux cibler les actions de développement. L objet de l étude est d identifier et d évaluer les pratiques paysannes mises en œuvre par les éleveurs du taurin Baoulé, dans la gestion individuelle ou collective des animaux pour assurer la conservation de cette race. Matériel et Méthodes Site d étude La région d étude au Sud-Ouest du pays est la partie burkinabé du pays «Lobi», nom de l ethnie dominante à cheval entre le Burkina Faso, la Côte d Ivoire et le Ghana. Elle est considérée comme le berceau du taurin Baoulé appelé encore bovin «Lobi». Elle est limitée à l Est par le fleuve Mouhoun, frontalier avec le Ghana, à l Ouest par la province de la Comoé, au Nord par celle du Bougouriba et au Sud par la Côte d Ivoire (Figure 2). Elle est située entre 9 25 et de latitude nord et entre 2 30 et 4 de longitude ouest et couvre km², soit 3,2% du territoire national (Kienou, Sanou et van Bronswijk, 1996). Elle comprend les provinces du Poni et Figure 2. Figure 1. Troupeau taurin Baoulé dans la zone de Gaoua (cliché: MOPATE). Situation de la zone d étude et principale aire de répartition du taurin Baoulé au Burkina Faso, en Côte d Ivoire et au Ghana en Afrique de l Ouest. (Fond gratuit de Carte Google Map).

166 Evaluation des pratiques paysannes de conservation in situ du taurin Baoulé 173 du Noumbiel avec respectivement Gaoua et Batié pour chefslieux. La région est à une altitude moyenne de 400 m. Le climat est de type soudanien, caractérisé par l alternance d une période pluvieuse de mai à octobre et d une période sèche de novembre à avril (Guinko, 1984). Les températures diurnes moyennes varient entre 24,9 C et 30,2 C, avec une amplitude moyenne de 5,3 C. C est une zone agroclimatique subhumide avec une pluviométrie variant de 900 mm à 1400 mm entre 1992 et 2001 (Lankoandé, 2002). Echantillonnage et méthodes de collecte de données La base de sondage a été constituée de 636 propriétaires détenteurs surtout de bovin Baoulé, recensés dans les deux provinces. Sur cet effectif, 52% ont été choisis au hasard dans 54 villages ciblés, soit 7% du total de 769 villages dans les deux provinces. Les axes principaux à partir de Gaoua et Batié ont été concernés dans un rayon moyen de 25 km autour de chaque centre, afin d accroître la variabilité des situations tout en limitant le coût de la logistique. L objectif final a été d établir un échantillon assez large (>300 répondants) pour améliorer la représentativité de la population détentrice de bétail Baoulé et prendre en compte les refus et absences au démarrage des enquêtes. L étude a été réalisée par enquête informelle et formelle complémentaires et appropriées pour améliorer la connaissance des systèmes d exploitation agricole (Hubert, 1992; Vianney et Palm, 1999). Ces enquêtes ont été menées en deux phases distinctes: Une pré-enquête pour recenser les propriétaires du taurin Baoulé afin de disposer d une base de sondage; consulter et discuter avec des groupes de personnes afin de connaître le milieu et les pratiques de production et réaliser une étude bibliographique sur le milieu. Une enquête (formelle) transversale et rétrospective, doublée des visites multiples pour améliorer les estimateurs. Des interviews et observations auprès d un échantillon d exploitants propriétaires du taurin Baoulé ont été menées. Les questions fermées ont porté sur les pratiques d élevage liées à la conservation in situ. Ils est agit ici d identifier au cours de l année qui s est écoulée (12 derniers mois), les actes posés par l éleveur en terme de pratiques réalisées ou des événements observés en relation avec la conservation de son taurin Baoulé. Le nombre d animaux concernés et les fréquences de ces événements et actes posés. Ainsi, quatre groupes de pratiques ont été visés: Pratiques de conduite et de gestion: elles renfermaient le pâturage, l abreuvement, la traite, la traction animale, la complémentation minérale, en fourrages/résidus et le parcage des animaux. Pratiques de reproduction: elles ont été subdivisées, en événements observés relatifs aux avortements survenus, aux morts-nés, aux naissances vivantes, aux mortalités naturelles et en un témoignage de l éleveur. Ces témoignages ont concerné la monte d une vache Baoulé par un taureau Baoulé, d une vache Baoulé par un taureau Zébu ou Métis, un croisement favorisé de façon active entre Zébu ou Métis et Baoulé, tout comme l abattage, la réforme, l achat de Zébu. Les croisements passifs (mélange des troupeaux, divagation et confiage) ont été concernés, tout comme les actes encourageant la conservation du Baoulé par l achat d un taureau pur, la castration des mâles de race Zébu ou Métis du troupeau. Pratiques sanitaires: elles ont été orientées sur la participation à la lutte contre les glossines, les traitements aux trypanocides, aux anthelmintiques, aux acaricides et aux soins traditionnels. Activités sociales liées à l utilisation des animaux tels que l abattage des taureaux pour les sacrifices, ceux remis en dot, en dons (divers), présentés aux concours d éleveurs, les vols enregistrés et les contacts ou l accessibilité à des agents d élevage. Analyse de données Les saisies et les traitements ont été effectués par le logiciel Statistical Package of Social Science (SPSS). Un codage en modalités de quelques variables issues des questions ouvertes a été effectué. Les pratiques (événements et actes) ont été quantifiées par des coefficients allant de 3 à + 3, selon le poids «objectif» que la pratique accomplie apporte en termes d impact favorable ou pas pour la conservation in situ de la race. L importance d un acte est la variable quantitative résultante de la multiplication du coefficient affecté à l acte (variant de 3 à + 3 suivant sa contribution à la conservation), par la modalité de la fréquence de l acte posé (aucune fois = 0; moins fréquent = 1; fréquent = 2 ou très fréquent = 3). Les avortements, naissances vivantes et mortalités qui sont considérés comme plus subis par les chefs d exploitations et les actes posés (conduite au pâturage et abreuvement) régulièrement réalisés dans la quasi totalité des exploitations et touchant également la quasi totalité des effectifs, n ont pas été pris en compte. Seule l importance des autres actes de conduite et gestion (traite, complémentations, traction et parcage), de reproduction (croisements) et de conservation, de soins de santé et d activités sociales a été considérée. Une agrégation générale de l importance des ces actes posés seuls et des pratiques (événements et actes compris) a été effectuée, afin de tirer des bilans généraux (positifs ou négatifs) de la conservation au niveau de la région. Un premier tri à plat a été effectué pour sélectionner des variables à croiser, à soumettre en analyse de variance (ANOVA). Le test de Bonferroni a été appliqué pour les comparaisons multiples (plus de deux groupes) des moyennes. La régression linéaire simple des déterminants de la participation ou contribution financière à un éventuel projet de conservation et de l importance des actes posés par les Chefs d Exploitation a été effectuée. Pour cela, la méthode d estimation par les moindres carrés ordinaires (MCO) a été utilisée. L adéquation d ensemble du

167 174 L.Y. Mopate et al. Tableau 1. Importance des actes entre les Chefs d Exploitation (CE) avec troupeaux purs et mixtes et entre ceux décidant seul de la gestion unique et à plusieurs. Importance des actes Type de troupeau Centre de décision Pur Mixte P Unique Multiple P N = 287 N = 38 N = 278 N = 47 Conduite 2,83 ± 3,18 7,81 ± 3,91 *** 3,65 ± 3,62 2,03 ± 3,49 ** Reproduction 3,27 ± 3,54 4,73 ± 6,34 *** 2,16 ± 4,83 3,36 ± 4,83 ns Santé 1,04 ± 1,72 4,57 ± 2,07 *** 1,56 ± 2,17 0,82 ± 1,44 * Activités sociales 0,39 ± 2,06 1,89 ± 2,58 *** 0,70 ± 2,16 0,23 ± 2,07 ** ***p < 0,001; **p < 0,01; *p < 0,05; ns = non significatif. modèle est mesurée par le coefficient de détermination (R²), le R² ajusté et le test de Fisher. Le modèle fonctionnel obtenu permet de légitimer l impact des variables explicatives considérées prises individuellement sur la participation ou l importance des actes posés qui sont des variables à expliquer. On considère l hypothèse alternative qu au moins un des coefficients du modèle est différent de zéro. Les variables explicatives des déterminants de la participation financière ont été choisies parmi les données socio-économiques des CE et de leur exploitation. Quant à celles des déterminants de l importance des actes des CE, elles découlent des critères retenus pour leur analyse. Résultats Importance des actes posés par les éleveurs Troupeaux purs ou mixtes et prise des décisions de gestion unique ou multiple Les moyennes de l importance des actes de conduite et gestion, des soins de santé et des activités sociales sont significativement plus élevées dans les exploitations avec des troupeaux mixtes que dans celles avec des troupeaux purs. En revanche, les actes en matière de reproduction ont une moyenne significativement plus élevée dans les troupeaux purs (Tableau 1); où il y a absence des croisements avec d autres races ou des actes favorisant activement ou passivement ces croisements. Pour la nature des centres de décision, les moyennes sont plus élevées et significatives pour les décisions de gestion relevant uniquement du propriétaire en matière de conduite et gestion, de soins de santé et des activités sociales. Degré de participation 1 à la conservation et appartenance ou pas à un groupement Pour les actes de conduite et gestion, des soins de santé animale et des activités sociales liées à l élevage du Baoulé, l importance de ces actes est significativement plus élevée chez les CE ayant annoncé une participation élevée à un 1 Le degré de participation à un éventuel projet de conservation du taurin Baoulé est la volonté exprimé par les chefs d exploitation d y contribuer de façon élevée (au moins F CFA) ou faible (moins de F CFA) à ce projet dans les deux provinces concernées. 1U$ = 500 F CFA projet éventuel de conservation du Baoulé. Il en est de même des actes de conduite et gestion et les soins de santé des CE membres d un groupement (Tableau 2). Niveau d alphabétisation/d instruction des CE et leur catégorie d âge Les CE alphabétisés et scolarisés ont significativement les moyennes les plus élevées pour les actes de conduite et gestion, de soins de santé et des activités sociales. En revanche, la moyenne des actes liés à la reproduction pour ce groupe est plus faible et négative que celle des non-alphabétisés (Tableau 3). Les CE âgés de moins de 40 ans posent plus d actes des soins de santé que ceux de 40 ans et au-delà, dont les actes de reproduction sont significativement plus élevés. Durée dans la pratique de l élevage et appartenance ou pas à l ethnie Lobi Les exploitations qui ont moins de 15 ans de pratique de l élevage bovins ont des moyennes de l importance des actes significativement plus élevées en conduite et gestion et en activités sociales (Tableau 4). Le groupe Non-Lobi (migrants) pose significativement plus des actes de conduite, de santé et des activités sociales. En revanche, les actes liés à la reproduction des bovins y sont négatifs. Le groupe Lobi a la moyenne la plus élevée pour les actes de reproduction. Possession ou pas d un attelage dans l exploitation et provinces concernées Les chefs d exploitation (CE) possédant un attelage dans leur exploitation ont significativement les moyennes les plus élevées pour les actes de conduite et gestion, de soins de santé et des activités sociales (Tableau 5). En revanche, ceux sans attelage ont la moyenne de l importance des actes de reproduction plus élevée. Selon les provinces, seuls les actes de reproduction montrent une différence significative. Noumbiel a effectivement la moyenne la plus élevée par rapport à Poni. Déterminants de l importance des actes posés par les Chefs d Exploitation Le test de Fisher (F) pour le modèle de régression linéaire simple utilisé est significatif au seuil de 1%. Ce qui indique que l ensemble de coefficients du modèle est

168 Evaluation des pratiques paysannes de conservation in situ du taurin Baoulé 175 Tableau 2. Importance des actes entre CE annonçant une participation élevée et moins élevée et entre CE appartenant à un groupement ou pas. Importance des actes Participation à la conservation Membre de groupement Élevée Faible P Oui Non P N = 137 N = 188 N = 208 N = 117 Conduite 4,73 ± 3,82 2,45 ± 3,18 *** 4,11 ± 3,67 2,17 ± 3,24 *** Reproduction 2,03 ± 4,66 2,56 ± 4,77 ns 2,29 ± 5,13 2,41 ± 3,90 ns Santé 2,16 ± 2,44 0,95 ± 1,67 *** 1,83 ± 2,25 0,80 ± 1,54 *** Activités sociales 1,02 ± 1,92 0,22 ± 2,29 ** 0,57 ± 2,51 0,54 ± 1,41 ns ***p < 0,001; **p < 0,01; ns = non significatif. Tableau 3. Importance des actes posés entre CE alphabétisés ou pas et entre CE de moins de 40 ans et de 40 ans et plus. Importance des actes Alphabétisation Catégorie d âge Oui Non P Moins de 40 ans 40 ans et plus P N = 47 N = 278 N = 208 N = 117 Conduite 6,19 ± 4,11 2,94 ± 3,34 *** 4,07 ± 4,12 3,29 ± 3,52 ns Reproduction 0,48 ± 6,55 2,82 ± 4,17 *** 1,09 ± 4,06 2,57 ± 4,81 * Santé 3,40 ± 2,54 1,13 ± 1,82 *** 1,92 ± 2,50 1,37 ± 2,00 * Activités sociales 1,65 ± 2,38 0,38 ± 2,09 *** 0,98 ± 1,70 0,48 ± 1,70 ns ***p < 0,001; **p < 0,01; *p < 0,05. Tableau 4. Importance des actes des CE de moins de 15 ans de pratique d élevage et de 15 ans plus et entre ceux Lobi et Non-Lobi. Importance des actes Durée dans la pratique d élevage Groupe ethnique de 15 ans 15 ans et + P Lobi Non-Lobi P N = 173 N = 152 N = 313 N = 12 Conduite 3,93 ± 3,85 2,82 ± 3,30 ** 3,25 ± 3,56 7,66 ± 3,31 ** Reproduction 2,30 ± 4,21 2,38 ± 5,26 ns 2,66 ± 4,39 6,08 ± 5,58 ** Santé 1,63 ± 2,20 1,28 ± 1,96 ns 1,31 ± 1,98 5,16 ± 1,58 ** Activités sociales 0,79 ± 1,89 0,30 ± 2,4 * 0,50 ± 2,16 2,08 ± 2,19 ** ***p < 0,001; **p < 0,01; *p < 0,05. Tableau 5. Importance des actes entre CE possédant un attelage et ceux n en possédant pas et entre les provinces. Importance des actes Possession d attelage Province Oui Non P Poni Noumbiel P N = 121 N = 200 N = 168 N = 157 Conduite 6,45 ± 3,41 1,61 ± 2,35 *** 3,67 ± 3,72 3,14 ± 3,54 ns Reproduction 0,69 ± 5,59 3,31 ± 3,72 *** 1,76 ± 5,45 2,96 ± 3,72 * Santé 2,85 ± 2,16 0,63 ± 1,55 *** 1,43 ± 2,17 1,49 ± 2,01 ns Activités sociales 1,45 ± 2,12 0,03 ± 2,03 *** 0,38 ± 2,25 0,76 ± 2,07 ns ***p < 0,001; *p < 0,05; ns = non significatif. significativement différent de zéro. L équation du modèle de l importance des actes (IMPACTES) est: Impactes = ß0 + ß1 AGE + ß2 ATT + ß3 MEM + ß4 PTP + ß5 TTB + μ où les ß sont des paramètres à estimer et μ le terme d erreur ou de perturbation aléatoire. AGE = âge; ATT = attelage; MEM = membre; PTP = participation; TTP = total bovin La valeur de R² ajusté montre que les variables considérées expliquent 28% de l importance des actes (Tableau 6). Les variables attelage, participation et membre ont une influence significative et positive sur l importance des actes. Elles ont une influence individuelle respective de 99% et 95% dans l importance des actes. En revanche, l âge et le nombre total de bovins de l exploitation ont une influence significative mais négative. Ces variables influencent respectivement l importance des actes dans 90% et 95% des cas.

169 176 L.Y. Mopate et al. Tableau 6. Estimation de l importance des actes des CE dans l ensemble des deux régions. Variable Coefficients t de student Constance 3,625** 2,418 Age (AGE) 0,040* 1,711 Attelage (ATT) 4,988*** 7,979 Membre (MEM) binaire 1,402** 2,226 Participation (PTP) binaire 0,004*** 4,559 Total bovin (TTB) 0,048** 2,188 F 22,342*** R² 0,297 R² ajusté 0,283 ***p < 0,001; **p < 0,01; *p < 0,05. Bilan des actes et des pratiques (événements et actes) au niveau régional La moyenne de l importance générale des actes est de 7,78 ± 6,28, avec un minimum = 5 et un maximum = 30. Celle des pratiques (événements et actes posés) est de 8,83 ± 6,93, avec un minimum = 8 et un maximum = 34. Cette moyenne générale positive indique que les actes posés participent globalement à la conservation du Baoulé. La différence de moyenne entre la province du Poni (7,52 ± 5,88) et celle du Noumbiel (8,35 ± 6,68) n est pas significative au seuil de 5%. Il en est de même de celle des pratiques à Poni (8,56 ± 6,69) et à Noumbiel (9,66 ± 7,44). Corrélation de l importance des actes entre eux Les actes de conduite et gestion sont corrélés positivement aux actes de santé et des activités sociales. Il en est de même entre actes de santé et activités sociales (Tableau 7). En revanche, les actes liés à la reproduction sont négativement corrélés avec tous les autres actes. Discussion Les variables quantitatives obtenues par transformation des modalités des fréquences des actes et des coefficients qui y sont affectés sont une nouvelle approche permettant de quantifier les pratiques des producteurs. Ici, il s agit des pratiques de conservation des races animales, notamment le bovin Baoulé, en milieu naturel. Les coefficients affectés bien que subjectifs, restent basés sur une connaissance des actes et leurs poids respectifs dans la conservation in situ de la race. Toutefois, ils gagneraient à être davantage améliorés par une investigation préliminaire auprès des éleveurs et autres producteurs ou agents de terrain. Ce qui permettrait d élucider le poids réel accordé à tel ou tel acte préjudiciable ou favorable à la préservation de la race, afin de parvenir à une certaine objectivité. L aspect quantitatif des coefficients permet de faire un bilan global (négatif ou positif) des pratiques conservatoires d une race dans une zone, de discriminer les acteurs, d identifier les liaisons ou les déterminants de certains actes posés et leur impact sur la conservation en milieu naturel. Le bilan global positif des actes seuls et celui des pratiques en pays Lobi burkinabè montre un faible développement du métissage du Baoulé. Les actes de croisements interraciales (Zébus/Métis X Baoulés) ou ceux favorisant activement ou passivement ces croisements y sont très faibles. La moyenne des actes de reproduction plus élevée à Noumbiel a été en rapport avec la faible présence des métis dans cette province. La conservation de la race Baoulé y serait mieux assurée. L importance des actes (conduite, santé et activités sociales) plus marquée chez les Chef d Exploitation (CE) avec des troupeaux mixtes serait liée à leur pouvoir d achat. En effet, seuls les moyens financiers permettent de complémenter (achat de minéraux et fourrages) les animaux, de côtoyer les agents vétérinaires pour leurs soins et de s équiper en traction animale. Les corrélations positives entre les actes de conduite et de santé (r = 0,607), entre actes de santé et activités sociales (r = 0,500) et entre la conduite et les activités sociales (r = 0,379) expliquent les liens entre ces actes. Les contacts pris avec les agents de l élevage expliquent l importance des activités sociales de ceux avec des troupeaux mixtes. Une certaine similarité se dégage dans l importance des actes posés entre ce groupe et les migrants. Ces derniers ont surtout des troupeaux mixtes et disposent également des moyens. Cette possibilité explique aussi l importance élevée de ces actes chez les CE qui possèdent un attelage. Néanmoins, les actes de reproduction plus faible dans ce groupe témoignent de la pratique du métissage. Il en est de même des CE avec des troupeaux mixtes et les migrants, où la cohabitation de plusieurs races dans les exploitations entraîne des croisements actifs et passifs, préjudiciables à la préservation. La préférence affichée des autochtones du pays Lobi burkinabè pour le Baoulé (Tano et al., 2001) et la réticence des Lobi ivoiriens (Camus, Landais Tableau 7. Corrélation de l importance des actes entre eux. Actes Conduite/Gestion Reproduction Santé Activités sociales Conduite/Gestion 1,000 0,366*** 0,607*** 0,379 Reproduction 0,366*** 1,000 0,391*** 0,231 Santé 0,607*** 0,391*** 1,000 0,500 Activités sociales 0,379*** 0,231*** 0,500*** 1,000 ***p < 0,001.

170 Evaluation des pratiques paysannes de conservation in situ du taurin Baoulé 177 et Poivey, 1981) et Burkinabé (Maillard et al., 1992) au métissage contribuent à accréditer le faible penchant de la grande majorité des Lobi pour le métissage. Toutefois, l émergence des autochtones avec des troupeaux mixtes (à visées spéculatives) constitue des menaces réelles qui planent sur le Baoulé (Mopaté, 2003). En effet, des études récentes dans les régions Nord de la Côte d Ivoire proche de la province de Noumbiel, indiquent que 86 à 90 % des élevages présentent des signes de métissage (Sokouri et al., 2007; Sokouri et al., 2009). L amélioration du format des Baoulé qui diffère significativement du Méré (croisé Zébu et Baoulé) en est la raison. Ce Métis utilisé comme animal de trait, se vend deux fois plus que le Baoulé. La prise des décisions de gestion par une seule personne a une influence positive sur les actes de conduite, de santé et les activités sociales. Ce qui confirme que les pratiques restent liées à l opérateur (Teissier, 1979) qui agit selon les références qu il s est construites par expériences personnelles et celles communiquées par les membres de sa communauté (Landais, 1992; Pleine, Schmithüsen et Sorg, 1995; Lalba, Badini et Kleene, 1999). On ressent cette influence globalement au niveau de la région où domine le groupe ethnique Lobi. Les membres de ce groupe appartiennent à la même communauté et gèrent seuls en majorité leur troupeau. Les actes de reproduction positifs chez les Lobi indiquent qu ils préservent mieux la race Baoulé du métissage, contrairement aux Non-Lobi caractérisés par des actes de reproduction négatifs. Ces actes affectent gravement la conservation de la race dans ce groupe. Les peuls gardiens des bovins Baoulé des autochtones appartiennent au groupe Non-Lobi. La multiplication des centres de décision de gestion par des acteurs (propriétaire, copropriétaire et gérant) appartenant à différentes ethnies induirait des actes contradictoires qui ne favorisent pas la préservation de la race. L importance des actes (conduite, santé et activités sociales) plus élevée chez les CE ayant annoncé une participation élevée à la conservation du Baoulé a été probablement liée aux moyens, aux objectifs et à la préférence accordée à cette race. Ces considérations liées à l éleveur influencent la gestion du troupeau (Lalba, Badini et Kleene, 1999). Les actes de conduite et de santé élevés des CE membres des groupements contrairement aux autres, montrent que l ouverture à la vie associative est un facteur positif de conservation de la race Baoulé. Ceux qui y adhèrent marquent leur volonté de progrès social. On note l influence de l ouverture d esprit chez les alphabétisés qui se sont démarqués par des actes (conduite, santé et activités sociales) plus fréquents que chez les analphabètes, favorables à la conservation. Le souci d amélioration de l élevage du Baoulé à travers les actes posés explique probablement cette différence. Il en été de même des CE qui pratiquent l élevage du Baoulé depuis moins de 15 ans avec des actes de conduite et activités sociales plus élevés, favorisant la conservation que ceux de 15 ans et plus. Néanmoins, les actes de reproduction faibles et négatifs des alphabétisés indiquent la pratique du métissage, préjudiciable à la préservation de la race malgré l importance notoire des autres actes posés. L influence de l âge sur les actes de reproduction montre que les CE plus âgés (de 40 ans et plus) préserveraient mieux la race Baoulé que ceux de moins de 40 ans. On peut faire des rapprochements sur les actes de reproduction avec le groupe des CE alphabétisés qui se retrouvent surtout dans la classe d âge de moins de 40 ans. Ce groupe a une moyenne des actes de reproduction très faible et négative indiquant une tendance au métissage. Les déterminants de l importance des actes par l analyse de régression renforcent les critères analysés et confirment que les facteurs socio-économiques de l exploitation conditionnent la conservation in situ de la race. Conclusion Les pratiques d élevage du taurin Baoulé mises en œuvre en Pays Lobi du Burkina Faso participent globalement à la préservation de cette race, au regard du bilan positif des actes et pratiques conservatoires. Les troupeaux Baoulés purs majoritaires en gardiennage individuel et la castration systématique des mâles d attelage achetés sont des stratégies de conservation mises en œuvre. L importance des actes et leurs déterminants indiquent l influence des facteurs socio-économiques qui concourent à la préservation ou pas du Baoulé. Les actes de conduite, santé et activités sociales participent à la préservation chez les CE ayant annoncé une participation élevée, ceux gérant seuls leur élevage, ceux âgés et ceux membres d un groupement. Leur importance plus nette chez ceux avec des troupeaux mixtes, les migrants, les alphabétisés et ceux avec attelage ne favorise pas la conservation car l importance des actes de reproduction y est faible ou négative. La stratégie d accumulation de richesse par la constitution des troupeaux mixtes s amplifiera dans la région, augmentant ainsi la menace d extinction du Baoulé. Les comportements et intentions des chefs d exploitation tendent vers l adoption du métissage dans l avenir. Le conservatisme Lobi a contribué à préserver le Baoulé du métissage. Mais des changements socioculturels et économiques prévisibles modifieront le comportement de la jeune génération vis à vis de ce bétail. Ainsi, l élevage du Baoulé ne s inscrira pas seulement dans une optique socioculturelle ayant contribué à sa préservation. Il faut à l instar du taurin N Dama qui fédère le Mali, la Guinée, le Gambie et le Sénégal autour d un projet de développement et de préservation, envisager des actions similaires pour le taurin Baoulé en pays Lobi. Les Lobi, principaux éleveurs de taurin Baoulé, se partagent entre le Burkina Faso, la Côte d Ivoire et le Ghana. Les facteurs ethnique et linguistique constituent un atout capital, favorisant l organisation des éleveurs autour du taurin Baoulé pour sa préservation.

171 178 L.Y. Mopate et al. Remerciement Les auteurs tiennent à remercier les différents bailleurs qui ont contribué financièrement à la réalisation de cette étude, ainsi que les responsables scientifiques du CIRDES et de l UPB à différents niveaux pour leurs appuis multiformes, les enquêteurs et les agro-éleveurs qui n ont ménagés leur temps pour ce travail. Références bibliographiques Achukwi, M.D., Musongong, G.A Trypanosomosis in the Doayo/ Namchi (Bos taurus) and zebu White Fulani (Bos indicus) cattle in Faro Division, North Cameroon. Journal of Applied Biosciences 15: Camus, E., Landais, E., Poivey, J.P Structure génétique du cheptel bovin sédentaire du Nord de la Côte d Ivoire. Perspectives d avenir en fonction de la diffusion croissante du sang zébu. Revue d Elevage et de Médecine Vétérinaire des Pays Tropicaux 34(2): Chupin, D Le rôle de la biotechnologie de la reproduction pour la conservation des ressources génétiques animales. Animal Genetic Resources Information 14: CIPEA (Centre international pour l élevage en Afrique) Le bétail trypanotolérant d Afrique occidentale et centrale. Tome I. Situation général, Étude de systèmes, Addis - Abeba, Ethiopie, 155. Hubert, J.P Enquête informelle versus enquête formelle? Cas des paysannats de Impo au Nord du Burundi. Annales de Gembloux 98: Guinko, S Végétation de la Haute-Volta. Thèse de Doctorat d État ès sciences naturelles. Université de Bordeau III, France, annexes. Kienou, J.D., Sanou, H., van Bronswijk, P Monographie provinciale du Poni, étude du milieu. Ministère de l Économie et des Finances, Ouagadougou (Burkina Faso), 176. Lalba, A., Badini, O., Kleene, P Etablissement des priorités géographiques de recherche et développement en élevage par combinaison d études de diagnostic informelles et formelles: Burkina Faso. In: Roeleveld et Broek (éds.), les systèmes d élevage: orienter la recherche. Institut royal des tropiques et Centre technique de coopération agricole et rurale (Cta), Landais, E Analyse des systèmes d élevage bovin sédentaire du Nord de la Côte d Ivoire. Thèse Doctorat d Etat IEMVT, Maisons-Alfort, France. 2 vol., 759. Landais, E Principes de modélisation des systèmes d élevage: approches graphiques. Cahiers de Recherche-Développement 32(3): Landais, E., Balent, G Introduction à l étude des systèmes d élevage extensif. In: Landais (éd.), pratiques d élevage extensif : identifier, modéliser, évaluer. INRA, Études et Recherches sur les Systèmes Agraires et le Développement 27: Lankoandé, Y.F Développement des bovins trypanotolérants au Burkina Faso: défis - potentialités - opportunités. Mémoire d Ingénieur de développement rural, option Elevage, IDR - Université Polytechnique de Bobo-Dioulasso (Burkina Faso), 68 + annexes. Lhoste, Ph Sélection génétique. In: Daget et Godron (coordonnateurs), Pastoralisme: Troupeaux, espaces et sociétés. Hatier - Aupelf - Uref (éds.), Maillard, J.C., Congo, I., Bassinga, A., Cuveillier, J.F Immunogénétique du taurin Baoulé en Pays Lobi (Burkina Faso). 1. Environnement de cette population trypanotolérante. Revue d Elevage et de Médecine Vétérinaire des Pays Tropicaux, 45(1): Moazami Goudari, K., Belemsaga, D.M., Geriotti, G., Laloë, D., Fagbohoun, F., Kouagou, N. T., Sdibé, I., Codja, V., Crimella, M.C., Grosclaude, F., Touré, S.M Caractérisation de la race Somba à l aide des marqueurs moléculaires. Revue d Elevage et de Médecine Vétérinaire des Pays Tropicaux, 54(2): Mopaté, L.Y Conservation In Situ du bovin Baoulé: pratiques paysannes et stratégies en «Pays Lobi» du Burkina Faso. Mémoire du D. E.A en Gestion Intégrée des Ressources Animales, option «Systèmes de Productions Animales». Institut de Développement Rural (IDR), l Université Polytechnique de Bobo-Dioulasso (UPB), Burkina Faso, 89. Ouédraogo, D Analyse socio-économique des pratiques de gestion de la trypanosomose animale et les facteurs associés au développement de la chimiorésistance dans la province du Kénédougou (Burkina Faso). Thèse de Doctorat Unique ès Sciences Économiques, UFR/SEG, Université de Ouagadougou, 210. Pleine, T., Schmithüsen, F., Sorg, J-P Logiques paysannes et reboisement: le cas d Andohavary sur les hauts plateaux de Madagascar. Cahiers de Recherche-Développement, 42: Rege, J.E.O., Aboagye, G.S., Tawah, C.L Shorthorn cattle in West and Central Africa. I. Origin, distribution, classification and population statistics. World Animal Review 78(1): Sokouri, D.P., Loukou, N.E., Yapi-Gnaoré, C.V., Mondeil, F., Gnangbe, F Caractérisation phénotypique des bovins à viande (Bos taurus et Bos indicus) au centre (Bouaké) et au nord (Korhogo) de la Côte d Ivoire. AGRI 40: Sokouri, D.P., Yapi-Gnaoré, C.V., N Guett, A.S.P., Loukou, N.E., Kouao, B.J., Touré, G., Sangaré, A., Kouassi, A Utilisation et gestion des races taurines locales sous la pression des croisements avec les zébus dans les régions Centre et Nord de la Côte d Ivoire. Journal of Animal & Plant Sciences 5(2): Sigué, H., Kamuanga, M.J.B Le secteur de l élevage et le politique de la santé animale au Burkina Faso: Revue et perspectives. Document de travail 2, 36. Tano, K., Kamuanga, M.J.B., Faminow, M.D., Swallow, B.M Adoption and demand for trypanotolerant cattle in the subhumide zone of West Africa. Journal of Agriculture and Environment for International Development 95(2 3): Thiombiano, D La banque de semence du Centre de Recherche sur la Trypanosomose Animale (CRTA). In: Chupin, Wagner et Wilson (éds.), amélioration génétique des bovins en Afrique de l Ouest. FAO, Collection Production et Santé Animal 110. Actes de l atelier tenu à Banjul (Gambie) en octobre 1992, Teissier, J.H Relations entre techniques et pratiques. Bulletin INRAP 38. Vianney Labé, Rudy Palm Statistique, empirique, informelle: quelle enquête pour la collecte d information sur les exploitations agricoles. Cahiers Agricultures 8(5):

172 Animal Genetic Resources, 2014, 54, Food and Agriculture Organization of the United Nations, 2014 doi: /s Influence of coat colour on Chamarita sheep phenotypes, behaviour, welfare and performances M. Pascual-Alonso 1, G.C. Miranda-de la Lama 2, L. Aguayo-Ulloa 1, M. Villarroel 3, S. Alierta 1 and G.A. Maria 1 1 Department of Animal Production and Food Science, Faculty of Veterinary Medicine, University of Zaragoza, Miguel Servet 177, E-50013, (50013) Zaragoza, Spain; 2 Group of Animal Welfare and Sustainable Livestock Production, Department of Food Science, Metropolitan Autonomous University, UAM-Lerma, State of México, México; 3 Department of Animal Science, E.T.S.I.A. Polytechnic University of Madrid, Madrid, Spain Summary Many rustic breeds under low levels of genetic selection maintain variations in coat colour, which local farmers link to production traits, but few quantitative analyses have been made of the effect of coat colour on several zootechnical parameters of importance. The aim of the study was to describe differences in morphology, production, behaviour and welfare of adult sheep ewes (n = 50) in a Chamarita breed flock in terms of coat colour. The wool and skin colour of Chamarita sheep are quite variable, with white, black and galana sheep. Morphological measurements (weight, body length, height at withers and thorax circumference) were not significantly different for different coat colours but white sheep tended to be larger. Average litter size (1.29 lambs per ewe) was also similar, but white coated sheep gave birth to larger lambs. White-mother lambs also had a higher average daily growth during the milking period, although final weaning weight was similar to black ones. Most (>70 percent) of all the aggressive interactions observed were performed by white sheep, while affiliative interactions (friendly interactions that promote group cohesion) were similar among coat colour. Regarding the welfare assessment, most indicators were similar among sheep with different coat colour except for glucose in blood plasma, which was higher in white sheep. Chamarita ewes are relatively small compared with other breeds, and well adapted to their environment, including the production system and harder climatic conditions. Their performance is within the average of local breeds and still has margins for improvement, which underline the importance of their conservation. Keywords: Chamarita ewes, coat colour, morphology, social interactions, stress response Résumé Beaucoup de races considérées comme rustiques et ayant été soumises à une faible pression de sélection génétique conservent des variations dans la couleur de leur robe que certains éleveurs associent à des caractères productifs. Cependant, l effet de la couleur de la robe sur des paramètres zootechniques d importance n a guère été étudié. L objectif de cette étude a été de décrire les différences en morphologie, production, comportement et bien-être selon la couleur de la robe de 50 brebis adultes de la race Chamarita. Les mesures morphologiques (poids, longueur, hauteur au garrot, circonférence thoracique) n ont pas différé significativement entre les couleurs de robe même si les brebis blanches ont tendu à être plus grandes. La prolificité (1.29 agneaux/brebis) a aussi été similaire mais les brebis blanches ont eu des agneaux plus grands. Les agneaux de mères blanches ont aussi eu un gain moyen quotidien plus élevé, bien que le poids final ait été similaire à celui des agneaux de mères noires. La plupart (>70 pour cent) des interactions agressives observées ont été réalisées par des brebis blanches, alors que le pourcentage d interactions affiliatives a été similaire pour les deux couleurs de robe. En ce qui concerne l évaluation du bien-être, aucune différence n a été décelée entre les couleurs de robe pour la plupart des indicateurs, hormis le glucose sanguin qui a été plus élevé chez les brebis blanches. Les brebis Chamaritas sont relativement petites par rapport à d autres races et sont bien adaptées à leur environnement productif. Leurs performances productives sont satisfaisantes et peuvent encore être améliorées, comme quoi la conservation de cette race s avère importante. Mots-clés: brebis Chamaritas, couleur de robe, morphologie, réponse au stress, interactions sociales Resumen Muchas razas consideradas rústicas y sometidas a poca presión de selección genética mantienen variaciones en el color de capa que algunos ganaderos relacionan con caracteres productivos, pero se han realizado pocos estudios respecto al efecto del color de capa en algunos parámetros zootécnicos de importancia. El objetivo del presente estudio es describir diferencias en morfología, producción, comportamiento y bienestar de ovejas adultas (n = 50) de raza Chamarita en cuanto a color de capa. Las medidas morfológicas (peso, largura, altura a la cruz, circunferencia torácica) no resultaron significativamente diferentes entre colores de capa aunque las ovejas blancas tendían a ser más grandes. La prolificidad (1.29 corderos/oveja) también fue similar pero las ovejas blancas tuvieron corderos más grandes. Los corderos de madres blancas también tuvieron una mayor ganancia diaria (ADG), aunque el peso final fue similar al de los corderos de madres negras. La mayoría (>70 percent) de las interacciones agresivas observadas fueron ejecutadas Correspondence to: M. Pascual-Alonso mpascual@unizar.es 179

173 180 M. Pascual-Alonso et al. por ovejas blancas, mientras que las interacciones afiliativas resultaron similares en ambos colores de capa. En cuanto a la evaluación del bienestar, la mayoría de los indicadores no presentaron diferencias entre colores de capa excepto la glucosa en sangre que fue mayor en las ovejas blancas. Las ovejas Chamaritas son relativamente pequeñas comparadas con otras razas, y están bien adaptadas a su ambiente productivo. Sus caracteres productivos son razonables con un buen margen de mejoría, demostrando la importancia de su conservación. Palabras clave: ovejas Chamaritas, color de capa, morfología, respuesta de estrés, interacciones sociales Submitted 25 September 2013; accepted 16 January 2014 Introduction Many local breeds under low selection pressure maintain large variations in coat colour, which the farmers link to production traits, but a few quantitative analyses have been made to substantiate the effect of coat colour on differences in morphology, production, behaviour and welfare. Chamarita sheep are a breed from La Rioja (Spain), included in the Official Catalogue of Spanish Livestock Breeds under Endangered Native Breeds. The Chamarita population size is about sheep and it is officially recognized as an endangered breed to be preserved. The herd book was created in 2007 by the Chamarita Sheep Association of La Rioja (or AROCHA as it is known by its Spanish acronym), which has an official recognition to promote and control the breed. Most herds are found in the mountains of lower La Rioja and in the Cidacos and Linares valleys (Warren, 2011). Animals are small in size (adult rams kg, ewes kg; Barrio, Falceto and Doménech, 1991), probably as an adaptive response to their hard living conditions and in their extensive and sustainable production system, which uses local resources (Álvarez and Arruga, 2007). The wool and skin colour of Chamarita sheep are quite variable, with white, black (roya) and rarely galana (a combination of white, black and brown skin with white and black/roya wool) coats (Figure 1). The roya sheep is born black but their wool turns in reddish-black with age, looking like brown. This type of coat is locally called roya. The black skin is mainly located in the face. It is likely that, as with other breeds in the region, breeders have selected for hardiness (through fertility and morphological characteristics) and adaptation to the environment and less for pure white wool. Nowadays there are many black sheep in herds, since it is associated with positive zootechnical traits (Doménech et al., 1992). Black animals usually have a white spot on the nape (crowned), and another at the distal end of the tail ( puntiblancos). Sometimes, the spot of the neck extends over the head, forehead and nasal area (caretos). Currently, Chamarita sheep are mostly used for meat production, both for suckling lamb (12 14 kg live weight) and a weaned lamb with a short indoor fattening period (20 22 kg live weight; Doménech et al., 1992). The Chamarito lamb was recognized as a quality brand in Figure 1. Differences in coat colour in Chamarita sheep.

174 Influence of coat colour on Chamarita sheep phenotypes, behaviour, welfare and performances 181 April 2010 (PROCORCHA). Thus, farmers now used a production system that ensures compliance with strict animal welfare health programmes, respecting the origin of sheep fed on extensive pastures and lambs produced in La Rioja on their mother s milk. In this study, we describe the morphological, productive and welfare traits of adult female Chamarita sheep and the influence of coat colour on these traits. Materials and methods A flock of 50 Chamarita adult ewes were transported from La Rioja to the Animal Experimentation Support Service (SAEA) of the University of Zaragoza, Aragón (41 41_N). All sheep used were raised, transported and slaughtered according to current regulations of the European Community Commission (1986) for Scientific Procedure Establishments. All the protocols were approved by the Animal Experimentation Ethics Committee of the University of Zaragoza. This study was the collaboration between the Autonomous Regions of Aragon and La Rioja (Spain). Animals The adult multiparous ewes were housed in pens during pregnancy and lactation (2 m 2 per ewe), fed twice a day (at and 15.00) with pellet concentrate (11.5 MJ ME per kg DM and 15.5 percent crude protein; 0.3 kg per ewe) and ad libitum lucerne chaff (Medicagosativa). The pen was equipped with a metallic water trough (1.5 m 0.60 m) and two metallic feeders (4.5 m 0.80 m, 27 cm per ewe) and a lick stone for minerals. Most ewes were inseminated by controlled natural mating before leaving the source farm and lambed at the University a few weeks after arrival. The lambs were kept together in the same pen with their mothers (2 m 2 per sheep and lamb). Flock prolificacy was calculated as the total number of lambs born per total ewes lambing. Lambs were weighed at birth (BW) and at weaning (WW). Pre-weaning average daily gain (ADG) was estimated by the difference WW BW divided by the total milking period (30 days). Morphological measurements of ewes One day before blood sampling, body length (BL), height at withers (HW) and thorax circumference (TC) were measured as in Miranda-de la Lama et al. (2011), and each ewe was weighed using a portable digital weighing scale. Social behaviour Social behaviour was evaluated by direct observation. All ewes were individually identified and marked with 30 cm-high numbers and letters painted on the sides and rump with washable paint for sheep marking (PeintureMarquage Mouton ). A platform with a seat 3 m above the ground wasusedtoobservetheflock from a distance. The ewes were observed 6 h daily, from 8:00 to 10:00 h, 12:00 to 14:00 h and 16:00 to 18:00 h for 16 consecutive days (96 h of observation) by the same trained observer. A behaviour sampling technique was used to record all social interactions (agonistic and non-agonistic behaviour). Agonistic interactions with contact included butts (when the ewe used the front of her head to make contact with another ewe), pushes (when a ewe used other parts of her body to make contact with another ewe) and bites (when one ewe bit another ewe s body using her teeth). Agonistic interactions without contact included threats (when a ewe turned towards or approached another individual with her head down and then lunged without making contact), and chase (when a ewe actively moved towards another individual, causing the latter to walk or run away). Non-agonistic interactions with contact included licking (when a ewe passed her tongue over the body of another individual) and grooming (when a ewe groomed another ewe s body using her teeth). Non-agonistic interactions without contact included sniffing (when a ewe sniffed another ewe s body) and the flehmen response (when a ewe retracted the upper lip, wrinkled the nose and bared the gums in the presence of another ewe). Physiological welfare indicators Blood samples were taken by jugular venipuncture with vacuum tubes during the dry period to evaluate physiological responses to stress (two 10 ml tubes per animal, with and without anticoagulant, Ethylenediaminetetraacetic acid EDTA-K3). Blood was sampled using the necessary precautions to avoid sampling error on stress indicators. Samples were kept on ice for a maximum of 2 h and taken to the laboratory for routine haematological measurements. EDTA plasma and serum were centrifuged at rpm for 10 min and aliquots were frozen and kept at 30 C until analysed. An automatic particle counter (Microcell counter F-800 and auto dilutor AD-260, SysmexTM both) was used to count red blood cells (RBC) and white blood cells (WBC) (number per mm 3 ), haemoglobin (g/dl) and haematocrit (percent). The leukocyte formula was estimated from blood swabs on clean slides. Staining was performed by the rapid panoptic method using dyes from QuímicaClinicaAplicada Inc. With an optic immersion microscope we counted and identified 100 leucocytes per sample (neutrophils, lymphocytes, eosinophils, basophils and monocytes). The neutrophil/lymphocyte ratio (N/L) was used an indicator of chronic stress (Lawrence and Rushen, 1993). Serum samples were used to determine the concentration of glucose (mg/dl, Ref. Glucose AE2-17), and the activity of creatine kinase (CK) (UI/L) (Ref. CK. NAC AE1-13) with a multianalyser ACE (Clinical Chemistry System) and reagents from Alfa Wasserman. Serum concentration of non-esterified fatty acid (NEFA) levels was analysed by a multianalyser ACE (Clinical Chemistry System of the Alfa Wasserman), with commercial

175 182 M. Pascual-Alonso et al. kits (NEFA C Ref of the Wako). The concentration of cortisol was determined from plasma (EDTA-K3) by enzyme immunoassay using an in home-kit (validated by Chacón et al., 2004). Each sample was determined in duplicate from 50 μl of plasma and the results were expressed in nmol/l, with the corresponding controls. Variation coefficients of the analysis, inter- and intra-assay, were 7 and 8 percent, respectively. The concentration of lactate was determined using a Sigma Diagnostic kit (lactate no ) and spectrophotometer (Lambda 5, Perkin Elmer). Data analysis Data were analysed using the least squares methods of the GLM procedure using SAS/STAT (9.1 SAS Inst. Inc., Cary, NC, USA) by SAS (1998), fitting a one-way model with a fixed effect of coat colour (two levels) within the ewe data. The general representation of the model used was: y = Xb+e, where y was an N 1 vector of records, b denoted the fixed effect in the model with the association matrix X and e was the vector of residual effects. A probability of P < 0.05 values was considered statistically significant. Results and discussion The proportion of coat colours frequencies, according to the association of breeders statistics, are 45 percent white coat and 55 percent non-white coat (including the minority galana coat). According with the genetic study performed by Álvarez and Arruga (2007), the Chamarita phenotypes are considered to be as a unique population. Morphological measurements Coat colour had little effect on morphological measurements (Table 1). Live weight of the adult ewes averaged (±6.74) kg with no significant differences between white and black sheep. The average TC was (±4.58) cm, HW (±3.7) cm and BL from chest to tail (±4.25) cm. The average live weight and HW were higher than described by Barrio, Falceto and Doménech (1991) (live weight 36.6 kg, HW 59.8 cm). Chamarita sheep may have a lower body weight after food shortages in extensive systems, which was not the case in our experimental herd where sheep were fed twice daily with concentrate and had water and straw ad libitum. However, BL was shorter than in Barrio, Falceto and Doménech (1991) (90.5 cm), but the latter authors did not find differences between body measurements of white and black sheep. Although body weight was not significantly higher in white ewes, there was a tendency for black sheep to be lighter. The small size of Chamarita ewes compared with more productive breeds (e.g. Rasa aragonesa: kg; Manchega: weighing over 45 kg; Merino: kg; and Table 1. Least-square means (±S.E.) for the effect of coat colour on morphological, physiological and behavioural traits in Chamarita sheep. Whole flock White ewes Black ewes Morphological measurements Bodyweight (kg) ± ± ± 1.16 TC (cm) ± ± ± 0.79 HW (cm) ± ± ± 0.64 BL (cm) ± ± ± 0.74 Physiological measurements Cortisol (nmol/l) ± ± ± 5.14 Glucose (mg/dl) ± ± 8.48 a ± 5.72 b Lactate (mg/dl) ± ± ± 2.45 NEFA (nmol/l) ± ± ± CK (UI/l) ± ± ± Ratio N/L 0.81 ± ± ± 0.06 WBC (10 3 /mm 3 ) 8.14 ± ± ± 0.7 RBC (10 3 /mm 3 ) 8.85 ± ± ± 0.23 HTC (%) 28.8 ± ± ± 0.75 Social interactions Agonistic ± ± a ± b Affiliative ± ± ± 3.79 Different letters (a, b) represent significant differences ( p < 0.05) between groups. NEFA, non-esterified fatty acid; CK, creatine kinase; Ratio N/L, neutrophil/lymphocyte ratio; WBC, white blood cells; RBC, red blood cells; HCT, haematocrit. Assaf: kg) is consistent with the morphology and weights of most breeds considered rustic and well adapted to a harsh environment (Catalogue Livestock Breeds in Spain). Such is the case of Ojalada (between 35 and 45 kg of body weight), Mallorquina (30 40 kg of body weight) or Gallega (25 35 kg in the mountain ecotype). This adaptation is due to food shortages, thereby reducing their energy needs (Álvarez and Arruga, 2007), allowing them to make use the resources at their disposal. In that regard, black sheep that tended to be smaller than white sheep may be more adapted to a harsh environment than white sheep (Álvarez and Arruga, 2007). Productive traits The average litter size of the whole flock was 1.29 (±0.36) lambs born per ewe lambing with no significant differences between coat colours (Table 2). Litter size is slightly higher than in Doménech et al. (1992), who found a tendency for single births (1.1 lambs per ewe). However, Table 2. Least-square means (±S.E.) for the effect of coat colour on productive traits in Chamarita sheep. Whole flock White ewes Black ewes Prolificacy 1.29 ± ± ± 0.07 BW of lambs (kg) 3.63 ± ± 0.16 a 3.53 ± 0.1 b ADG of lambs (g) ± ± a ± 8.73 b WW of lambs (kg) ± ± ± 0.39 Different letters (a, b) represent significant differences ( p < 0.05) between groups. BLW, birth weight; ADG, average daily gain during the milking period; WLW, weaning weight.

176 Influence of coat colour on Chamarita sheep phenotypes, behaviour, welfare and performances 183 the latter authors suggested that an increased food supply could increase prolificacy to 1.5 lambs per ewe and delivery. These data confirm the potential for improvement of the Chamarita sheep when environmental conditions are optimal. The lambs weighed 3.63 (±0.59) kg at birth, and those born to white mothers were significantly (P 0.05) heavier than these born (+9 percent) black ones. Balda, Chavarri and Doménech (1981) reported a birth weight of 2.6 kg in pure lines and 3.6 kg in industrial crossings. In our study, the herd was pure and the crosses were with Chamarito males, so probably the increased birth weight in lambs may be due to higher feeding rate in our experimental group. The fact that lambs from white females weighed more at birth than those from black females is probably due to larger body size of their mothers is associated with higher consumption of concentrate. This hypothesis partially corroborates the data on blood glucose levels and social interactions of the herd, which suggest that white sheep were dominant (see below). Similarly, lambs from white ewes had a higher ADG during the milking period (+14.7 percent, P 0.05), compared with the ADG of the whole herd (±48.56) g. The ADG data support the hypothesis that white females could eat more and therefore produce more milk for their lambs. However, the average weaning weight of the lambs (11.95 ± 2.12 kg) was not significantly different among different coat coloured mothers. That implies that a homogeneous product can be obtained at weaning regardless of coat colour. Social behaviour Of the (±107.03) aggressive interactions observed (Table 1), most were performed by white sheep (+71.5 percent, P 0, 05). That suggests a higher dominance status for white sheep (Orgeur, Mimouni and Signoret, 1990; Barroso, Alados and Boza, 2000), tied with the fact that they were heavier and larger (Miranda-de Lama et al., 2011). Dominance is a mechanism that regulates social behaviour and priority access to available resources, since high dominance animals have priority access to resources in intensive production conditions (Barroso, Alados and Boza, 2000). That coincides with the higher glucose levels in white sheep. The levels of aggressive interactions observed within a stable herd are consistent with the idea of Doménech et al. (1992) thatchamarita behaviour is more like goat behaviour, with a lively temperament and instinct for survival, linked to foraging behaviour and browsing. Affiliative interactions totalled (±21.55) for the whole herd, with no significant differences regarding coat colour. Affiliations are a very important part of the social cohesion of a group (Miranda-de la Lama and Mattiello, 2010), but under conditions of social competition for limited resources, animals tend to do without them (Miranda-de la Lama et al., 2011). It is likely that an adaptation to very scarce resources has led to Chamarita to favour aggressive interactions in detriment of affiliative ones in the context of the social group. Physiological welfare indicators Physiological and haematological variables can be used to assess the welfare of animals subject to acute and/or chronic stress (Barnett and Hemsworth, 1990). The results of physiological measurements are presented in Table 1. Cortisol levels averaged (±29.2) nmol/l, glucose (±33.91) mg/dl, lactate (±13.91) mg/dl, NEFA (±0.098) nmol/l and CK (±77.6) IU/l. The N/L ratio averaged 0.81 (±0.37), which suggests there was no immune suppression (Lawrence and Rushen, 1993). None of the sheep had clinical signs of illness. The WBC count was 8.14 (±4) 10 3 /mm 3, RBC count 8.85 (±1.28) 10 3 /mm 3, and haematocrit percentage was 28.8 (±4.27). All physiological measures were within normal ranges for sheep during routine handling procedures and do not suggest that welfare was compromised (Hargreaves and Hutson, 1990). Glucose was the only physiological variable that was significantly higher (P 0.05) in white sheep (by 31.1 percent) than black sheep. Plasma glucose levels increase after acute stress when adrenaline induces muscle glycogen catabolism for its use in gluconeogenesis in the liver (Apple et al., 1995). However, in our case, the higher glucose levels in white sheep could have been the result of their higher feed intake. White sheep tended to be heavier and be the dominant animals in the herd, as reflected in levels of aggressive social interactions, all of which imply higher energy costs. However, we did not find a difference in stress levels between black and white sheep, probably since the group was quite stable and had abundant resources. Conclusions White ewes of the Chamarita breed are slightly larger than the black-coated ewes and give birth to larger lambs. They also behaved more aggressively and had higher plasma concentrations of glucose, underlying their more dominant character. Overall, the productive traits of Chamarita ewes and lambs were quite good compared with other sheep and taking into consideration their small size. Owing to their strong adaptation to the harsh environment of La Rioja sierra, using grazing resources within a sustainable production system, Chamarita breed should be preserved. Appropriate selection and conservation practises will help to improve their performance and help contribute to rural development and sustaining the local agro-ecosystem, preserving a valuable cultural and gastronomic heritage of the Community of La Rioja. Acknowledgements This study was funded by the Spanish Ministry of Science and Innovation (projects AGL /GAN and AGL ). Many thanks to the Autonomous

177 184 M. Pascual-Alonso et al. Community of La Rioja for the Scholarship PhD Programme (Programa de AyudasPredoctoralesdelGobierno de la Rioja) of M. Pascual-Alonso. Special thanks to AROCHA and PROCORCHA associations for their support and their invaluable help. This appreciation is special for the veterinary responsible of these associations, Don José Antonio González. Thank you very much to the staff of the Animal Research Centre of the University of Zaragoza (SAEA). References Álvarez, L. & Arruga, M.V Caracterización de la oveja Chamarita mediante marcadores microsatélites. Estructura genética, genotipado del gen Scrapie y creación de un banco de ADN. Beca de estudios Arnedanos. Apple, J.K., Dikeman, M.E., Minton, J.E., McMurphy, R.M., Fedde, M.R., Leith, D.E. & Unruh, J.A Effects of restraint and isolation stress and epidural blockade on endocrine and blood metabolite status, muscle glycogen metabolism, and incidence of dark-cutting longissimus muscle of sheep. J. Anim. Sci. 7: Balda, I., Chavarri, J.B. & Doménech, J.M La oveja Chamarita I Jornadas Técnicas de ganado ovino en La Rioja-Logroño. Barnett, J.L. & Hemsworth, P.H The validity of physiological and behavioural measures of animal welfare. Appl. Anim. Behav. Sci. 25: Barrio, A.R., Falceto, M.D. & Doménech, J.M La raza Chamarita: Estudio etnológico y morfométrico. XVI Jornadas Científicas de la S.E.O.C. Barroso, F.G., Alados, C.L. & Boza, J Social hierarchy in the domestic goat: effect on food habits and production. Appl. Anim. Behav. Sci. 69: Chacón, G., García-Belenguer, S., Illera, J.C. & Palacio, J Validation of an EIA [enzyme immunoassay] technique for the determination of salivary cortisol in cattle. Span. J. Agric. Res.2: Doménech, J.M., Barrio, A.R., Falceto, M.D. & González Jiménez, J.A La oveja Chamarita. Serie Estudios no 25, Gobierno de La Rioja. Logroño, La Rioja, España. European Commission Scientific procedure and breeding of animals for use in scientific procedure establishments. Directive 86/609/ ECC Hargreaves, A.L. & Hutson, G.D The effect of gentling on heart rate, flight distance and aversion of sheep to a handling procedure. Appl. Anim. Behav. Sci. 26: Lawrence, A.B. & Rushen, J Stereotypic animal behaviour: fundamental and application to welfare. UK, CAB International. Miranda-de la Lama, G.C. & Mattiello, S The importance of social behaviour for goat welfare in livestock farming. Small Rumin. Res. 90: Miranda-de la Lama, G.C., Sepúlveda, W.S., Montaldo, H.H., María, G.A. & Galindo, F Social strategies associated with identity profiles in dairy goats. Appl. Anim. Behav. Sci. 134: Orgeur, P., Mimouni, P. & Signoret, J.P The influence of rearing conditions on the social relationships of young male goats, Capra hircus. Appl. Anim. Behav. Sci. 27: Statistical Analysis System Institute SAS/STATS user s guide (Release 6.03). Cary, NC, SAS Institute. Warren, J Chamarito Lamb Promoted among Restaurant Owners and Butchers.

178 Animal Genetic Resources, 2014, 54, 185. Food and Agriculture Organization of the United Nations, 2014 Recent Publication The camels of Kumbhalgarh. A biodiversity treasure I. Köhler-Rollefson, H. Singh Rathore and A. Rollefson Edited by K. Hardy Lokhit Pashu-Palak Sansthan, Sadri Published in 2013, pp 48. ISBN Available at Camels_Of_Kumbhalgarh_web.pdf doi: /s This publication describes the roles of the camels in the Kumbhalgarh region in Rajastan, India, ecologically sustainable camel herding system of unique biocultural heritage. The publication highlights the importance of the camel herding in the culture and livelihoods of the Raika people, traditional keepers of the camels. The two camel breeds present in the Kumbhalgarh are one-humped or dromedary camels (Camelus dromedarius), sub-set of the Marwari breed. The book provides descriptions of the camels, their feeding behaviour (the book includes an appendix of preferred forage plants containing their medicinal and other uses) and their roles in the functioning of the agro-ecological system, daily routine of the herders and the migration routes, as well as the importance of the traditional knowledge about the ecosystem. The annual cycle of the camel breeders and the management activities are presented, highlighting the unique character of this breeding system. However, a number of problems are threatening this system and these problems are named and discussed. Due to declining incomes and camel populations, urban migration of the members of the herder families and the plans of Rajastan government to declare the part of traditional herding lands as a National Park, this system is threatened by the loss of its agro-ecological and socio-cultural heritage values. However, the authors describe the options which can ensure that the herding system can exist as an integral part of the planned National Park, attract visitors interested in ecological tourism and provide a diversity of camel products such as camel milk, camel milk soap and camel dung paper. The publication concludes with presenting the legal documents which are important to ensure the camel breeders rights and recommendations for enhancing the potential of the camels to contribute to the diversity of the Kumbhalgarh Protected Area. 185

179 Animal Genetic Resources, 2014, 54, 187. Food and Agriculture Organization of the United Nations, 2014 Recent Publication Caractérisation phénotypique des ressources génétiques animales Directives FAO sur la production et la santé animales No. 11 Publication de la version française en 201, pp E-ISBN Available at doi: /s Les Directives sur la caractérisation phénotypique des ressources génétiques animales ont été publiées en français en Ces Directives font partie d une série de publication produites par la FAO pour aider les pays à mettre en œuvre le Plan d action mondial pour les ressources zoogénétiques, qui reconnaît qu «une bonne compréhension des caractéristiques des races est nécessaire pour guider la prise de décision en matière de programmes de développement et de sélection des animaux d élevage» (FAO, 2007). Ces directives sur la caractérisation phénotypique des ressources génétiques animales s adressent à la Priorité stratégique 1 du Plan d action mondial: «Caractérisation, inventaire et surveillance des évolutions et des risques associés». Elles complètent, en particulier, les directives sur la caractérisation génétique moléculaire et sur la réalisation d enquêtes et de suivi pour les ressources génétiques animales. En effet, les trois types d informations phénotypique, génétique et historique sont nécessaires pour caractériser les ressources génétiques animales. Les informations fournies par les études de caractérisation sont primordiales pour planifier la gestion des ressources génétiques animales au niveau local, national, régional et mondial. Ces directives permettent d identifier des différentes races et de décrire de leurs caractéristiques externes et de production dans un environnement et un cadre de gestion donnés, en tenant compte des facteurs socio-économiques qui les affectent. Pour cela, elles se concentrent sur la collecte et l utilisation de l information phénotypique. Par ailleurs, les activités de caractérisation phénotypiques sont difficiles à mettre en œuvre d un point de vue logistique et technique. Ainsi, les directives donnent des conseils sur la façon de mener une étude de caractérisation phénotypique bien ciblée et rentable qui contribue à l amélioration de la gestion des ressources génétiques animales dans le cadre de la mise en œuvre du Plan d action mondial au niveau des pays. De plus, un aperçu des concepts et des approches qui soutiennent les études de caractérisation phénotypique est suivi de conseils pratiques sur la planification et la mise en œuvre des activités sur le terrain, ainsi que sur la gestion et l analyse de données. Enfin, les annexes comprennent des formats génériques de collecte de données pour la caractérisation phénotypique des principales espèces d élevage, ainsi qu un cadre pour l enregistrement de données sur les milieux de production des races. 187

180 Animal Genetic Resources, 2014, 54, 189. Food and Agriculture Organization of the United Nations, 2014 Recent Publication Cattle husbandry in Eastern Europe and China. Structure, development paths and optimisation Edited by A. Kuipers, A. Rozstalnyy and G. Keane Wageningen Academic Publishers Published in 2014, pp. 280 ISBN doi: / doi: /s x This book presents an analysis of the dairy and associated sector developments in Eastern Europe and China. The transition in cattle husbandry in Eastern Europe, resulting in enormous structural changes, but not in an increased production volume, is different from emerging countries, like China, where traditional small scale farming goes together with the remodelling or start of mega farms. Capacity building by means of cooperation appears to be hindered by historical experience. Farm development paths were studied in more depth in Poland, Lithuania and Slovenia using detailed farmer surveys. Farmers strategies, availability of resources, and opportunities and threats were analyzed, and interactive group trainings in strategic management were part of the analysis. The results are presented here. Although farmers showed similar wishes concerning farm development, the local environment and policies determines the degree of achievement. Farm fragmentation, the high percentage of rented land and availability of suitable labour are major constraints. Future EU policies are a concern but new technologies are embraced. This book is a must for those interested in the transition in Eastern Europe. It is indispensible to consultants, marketers, companies, farm leaders and government officials in agriculture. 189